diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,105201 @@ +{ + "best_metric": 0.2650785446166992, + "best_model_checkpoint": "./w2v-bert-2.0-hausa_naijavoices_100h/checkpoint-10000", + "epoch": 41.899441340782126, + "eval_steps": 1000, + "global_step": 15000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002793296089385475, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 7.8652, + "step": 1 + }, + { + "epoch": 0.00558659217877095, + "grad_norm": 19.80986213684082, + "learning_rate": 3.35195530726257e-09, + "loss": 7.2158, + "step": 2 + }, + { + "epoch": 0.008379888268156424, + "grad_norm": 20.03233528137207, + "learning_rate": 6.70391061452514e-09, + "loss": 7.2505, + "step": 3 + }, + { + "epoch": 0.0111731843575419, + "grad_norm": 19.507043838500977, + "learning_rate": 1.005586592178771e-08, + "loss": 7.089, + "step": 4 + }, + { + "epoch": 0.013966480446927373, + "grad_norm": 19.127351760864258, + "learning_rate": 1.340782122905028e-08, + "loss": 6.9385, + "step": 5 + }, + { + "epoch": 0.01675977653631285, + "grad_norm": 19.433597564697266, + "learning_rate": 1.6759776536312847e-08, + "loss": 6.9998, + "step": 6 + }, + { + "epoch": 0.019553072625698324, + "grad_norm": 19.153995513916016, + "learning_rate": 2.011173184357542e-08, + "loss": 6.9769, + "step": 7 + }, + { + "epoch": 0.0223463687150838, + "grad_norm": 19.540985107421875, + "learning_rate": 2.3463687150837986e-08, + "loss": 6.9788, + "step": 8 + }, + { + "epoch": 0.025139664804469275, + "grad_norm": 19.67742347717285, + "learning_rate": 2.681564245810056e-08, + "loss": 7.0126, + "step": 9 + }, + { + "epoch": 0.027932960893854747, + "grad_norm": 19.980894088745117, + "learning_rate": 3.016759776536313e-08, + "loss": 7.061, + "step": 10 + }, + { + "epoch": 0.030726256983240222, + "grad_norm": 19.15444564819336, + "learning_rate": 3.3519553072625695e-08, + "loss": 6.8735, + "step": 11 + }, + { + "epoch": 0.0335195530726257, + "grad_norm": 19.586952209472656, + "learning_rate": 3.687150837988827e-08, + "loss": 6.7931, + "step": 12 + }, + { + "epoch": 0.036312849162011177, + "grad_norm": 19.675617218017578, + "learning_rate": 4.022346368715084e-08, + "loss": 6.9565, + "step": 13 + }, + { + "epoch": 0.03910614525139665, + "grad_norm": 18.771188735961914, + "learning_rate": 4.357541899441341e-08, + "loss": 6.7402, + "step": 14 + }, + { + "epoch": 0.04189944134078212, + "grad_norm": 19.51713752746582, + "learning_rate": 4.692737430167597e-08, + "loss": 6.907, + "step": 15 + }, + { + "epoch": 0.0446927374301676, + "grad_norm": 19.655838012695312, + "learning_rate": 5.027932960893855e-08, + "loss": 6.9306, + "step": 16 + }, + { + "epoch": 0.04748603351955307, + "grad_norm": 20.07324981689453, + "learning_rate": 5.363128491620112e-08, + "loss": 6.978, + "step": 17 + }, + { + "epoch": 0.05027932960893855, + "grad_norm": 18.614059448242188, + "learning_rate": 5.698324022346369e-08, + "loss": 6.6154, + "step": 18 + }, + { + "epoch": 0.05307262569832402, + "grad_norm": 18.90323257446289, + "learning_rate": 6.033519553072626e-08, + "loss": 6.6767, + "step": 19 + }, + { + "epoch": 0.055865921787709494, + "grad_norm": 19.171218872070312, + "learning_rate": 6.368715083798884e-08, + "loss": 6.7296, + "step": 20 + }, + { + "epoch": 0.05865921787709497, + "grad_norm": 18.7718505859375, + "learning_rate": 6.703910614525139e-08, + "loss": 6.6178, + "step": 21 + }, + { + "epoch": 0.061452513966480445, + "grad_norm": 18.965999603271484, + "learning_rate": 7.039106145251397e-08, + "loss": 6.6314, + "step": 22 + }, + { + "epoch": 0.06424581005586592, + "grad_norm": 17.8748836517334, + "learning_rate": 7.374301675977654e-08, + "loss": 6.3603, + "step": 23 + }, + { + "epoch": 0.0670391061452514, + "grad_norm": 18.486740112304688, + "learning_rate": 7.70949720670391e-08, + "loss": 6.4326, + "step": 24 + }, + { + "epoch": 0.06983240223463687, + "grad_norm": 17.962934494018555, + "learning_rate": 8.044692737430168e-08, + "loss": 6.1237, + "step": 25 + }, + { + "epoch": 0.07262569832402235, + "grad_norm": 21.872880935668945, + "learning_rate": 8.379888268156424e-08, + "loss": 7.8487, + "step": 26 + }, + { + "epoch": 0.07541899441340782, + "grad_norm": 20.045007705688477, + "learning_rate": 8.715083798882682e-08, + "loss": 7.191, + "step": 27 + }, + { + "epoch": 0.0782122905027933, + "grad_norm": 19.554519653320312, + "learning_rate": 9.05027932960894e-08, + "loss": 7.024, + "step": 28 + }, + { + "epoch": 0.08100558659217877, + "grad_norm": 20.75141716003418, + "learning_rate": 9.385474860335195e-08, + "loss": 7.2539, + "step": 29 + }, + { + "epoch": 0.08379888268156424, + "grad_norm": 19.710092544555664, + "learning_rate": 9.720670391061452e-08, + "loss": 6.9741, + "step": 30 + }, + { + "epoch": 0.08659217877094973, + "grad_norm": 20.012550354003906, + "learning_rate": 1.005586592178771e-07, + "loss": 7.0224, + "step": 31 + }, + { + "epoch": 0.0893854748603352, + "grad_norm": 19.30960464477539, + "learning_rate": 1.0391061452513966e-07, + "loss": 6.8263, + "step": 32 + }, + { + "epoch": 0.09217877094972067, + "grad_norm": 19.61741065979004, + "learning_rate": 1.0726256983240224e-07, + "loss": 6.8718, + "step": 33 + }, + { + "epoch": 0.09497206703910614, + "grad_norm": 20.285778045654297, + "learning_rate": 1.1061452513966481e-07, + "loss": 6.9991, + "step": 34 + }, + { + "epoch": 0.09776536312849161, + "grad_norm": 19.67317008972168, + "learning_rate": 1.1396648044692738e-07, + "loss": 6.8747, + "step": 35 + }, + { + "epoch": 0.1005586592178771, + "grad_norm": 19.69019889831543, + "learning_rate": 1.1731843575418995e-07, + "loss": 6.8254, + "step": 36 + }, + { + "epoch": 0.10335195530726257, + "grad_norm": 19.811420440673828, + "learning_rate": 1.2067039106145253e-07, + "loss": 6.8285, + "step": 37 + }, + { + "epoch": 0.10614525139664804, + "grad_norm": 20.398401260375977, + "learning_rate": 1.240223463687151e-07, + "loss": 6.9878, + "step": 38 + }, + { + "epoch": 0.10893854748603352, + "grad_norm": 19.154117584228516, + "learning_rate": 1.2737430167597768e-07, + "loss": 6.6629, + "step": 39 + }, + { + "epoch": 0.11173184357541899, + "grad_norm": 19.64495277404785, + "learning_rate": 1.3072625698324022e-07, + "loss": 6.727, + "step": 40 + }, + { + "epoch": 0.11452513966480447, + "grad_norm": 19.511510848999023, + "learning_rate": 1.3407821229050278e-07, + "loss": 6.6674, + "step": 41 + }, + { + "epoch": 0.11731843575418995, + "grad_norm": 19.7799072265625, + "learning_rate": 1.3743016759776537e-07, + "loss": 6.7227, + "step": 42 + }, + { + "epoch": 0.12011173184357542, + "grad_norm": 19.206729888916016, + "learning_rate": 1.4078212290502793e-07, + "loss": 6.5941, + "step": 43 + }, + { + "epoch": 0.12290502793296089, + "grad_norm": 19.65355682373047, + "learning_rate": 1.441340782122905e-07, + "loss": 6.6932, + "step": 44 + }, + { + "epoch": 0.12569832402234637, + "grad_norm": 20.070892333984375, + "learning_rate": 1.4748603351955308e-07, + "loss": 6.62, + "step": 45 + }, + { + "epoch": 0.12849162011173185, + "grad_norm": 19.747493743896484, + "learning_rate": 1.5083798882681565e-07, + "loss": 6.6005, + "step": 46 + }, + { + "epoch": 0.13128491620111732, + "grad_norm": 18.991737365722656, + "learning_rate": 1.541899441340782e-07, + "loss": 6.4281, + "step": 47 + }, + { + "epoch": 0.1340782122905028, + "grad_norm": 18.761545181274414, + "learning_rate": 1.575418994413408e-07, + "loss": 6.3675, + "step": 48 + }, + { + "epoch": 0.13687150837988826, + "grad_norm": 18.168710708618164, + "learning_rate": 1.6089385474860336e-07, + "loss": 6.1896, + "step": 49 + }, + { + "epoch": 0.13966480446927373, + "grad_norm": 17.709774017333984, + "learning_rate": 1.6424581005586593e-07, + "loss": 5.9514, + "step": 50 + }, + { + "epoch": 0.1424581005586592, + "grad_norm": 22.716880798339844, + "learning_rate": 1.675977653631285e-07, + "loss": 7.6307, + "step": 51 + }, + { + "epoch": 0.1452513966480447, + "grad_norm": 20.577722549438477, + "learning_rate": 1.7094972067039108e-07, + "loss": 6.9813, + "step": 52 + }, + { + "epoch": 0.14804469273743018, + "grad_norm": 20.145862579345703, + "learning_rate": 1.7430167597765364e-07, + "loss": 6.8188, + "step": 53 + }, + { + "epoch": 0.15083798882681565, + "grad_norm": 20.239681243896484, + "learning_rate": 1.776536312849162e-07, + "loss": 6.8036, + "step": 54 + }, + { + "epoch": 0.15363128491620112, + "grad_norm": 20.943864822387695, + "learning_rate": 1.810055865921788e-07, + "loss": 6.9366, + "step": 55 + }, + { + "epoch": 0.1564245810055866, + "grad_norm": 20.45275115966797, + "learning_rate": 1.8435754189944133e-07, + "loss": 6.7846, + "step": 56 + }, + { + "epoch": 0.15921787709497207, + "grad_norm": 20.86522102355957, + "learning_rate": 1.877094972067039e-07, + "loss": 6.8089, + "step": 57 + }, + { + "epoch": 0.16201117318435754, + "grad_norm": 21.111095428466797, + "learning_rate": 1.9106145251396648e-07, + "loss": 6.8436, + "step": 58 + }, + { + "epoch": 0.164804469273743, + "grad_norm": 20.26334571838379, + "learning_rate": 1.9441340782122904e-07, + "loss": 6.6349, + "step": 59 + }, + { + "epoch": 0.16759776536312848, + "grad_norm": 21.028430938720703, + "learning_rate": 1.977653631284916e-07, + "loss": 6.7612, + "step": 60 + }, + { + "epoch": 0.17039106145251395, + "grad_norm": 20.546945571899414, + "learning_rate": 2.011173184357542e-07, + "loss": 6.6866, + "step": 61 + }, + { + "epoch": 0.17318435754189945, + "grad_norm": 20.8492374420166, + "learning_rate": 2.0446927374301676e-07, + "loss": 6.6921, + "step": 62 + }, + { + "epoch": 0.17597765363128492, + "grad_norm": 20.75248146057129, + "learning_rate": 2.0782122905027932e-07, + "loss": 6.6165, + "step": 63 + }, + { + "epoch": 0.1787709497206704, + "grad_norm": 20.429805755615234, + "learning_rate": 2.111731843575419e-07, + "loss": 6.5063, + "step": 64 + }, + { + "epoch": 0.18156424581005587, + "grad_norm": 20.115514755249023, + "learning_rate": 2.1452513966480447e-07, + "loss": 6.4592, + "step": 65 + }, + { + "epoch": 0.18435754189944134, + "grad_norm": 20.894168853759766, + "learning_rate": 2.1787709497206704e-07, + "loss": 6.5696, + "step": 66 + }, + { + "epoch": 0.1871508379888268, + "grad_norm": 20.521568298339844, + "learning_rate": 2.2122905027932963e-07, + "loss": 6.4332, + "step": 67 + }, + { + "epoch": 0.18994413407821228, + "grad_norm": 21.15107536315918, + "learning_rate": 2.245810055865922e-07, + "loss": 6.523, + "step": 68 + }, + { + "epoch": 0.19273743016759776, + "grad_norm": 20.667062759399414, + "learning_rate": 2.2793296089385475e-07, + "loss": 6.4367, + "step": 69 + }, + { + "epoch": 0.19553072625698323, + "grad_norm": 20.02320671081543, + "learning_rate": 2.3128491620111734e-07, + "loss": 6.2834, + "step": 70 + }, + { + "epoch": 0.19832402234636873, + "grad_norm": 20.006704330444336, + "learning_rate": 2.346368715083799e-07, + "loss": 6.2239, + "step": 71 + }, + { + "epoch": 0.2011173184357542, + "grad_norm": 22.127769470214844, + "learning_rate": 2.3798882681564244e-07, + "loss": 6.1406, + "step": 72 + }, + { + "epoch": 0.20391061452513967, + "grad_norm": 19.56688117980957, + "learning_rate": 2.4134078212290506e-07, + "loss": 6.0862, + "step": 73 + }, + { + "epoch": 0.20670391061452514, + "grad_norm": 19.709672927856445, + "learning_rate": 2.446927374301676e-07, + "loss": 6.0997, + "step": 74 + }, + { + "epoch": 0.20949720670391062, + "grad_norm": 19.160417556762695, + "learning_rate": 2.480446927374302e-07, + "loss": 5.8591, + "step": 75 + }, + { + "epoch": 0.2122905027932961, + "grad_norm": 25.197885513305664, + "learning_rate": 2.5139664804469275e-07, + "loss": 7.4639, + "step": 76 + }, + { + "epoch": 0.21508379888268156, + "grad_norm": 22.145742416381836, + "learning_rate": 2.5474860335195536e-07, + "loss": 6.7441, + "step": 77 + }, + { + "epoch": 0.21787709497206703, + "grad_norm": 21.72708511352539, + "learning_rate": 2.5810055865921787e-07, + "loss": 6.5987, + "step": 78 + }, + { + "epoch": 0.2206703910614525, + "grad_norm": 21.904579162597656, + "learning_rate": 2.6145251396648043e-07, + "loss": 6.5743, + "step": 79 + }, + { + "epoch": 0.22346368715083798, + "grad_norm": 21.911531448364258, + "learning_rate": 2.6480446927374305e-07, + "loss": 6.5286, + "step": 80 + }, + { + "epoch": 0.22625698324022347, + "grad_norm": 22.572294235229492, + "learning_rate": 2.6815642458100556e-07, + "loss": 6.6001, + "step": 81 + }, + { + "epoch": 0.22905027932960895, + "grad_norm": 21.975139617919922, + "learning_rate": 2.715083798882682e-07, + "loss": 6.4497, + "step": 82 + }, + { + "epoch": 0.23184357541899442, + "grad_norm": 21.721298217773438, + "learning_rate": 2.7486033519553074e-07, + "loss": 6.3596, + "step": 83 + }, + { + "epoch": 0.2346368715083799, + "grad_norm": 21.47066879272461, + "learning_rate": 2.782122905027933e-07, + "loss": 6.2913, + "step": 84 + }, + { + "epoch": 0.23743016759776536, + "grad_norm": 22.684877395629883, + "learning_rate": 2.8156424581005586e-07, + "loss": 6.4388, + "step": 85 + }, + { + "epoch": 0.24022346368715083, + "grad_norm": 22.228368759155273, + "learning_rate": 2.849162011173185e-07, + "loss": 6.3357, + "step": 86 + }, + { + "epoch": 0.2430167597765363, + "grad_norm": 23.088401794433594, + "learning_rate": 2.88268156424581e-07, + "loss": 6.4428, + "step": 87 + }, + { + "epoch": 0.24581005586592178, + "grad_norm": 22.520559310913086, + "learning_rate": 2.916201117318436e-07, + "loss": 6.3404, + "step": 88 + }, + { + "epoch": 0.24860335195530725, + "grad_norm": 23.169347763061523, + "learning_rate": 2.9497206703910617e-07, + "loss": 6.2085, + "step": 89 + }, + { + "epoch": 0.25139664804469275, + "grad_norm": 22.493738174438477, + "learning_rate": 2.9832402234636873e-07, + "loss": 6.2382, + "step": 90 + }, + { + "epoch": 0.2541899441340782, + "grad_norm": 21.95569610595703, + "learning_rate": 3.016759776536313e-07, + "loss": 6.1045, + "step": 91 + }, + { + "epoch": 0.2569832402234637, + "grad_norm": 23.02143669128418, + "learning_rate": 3.050279329608939e-07, + "loss": 6.2332, + "step": 92 + }, + { + "epoch": 0.25977653631284914, + "grad_norm": 22.416419982910156, + "learning_rate": 3.083798882681564e-07, + "loss": 6.1144, + "step": 93 + }, + { + "epoch": 0.26256983240223464, + "grad_norm": 22.419029235839844, + "learning_rate": 3.1173184357541904e-07, + "loss": 6.0606, + "step": 94 + }, + { + "epoch": 0.26536312849162014, + "grad_norm": 21.63622283935547, + "learning_rate": 3.150837988826816e-07, + "loss": 5.934, + "step": 95 + }, + { + "epoch": 0.2681564245810056, + "grad_norm": 23.216066360473633, + "learning_rate": 3.184357541899441e-07, + "loss": 6.0766, + "step": 96 + }, + { + "epoch": 0.2709497206703911, + "grad_norm": 21.40949058532715, + "learning_rate": 3.217877094972067e-07, + "loss": 5.83, + "step": 97 + }, + { + "epoch": 0.2737430167597765, + "grad_norm": 21.275466918945312, + "learning_rate": 3.251396648044693e-07, + "loss": 5.7788, + "step": 98 + }, + { + "epoch": 0.276536312849162, + "grad_norm": 21.06237030029297, + "learning_rate": 3.2849162011173185e-07, + "loss": 5.6801, + "step": 99 + }, + { + "epoch": 0.27932960893854747, + "grad_norm": 21.72116470336914, + "learning_rate": 3.318435754189944e-07, + "loss": 5.6486, + "step": 100 + }, + { + "epoch": 0.28212290502793297, + "grad_norm": 28.859825134277344, + "learning_rate": 3.35195530726257e-07, + "loss": 6.8866, + "step": 101 + }, + { + "epoch": 0.2849162011173184, + "grad_norm": 25.7071475982666, + "learning_rate": 3.3854748603351954e-07, + "loss": 6.379, + "step": 102 + }, + { + "epoch": 0.2877094972067039, + "grad_norm": 25.998271942138672, + "learning_rate": 3.4189944134078215e-07, + "loss": 6.3432, + "step": 103 + }, + { + "epoch": 0.2905027932960894, + "grad_norm": 25.658390045166016, + "learning_rate": 3.452513966480447e-07, + "loss": 6.2206, + "step": 104 + }, + { + "epoch": 0.29329608938547486, + "grad_norm": 24.727445602416992, + "learning_rate": 3.486033519553073e-07, + "loss": 6.0769, + "step": 105 + }, + { + "epoch": 0.29608938547486036, + "grad_norm": 25.908058166503906, + "learning_rate": 3.5195530726256984e-07, + "loss": 6.156, + "step": 106 + }, + { + "epoch": 0.2988826815642458, + "grad_norm": 25.290451049804688, + "learning_rate": 3.553072625698324e-07, + "loss": 6.0467, + "step": 107 + }, + { + "epoch": 0.3016759776536313, + "grad_norm": 24.850561141967773, + "learning_rate": 3.5865921787709497e-07, + "loss": 5.9472, + "step": 108 + }, + { + "epoch": 0.30446927374301674, + "grad_norm": 24.252979278564453, + "learning_rate": 3.620111731843576e-07, + "loss": 5.8208, + "step": 109 + }, + { + "epoch": 0.30726256983240224, + "grad_norm": 25.877660751342773, + "learning_rate": 3.653631284916201e-07, + "loss": 5.9759, + "step": 110 + }, + { + "epoch": 0.3100558659217877, + "grad_norm": 26.26590919494629, + "learning_rate": 3.6871508379888266e-07, + "loss": 5.957, + "step": 111 + }, + { + "epoch": 0.3128491620111732, + "grad_norm": 25.65629005432129, + "learning_rate": 3.720670391061453e-07, + "loss": 5.858, + "step": 112 + }, + { + "epoch": 0.31564245810055863, + "grad_norm": 26.053638458251953, + "learning_rate": 3.754189944134078e-07, + "loss": 5.8499, + "step": 113 + }, + { + "epoch": 0.31843575418994413, + "grad_norm": 25.392114639282227, + "learning_rate": 3.787709497206704e-07, + "loss": 5.7229, + "step": 114 + }, + { + "epoch": 0.32122905027932963, + "grad_norm": 25.433963775634766, + "learning_rate": 3.8212290502793296e-07, + "loss": 5.7061, + "step": 115 + }, + { + "epoch": 0.3240223463687151, + "grad_norm": 24.9486083984375, + "learning_rate": 3.854748603351955e-07, + "loss": 5.6012, + "step": 116 + }, + { + "epoch": 0.3268156424581006, + "grad_norm": 25.850622177124023, + "learning_rate": 3.888268156424581e-07, + "loss": 5.6477, + "step": 117 + }, + { + "epoch": 0.329608938547486, + "grad_norm": 25.277982711791992, + "learning_rate": 3.921787709497207e-07, + "loss": 5.5669, + "step": 118 + }, + { + "epoch": 0.3324022346368715, + "grad_norm": 25.602333068847656, + "learning_rate": 3.955307262569832e-07, + "loss": 5.5535, + "step": 119 + }, + { + "epoch": 0.33519553072625696, + "grad_norm": 26.293222427368164, + "learning_rate": 3.9888268156424583e-07, + "loss": 5.5815, + "step": 120 + }, + { + "epoch": 0.33798882681564246, + "grad_norm": 26.32081413269043, + "learning_rate": 4.022346368715084e-07, + "loss": 5.5168, + "step": 121 + }, + { + "epoch": 0.3407821229050279, + "grad_norm": 24.17746353149414, + "learning_rate": 4.0558659217877096e-07, + "loss": 5.3294, + "step": 122 + }, + { + "epoch": 0.3435754189944134, + "grad_norm": 24.004043579101562, + "learning_rate": 4.089385474860335e-07, + "loss": 5.2384, + "step": 123 + }, + { + "epoch": 0.3463687150837989, + "grad_norm": 24.090822219848633, + "learning_rate": 4.1229050279329613e-07, + "loss": 5.2789, + "step": 124 + }, + { + "epoch": 0.34916201117318435, + "grad_norm": 21.800615310668945, + "learning_rate": 4.1564245810055864e-07, + "loss": 5.0153, + "step": 125 + }, + { + "epoch": 0.35195530726256985, + "grad_norm": Infinity, + "learning_rate": 4.1564245810055864e-07, + "loss": 6.1091, + "step": 126 + }, + { + "epoch": 0.3547486033519553, + "grad_norm": 29.60270881652832, + "learning_rate": 4.189944134078212e-07, + "loss": 5.6752, + "step": 127 + }, + { + "epoch": 0.3575418994413408, + "grad_norm": 30.055761337280273, + "learning_rate": 4.223463687150838e-07, + "loss": 5.6863, + "step": 128 + }, + { + "epoch": 0.36033519553072624, + "grad_norm": 29.652475357055664, + "learning_rate": 4.2569832402234633e-07, + "loss": 5.6058, + "step": 129 + }, + { + "epoch": 0.36312849162011174, + "grad_norm": 28.582883834838867, + "learning_rate": 4.2905027932960895e-07, + "loss": 5.4857, + "step": 130 + }, + { + "epoch": 0.3659217877094972, + "grad_norm": 28.472801208496094, + "learning_rate": 4.324022346368715e-07, + "loss": 5.4199, + "step": 131 + }, + { + "epoch": 0.3687150837988827, + "grad_norm": 28.52730941772461, + "learning_rate": 4.357541899441341e-07, + "loss": 5.3921, + "step": 132 + }, + { + "epoch": 0.3715083798882682, + "grad_norm": 27.762935638427734, + "learning_rate": 4.3910614525139664e-07, + "loss": 5.2869, + "step": 133 + }, + { + "epoch": 0.3743016759776536, + "grad_norm": 28.015796661376953, + "learning_rate": 4.4245810055865925e-07, + "loss": 5.2532, + "step": 134 + }, + { + "epoch": 0.3770949720670391, + "grad_norm": 28.77267074584961, + "learning_rate": 4.4581005586592176e-07, + "loss": 5.2572, + "step": 135 + }, + { + "epoch": 0.37988826815642457, + "grad_norm": 28.106008529663086, + "learning_rate": 4.491620111731844e-07, + "loss": 5.2126, + "step": 136 + }, + { + "epoch": 0.38268156424581007, + "grad_norm": 27.434890747070312, + "learning_rate": 4.5251396648044694e-07, + "loss": 5.112, + "step": 137 + }, + { + "epoch": 0.3854748603351955, + "grad_norm": 26.91579246520996, + "learning_rate": 4.558659217877095e-07, + "loss": 5.0503, + "step": 138 + }, + { + "epoch": 0.388268156424581, + "grad_norm": 27.299861907958984, + "learning_rate": 4.5921787709497207e-07, + "loss": 5.0401, + "step": 139 + }, + { + "epoch": 0.39106145251396646, + "grad_norm": 24.966264724731445, + "learning_rate": 4.625698324022347e-07, + "loss": 4.906, + "step": 140 + }, + { + "epoch": 0.39385474860335196, + "grad_norm": 25.824954986572266, + "learning_rate": 4.659217877094972e-07, + "loss": 4.9065, + "step": 141 + }, + { + "epoch": 0.39664804469273746, + "grad_norm": 26.124618530273438, + "learning_rate": 4.692737430167598e-07, + "loss": 4.8957, + "step": 142 + }, + { + "epoch": 0.3994413407821229, + "grad_norm": 26.035594940185547, + "learning_rate": 4.7262569832402237e-07, + "loss": 4.8629, + "step": 143 + }, + { + "epoch": 0.4022346368715084, + "grad_norm": 24.190969467163086, + "learning_rate": 4.759776536312849e-07, + "loss": 4.7476, + "step": 144 + }, + { + "epoch": 0.40502793296089384, + "grad_norm": 24.30831527709961, + "learning_rate": 4.793296089385475e-07, + "loss": 4.7047, + "step": 145 + }, + { + "epoch": 0.40782122905027934, + "grad_norm": 24.03734016418457, + "learning_rate": 4.826815642458101e-07, + "loss": 4.6656, + "step": 146 + }, + { + "epoch": 0.4106145251396648, + "grad_norm": 22.20307159423828, + "learning_rate": 4.860335195530726e-07, + "loss": 4.5773, + "step": 147 + }, + { + "epoch": 0.4134078212290503, + "grad_norm": 20.917247772216797, + "learning_rate": 4.893854748603352e-07, + "loss": 4.4776, + "step": 148 + }, + { + "epoch": 0.41620111731843573, + "grad_norm": 21.282203674316406, + "learning_rate": 4.927374301675977e-07, + "loss": 4.4688, + "step": 149 + }, + { + "epoch": 0.41899441340782123, + "grad_norm": 18.756784439086914, + "learning_rate": 4.960893854748604e-07, + "loss": 4.3891, + "step": 150 + }, + { + "epoch": 0.42178770949720673, + "grad_norm": 31.841278076171875, + "learning_rate": 4.99441340782123e-07, + "loss": 4.924, + "step": 151 + }, + { + "epoch": 0.4245810055865922, + "grad_norm": 26.897445678710938, + "learning_rate": 5.027932960893855e-07, + "loss": 4.6871, + "step": 152 + }, + { + "epoch": 0.4273743016759777, + "grad_norm": 24.674047470092773, + "learning_rate": 5.06145251396648e-07, + "loss": 4.5651, + "step": 153 + }, + { + "epoch": 0.4301675977653631, + "grad_norm": 25.739898681640625, + "learning_rate": 5.094972067039107e-07, + "loss": 4.5789, + "step": 154 + }, + { + "epoch": 0.4329608938547486, + "grad_norm": 22.770484924316406, + "learning_rate": 5.128491620111732e-07, + "loss": 4.4441, + "step": 155 + }, + { + "epoch": 0.43575418994413406, + "grad_norm": 23.948375701904297, + "learning_rate": 5.162011173184357e-07, + "loss": 4.4682, + "step": 156 + }, + { + "epoch": 0.43854748603351956, + "grad_norm": 22.47032928466797, + "learning_rate": 5.195530726256984e-07, + "loss": 4.3849, + "step": 157 + }, + { + "epoch": 0.441340782122905, + "grad_norm": 20.41448211669922, + "learning_rate": 5.229050279329609e-07, + "loss": 4.3154, + "step": 158 + }, + { + "epoch": 0.4441340782122905, + "grad_norm": 19.884496688842773, + "learning_rate": 5.262569832402235e-07, + "loss": 4.2927, + "step": 159 + }, + { + "epoch": 0.44692737430167595, + "grad_norm": 19.0008544921875, + "learning_rate": 5.296089385474861e-07, + "loss": 4.2518, + "step": 160 + }, + { + "epoch": 0.44972067039106145, + "grad_norm": 18.71428680419922, + "learning_rate": 5.329608938547486e-07, + "loss": 4.222, + "step": 161 + }, + { + "epoch": 0.45251396648044695, + "grad_norm": 16.711612701416016, + "learning_rate": 5.363128491620111e-07, + "loss": 4.1722, + "step": 162 + }, + { + "epoch": 0.4553072625698324, + "grad_norm": 16.717391967773438, + "learning_rate": 5.396648044692738e-07, + "loss": 4.16, + "step": 163 + }, + { + "epoch": 0.4581005586592179, + "grad_norm": 15.959756851196289, + "learning_rate": 5.430167597765364e-07, + "loss": 4.1358, + "step": 164 + }, + { + "epoch": 0.46089385474860334, + "grad_norm": 14.417510986328125, + "learning_rate": 5.463687150837989e-07, + "loss": 4.0909, + "step": 165 + }, + { + "epoch": 0.46368715083798884, + "grad_norm": 12.826011657714844, + "learning_rate": 5.497206703910615e-07, + "loss": 4.0702, + "step": 166 + }, + { + "epoch": 0.4664804469273743, + "grad_norm": 13.086868286132812, + "learning_rate": 5.530726256983241e-07, + "loss": 4.0272, + "step": 167 + }, + { + "epoch": 0.4692737430167598, + "grad_norm": 11.886052131652832, + "learning_rate": 5.564245810055866e-07, + "loss": 4.012, + "step": 168 + }, + { + "epoch": 0.4720670391061452, + "grad_norm": 11.402913093566895, + "learning_rate": 5.597765363128492e-07, + "loss": 3.9883, + "step": 169 + }, + { + "epoch": 0.4748603351955307, + "grad_norm": 11.201967239379883, + "learning_rate": 5.631284916201117e-07, + "loss": 4.0029, + "step": 170 + }, + { + "epoch": 0.4776536312849162, + "grad_norm": 11.003469467163086, + "learning_rate": 5.664804469273743e-07, + "loss": 3.9836, + "step": 171 + }, + { + "epoch": 0.48044692737430167, + "grad_norm": 10.996882438659668, + "learning_rate": 5.69832402234637e-07, + "loss": 3.9433, + "step": 172 + }, + { + "epoch": 0.48324022346368717, + "grad_norm": 11.566518783569336, + "learning_rate": 5.731843575418995e-07, + "loss": 3.9513, + "step": 173 + }, + { + "epoch": 0.4860335195530726, + "grad_norm": 12.240893363952637, + "learning_rate": 5.76536312849162e-07, + "loss": 3.9372, + "step": 174 + }, + { + "epoch": 0.4888268156424581, + "grad_norm": 12.765401840209961, + "learning_rate": 5.798882681564246e-07, + "loss": 3.8632, + "step": 175 + }, + { + "epoch": 0.49162011173184356, + "grad_norm": 11.989778518676758, + "learning_rate": 5.832402234636872e-07, + "loss": 3.8779, + "step": 176 + }, + { + "epoch": 0.49441340782122906, + "grad_norm": 10.576717376708984, + "learning_rate": 5.865921787709497e-07, + "loss": 3.8372, + "step": 177 + }, + { + "epoch": 0.4972067039106145, + "grad_norm": 10.475711822509766, + "learning_rate": 5.899441340782123e-07, + "loss": 3.8063, + "step": 178 + }, + { + "epoch": 0.5, + "grad_norm": 10.24913501739502, + "learning_rate": 5.932960893854748e-07, + "loss": 3.7947, + "step": 179 + }, + { + "epoch": 0.5027932960893855, + "grad_norm": 10.285189628601074, + "learning_rate": 5.966480446927375e-07, + "loss": 3.7751, + "step": 180 + }, + { + "epoch": 0.505586592178771, + "grad_norm": 10.363862991333008, + "learning_rate": 6.000000000000001e-07, + "loss": 3.7556, + "step": 181 + }, + { + "epoch": 0.5083798882681564, + "grad_norm": 10.241353988647461, + "learning_rate": 6.033519553072626e-07, + "loss": 3.7343, + "step": 182 + }, + { + "epoch": 0.5111731843575419, + "grad_norm": 10.140996932983398, + "learning_rate": 6.067039106145251e-07, + "loss": 3.7089, + "step": 183 + }, + { + "epoch": 0.5139664804469274, + "grad_norm": 10.210321426391602, + "learning_rate": 6.100558659217878e-07, + "loss": 3.7049, + "step": 184 + }, + { + "epoch": 0.5167597765363129, + "grad_norm": 10.53237533569336, + "learning_rate": 6.134078212290503e-07, + "loss": 3.6723, + "step": 185 + }, + { + "epoch": 0.5195530726256983, + "grad_norm": 10.367179870605469, + "learning_rate": 6.167597765363128e-07, + "loss": 3.6731, + "step": 186 + }, + { + "epoch": 0.5223463687150838, + "grad_norm": 9.747180938720703, + "learning_rate": 6.201117318435755e-07, + "loss": 3.6399, + "step": 187 + }, + { + "epoch": 0.5251396648044693, + "grad_norm": 9.934556007385254, + "learning_rate": 6.234636871508381e-07, + "loss": 3.631, + "step": 188 + }, + { + "epoch": 0.5279329608938548, + "grad_norm": 9.799208641052246, + "learning_rate": 6.268156424581006e-07, + "loss": 3.6148, + "step": 189 + }, + { + "epoch": 0.5307262569832403, + "grad_norm": 9.911898612976074, + "learning_rate": 6.301675977653632e-07, + "loss": 3.5895, + "step": 190 + }, + { + "epoch": 0.5335195530726257, + "grad_norm": 9.212740898132324, + "learning_rate": 6.335195530726257e-07, + "loss": 3.5583, + "step": 191 + }, + { + "epoch": 0.5363128491620112, + "grad_norm": 9.001689910888672, + "learning_rate": 6.368715083798882e-07, + "loss": 3.5445, + "step": 192 + }, + { + "epoch": 0.5391061452513967, + "grad_norm": 9.300618171691895, + "learning_rate": 6.402234636871509e-07, + "loss": 3.5169, + "step": 193 + }, + { + "epoch": 0.5418994413407822, + "grad_norm": 8.590590476989746, + "learning_rate": 6.435754189944134e-07, + "loss": 3.5231, + "step": 194 + }, + { + "epoch": 0.5446927374301676, + "grad_norm": 8.971132278442383, + "learning_rate": 6.46927374301676e-07, + "loss": 3.4963, + "step": 195 + }, + { + "epoch": 0.547486033519553, + "grad_norm": 8.398039817810059, + "learning_rate": 6.502793296089386e-07, + "loss": 3.4898, + "step": 196 + }, + { + "epoch": 0.5502793296089385, + "grad_norm": 8.892096519470215, + "learning_rate": 6.536312849162012e-07, + "loss": 3.474, + "step": 197 + }, + { + "epoch": 0.553072625698324, + "grad_norm": 7.99041223526001, + "learning_rate": 6.569832402234637e-07, + "loss": 3.4624, + "step": 198 + }, + { + "epoch": 0.5558659217877095, + "grad_norm": 8.12684154510498, + "learning_rate": 6.603351955307263e-07, + "loss": 3.4587, + "step": 199 + }, + { + "epoch": 0.5586592178770949, + "grad_norm": 8.900531768798828, + "learning_rate": 6.636871508379888e-07, + "loss": 3.4191, + "step": 200 + }, + { + "epoch": 0.5614525139664804, + "grad_norm": 12.388908386230469, + "learning_rate": 6.670391061452514e-07, + "loss": 3.4585, + "step": 201 + }, + { + "epoch": 0.5642458100558659, + "grad_norm": 9.069743156433105, + "learning_rate": 6.70391061452514e-07, + "loss": 3.3942, + "step": 202 + }, + { + "epoch": 0.5670391061452514, + "grad_norm": 8.6379976272583, + "learning_rate": 6.737430167597766e-07, + "loss": 3.3726, + "step": 203 + }, + { + "epoch": 0.5698324022346368, + "grad_norm": 7.357120513916016, + "learning_rate": 6.770949720670391e-07, + "loss": 3.3492, + "step": 204 + }, + { + "epoch": 0.5726256983240223, + "grad_norm": 7.2352213859558105, + "learning_rate": 6.804469273743017e-07, + "loss": 3.3214, + "step": 205 + }, + { + "epoch": 0.5754189944134078, + "grad_norm": 6.187373161315918, + "learning_rate": 6.837988826815643e-07, + "loss": 3.3035, + "step": 206 + }, + { + "epoch": 0.5782122905027933, + "grad_norm": 5.8618645668029785, + "learning_rate": 6.871508379888268e-07, + "loss": 3.3151, + "step": 207 + }, + { + "epoch": 0.5810055865921788, + "grad_norm": 5.341691493988037, + "learning_rate": 6.905027932960894e-07, + "loss": 3.2707, + "step": 208 + }, + { + "epoch": 0.5837988826815642, + "grad_norm": 5.05117130279541, + "learning_rate": 6.938547486033519e-07, + "loss": 3.2774, + "step": 209 + }, + { + "epoch": 0.5865921787709497, + "grad_norm": 5.515730857849121, + "learning_rate": 6.972067039106146e-07, + "loss": 3.2387, + "step": 210 + }, + { + "epoch": 0.5893854748603352, + "grad_norm": 5.579941272735596, + "learning_rate": 7.005586592178771e-07, + "loss": 3.2309, + "step": 211 + }, + { + "epoch": 0.5921787709497207, + "grad_norm": 5.205024242401123, + "learning_rate": 7.039106145251397e-07, + "loss": 3.2432, + "step": 212 + }, + { + "epoch": 0.5949720670391061, + "grad_norm": 5.6559319496154785, + "learning_rate": 7.072625698324022e-07, + "loss": 3.2311, + "step": 213 + }, + { + "epoch": 0.5977653631284916, + "grad_norm": 10.072444915771484, + "learning_rate": 7.106145251396648e-07, + "loss": 3.2296, + "step": 214 + }, + { + "epoch": 0.6005586592178771, + "grad_norm": 5.75700569152832, + "learning_rate": 7.139664804469274e-07, + "loss": 3.2185, + "step": 215 + }, + { + "epoch": 0.6033519553072626, + "grad_norm": 5.9261040687561035, + "learning_rate": 7.173184357541899e-07, + "loss": 3.1983, + "step": 216 + }, + { + "epoch": 0.6061452513966481, + "grad_norm": 4.855809211730957, + "learning_rate": 7.206703910614524e-07, + "loss": 3.1761, + "step": 217 + }, + { + "epoch": 0.6089385474860335, + "grad_norm": 4.909560203552246, + "learning_rate": 7.240223463687152e-07, + "loss": 3.1839, + "step": 218 + }, + { + "epoch": 0.611731843575419, + "grad_norm": 4.325473308563232, + "learning_rate": 7.273743016759777e-07, + "loss": 3.1623, + "step": 219 + }, + { + "epoch": 0.6145251396648045, + "grad_norm": 4.5805864334106445, + "learning_rate": 7.307262569832402e-07, + "loss": 3.1562, + "step": 220 + }, + { + "epoch": 0.61731843575419, + "grad_norm": 4.084714889526367, + "learning_rate": 7.340782122905028e-07, + "loss": 3.1628, + "step": 221 + }, + { + "epoch": 0.6201117318435754, + "grad_norm": 4.171990871429443, + "learning_rate": 7.374301675977653e-07, + "loss": 3.1514, + "step": 222 + }, + { + "epoch": 0.6229050279329609, + "grad_norm": 4.16370964050293, + "learning_rate": 7.407821229050279e-07, + "loss": 3.139, + "step": 223 + }, + { + "epoch": 0.6256983240223464, + "grad_norm": 3.9163060188293457, + "learning_rate": 7.441340782122905e-07, + "loss": 3.1305, + "step": 224 + }, + { + "epoch": 0.6284916201117319, + "grad_norm": 4.226376533508301, + "learning_rate": 7.474860335195531e-07, + "loss": 3.1058, + "step": 225 + }, + { + "epoch": 0.6312849162011173, + "grad_norm": 13.980515480041504, + "learning_rate": 7.508379888268156e-07, + "loss": 3.1675, + "step": 226 + }, + { + "epoch": 0.6340782122905028, + "grad_norm": 8.082894325256348, + "learning_rate": 7.541899441340783e-07, + "loss": 3.0988, + "step": 227 + }, + { + "epoch": 0.6368715083798883, + "grad_norm": 8.367555618286133, + "learning_rate": 7.575418994413408e-07, + "loss": 3.1129, + "step": 228 + }, + { + "epoch": 0.6396648044692738, + "grad_norm": 6.037868499755859, + "learning_rate": 7.608938547486033e-07, + "loss": 3.0652, + "step": 229 + }, + { + "epoch": 0.6424581005586593, + "grad_norm": 5.335470199584961, + "learning_rate": 7.642458100558659e-07, + "loss": 3.0464, + "step": 230 + }, + { + "epoch": 0.6452513966480447, + "grad_norm": 3.876671075820923, + "learning_rate": 7.675977653631285e-07, + "loss": 3.0362, + "step": 231 + }, + { + "epoch": 0.6480446927374302, + "grad_norm": 4.040086269378662, + "learning_rate": 7.70949720670391e-07, + "loss": 3.0345, + "step": 232 + }, + { + "epoch": 0.6508379888268156, + "grad_norm": 3.3280045986175537, + "learning_rate": 7.743016759776537e-07, + "loss": 3.0273, + "step": 233 + }, + { + "epoch": 0.6536312849162011, + "grad_norm": 5.24937629699707, + "learning_rate": 7.776536312849162e-07, + "loss": 3.0235, + "step": 234 + }, + { + "epoch": 0.6564245810055865, + "grad_norm": 5.83313512802124, + "learning_rate": 7.810055865921788e-07, + "loss": 3.0296, + "step": 235 + }, + { + "epoch": 0.659217877094972, + "grad_norm": 6.986160755157471, + "learning_rate": 7.843575418994414e-07, + "loss": 3.0073, + "step": 236 + }, + { + "epoch": 0.6620111731843575, + "grad_norm": 4.929813385009766, + "learning_rate": 7.877094972067039e-07, + "loss": 3.0155, + "step": 237 + }, + { + "epoch": 0.664804469273743, + "grad_norm": 5.291236400604248, + "learning_rate": 7.910614525139664e-07, + "loss": 3.013, + "step": 238 + }, + { + "epoch": 0.6675977653631285, + "grad_norm": 3.8202602863311768, + "learning_rate": 7.94413407821229e-07, + "loss": 3.0017, + "step": 239 + }, + { + "epoch": 0.6703910614525139, + "grad_norm": 3.7304632663726807, + "learning_rate": 7.977653631284917e-07, + "loss": 2.9769, + "step": 240 + }, + { + "epoch": 0.6731843575418994, + "grad_norm": 3.027048349380493, + "learning_rate": 8.011173184357542e-07, + "loss": 3.0037, + "step": 241 + }, + { + "epoch": 0.6759776536312849, + "grad_norm": 3.700498342514038, + "learning_rate": 8.044692737430168e-07, + "loss": 2.9856, + "step": 242 + }, + { + "epoch": 0.6787709497206704, + "grad_norm": 2.9869723320007324, + "learning_rate": 8.078212290502793e-07, + "loss": 2.9851, + "step": 243 + }, + { + "epoch": 0.6815642458100558, + "grad_norm": 4.070263385772705, + "learning_rate": 8.111731843575419e-07, + "loss": 2.9869, + "step": 244 + }, + { + "epoch": 0.6843575418994413, + "grad_norm": 2.795139789581299, + "learning_rate": 8.145251396648045e-07, + "loss": 2.9529, + "step": 245 + }, + { + "epoch": 0.6871508379888268, + "grad_norm": 3.5354831218719482, + "learning_rate": 8.17877094972067e-07, + "loss": 2.9791, + "step": 246 + }, + { + "epoch": 0.6899441340782123, + "grad_norm": 2.4891409873962402, + "learning_rate": 8.212290502793295e-07, + "loss": 2.9541, + "step": 247 + }, + { + "epoch": 0.6927374301675978, + "grad_norm": 2.929004430770874, + "learning_rate": 8.245810055865923e-07, + "loss": 2.9708, + "step": 248 + }, + { + "epoch": 0.6955307262569832, + "grad_norm": 2.911832571029663, + "learning_rate": 8.279329608938548e-07, + "loss": 2.9494, + "step": 249 + }, + { + "epoch": 0.6983240223463687, + "grad_norm": 5.461743354797363, + "learning_rate": 8.312849162011173e-07, + "loss": 2.9225, + "step": 250 + }, + { + "epoch": 0.7011173184357542, + "grad_norm": 11.146185874938965, + "learning_rate": 8.346368715083799e-07, + "loss": 2.9543, + "step": 251 + }, + { + "epoch": 0.7039106145251397, + "grad_norm": 6.480380535125732, + "learning_rate": 8.379888268156424e-07, + "loss": 2.9118, + "step": 252 + }, + { + "epoch": 0.7067039106145251, + "grad_norm": 5.089816093444824, + "learning_rate": 8.41340782122905e-07, + "loss": 2.8899, + "step": 253 + }, + { + "epoch": 0.7094972067039106, + "grad_norm": 3.555894136428833, + "learning_rate": 8.446927374301676e-07, + "loss": 2.9015, + "step": 254 + }, + { + "epoch": 0.7122905027932961, + "grad_norm": 2.6799733638763428, + "learning_rate": 8.480446927374302e-07, + "loss": 2.883, + "step": 255 + }, + { + "epoch": 0.7150837988826816, + "grad_norm": 3.125065803527832, + "learning_rate": 8.513966480446927e-07, + "loss": 2.8557, + "step": 256 + }, + { + "epoch": 0.7178770949720671, + "grad_norm": 4.432955265045166, + "learning_rate": 8.547486033519554e-07, + "loss": 2.8932, + "step": 257 + }, + { + "epoch": 0.7206703910614525, + "grad_norm": 6.380085468292236, + "learning_rate": 8.581005586592179e-07, + "loss": 2.8731, + "step": 258 + }, + { + "epoch": 0.723463687150838, + "grad_norm": 4.993497371673584, + "learning_rate": 8.614525139664804e-07, + "loss": 2.8623, + "step": 259 + }, + { + "epoch": 0.7262569832402235, + "grad_norm": 4.5500006675720215, + "learning_rate": 8.64804469273743e-07, + "loss": 2.8757, + "step": 260 + }, + { + "epoch": 0.729050279329609, + "grad_norm": 2.299506187438965, + "learning_rate": 8.681564245810056e-07, + "loss": 2.8663, + "step": 261 + }, + { + "epoch": 0.7318435754189944, + "grad_norm": 2.540473222732544, + "learning_rate": 8.715083798882681e-07, + "loss": 2.8655, + "step": 262 + }, + { + "epoch": 0.7346368715083799, + "grad_norm": 2.9556398391723633, + "learning_rate": 8.748603351955308e-07, + "loss": 2.8639, + "step": 263 + }, + { + "epoch": 0.7374301675977654, + "grad_norm": 3.4782748222351074, + "learning_rate": 8.782122905027933e-07, + "loss": 2.8564, + "step": 264 + }, + { + "epoch": 0.7402234636871509, + "grad_norm": 3.001394748687744, + "learning_rate": 8.815642458100559e-07, + "loss": 2.854, + "step": 265 + }, + { + "epoch": 0.7430167597765364, + "grad_norm": 2.0149693489074707, + "learning_rate": 8.849162011173185e-07, + "loss": 2.8386, + "step": 266 + }, + { + "epoch": 0.7458100558659218, + "grad_norm": 2.475633144378662, + "learning_rate": 8.88268156424581e-07, + "loss": 2.8543, + "step": 267 + }, + { + "epoch": 0.7486033519553073, + "grad_norm": 3.3752851486206055, + "learning_rate": 8.916201117318435e-07, + "loss": 2.8488, + "step": 268 + }, + { + "epoch": 0.7513966480446927, + "grad_norm": 4.068490982055664, + "learning_rate": 8.949720670391061e-07, + "loss": 2.8543, + "step": 269 + }, + { + "epoch": 0.7541899441340782, + "grad_norm": 2.4369142055511475, + "learning_rate": 8.983240223463688e-07, + "loss": 2.8597, + "step": 270 + }, + { + "epoch": 0.7569832402234636, + "grad_norm": 5.3020853996276855, + "learning_rate": 9.016759776536313e-07, + "loss": 2.8617, + "step": 271 + }, + { + "epoch": 0.7597765363128491, + "grad_norm": 2.1423234939575195, + "learning_rate": 9.050279329608939e-07, + "loss": 2.8604, + "step": 272 + }, + { + "epoch": 0.7625698324022346, + "grad_norm": 2.131009101867676, + "learning_rate": 9.083798882681564e-07, + "loss": 2.8463, + "step": 273 + }, + { + "epoch": 0.7653631284916201, + "grad_norm": 2.1152291297912598, + "learning_rate": 9.11731843575419e-07, + "loss": 2.8015, + "step": 274 + }, + { + "epoch": 0.7681564245810056, + "grad_norm": 2.613320827484131, + "learning_rate": 9.150837988826816e-07, + "loss": 2.7956, + "step": 275 + }, + { + "epoch": 0.770949720670391, + "grad_norm": 13.962514877319336, + "learning_rate": 9.184357541899441e-07, + "loss": 2.8869, + "step": 276 + }, + { + "epoch": 0.7737430167597765, + "grad_norm": 8.31851577758789, + "learning_rate": 9.217877094972066e-07, + "loss": 2.8341, + "step": 277 + }, + { + "epoch": 0.776536312849162, + "grad_norm": 5.490304946899414, + "learning_rate": 9.251396648044694e-07, + "loss": 2.7974, + "step": 278 + }, + { + "epoch": 0.7793296089385475, + "grad_norm": 2.2948946952819824, + "learning_rate": 9.284916201117319e-07, + "loss": 2.8024, + "step": 279 + }, + { + "epoch": 0.7821229050279329, + "grad_norm": 2.867436170578003, + "learning_rate": 9.318435754189944e-07, + "loss": 2.7926, + "step": 280 + }, + { + "epoch": 0.7849162011173184, + "grad_norm": 6.0475053787231445, + "learning_rate": 9.35195530726257e-07, + "loss": 2.7964, + "step": 281 + }, + { + "epoch": 0.7877094972067039, + "grad_norm": 5.882787227630615, + "learning_rate": 9.385474860335196e-07, + "loss": 2.7968, + "step": 282 + }, + { + "epoch": 0.7905027932960894, + "grad_norm": 5.171096324920654, + "learning_rate": 9.41899441340782e-07, + "loss": 2.7926, + "step": 283 + }, + { + "epoch": 0.7932960893854749, + "grad_norm": 3.442181348800659, + "learning_rate": 9.452513966480447e-07, + "loss": 2.7817, + "step": 284 + }, + { + "epoch": 0.7960893854748603, + "grad_norm": 1.6770849227905273, + "learning_rate": 9.486033519553074e-07, + "loss": 2.7901, + "step": 285 + }, + { + "epoch": 0.7988826815642458, + "grad_norm": 2.9833269119262695, + "learning_rate": 9.519553072625698e-07, + "loss": 2.7988, + "step": 286 + }, + { + "epoch": 0.8016759776536313, + "grad_norm": 2.9734139442443848, + "learning_rate": 9.553072625698324e-07, + "loss": 2.7946, + "step": 287 + }, + { + "epoch": 0.8044692737430168, + "grad_norm": 3.457808256149292, + "learning_rate": 9.58659217877095e-07, + "loss": 2.8042, + "step": 288 + }, + { + "epoch": 0.8072625698324022, + "grad_norm": 3.2319424152374268, + "learning_rate": 9.620111731843576e-07, + "loss": 2.7911, + "step": 289 + }, + { + "epoch": 0.8100558659217877, + "grad_norm": 1.4024688005447388, + "learning_rate": 9.653631284916202e-07, + "loss": 2.7872, + "step": 290 + }, + { + "epoch": 0.8128491620111732, + "grad_norm": 2.7842049598693848, + "learning_rate": 9.687150837988826e-07, + "loss": 2.7886, + "step": 291 + }, + { + "epoch": 0.8156424581005587, + "grad_norm": 3.661363124847412, + "learning_rate": 9.720670391061452e-07, + "loss": 2.7805, + "step": 292 + }, + { + "epoch": 0.8184357541899442, + "grad_norm": 3.13507080078125, + "learning_rate": 9.754189944134079e-07, + "loss": 2.7846, + "step": 293 + }, + { + "epoch": 0.8212290502793296, + "grad_norm": 1.4973236322402954, + "learning_rate": 9.787709497206705e-07, + "loss": 2.7834, + "step": 294 + }, + { + "epoch": 0.8240223463687151, + "grad_norm": 1.5215781927108765, + "learning_rate": 9.82122905027933e-07, + "loss": 2.768, + "step": 295 + }, + { + "epoch": 0.8268156424581006, + "grad_norm": 2.0090668201446533, + "learning_rate": 9.854748603351955e-07, + "loss": 2.7771, + "step": 296 + }, + { + "epoch": 0.8296089385474861, + "grad_norm": 1.5648643970489502, + "learning_rate": 9.888268156424581e-07, + "loss": 2.788, + "step": 297 + }, + { + "epoch": 0.8324022346368715, + "grad_norm": 1.6587365865707397, + "learning_rate": 9.921787709497207e-07, + "loss": 2.7529, + "step": 298 + }, + { + "epoch": 0.835195530726257, + "grad_norm": 3.5992074012756348, + "learning_rate": 9.955307262569831e-07, + "loss": 2.7948, + "step": 299 + }, + { + "epoch": 0.8379888268156425, + "grad_norm": 4.357047080993652, + "learning_rate": 9.98882681564246e-07, + "loss": 2.7664, + "step": 300 + }, + { + "epoch": 0.840782122905028, + "grad_norm": 14.75844669342041, + "learning_rate": 1.0022346368715084e-06, + "loss": 2.8342, + "step": 301 + }, + { + "epoch": 0.8435754189944135, + "grad_norm": 9.005867004394531, + "learning_rate": 1.005586592178771e-06, + "loss": 2.7803, + "step": 302 + }, + { + "epoch": 0.8463687150837989, + "grad_norm": 6.579145908355713, + "learning_rate": 1.0089385474860336e-06, + "loss": 2.7546, + "step": 303 + }, + { + "epoch": 0.8491620111731844, + "grad_norm": 3.551954746246338, + "learning_rate": 1.012290502793296e-06, + "loss": 2.7646, + "step": 304 + }, + { + "epoch": 0.8519553072625698, + "grad_norm": 2.1725635528564453, + "learning_rate": 1.0156424581005586e-06, + "loss": 2.7493, + "step": 305 + }, + { + "epoch": 0.8547486033519553, + "grad_norm": 3.9388949871063232, + "learning_rate": 1.0189944134078214e-06, + "loss": 2.7514, + "step": 306 + }, + { + "epoch": 0.8575418994413407, + "grad_norm": 3.5476126670837402, + "learning_rate": 1.0223463687150838e-06, + "loss": 2.7381, + "step": 307 + }, + { + "epoch": 0.8603351955307262, + "grad_norm": 2.508385419845581, + "learning_rate": 1.0256983240223465e-06, + "loss": 2.7492, + "step": 308 + }, + { + "epoch": 0.8631284916201117, + "grad_norm": 1.6094602346420288, + "learning_rate": 1.0290502793296089e-06, + "loss": 2.7328, + "step": 309 + }, + { + "epoch": 0.8659217877094972, + "grad_norm": 2.497408628463745, + "learning_rate": 1.0324022346368715e-06, + "loss": 2.738, + "step": 310 + }, + { + "epoch": 0.8687150837988827, + "grad_norm": 2.9154629707336426, + "learning_rate": 1.035754189944134e-06, + "loss": 2.7206, + "step": 311 + }, + { + "epoch": 0.8715083798882681, + "grad_norm": 4.531601905822754, + "learning_rate": 1.0391061452513967e-06, + "loss": 2.7293, + "step": 312 + }, + { + "epoch": 0.8743016759776536, + "grad_norm": 2.6571145057678223, + "learning_rate": 1.0424581005586593e-06, + "loss": 2.7193, + "step": 313 + }, + { + "epoch": 0.8770949720670391, + "grad_norm": 1.3236656188964844, + "learning_rate": 1.0458100558659217e-06, + "loss": 2.7384, + "step": 314 + }, + { + "epoch": 0.8798882681564246, + "grad_norm": 3.380000591278076, + "learning_rate": 1.0491620111731844e-06, + "loss": 2.7436, + "step": 315 + }, + { + "epoch": 0.88268156424581, + "grad_norm": 5.088804721832275, + "learning_rate": 1.052513966480447e-06, + "loss": 2.7443, + "step": 316 + }, + { + "epoch": 0.8854748603351955, + "grad_norm": 4.631897449493408, + "learning_rate": 1.0558659217877094e-06, + "loss": 2.7498, + "step": 317 + }, + { + "epoch": 0.888268156424581, + "grad_norm": 2.65725040435791, + "learning_rate": 1.0592178770949722e-06, + "loss": 2.7444, + "step": 318 + }, + { + "epoch": 0.8910614525139665, + "grad_norm": 2.0492613315582275, + "learning_rate": 1.0625698324022348e-06, + "loss": 2.7228, + "step": 319 + }, + { + "epoch": 0.8938547486033519, + "grad_norm": 1.6357754468917847, + "learning_rate": 1.0659217877094972e-06, + "loss": 2.7197, + "step": 320 + }, + { + "epoch": 0.8966480446927374, + "grad_norm": 3.113259792327881, + "learning_rate": 1.0692737430167598e-06, + "loss": 2.7238, + "step": 321 + }, + { + "epoch": 0.8994413407821229, + "grad_norm": 3.3314061164855957, + "learning_rate": 1.0726256983240222e-06, + "loss": 2.7269, + "step": 322 + }, + { + "epoch": 0.9022346368715084, + "grad_norm": 2.330226182937622, + "learning_rate": 1.0759776536312849e-06, + "loss": 2.7114, + "step": 323 + }, + { + "epoch": 0.9050279329608939, + "grad_norm": 2.3664584159851074, + "learning_rate": 1.0793296089385477e-06, + "loss": 2.735, + "step": 324 + }, + { + "epoch": 0.9078212290502793, + "grad_norm": 5.604909896850586, + "learning_rate": 1.08268156424581e-06, + "loss": 2.7073, + "step": 325 + }, + { + "epoch": 0.9106145251396648, + "grad_norm": 8.01389217376709, + "learning_rate": 1.0860335195530727e-06, + "loss": 2.7317, + "step": 326 + }, + { + "epoch": 0.9134078212290503, + "grad_norm": 6.145814895629883, + "learning_rate": 1.0893854748603353e-06, + "loss": 2.7187, + "step": 327 + }, + { + "epoch": 0.9162011173184358, + "grad_norm": 3.266411066055298, + "learning_rate": 1.0927374301675977e-06, + "loss": 2.7004, + "step": 328 + }, + { + "epoch": 0.9189944134078212, + "grad_norm": 2.0981202125549316, + "learning_rate": 1.0960893854748603e-06, + "loss": 2.6909, + "step": 329 + }, + { + "epoch": 0.9217877094972067, + "grad_norm": 3.9600586891174316, + "learning_rate": 1.099441340782123e-06, + "loss": 2.7047, + "step": 330 + }, + { + "epoch": 0.9245810055865922, + "grad_norm": 4.372499465942383, + "learning_rate": 1.1027932960893856e-06, + "loss": 2.6975, + "step": 331 + }, + { + "epoch": 0.9273743016759777, + "grad_norm": 3.3100552558898926, + "learning_rate": 1.1061452513966482e-06, + "loss": 2.6921, + "step": 332 + }, + { + "epoch": 0.9301675977653632, + "grad_norm": 4.31013822555542, + "learning_rate": 1.1094972067039106e-06, + "loss": 2.6901, + "step": 333 + }, + { + "epoch": 0.9329608938547486, + "grad_norm": 2.736295223236084, + "learning_rate": 1.1128491620111732e-06, + "loss": 2.6851, + "step": 334 + }, + { + "epoch": 0.9357541899441341, + "grad_norm": 2.497781276702881, + "learning_rate": 1.1162011173184356e-06, + "loss": 2.6891, + "step": 335 + }, + { + "epoch": 0.9385474860335196, + "grad_norm": 1.7287952899932861, + "learning_rate": 1.1195530726256984e-06, + "loss": 2.69, + "step": 336 + }, + { + "epoch": 0.9413407821229051, + "grad_norm": 2.1857104301452637, + "learning_rate": 1.122905027932961e-06, + "loss": 2.669, + "step": 337 + }, + { + "epoch": 0.9441340782122905, + "grad_norm": 2.5223312377929688, + "learning_rate": 1.1262569832402235e-06, + "loss": 2.6631, + "step": 338 + }, + { + "epoch": 0.946927374301676, + "grad_norm": 1.8268535137176514, + "learning_rate": 1.129608938547486e-06, + "loss": 2.6851, + "step": 339 + }, + { + "epoch": 0.9497206703910615, + "grad_norm": 1.8221303224563599, + "learning_rate": 1.1329608938547487e-06, + "loss": 2.6582, + "step": 340 + }, + { + "epoch": 0.952513966480447, + "grad_norm": 2.0452535152435303, + "learning_rate": 1.136312849162011e-06, + "loss": 2.6765, + "step": 341 + }, + { + "epoch": 0.9553072625698324, + "grad_norm": 2.438478469848633, + "learning_rate": 1.139664804469274e-06, + "loss": 2.6594, + "step": 342 + }, + { + "epoch": 0.9581005586592178, + "grad_norm": 1.9173520803451538, + "learning_rate": 1.1430167597765363e-06, + "loss": 2.6506, + "step": 343 + }, + { + "epoch": 0.9608938547486033, + "grad_norm": 2.350595474243164, + "learning_rate": 1.146368715083799e-06, + "loss": 2.6749, + "step": 344 + }, + { + "epoch": 0.9636871508379888, + "grad_norm": 2.3652219772338867, + "learning_rate": 1.1497206703910616e-06, + "loss": 2.6677, + "step": 345 + }, + { + "epoch": 0.9664804469273743, + "grad_norm": 3.003553628921509, + "learning_rate": 1.153072625698324e-06, + "loss": 2.6729, + "step": 346 + }, + { + "epoch": 0.9692737430167597, + "grad_norm": 2.140420436859131, + "learning_rate": 1.1564245810055866e-06, + "loss": 2.6405, + "step": 347 + }, + { + "epoch": 0.9720670391061452, + "grad_norm": 2.5123233795166016, + "learning_rate": 1.1597765363128492e-06, + "loss": 2.6479, + "step": 348 + }, + { + "epoch": 0.9748603351955307, + "grad_norm": 2.331413507461548, + "learning_rate": 1.1631284916201118e-06, + "loss": 2.6471, + "step": 349 + }, + { + "epoch": 0.9776536312849162, + "grad_norm": 2.3850414752960205, + "learning_rate": 1.1664804469273744e-06, + "loss": 2.6173, + "step": 350 + }, + { + "epoch": 0.9804469273743017, + "grad_norm": 7.44425630569458, + "learning_rate": 1.1698324022346368e-06, + "loss": 2.6451, + "step": 351 + }, + { + "epoch": 0.9832402234636871, + "grad_norm": 4.640804290771484, + "learning_rate": 1.1731843575418994e-06, + "loss": 2.6187, + "step": 352 + }, + { + "epoch": 0.9860335195530726, + "grad_norm": 5.316421985626221, + "learning_rate": 1.176536312849162e-06, + "loss": 2.5743, + "step": 353 + }, + { + "epoch": 0.9888268156424581, + "grad_norm": 3.1345438957214355, + "learning_rate": 1.1798882681564247e-06, + "loss": 2.5968, + "step": 354 + }, + { + "epoch": 0.9916201117318436, + "grad_norm": 2.5556282997131348, + "learning_rate": 1.1832402234636873e-06, + "loss": 2.5902, + "step": 355 + }, + { + "epoch": 0.994413407821229, + "grad_norm": 2.9765827655792236, + "learning_rate": 1.1865921787709497e-06, + "loss": 2.6014, + "step": 356 + }, + { + "epoch": 0.9972067039106145, + "grad_norm": 4.262032985687256, + "learning_rate": 1.1899441340782123e-06, + "loss": 2.6093, + "step": 357 + }, + { + "epoch": 1.0, + "grad_norm": 4.188101768493652, + "learning_rate": 1.193296089385475e-06, + "loss": 2.5773, + "step": 358 + }, + { + "epoch": 1.0027932960893855, + "grad_norm": 3.8787593841552734, + "learning_rate": 1.1966480446927373e-06, + "loss": 2.6189, + "step": 359 + }, + { + "epoch": 1.005586592178771, + "grad_norm": 1.6676547527313232, + "learning_rate": 1.2000000000000002e-06, + "loss": 2.5751, + "step": 360 + }, + { + "epoch": 1.0083798882681565, + "grad_norm": 1.8850034475326538, + "learning_rate": 1.2033519553072626e-06, + "loss": 2.5669, + "step": 361 + }, + { + "epoch": 1.011173184357542, + "grad_norm": 1.7757771015167236, + "learning_rate": 1.2067039106145252e-06, + "loss": 2.5703, + "step": 362 + }, + { + "epoch": 1.0139664804469273, + "grad_norm": 1.948617935180664, + "learning_rate": 1.2100558659217878e-06, + "loss": 2.5473, + "step": 363 + }, + { + "epoch": 1.0167597765363128, + "grad_norm": 1.9708824157714844, + "learning_rate": 1.2134078212290502e-06, + "loss": 2.5569, + "step": 364 + }, + { + "epoch": 1.0195530726256983, + "grad_norm": 1.8838140964508057, + "learning_rate": 1.2167597765363128e-06, + "loss": 2.5196, + "step": 365 + }, + { + "epoch": 1.0223463687150838, + "grad_norm": 1.7208871841430664, + "learning_rate": 1.2201117318435756e-06, + "loss": 2.5112, + "step": 366 + }, + { + "epoch": 1.0251396648044693, + "grad_norm": 1.6913851499557495, + "learning_rate": 1.223463687150838e-06, + "loss": 2.4787, + "step": 367 + }, + { + "epoch": 1.0279329608938548, + "grad_norm": 2.0982937812805176, + "learning_rate": 1.2268156424581007e-06, + "loss": 2.5157, + "step": 368 + }, + { + "epoch": 1.0307262569832403, + "grad_norm": 1.8002331256866455, + "learning_rate": 1.230167597765363e-06, + "loss": 2.4985, + "step": 369 + }, + { + "epoch": 1.0335195530726258, + "grad_norm": 1.5072757005691528, + "learning_rate": 1.2335195530726257e-06, + "loss": 2.4897, + "step": 370 + }, + { + "epoch": 1.0363128491620113, + "grad_norm": 1.8523015975952148, + "learning_rate": 1.2368715083798883e-06, + "loss": 2.4945, + "step": 371 + }, + { + "epoch": 1.0391061452513966, + "grad_norm": 2.3207387924194336, + "learning_rate": 1.240223463687151e-06, + "loss": 2.473, + "step": 372 + }, + { + "epoch": 1.041899441340782, + "grad_norm": 2.477930784225464, + "learning_rate": 1.2435754189944135e-06, + "loss": 2.4753, + "step": 373 + }, + { + "epoch": 1.0446927374301676, + "grad_norm": 2.0620391368865967, + "learning_rate": 1.2469273743016761e-06, + "loss": 2.4674, + "step": 374 + }, + { + "epoch": 1.047486033519553, + "grad_norm": 2.6486380100250244, + "learning_rate": 1.2502793296089385e-06, + "loss": 2.446, + "step": 375 + }, + { + "epoch": 1.0502793296089385, + "grad_norm": 2.0417590141296387, + "learning_rate": 1.2536312849162012e-06, + "loss": 2.4278, + "step": 376 + }, + { + "epoch": 1.053072625698324, + "grad_norm": 2.2296552658081055, + "learning_rate": 1.2569832402234636e-06, + "loss": 2.4754, + "step": 377 + }, + { + "epoch": 1.0558659217877095, + "grad_norm": 2.0083444118499756, + "learning_rate": 1.2603351955307264e-06, + "loss": 2.434, + "step": 378 + }, + { + "epoch": 1.058659217877095, + "grad_norm": 2.2740824222564697, + "learning_rate": 1.263687150837989e-06, + "loss": 2.4171, + "step": 379 + }, + { + "epoch": 1.0614525139664805, + "grad_norm": 2.5654807090759277, + "learning_rate": 1.2670391061452514e-06, + "loss": 2.4419, + "step": 380 + }, + { + "epoch": 1.0642458100558658, + "grad_norm": 3.1536288261413574, + "learning_rate": 1.270391061452514e-06, + "loss": 2.4298, + "step": 381 + }, + { + "epoch": 1.0670391061452513, + "grad_norm": 3.1465072631835938, + "learning_rate": 1.2737430167597764e-06, + "loss": 2.4511, + "step": 382 + }, + { + "epoch": 1.0698324022346368, + "grad_norm": 3.1121699810028076, + "learning_rate": 1.277094972067039e-06, + "loss": 2.4372, + "step": 383 + }, + { + "epoch": 1.0726256983240223, + "grad_norm": 6.384245872497559, + "learning_rate": 1.2804469273743019e-06, + "loss": 2.4105, + "step": 384 + }, + { + "epoch": 1.0754189944134078, + "grad_norm": 4.305026054382324, + "learning_rate": 1.2837988826815643e-06, + "loss": 2.3692, + "step": 385 + }, + { + "epoch": 1.0782122905027933, + "grad_norm": 3.4388139247894287, + "learning_rate": 1.287150837988827e-06, + "loss": 2.3404, + "step": 386 + }, + { + "epoch": 1.0810055865921788, + "grad_norm": 2.009432077407837, + "learning_rate": 1.2905027932960895e-06, + "loss": 2.3367, + "step": 387 + }, + { + "epoch": 1.0837988826815643, + "grad_norm": 2.672049045562744, + "learning_rate": 1.293854748603352e-06, + "loss": 2.3268, + "step": 388 + }, + { + "epoch": 1.0865921787709498, + "grad_norm": 3.756610155105591, + "learning_rate": 1.2972067039106145e-06, + "loss": 2.3176, + "step": 389 + }, + { + "epoch": 1.089385474860335, + "grad_norm": 2.3270421028137207, + "learning_rate": 1.3005586592178771e-06, + "loss": 2.312, + "step": 390 + }, + { + "epoch": 1.0921787709497206, + "grad_norm": 2.206343650817871, + "learning_rate": 1.3039106145251398e-06, + "loss": 2.2986, + "step": 391 + }, + { + "epoch": 1.094972067039106, + "grad_norm": 2.40895414352417, + "learning_rate": 1.3072625698324024e-06, + "loss": 2.2754, + "step": 392 + }, + { + "epoch": 1.0977653631284916, + "grad_norm": 2.135533571243286, + "learning_rate": 1.3106145251396648e-06, + "loss": 2.2839, + "step": 393 + }, + { + "epoch": 1.100558659217877, + "grad_norm": 3.0056819915771484, + "learning_rate": 1.3139664804469274e-06, + "loss": 2.2634, + "step": 394 + }, + { + "epoch": 1.1033519553072626, + "grad_norm": 4.119938850402832, + "learning_rate": 1.3173184357541898e-06, + "loss": 2.2603, + "step": 395 + }, + { + "epoch": 1.106145251396648, + "grad_norm": 2.9219560623168945, + "learning_rate": 1.3206703910614526e-06, + "loss": 2.2772, + "step": 396 + }, + { + "epoch": 1.1089385474860336, + "grad_norm": 2.6755149364471436, + "learning_rate": 1.3240223463687152e-06, + "loss": 2.2526, + "step": 397 + }, + { + "epoch": 1.111731843575419, + "grad_norm": 2.001007318496704, + "learning_rate": 1.3273743016759777e-06, + "loss": 2.2393, + "step": 398 + }, + { + "epoch": 1.1145251396648044, + "grad_norm": 2.906670570373535, + "learning_rate": 1.3307262569832403e-06, + "loss": 2.2534, + "step": 399 + }, + { + "epoch": 1.1173184357541899, + "grad_norm": 2.5072031021118164, + "learning_rate": 1.3340782122905029e-06, + "loss": 2.2203, + "step": 400 + }, + { + "epoch": 1.1201117318435754, + "grad_norm": 2.830618143081665, + "learning_rate": 1.3374301675977653e-06, + "loss": 2.2226, + "step": 401 + }, + { + "epoch": 1.1229050279329609, + "grad_norm": 2.6095974445343018, + "learning_rate": 1.340782122905028e-06, + "loss": 2.2256, + "step": 402 + }, + { + "epoch": 1.1256983240223464, + "grad_norm": 2.3497772216796875, + "learning_rate": 1.3441340782122905e-06, + "loss": 2.2371, + "step": 403 + }, + { + "epoch": 1.1284916201117319, + "grad_norm": 2.540508270263672, + "learning_rate": 1.3474860335195531e-06, + "loss": 2.2598, + "step": 404 + }, + { + "epoch": 1.1312849162011174, + "grad_norm": 2.0796802043914795, + "learning_rate": 1.3508379888268158e-06, + "loss": 2.1843, + "step": 405 + }, + { + "epoch": 1.1340782122905029, + "grad_norm": 3.1855361461639404, + "learning_rate": 1.3541899441340782e-06, + "loss": 2.2102, + "step": 406 + }, + { + "epoch": 1.1368715083798882, + "grad_norm": 17.37340545654297, + "learning_rate": 1.3575418994413408e-06, + "loss": 2.2032, + "step": 407 + }, + { + "epoch": 1.1396648044692737, + "grad_norm": 3.15460467338562, + "learning_rate": 1.3608938547486034e-06, + "loss": 2.2556, + "step": 408 + }, + { + "epoch": 1.1424581005586592, + "grad_norm": 5.4071044921875, + "learning_rate": 1.364245810055866e-06, + "loss": 2.1445, + "step": 409 + }, + { + "epoch": 1.1452513966480447, + "grad_norm": 3.938843011856079, + "learning_rate": 1.3675977653631286e-06, + "loss": 2.1102, + "step": 410 + }, + { + "epoch": 1.1480446927374302, + "grad_norm": 3.015289783477783, + "learning_rate": 1.370949720670391e-06, + "loss": 2.0782, + "step": 411 + }, + { + "epoch": 1.1508379888268156, + "grad_norm": 2.453061819076538, + "learning_rate": 1.3743016759776536e-06, + "loss": 2.0661, + "step": 412 + }, + { + "epoch": 1.1536312849162011, + "grad_norm": 2.700450897216797, + "learning_rate": 1.3776536312849163e-06, + "loss": 2.0578, + "step": 413 + }, + { + "epoch": 1.1564245810055866, + "grad_norm": 2.391179323196411, + "learning_rate": 1.3810055865921789e-06, + "loss": 2.0137, + "step": 414 + }, + { + "epoch": 1.1592178770949721, + "grad_norm": 2.376694679260254, + "learning_rate": 1.3843575418994415e-06, + "loss": 2.0259, + "step": 415 + }, + { + "epoch": 1.1620111731843576, + "grad_norm": 2.650980234146118, + "learning_rate": 1.3877094972067039e-06, + "loss": 2.0157, + "step": 416 + }, + { + "epoch": 1.164804469273743, + "grad_norm": 3.7345833778381348, + "learning_rate": 1.3910614525139665e-06, + "loss": 1.9962, + "step": 417 + }, + { + "epoch": 1.1675977653631284, + "grad_norm": 2.3333022594451904, + "learning_rate": 1.3944134078212291e-06, + "loss": 2.0153, + "step": 418 + }, + { + "epoch": 1.170391061452514, + "grad_norm": 2.848403215408325, + "learning_rate": 1.3977653631284915e-06, + "loss": 1.9865, + "step": 419 + }, + { + "epoch": 1.1731843575418994, + "grad_norm": 11.284046173095703, + "learning_rate": 1.4011173184357541e-06, + "loss": 1.9427, + "step": 420 + }, + { + "epoch": 1.175977653631285, + "grad_norm": 3.962151527404785, + "learning_rate": 1.404469273743017e-06, + "loss": 1.9384, + "step": 421 + }, + { + "epoch": 1.1787709497206704, + "grad_norm": 28.22885513305664, + "learning_rate": 1.4078212290502794e-06, + "loss": 1.9658, + "step": 422 + }, + { + "epoch": 1.181564245810056, + "grad_norm": 2.8012218475341797, + "learning_rate": 1.411173184357542e-06, + "loss": 1.9141, + "step": 423 + }, + { + "epoch": 1.1843575418994414, + "grad_norm": 4.126655578613281, + "learning_rate": 1.4145251396648044e-06, + "loss": 1.9485, + "step": 424 + }, + { + "epoch": 1.1871508379888267, + "grad_norm": 3.150963544845581, + "learning_rate": 1.417877094972067e-06, + "loss": 1.9262, + "step": 425 + }, + { + "epoch": 1.1899441340782122, + "grad_norm": 3.994974374771118, + "learning_rate": 1.4212290502793296e-06, + "loss": 1.9049, + "step": 426 + }, + { + "epoch": 1.1927374301675977, + "grad_norm": 2.680233955383301, + "learning_rate": 1.4245810055865922e-06, + "loss": 1.9279, + "step": 427 + }, + { + "epoch": 1.1955307262569832, + "grad_norm": 3.501180648803711, + "learning_rate": 1.4279329608938549e-06, + "loss": 1.9344, + "step": 428 + }, + { + "epoch": 1.1983240223463687, + "grad_norm": 4.773406028747559, + "learning_rate": 1.4312849162011173e-06, + "loss": 1.9276, + "step": 429 + }, + { + "epoch": 1.2011173184357542, + "grad_norm": 3.9312944412231445, + "learning_rate": 1.4346368715083799e-06, + "loss": 1.8876, + "step": 430 + }, + { + "epoch": 1.2039106145251397, + "grad_norm": 2.6023378372192383, + "learning_rate": 1.4379888268156425e-06, + "loss": 1.9333, + "step": 431 + }, + { + "epoch": 1.2067039106145252, + "grad_norm": 3.8897290229797363, + "learning_rate": 1.441340782122905e-06, + "loss": 1.8985, + "step": 432 + }, + { + "epoch": 1.2094972067039107, + "grad_norm": 3.6419527530670166, + "learning_rate": 1.4446927374301677e-06, + "loss": 2.018, + "step": 433 + }, + { + "epoch": 1.2122905027932962, + "grad_norm": 5.602702617645264, + "learning_rate": 1.4480446927374303e-06, + "loss": 1.8174, + "step": 434 + }, + { + "epoch": 1.2150837988826815, + "grad_norm": 4.6573381423950195, + "learning_rate": 1.4513966480446927e-06, + "loss": 1.7895, + "step": 435 + }, + { + "epoch": 1.217877094972067, + "grad_norm": 3.234393835067749, + "learning_rate": 1.4547486033519554e-06, + "loss": 1.7371, + "step": 436 + }, + { + "epoch": 1.2206703910614525, + "grad_norm": 2.8599212169647217, + "learning_rate": 1.4581005586592178e-06, + "loss": 1.7472, + "step": 437 + }, + { + "epoch": 1.223463687150838, + "grad_norm": 11.154973983764648, + "learning_rate": 1.4614525139664804e-06, + "loss": 1.7269, + "step": 438 + }, + { + "epoch": 1.2262569832402235, + "grad_norm": 3.9509241580963135, + "learning_rate": 1.4648044692737432e-06, + "loss": 1.7232, + "step": 439 + }, + { + "epoch": 1.229050279329609, + "grad_norm": 3.2199883460998535, + "learning_rate": 1.4681564245810056e-06, + "loss": 1.7065, + "step": 440 + }, + { + "epoch": 1.2318435754189945, + "grad_norm": 4.770200252532959, + "learning_rate": 1.4715083798882682e-06, + "loss": 1.6844, + "step": 441 + }, + { + "epoch": 1.23463687150838, + "grad_norm": 2.3367488384246826, + "learning_rate": 1.4748603351955306e-06, + "loss": 1.6746, + "step": 442 + }, + { + "epoch": 1.2374301675977653, + "grad_norm": 2.442875862121582, + "learning_rate": 1.4782122905027932e-06, + "loss": 1.6703, + "step": 443 + }, + { + "epoch": 1.2402234636871508, + "grad_norm": 3.103874444961548, + "learning_rate": 1.4815642458100559e-06, + "loss": 1.7114, + "step": 444 + }, + { + "epoch": 1.2430167597765363, + "grad_norm": 6.310858726501465, + "learning_rate": 1.4849162011173185e-06, + "loss": 1.6606, + "step": 445 + }, + { + "epoch": 1.2458100558659218, + "grad_norm": 2.832207441329956, + "learning_rate": 1.488268156424581e-06, + "loss": 1.6579, + "step": 446 + }, + { + "epoch": 1.2486033519553073, + "grad_norm": 2.9949917793273926, + "learning_rate": 1.4916201117318437e-06, + "loss": 1.6452, + "step": 447 + }, + { + "epoch": 1.2513966480446927, + "grad_norm": 6.813316345214844, + "learning_rate": 1.4949720670391061e-06, + "loss": 1.6274, + "step": 448 + }, + { + "epoch": 1.2541899441340782, + "grad_norm": 4.267480373382568, + "learning_rate": 1.4983240223463687e-06, + "loss": 1.6345, + "step": 449 + }, + { + "epoch": 1.2569832402234637, + "grad_norm": 5.631314754486084, + "learning_rate": 1.5016759776536311e-06, + "loss": 1.5913, + "step": 450 + }, + { + "epoch": 1.2597765363128492, + "grad_norm": 4.5266594886779785, + "learning_rate": 1.505027932960894e-06, + "loss": 1.6417, + "step": 451 + }, + { + "epoch": 1.2625698324022347, + "grad_norm": 3.0009119510650635, + "learning_rate": 1.5083798882681566e-06, + "loss": 1.6307, + "step": 452 + }, + { + "epoch": 1.2653631284916202, + "grad_norm": 4.690017223358154, + "learning_rate": 1.511731843575419e-06, + "loss": 1.6528, + "step": 453 + }, + { + "epoch": 1.2681564245810055, + "grad_norm": 3.483548164367676, + "learning_rate": 1.5150837988826816e-06, + "loss": 1.615, + "step": 454 + }, + { + "epoch": 1.270949720670391, + "grad_norm": 3.279832601547241, + "learning_rate": 1.518435754189944e-06, + "loss": 1.608, + "step": 455 + }, + { + "epoch": 1.2737430167597765, + "grad_norm": 5.070201396942139, + "learning_rate": 1.5217877094972066e-06, + "loss": 1.6601, + "step": 456 + }, + { + "epoch": 1.276536312849162, + "grad_norm": 3.9738433361053467, + "learning_rate": 1.5251396648044694e-06, + "loss": 1.6852, + "step": 457 + }, + { + "epoch": 1.2793296089385475, + "grad_norm": 11.614875793457031, + "learning_rate": 1.5284916201117318e-06, + "loss": 1.7458, + "step": 458 + }, + { + "epoch": 1.282122905027933, + "grad_norm": 4.564628601074219, + "learning_rate": 1.5318435754189945e-06, + "loss": 1.5058, + "step": 459 + }, + { + "epoch": 1.2849162011173183, + "grad_norm": 4.071938514709473, + "learning_rate": 1.535195530726257e-06, + "loss": 1.44, + "step": 460 + }, + { + "epoch": 1.2877094972067038, + "grad_norm": 3.009357452392578, + "learning_rate": 1.5385474860335195e-06, + "loss": 1.4283, + "step": 461 + }, + { + "epoch": 1.2905027932960893, + "grad_norm": 2.3847157955169678, + "learning_rate": 1.541899441340782e-06, + "loss": 1.4205, + "step": 462 + }, + { + "epoch": 1.2932960893854748, + "grad_norm": 3.3101747035980225, + "learning_rate": 1.5452513966480447e-06, + "loss": 1.3939, + "step": 463 + }, + { + "epoch": 1.2960893854748603, + "grad_norm": 3.9084420204162598, + "learning_rate": 1.5486033519553073e-06, + "loss": 1.4118, + "step": 464 + }, + { + "epoch": 1.2988826815642458, + "grad_norm": 2.308286190032959, + "learning_rate": 1.55195530726257e-06, + "loss": 1.418, + "step": 465 + }, + { + "epoch": 1.3016759776536313, + "grad_norm": 2.2199997901916504, + "learning_rate": 1.5553072625698324e-06, + "loss": 1.3844, + "step": 466 + }, + { + "epoch": 1.3044692737430168, + "grad_norm": 3.087280035018921, + "learning_rate": 1.558659217877095e-06, + "loss": 1.3897, + "step": 467 + }, + { + "epoch": 1.3072625698324023, + "grad_norm": 2.7029004096984863, + "learning_rate": 1.5620111731843576e-06, + "loss": 1.3701, + "step": 468 + }, + { + "epoch": 1.3100558659217878, + "grad_norm": 3.2159879207611084, + "learning_rate": 1.5653631284916202e-06, + "loss": 1.3509, + "step": 469 + }, + { + "epoch": 1.3128491620111733, + "grad_norm": 4.119954586029053, + "learning_rate": 1.5687150837988828e-06, + "loss": 1.4155, + "step": 470 + }, + { + "epoch": 1.3156424581005586, + "grad_norm": 8.260847091674805, + "learning_rate": 1.5720670391061452e-06, + "loss": 1.3806, + "step": 471 + }, + { + "epoch": 1.318435754189944, + "grad_norm": 2.6251444816589355, + "learning_rate": 1.5754189944134078e-06, + "loss": 1.3382, + "step": 472 + }, + { + "epoch": 1.3212290502793296, + "grad_norm": 4.167954444885254, + "learning_rate": 1.5787709497206705e-06, + "loss": 1.3481, + "step": 473 + }, + { + "epoch": 1.324022346368715, + "grad_norm": 4.6547441482543945, + "learning_rate": 1.5821229050279329e-06, + "loss": 1.3506, + "step": 474 + }, + { + "epoch": 1.3268156424581006, + "grad_norm": 5.066556930541992, + "learning_rate": 1.5854748603351957e-06, + "loss": 1.3362, + "step": 475 + }, + { + "epoch": 1.329608938547486, + "grad_norm": 4.714915752410889, + "learning_rate": 1.588826815642458e-06, + "loss": 1.3465, + "step": 476 + }, + { + "epoch": 1.3324022346368716, + "grad_norm": 3.8189709186553955, + "learning_rate": 1.5921787709497207e-06, + "loss": 1.3451, + "step": 477 + }, + { + "epoch": 1.3351955307262569, + "grad_norm": 3.3221728801727295, + "learning_rate": 1.5955307262569833e-06, + "loss": 1.3286, + "step": 478 + }, + { + "epoch": 1.3379888268156424, + "grad_norm": 5.82459020614624, + "learning_rate": 1.5988826815642457e-06, + "loss": 1.3671, + "step": 479 + }, + { + "epoch": 1.3407821229050279, + "grad_norm": 5.4333295822143555, + "learning_rate": 1.6022346368715083e-06, + "loss": 1.3743, + "step": 480 + }, + { + "epoch": 1.3435754189944134, + "grad_norm": 3.4417226314544678, + "learning_rate": 1.6055865921787712e-06, + "loss": 1.3897, + "step": 481 + }, + { + "epoch": 1.3463687150837989, + "grad_norm": 5.166982650756836, + "learning_rate": 1.6089385474860336e-06, + "loss": 1.4063, + "step": 482 + }, + { + "epoch": 1.3491620111731844, + "grad_norm": 3.7908051013946533, + "learning_rate": 1.6122905027932962e-06, + "loss": 1.5673, + "step": 483 + }, + { + "epoch": 1.3519553072625698, + "grad_norm": 4.9061760902404785, + "learning_rate": 1.6156424581005586e-06, + "loss": 1.2128, + "step": 484 + }, + { + "epoch": 1.3547486033519553, + "grad_norm": 3.0365254878997803, + "learning_rate": 1.6189944134078212e-06, + "loss": 1.1542, + "step": 485 + }, + { + "epoch": 1.3575418994413408, + "grad_norm": 3.0932672023773193, + "learning_rate": 1.6223463687150838e-06, + "loss": 1.1467, + "step": 486 + }, + { + "epoch": 1.3603351955307263, + "grad_norm": 3.1406025886535645, + "learning_rate": 1.6256983240223464e-06, + "loss": 1.1401, + "step": 487 + }, + { + "epoch": 1.3631284916201118, + "grad_norm": 2.439619302749634, + "learning_rate": 1.629050279329609e-06, + "loss": 1.1215, + "step": 488 + }, + { + "epoch": 1.3659217877094971, + "grad_norm": 3.2568960189819336, + "learning_rate": 1.6324022346368715e-06, + "loss": 1.135, + "step": 489 + }, + { + "epoch": 1.3687150837988826, + "grad_norm": 3.0093507766723633, + "learning_rate": 1.635754189944134e-06, + "loss": 1.1123, + "step": 490 + }, + { + "epoch": 1.3715083798882681, + "grad_norm": 2.181431293487549, + "learning_rate": 1.6391061452513967e-06, + "loss": 1.1262, + "step": 491 + }, + { + "epoch": 1.3743016759776536, + "grad_norm": 3.0067858695983887, + "learning_rate": 1.642458100558659e-06, + "loss": 1.1087, + "step": 492 + }, + { + "epoch": 1.3770949720670391, + "grad_norm": 3.1210923194885254, + "learning_rate": 1.645810055865922e-06, + "loss": 1.1094, + "step": 493 + }, + { + "epoch": 1.3798882681564246, + "grad_norm": 2.273655891418457, + "learning_rate": 1.6491620111731845e-06, + "loss": 1.0794, + "step": 494 + }, + { + "epoch": 1.3826815642458101, + "grad_norm": 3.0309534072875977, + "learning_rate": 1.652513966480447e-06, + "loss": 1.1155, + "step": 495 + }, + { + "epoch": 1.3854748603351954, + "grad_norm": 3.1782596111297607, + "learning_rate": 1.6558659217877096e-06, + "loss": 1.0654, + "step": 496 + }, + { + "epoch": 1.388268156424581, + "grad_norm": 6.204005241394043, + "learning_rate": 1.659217877094972e-06, + "loss": 1.0907, + "step": 497 + }, + { + "epoch": 1.3910614525139664, + "grad_norm": 4.872853755950928, + "learning_rate": 1.6625698324022346e-06, + "loss": 1.0536, + "step": 498 + }, + { + "epoch": 1.393854748603352, + "grad_norm": 3.0108425617218018, + "learning_rate": 1.6659217877094974e-06, + "loss": 1.0922, + "step": 499 + }, + { + "epoch": 1.3966480446927374, + "grad_norm": 8.511221885681152, + "learning_rate": 1.6692737430167598e-06, + "loss": 1.1465, + "step": 500 + }, + { + "epoch": 1.399441340782123, + "grad_norm": 3.4301869869232178, + "learning_rate": 1.6726256983240224e-06, + "loss": 1.1524, + "step": 501 + }, + { + "epoch": 1.4022346368715084, + "grad_norm": 2.849755048751831, + "learning_rate": 1.6759776536312848e-06, + "loss": 1.0893, + "step": 502 + }, + { + "epoch": 1.405027932960894, + "grad_norm": 3.152590036392212, + "learning_rate": 1.6793296089385474e-06, + "loss": 1.09, + "step": 503 + }, + { + "epoch": 1.4078212290502794, + "grad_norm": 3.8768482208251953, + "learning_rate": 1.68268156424581e-06, + "loss": 1.1744, + "step": 504 + }, + { + "epoch": 1.410614525139665, + "grad_norm": 4.2038655281066895, + "learning_rate": 1.6860335195530727e-06, + "loss": 1.1626, + "step": 505 + }, + { + "epoch": 1.4134078212290504, + "grad_norm": 3.793242931365967, + "learning_rate": 1.6893854748603353e-06, + "loss": 1.1415, + "step": 506 + }, + { + "epoch": 1.4162011173184357, + "grad_norm": 4.308712482452393, + "learning_rate": 1.692737430167598e-06, + "loss": 1.2209, + "step": 507 + }, + { + "epoch": 1.4189944134078212, + "grad_norm": 4.848049640655518, + "learning_rate": 1.6960893854748603e-06, + "loss": 1.3859, + "step": 508 + }, + { + "epoch": 1.4217877094972067, + "grad_norm": 3.366452693939209, + "learning_rate": 1.699441340782123e-06, + "loss": 0.9221, + "step": 509 + }, + { + "epoch": 1.4245810055865922, + "grad_norm": 2.1573245525360107, + "learning_rate": 1.7027932960893853e-06, + "loss": 0.8765, + "step": 510 + }, + { + "epoch": 1.4273743016759777, + "grad_norm": 3.3252761363983154, + "learning_rate": 1.7061452513966482e-06, + "loss": 0.8674, + "step": 511 + }, + { + "epoch": 1.4301675977653632, + "grad_norm": 2.922654390335083, + "learning_rate": 1.7094972067039108e-06, + "loss": 0.8747, + "step": 512 + }, + { + "epoch": 1.4329608938547487, + "grad_norm": 2.5519540309906006, + "learning_rate": 1.7128491620111732e-06, + "loss": 0.9139, + "step": 513 + }, + { + "epoch": 1.435754189944134, + "grad_norm": 2.7157979011535645, + "learning_rate": 1.7162011173184358e-06, + "loss": 0.8349, + "step": 514 + }, + { + "epoch": 1.4385474860335195, + "grad_norm": 2.3243634700775146, + "learning_rate": 1.7195530726256984e-06, + "loss": 0.9098, + "step": 515 + }, + { + "epoch": 1.441340782122905, + "grad_norm": 2.354066848754883, + "learning_rate": 1.7229050279329608e-06, + "loss": 0.8837, + "step": 516 + }, + { + "epoch": 1.4441340782122905, + "grad_norm": 2.5841190814971924, + "learning_rate": 1.7262569832402236e-06, + "loss": 0.8587, + "step": 517 + }, + { + "epoch": 1.446927374301676, + "grad_norm": 2.3069067001342773, + "learning_rate": 1.729608938547486e-06, + "loss": 0.8734, + "step": 518 + }, + { + "epoch": 1.4497206703910615, + "grad_norm": 4.320409297943115, + "learning_rate": 1.7329608938547487e-06, + "loss": 0.949, + "step": 519 + }, + { + "epoch": 1.452513966480447, + "grad_norm": 5.542835235595703, + "learning_rate": 1.7363128491620113e-06, + "loss": 0.8646, + "step": 520 + }, + { + "epoch": 1.4553072625698324, + "grad_norm": 2.6305816173553467, + "learning_rate": 1.7396648044692737e-06, + "loss": 0.9142, + "step": 521 + }, + { + "epoch": 1.458100558659218, + "grad_norm": 2.961203098297119, + "learning_rate": 1.7430167597765363e-06, + "loss": 0.8907, + "step": 522 + }, + { + "epoch": 1.4608938547486034, + "grad_norm": 4.158580780029297, + "learning_rate": 1.746368715083799e-06, + "loss": 0.8937, + "step": 523 + }, + { + "epoch": 1.463687150837989, + "grad_norm": 3.743701696395874, + "learning_rate": 1.7497206703910615e-06, + "loss": 0.9406, + "step": 524 + }, + { + "epoch": 1.4664804469273742, + "grad_norm": 3.2914726734161377, + "learning_rate": 1.7530726256983241e-06, + "loss": 0.9288, + "step": 525 + }, + { + "epoch": 1.4692737430167597, + "grad_norm": 3.3137688636779785, + "learning_rate": 1.7564245810055865e-06, + "loss": 0.8726, + "step": 526 + }, + { + "epoch": 1.4720670391061452, + "grad_norm": 9.583857536315918, + "learning_rate": 1.7597765363128492e-06, + "loss": 0.9538, + "step": 527 + }, + { + "epoch": 1.4748603351955307, + "grad_norm": 3.6832258701324463, + "learning_rate": 1.7631284916201118e-06, + "loss": 0.9711, + "step": 528 + }, + { + "epoch": 1.4776536312849162, + "grad_norm": 4.1859588623046875, + "learning_rate": 1.7664804469273744e-06, + "loss": 0.9074, + "step": 529 + }, + { + "epoch": 1.4804469273743017, + "grad_norm": 5.439997673034668, + "learning_rate": 1.769832402234637e-06, + "loss": 0.9673, + "step": 530 + }, + { + "epoch": 1.4832402234636872, + "grad_norm": 3.5975241661071777, + "learning_rate": 1.7731843575418994e-06, + "loss": 0.9423, + "step": 531 + }, + { + "epoch": 1.4860335195530725, + "grad_norm": 5.502966403961182, + "learning_rate": 1.776536312849162e-06, + "loss": 1.0808, + "step": 532 + }, + { + "epoch": 1.488826815642458, + "grad_norm": 5.888707637786865, + "learning_rate": 1.7798882681564246e-06, + "loss": 1.2279, + "step": 533 + }, + { + "epoch": 1.4916201117318435, + "grad_norm": 6.175964832305908, + "learning_rate": 1.783240223463687e-06, + "loss": 0.7647, + "step": 534 + }, + { + "epoch": 1.494413407821229, + "grad_norm": 3.155059814453125, + "learning_rate": 1.7865921787709499e-06, + "loss": 0.7267, + "step": 535 + }, + { + "epoch": 1.4972067039106145, + "grad_norm": 2.1428418159484863, + "learning_rate": 1.7899441340782123e-06, + "loss": 0.6997, + "step": 536 + }, + { + "epoch": 1.5, + "grad_norm": 2.0101194381713867, + "learning_rate": 1.793296089385475e-06, + "loss": 0.7625, + "step": 537 + }, + { + "epoch": 1.5027932960893855, + "grad_norm": 2.359821319580078, + "learning_rate": 1.7966480446927375e-06, + "loss": 0.6992, + "step": 538 + }, + { + "epoch": 1.505586592178771, + "grad_norm": 6.037433624267578, + "learning_rate": 1.8e-06, + "loss": 0.74, + "step": 539 + }, + { + "epoch": 1.5083798882681565, + "grad_norm": 2.141446828842163, + "learning_rate": 1.8033519553072625e-06, + "loss": 0.7164, + "step": 540 + }, + { + "epoch": 1.511173184357542, + "grad_norm": 3.3127334117889404, + "learning_rate": 1.8067039106145254e-06, + "loss": 0.7025, + "step": 541 + }, + { + "epoch": 1.5139664804469275, + "grad_norm": 2.0130879878997803, + "learning_rate": 1.8100558659217878e-06, + "loss": 0.7381, + "step": 542 + }, + { + "epoch": 1.516759776536313, + "grad_norm": 3.337599992752075, + "learning_rate": 1.8134078212290504e-06, + "loss": 0.7756, + "step": 543 + }, + { + "epoch": 1.5195530726256983, + "grad_norm": 3.5758862495422363, + "learning_rate": 1.8167597765363128e-06, + "loss": 0.7551, + "step": 544 + }, + { + "epoch": 1.5223463687150838, + "grad_norm": 3.5906145572662354, + "learning_rate": 1.8201117318435754e-06, + "loss": 0.7617, + "step": 545 + }, + { + "epoch": 1.5251396648044693, + "grad_norm": 2.6607818603515625, + "learning_rate": 1.823463687150838e-06, + "loss": 0.6749, + "step": 546 + }, + { + "epoch": 1.5279329608938548, + "grad_norm": 2.3154947757720947, + "learning_rate": 1.8268156424581006e-06, + "loss": 0.7508, + "step": 547 + }, + { + "epoch": 1.5307262569832403, + "grad_norm": 3.212653398513794, + "learning_rate": 1.8301675977653632e-06, + "loss": 0.7767, + "step": 548 + }, + { + "epoch": 1.5335195530726256, + "grad_norm": 2.4277255535125732, + "learning_rate": 1.8335195530726257e-06, + "loss": 0.8314, + "step": 549 + }, + { + "epoch": 1.536312849162011, + "grad_norm": 4.34569787979126, + "learning_rate": 1.8368715083798883e-06, + "loss": 0.7672, + "step": 550 + }, + { + "epoch": 1.5391061452513966, + "grad_norm": 4.583427429199219, + "learning_rate": 1.8402234636871509e-06, + "loss": 0.7439, + "step": 551 + }, + { + "epoch": 1.541899441340782, + "grad_norm": 4.682043075561523, + "learning_rate": 1.8435754189944133e-06, + "loss": 0.8196, + "step": 552 + }, + { + "epoch": 1.5446927374301676, + "grad_norm": 4.559029579162598, + "learning_rate": 1.8469273743016761e-06, + "loss": 0.8106, + "step": 553 + }, + { + "epoch": 1.547486033519553, + "grad_norm": 5.649045467376709, + "learning_rate": 1.8502793296089387e-06, + "loss": 0.831, + "step": 554 + }, + { + "epoch": 1.5502793296089385, + "grad_norm": 4.170166969299316, + "learning_rate": 1.8536312849162011e-06, + "loss": 0.8771, + "step": 555 + }, + { + "epoch": 1.553072625698324, + "grad_norm": 4.922107696533203, + "learning_rate": 1.8569832402234638e-06, + "loss": 0.9717, + "step": 556 + }, + { + "epoch": 1.5558659217877095, + "grad_norm": 5.161901473999023, + "learning_rate": 1.8603351955307262e-06, + "loss": 0.9729, + "step": 557 + }, + { + "epoch": 1.558659217877095, + "grad_norm": 11.693743705749512, + "learning_rate": 1.8636871508379888e-06, + "loss": 1.0981, + "step": 558 + }, + { + "epoch": 1.5614525139664805, + "grad_norm": 3.376227617263794, + "learning_rate": 1.8670391061452516e-06, + "loss": 0.7138, + "step": 559 + }, + { + "epoch": 1.564245810055866, + "grad_norm": 2.6003496646881104, + "learning_rate": 1.870391061452514e-06, + "loss": 0.6256, + "step": 560 + }, + { + "epoch": 1.5670391061452515, + "grad_norm": 1.6465438604354858, + "learning_rate": 1.8737430167597766e-06, + "loss": 0.6329, + "step": 561 + }, + { + "epoch": 1.5698324022346368, + "grad_norm": 3.8619768619537354, + "learning_rate": 1.8770949720670392e-06, + "loss": 0.6551, + "step": 562 + }, + { + "epoch": 1.5726256983240223, + "grad_norm": 5.166996002197266, + "learning_rate": 1.8804469273743019e-06, + "loss": 0.6629, + "step": 563 + }, + { + "epoch": 1.5754189944134078, + "grad_norm": 3.14005446434021, + "learning_rate": 1.883798882681564e-06, + "loss": 0.6335, + "step": 564 + }, + { + "epoch": 1.5782122905027933, + "grad_norm": 2.742358684539795, + "learning_rate": 1.8871508379888269e-06, + "loss": 0.6816, + "step": 565 + }, + { + "epoch": 1.5810055865921788, + "grad_norm": 3.7622432708740234, + "learning_rate": 1.8905027932960895e-06, + "loss": 0.744, + "step": 566 + }, + { + "epoch": 1.583798882681564, + "grad_norm": 2.638221025466919, + "learning_rate": 1.893854748603352e-06, + "loss": 0.6449, + "step": 567 + }, + { + "epoch": 1.5865921787709496, + "grad_norm": 2.2801673412323, + "learning_rate": 1.8972067039106147e-06, + "loss": 0.6993, + "step": 568 + }, + { + "epoch": 1.589385474860335, + "grad_norm": 3.0075836181640625, + "learning_rate": 1.900558659217877e-06, + "loss": 0.6633, + "step": 569 + }, + { + "epoch": 1.5921787709497206, + "grad_norm": 2.5129292011260986, + "learning_rate": 1.9039106145251395e-06, + "loss": 0.6822, + "step": 570 + }, + { + "epoch": 1.594972067039106, + "grad_norm": 4.364952087402344, + "learning_rate": 1.9072625698324024e-06, + "loss": 0.6633, + "step": 571 + }, + { + "epoch": 1.5977653631284916, + "grad_norm": 4.462542533874512, + "learning_rate": 1.9106145251396648e-06, + "loss": 0.6803, + "step": 572 + }, + { + "epoch": 1.600558659217877, + "grad_norm": 4.640573024749756, + "learning_rate": 1.9139664804469274e-06, + "loss": 0.6711, + "step": 573 + }, + { + "epoch": 1.6033519553072626, + "grad_norm": 2.8808493614196777, + "learning_rate": 1.91731843575419e-06, + "loss": 0.7058, + "step": 574 + }, + { + "epoch": 1.606145251396648, + "grad_norm": 2.855750560760498, + "learning_rate": 1.9206703910614526e-06, + "loss": 0.699, + "step": 575 + }, + { + "epoch": 1.6089385474860336, + "grad_norm": 2.7278571128845215, + "learning_rate": 1.9240223463687152e-06, + "loss": 0.6946, + "step": 576 + }, + { + "epoch": 1.611731843575419, + "grad_norm": 33.33197784423828, + "learning_rate": 1.927374301675978e-06, + "loss": 0.7158, + "step": 577 + }, + { + "epoch": 1.6145251396648046, + "grad_norm": 2.9647481441497803, + "learning_rate": 1.9307262569832405e-06, + "loss": 0.75, + "step": 578 + }, + { + "epoch": 1.61731843575419, + "grad_norm": 3.809314489364624, + "learning_rate": 1.9340782122905026e-06, + "loss": 0.7861, + "step": 579 + }, + { + "epoch": 1.6201117318435754, + "grad_norm": 4.480356216430664, + "learning_rate": 1.9374301675977653e-06, + "loss": 0.8302, + "step": 580 + }, + { + "epoch": 1.6229050279329609, + "grad_norm": 3.6925835609436035, + "learning_rate": 1.940782122905028e-06, + "loss": 0.8646, + "step": 581 + }, + { + "epoch": 1.6256983240223464, + "grad_norm": 4.995118141174316, + "learning_rate": 1.9441340782122905e-06, + "loss": 0.9581, + "step": 582 + }, + { + "epoch": 1.6284916201117319, + "grad_norm": 5.886316776275635, + "learning_rate": 1.947486033519553e-06, + "loss": 1.1503, + "step": 583 + }, + { + "epoch": 1.6312849162011172, + "grad_norm": 3.0202009677886963, + "learning_rate": 1.9508379888268157e-06, + "loss": 0.5929, + "step": 584 + }, + { + "epoch": 1.6340782122905027, + "grad_norm": 3.692077398300171, + "learning_rate": 1.954189944134078e-06, + "loss": 0.6014, + "step": 585 + }, + { + "epoch": 1.6368715083798882, + "grad_norm": 2.0695741176605225, + "learning_rate": 1.957541899441341e-06, + "loss": 0.5707, + "step": 586 + }, + { + "epoch": 1.6396648044692737, + "grad_norm": 3.412735939025879, + "learning_rate": 1.9608938547486036e-06, + "loss": 0.5883, + "step": 587 + }, + { + "epoch": 1.6424581005586592, + "grad_norm": 2.7252159118652344, + "learning_rate": 1.964245810055866e-06, + "loss": 0.6087, + "step": 588 + }, + { + "epoch": 1.6452513966480447, + "grad_norm": 4.218059539794922, + "learning_rate": 1.967597765363129e-06, + "loss": 0.6536, + "step": 589 + }, + { + "epoch": 1.6480446927374302, + "grad_norm": 2.4523708820343018, + "learning_rate": 1.970949720670391e-06, + "loss": 0.6122, + "step": 590 + }, + { + "epoch": 1.6508379888268156, + "grad_norm": 1.808193564414978, + "learning_rate": 1.9743016759776536e-06, + "loss": 0.5769, + "step": 591 + }, + { + "epoch": 1.6536312849162011, + "grad_norm": 3.268369436264038, + "learning_rate": 1.9776536312849162e-06, + "loss": 0.5868, + "step": 592 + }, + { + "epoch": 1.6564245810055866, + "grad_norm": 2.4960384368896484, + "learning_rate": 1.981005586592179e-06, + "loss": 0.6622, + "step": 593 + }, + { + "epoch": 1.6592178770949721, + "grad_norm": 2.2760348320007324, + "learning_rate": 1.9843575418994415e-06, + "loss": 0.601, + "step": 594 + }, + { + "epoch": 1.6620111731843576, + "grad_norm": 2.9752089977264404, + "learning_rate": 1.9877094972067037e-06, + "loss": 0.6357, + "step": 595 + }, + { + "epoch": 1.6648044692737431, + "grad_norm": 2.4375922679901123, + "learning_rate": 1.9910614525139663e-06, + "loss": 0.6266, + "step": 596 + }, + { + "epoch": 1.6675977653631286, + "grad_norm": 3.6522600650787354, + "learning_rate": 1.994413407821229e-06, + "loss": 0.6634, + "step": 597 + }, + { + "epoch": 1.670391061452514, + "grad_norm": 6.655056953430176, + "learning_rate": 1.997765363128492e-06, + "loss": 0.6238, + "step": 598 + }, + { + "epoch": 1.6731843575418994, + "grad_norm": 2.622701406478882, + "learning_rate": 2.0011173184357545e-06, + "loss": 0.707, + "step": 599 + }, + { + "epoch": 1.675977653631285, + "grad_norm": 3.2559404373168945, + "learning_rate": 2.0044692737430167e-06, + "loss": 0.6875, + "step": 600 + }, + { + "epoch": 1.6787709497206704, + "grad_norm": 3.0451583862304688, + "learning_rate": 2.0078212290502793e-06, + "loss": 0.6627, + "step": 601 + }, + { + "epoch": 1.6815642458100557, + "grad_norm": 3.7823233604431152, + "learning_rate": 2.011173184357542e-06, + "loss": 0.6409, + "step": 602 + }, + { + "epoch": 1.6843575418994412, + "grad_norm": 8.39523983001709, + "learning_rate": 2.0145251396648046e-06, + "loss": 0.7287, + "step": 603 + }, + { + "epoch": 1.6871508379888267, + "grad_norm": 4.533437728881836, + "learning_rate": 2.017877094972067e-06, + "loss": 0.6856, + "step": 604 + }, + { + "epoch": 1.6899441340782122, + "grad_norm": 8.833052635192871, + "learning_rate": 2.02122905027933e-06, + "loss": 0.6935, + "step": 605 + }, + { + "epoch": 1.6927374301675977, + "grad_norm": 13.594083786010742, + "learning_rate": 2.024581005586592e-06, + "loss": 0.8373, + "step": 606 + }, + { + "epoch": 1.6955307262569832, + "grad_norm": 5.71712064743042, + "learning_rate": 2.0279329608938546e-06, + "loss": 0.8277, + "step": 607 + }, + { + "epoch": 1.6983240223463687, + "grad_norm": 7.775496006011963, + "learning_rate": 2.0312849162011172e-06, + "loss": 1.0771, + "step": 608 + }, + { + "epoch": 1.7011173184357542, + "grad_norm": 2.2539398670196533, + "learning_rate": 2.03463687150838e-06, + "loss": 0.581, + "step": 609 + }, + { + "epoch": 1.7039106145251397, + "grad_norm": 2.033440113067627, + "learning_rate": 2.037988826815643e-06, + "loss": 0.5603, + "step": 610 + }, + { + "epoch": 1.7067039106145252, + "grad_norm": 1.4637959003448486, + "learning_rate": 2.041340782122905e-06, + "loss": 0.5341, + "step": 611 + }, + { + "epoch": 1.7094972067039107, + "grad_norm": 2.0620572566986084, + "learning_rate": 2.0446927374301677e-06, + "loss": 0.5374, + "step": 612 + }, + { + "epoch": 1.7122905027932962, + "grad_norm": 4.592379570007324, + "learning_rate": 2.0480446927374303e-06, + "loss": 0.5716, + "step": 613 + }, + { + "epoch": 1.7150837988826817, + "grad_norm": 3.1672632694244385, + "learning_rate": 2.051396648044693e-06, + "loss": 0.5514, + "step": 614 + }, + { + "epoch": 1.7178770949720672, + "grad_norm": 2.562089681625366, + "learning_rate": 2.0547486033519555e-06, + "loss": 0.5458, + "step": 615 + }, + { + "epoch": 1.7206703910614525, + "grad_norm": 6.7326579093933105, + "learning_rate": 2.0581005586592177e-06, + "loss": 0.5633, + "step": 616 + }, + { + "epoch": 1.723463687150838, + "grad_norm": 2.505750894546509, + "learning_rate": 2.0614525139664804e-06, + "loss": 0.5558, + "step": 617 + }, + { + "epoch": 1.7262569832402235, + "grad_norm": 10.374075889587402, + "learning_rate": 2.064804469273743e-06, + "loss": 0.5797, + "step": 618 + }, + { + "epoch": 1.729050279329609, + "grad_norm": 3.768829345703125, + "learning_rate": 2.0681564245810056e-06, + "loss": 0.5454, + "step": 619 + }, + { + "epoch": 1.7318435754189943, + "grad_norm": 5.759853839874268, + "learning_rate": 2.071508379888268e-06, + "loss": 0.5928, + "step": 620 + }, + { + "epoch": 1.7346368715083798, + "grad_norm": 16.174375534057617, + "learning_rate": 2.0748603351955304e-06, + "loss": 0.5914, + "step": 621 + }, + { + "epoch": 1.7374301675977653, + "grad_norm": 2.7420239448547363, + "learning_rate": 2.0782122905027934e-06, + "loss": 0.6008, + "step": 622 + }, + { + "epoch": 1.7402234636871508, + "grad_norm": 3.682753324508667, + "learning_rate": 2.081564245810056e-06, + "loss": 0.6038, + "step": 623 + }, + { + "epoch": 1.7430167597765363, + "grad_norm": 3.2124063968658447, + "learning_rate": 2.0849162011173187e-06, + "loss": 0.6288, + "step": 624 + }, + { + "epoch": 1.7458100558659218, + "grad_norm": 3.5079596042633057, + "learning_rate": 2.0882681564245813e-06, + "loss": 0.5924, + "step": 625 + }, + { + "epoch": 1.7486033519553073, + "grad_norm": 5.659529209136963, + "learning_rate": 2.0916201117318435e-06, + "loss": 0.662, + "step": 626 + }, + { + "epoch": 1.7513966480446927, + "grad_norm": 6.706552505493164, + "learning_rate": 2.094972067039106e-06, + "loss": 0.6614, + "step": 627 + }, + { + "epoch": 1.7541899441340782, + "grad_norm": 2.6417179107666016, + "learning_rate": 2.0983240223463687e-06, + "loss": 0.6695, + "step": 628 + }, + { + "epoch": 1.7569832402234637, + "grad_norm": 6.503865718841553, + "learning_rate": 2.1016759776536313e-06, + "loss": 0.6429, + "step": 629 + }, + { + "epoch": 1.7597765363128492, + "grad_norm": 7.257603168487549, + "learning_rate": 2.105027932960894e-06, + "loss": 0.735, + "step": 630 + }, + { + "epoch": 1.7625698324022347, + "grad_norm": 4.758538246154785, + "learning_rate": 2.1083798882681566e-06, + "loss": 0.7662, + "step": 631 + }, + { + "epoch": 1.7653631284916202, + "grad_norm": 3.5461277961730957, + "learning_rate": 2.1117318435754187e-06, + "loss": 0.7482, + "step": 632 + }, + { + "epoch": 1.7681564245810057, + "grad_norm": 6.548026084899902, + "learning_rate": 2.1150837988826814e-06, + "loss": 1.0079, + "step": 633 + }, + { + "epoch": 1.770949720670391, + "grad_norm": 3.2100045680999756, + "learning_rate": 2.1184357541899444e-06, + "loss": 0.5569, + "step": 634 + }, + { + "epoch": 1.7737430167597765, + "grad_norm": 1.9920670986175537, + "learning_rate": 2.121787709497207e-06, + "loss": 0.5079, + "step": 635 + }, + { + "epoch": 1.776536312849162, + "grad_norm": 1.520889401435852, + "learning_rate": 2.1251396648044696e-06, + "loss": 0.5082, + "step": 636 + }, + { + "epoch": 1.7793296089385475, + "grad_norm": 1.939846396446228, + "learning_rate": 2.128491620111732e-06, + "loss": 0.5514, + "step": 637 + }, + { + "epoch": 1.7821229050279328, + "grad_norm": 1.9949557781219482, + "learning_rate": 2.1318435754189944e-06, + "loss": 0.5329, + "step": 638 + }, + { + "epoch": 1.7849162011173183, + "grad_norm": 2.6726491451263428, + "learning_rate": 2.135195530726257e-06, + "loss": 0.5287, + "step": 639 + }, + { + "epoch": 1.7877094972067038, + "grad_norm": 3.3091368675231934, + "learning_rate": 2.1385474860335197e-06, + "loss": 0.538, + "step": 640 + }, + { + "epoch": 1.7905027932960893, + "grad_norm": 3.034020185470581, + "learning_rate": 2.1418994413407823e-06, + "loss": 0.5528, + "step": 641 + }, + { + "epoch": 1.7932960893854748, + "grad_norm": 2.2050046920776367, + "learning_rate": 2.1452513966480445e-06, + "loss": 0.5503, + "step": 642 + }, + { + "epoch": 1.7960893854748603, + "grad_norm": 8.574934005737305, + "learning_rate": 2.148603351955307e-06, + "loss": 0.5881, + "step": 643 + }, + { + "epoch": 1.7988826815642458, + "grad_norm": 2.4102401733398438, + "learning_rate": 2.1519553072625697e-06, + "loss": 0.5309, + "step": 644 + }, + { + "epoch": 1.8016759776536313, + "grad_norm": 2.15828800201416, + "learning_rate": 2.1553072625698323e-06, + "loss": 0.5645, + "step": 645 + }, + { + "epoch": 1.8044692737430168, + "grad_norm": 2.4150679111480713, + "learning_rate": 2.1586592178770954e-06, + "loss": 0.5449, + "step": 646 + }, + { + "epoch": 1.8072625698324023, + "grad_norm": 2.3946781158447266, + "learning_rate": 2.1620111731843576e-06, + "loss": 0.5519, + "step": 647 + }, + { + "epoch": 1.8100558659217878, + "grad_norm": 3.8243486881256104, + "learning_rate": 2.16536312849162e-06, + "loss": 0.5931, + "step": 648 + }, + { + "epoch": 1.8128491620111733, + "grad_norm": 2.7094767093658447, + "learning_rate": 2.1687150837988828e-06, + "loss": 0.5095, + "step": 649 + }, + { + "epoch": 1.8156424581005588, + "grad_norm": 2.7087717056274414, + "learning_rate": 2.1720670391061454e-06, + "loss": 0.5728, + "step": 650 + }, + { + "epoch": 1.8184357541899443, + "grad_norm": 3.5444536209106445, + "learning_rate": 2.175418994413408e-06, + "loss": 0.5842, + "step": 651 + }, + { + "epoch": 1.8212290502793296, + "grad_norm": 3.20707368850708, + "learning_rate": 2.1787709497206706e-06, + "loss": 0.676, + "step": 652 + }, + { + "epoch": 1.824022346368715, + "grad_norm": 3.200573205947876, + "learning_rate": 2.182122905027933e-06, + "loss": 0.6491, + "step": 653 + }, + { + "epoch": 1.8268156424581006, + "grad_norm": 2.834055185317993, + "learning_rate": 2.1854748603351954e-06, + "loss": 0.714, + "step": 654 + }, + { + "epoch": 1.829608938547486, + "grad_norm": 3.2949283123016357, + "learning_rate": 2.188826815642458e-06, + "loss": 0.7074, + "step": 655 + }, + { + "epoch": 1.8324022346368714, + "grad_norm": 4.357840538024902, + "learning_rate": 2.1921787709497207e-06, + "loss": 0.7242, + "step": 656 + }, + { + "epoch": 1.8351955307262569, + "grad_norm": 15.892206192016602, + "learning_rate": 2.1955307262569833e-06, + "loss": 0.7813, + "step": 657 + }, + { + "epoch": 1.8379888268156424, + "grad_norm": 3.9403421878814697, + "learning_rate": 2.198882681564246e-06, + "loss": 0.9752, + "step": 658 + }, + { + "epoch": 1.8407821229050279, + "grad_norm": 2.1453473567962646, + "learning_rate": 2.2022346368715085e-06, + "loss": 0.4998, + "step": 659 + }, + { + "epoch": 1.8435754189944134, + "grad_norm": 3.6731009483337402, + "learning_rate": 2.205586592178771e-06, + "loss": 0.4787, + "step": 660 + }, + { + "epoch": 1.8463687150837989, + "grad_norm": 1.8011059761047363, + "learning_rate": 2.2089385474860338e-06, + "loss": 0.509, + "step": 661 + }, + { + "epoch": 1.8491620111731844, + "grad_norm": 1.9362812042236328, + "learning_rate": 2.2122905027932964e-06, + "loss": 0.4839, + "step": 662 + }, + { + "epoch": 1.8519553072625698, + "grad_norm": 1.5970649719238281, + "learning_rate": 2.2156424581005586e-06, + "loss": 0.5257, + "step": 663 + }, + { + "epoch": 1.8547486033519553, + "grad_norm": 1.8080413341522217, + "learning_rate": 2.218994413407821e-06, + "loss": 0.5016, + "step": 664 + }, + { + "epoch": 1.8575418994413408, + "grad_norm": 1.537537693977356, + "learning_rate": 2.222346368715084e-06, + "loss": 0.5297, + "step": 665 + }, + { + "epoch": 1.8603351955307263, + "grad_norm": 4.883070468902588, + "learning_rate": 2.2256983240223464e-06, + "loss": 0.5234, + "step": 666 + }, + { + "epoch": 1.8631284916201118, + "grad_norm": 1.8864290714263916, + "learning_rate": 2.229050279329609e-06, + "loss": 0.5089, + "step": 667 + }, + { + "epoch": 1.8659217877094973, + "grad_norm": 2.19586181640625, + "learning_rate": 2.2324022346368712e-06, + "loss": 0.4925, + "step": 668 + }, + { + "epoch": 1.8687150837988828, + "grad_norm": 1.971967339515686, + "learning_rate": 2.235754189944134e-06, + "loss": 0.4691, + "step": 669 + }, + { + "epoch": 1.8715083798882681, + "grad_norm": 2.3327836990356445, + "learning_rate": 2.239106145251397e-06, + "loss": 0.501, + "step": 670 + }, + { + "epoch": 1.8743016759776536, + "grad_norm": 1.9682302474975586, + "learning_rate": 2.2424581005586595e-06, + "loss": 0.5394, + "step": 671 + }, + { + "epoch": 1.8770949720670391, + "grad_norm": 1.7234278917312622, + "learning_rate": 2.245810055865922e-06, + "loss": 0.5911, + "step": 672 + }, + { + "epoch": 1.8798882681564246, + "grad_norm": 1.9518616199493408, + "learning_rate": 2.2491620111731843e-06, + "loss": 0.5285, + "step": 673 + }, + { + "epoch": 1.88268156424581, + "grad_norm": 2.3888907432556152, + "learning_rate": 2.252513966480447e-06, + "loss": 0.5545, + "step": 674 + }, + { + "epoch": 1.8854748603351954, + "grad_norm": 2.2294349670410156, + "learning_rate": 2.2558659217877095e-06, + "loss": 0.5315, + "step": 675 + }, + { + "epoch": 1.888268156424581, + "grad_norm": 4.011464595794678, + "learning_rate": 2.259217877094972e-06, + "loss": 0.5737, + "step": 676 + }, + { + "epoch": 1.8910614525139664, + "grad_norm": 6.335679054260254, + "learning_rate": 2.2625698324022348e-06, + "loss": 0.6131, + "step": 677 + }, + { + "epoch": 1.893854748603352, + "grad_norm": 3.0629546642303467, + "learning_rate": 2.2659217877094974e-06, + "loss": 0.6382, + "step": 678 + }, + { + "epoch": 1.8966480446927374, + "grad_norm": 2.539790391921997, + "learning_rate": 2.2692737430167596e-06, + "loss": 0.6268, + "step": 679 + }, + { + "epoch": 1.899441340782123, + "grad_norm": 4.425440311431885, + "learning_rate": 2.272625698324022e-06, + "loss": 0.7165, + "step": 680 + }, + { + "epoch": 1.9022346368715084, + "grad_norm": 5.979133605957031, + "learning_rate": 2.275977653631285e-06, + "loss": 0.6913, + "step": 681 + }, + { + "epoch": 1.905027932960894, + "grad_norm": 6.485985279083252, + "learning_rate": 2.279329608938548e-06, + "loss": 0.7045, + "step": 682 + }, + { + "epoch": 1.9078212290502794, + "grad_norm": 4.674729824066162, + "learning_rate": 2.2826815642458105e-06, + "loss": 0.8365, + "step": 683 + }, + { + "epoch": 1.910614525139665, + "grad_norm": 2.0277438163757324, + "learning_rate": 2.2860335195530726e-06, + "loss": 0.519, + "step": 684 + }, + { + "epoch": 1.9134078212290504, + "grad_norm": 1.9321633577346802, + "learning_rate": 2.2893854748603353e-06, + "loss": 0.4995, + "step": 685 + }, + { + "epoch": 1.916201117318436, + "grad_norm": 1.1962131261825562, + "learning_rate": 2.292737430167598e-06, + "loss": 0.4529, + "step": 686 + }, + { + "epoch": 1.9189944134078212, + "grad_norm": 5.037054538726807, + "learning_rate": 2.2960893854748605e-06, + "loss": 0.4586, + "step": 687 + }, + { + "epoch": 1.9217877094972067, + "grad_norm": 1.7302613258361816, + "learning_rate": 2.299441340782123e-06, + "loss": 0.4964, + "step": 688 + }, + { + "epoch": 1.9245810055865922, + "grad_norm": 2.0159072875976562, + "learning_rate": 2.3027932960893853e-06, + "loss": 0.4934, + "step": 689 + }, + { + "epoch": 1.9273743016759777, + "grad_norm": 1.7113699913024902, + "learning_rate": 2.306145251396648e-06, + "loss": 0.5329, + "step": 690 + }, + { + "epoch": 1.9301675977653632, + "grad_norm": 4.022100925445557, + "learning_rate": 2.3094972067039105e-06, + "loss": 0.4707, + "step": 691 + }, + { + "epoch": 1.9329608938547485, + "grad_norm": 9.428092956542969, + "learning_rate": 2.312849162011173e-06, + "loss": 0.4662, + "step": 692 + }, + { + "epoch": 1.935754189944134, + "grad_norm": 2.056180477142334, + "learning_rate": 2.3162011173184358e-06, + "loss": 0.4915, + "step": 693 + }, + { + "epoch": 1.9385474860335195, + "grad_norm": 2.3734748363494873, + "learning_rate": 2.3195530726256984e-06, + "loss": 0.4869, + "step": 694 + }, + { + "epoch": 1.941340782122905, + "grad_norm": 1.9549267292022705, + "learning_rate": 2.322905027932961e-06, + "loss": 0.4948, + "step": 695 + }, + { + "epoch": 1.9441340782122905, + "grad_norm": 2.4207677841186523, + "learning_rate": 2.3262569832402236e-06, + "loss": 0.5168, + "step": 696 + }, + { + "epoch": 1.946927374301676, + "grad_norm": 1.97707200050354, + "learning_rate": 2.3296089385474862e-06, + "loss": 0.5034, + "step": 697 + }, + { + "epoch": 1.9497206703910615, + "grad_norm": 2.54105806350708, + "learning_rate": 2.332960893854749e-06, + "loss": 0.5496, + "step": 698 + }, + { + "epoch": 1.952513966480447, + "grad_norm": 3.2706236839294434, + "learning_rate": 2.3363128491620115e-06, + "loss": 0.4929, + "step": 699 + }, + { + "epoch": 1.9553072625698324, + "grad_norm": 6.414819717407227, + "learning_rate": 2.3396648044692737e-06, + "loss": 0.5315, + "step": 700 + }, + { + "epoch": 1.958100558659218, + "grad_norm": 2.4252536296844482, + "learning_rate": 2.3430167597765363e-06, + "loss": 0.5221, + "step": 701 + }, + { + "epoch": 1.9608938547486034, + "grad_norm": 3.864797353744507, + "learning_rate": 2.346368715083799e-06, + "loss": 0.5573, + "step": 702 + }, + { + "epoch": 1.963687150837989, + "grad_norm": 6.227105617523193, + "learning_rate": 2.3497206703910615e-06, + "loss": 0.5591, + "step": 703 + }, + { + "epoch": 1.9664804469273744, + "grad_norm": 3.225536346435547, + "learning_rate": 2.353072625698324e-06, + "loss": 0.6533, + "step": 704 + }, + { + "epoch": 1.9692737430167597, + "grad_norm": 4.043196678161621, + "learning_rate": 2.3564245810055863e-06, + "loss": 0.6377, + "step": 705 + }, + { + "epoch": 1.9720670391061452, + "grad_norm": 8.007545471191406, + "learning_rate": 2.3597765363128493e-06, + "loss": 0.7008, + "step": 706 + }, + { + "epoch": 1.9748603351955307, + "grad_norm": 3.230891227722168, + "learning_rate": 2.363128491620112e-06, + "loss": 0.6913, + "step": 707 + }, + { + "epoch": 1.9776536312849162, + "grad_norm": 3.498626232147217, + "learning_rate": 2.3664804469273746e-06, + "loss": 0.8618, + "step": 708 + }, + { + "epoch": 1.9804469273743017, + "grad_norm": 3.6880428791046143, + "learning_rate": 2.369832402234637e-06, + "loss": 0.5485, + "step": 709 + }, + { + "epoch": 1.983240223463687, + "grad_norm": 2.3903260231018066, + "learning_rate": 2.3731843575418994e-06, + "loss": 0.4698, + "step": 710 + }, + { + "epoch": 1.9860335195530725, + "grad_norm": 1.5072993040084839, + "learning_rate": 2.376536312849162e-06, + "loss": 0.441, + "step": 711 + }, + { + "epoch": 1.988826815642458, + "grad_norm": 2.2366507053375244, + "learning_rate": 2.3798882681564246e-06, + "loss": 0.4758, + "step": 712 + }, + { + "epoch": 1.9916201117318435, + "grad_norm": 2.1876235008239746, + "learning_rate": 2.3832402234636872e-06, + "loss": 0.5529, + "step": 713 + }, + { + "epoch": 1.994413407821229, + "grad_norm": 3.1262052059173584, + "learning_rate": 2.38659217877095e-06, + "loss": 0.5167, + "step": 714 + }, + { + "epoch": 1.9972067039106145, + "grad_norm": 3.906324863433838, + "learning_rate": 2.389944134078212e-06, + "loss": 0.6161, + "step": 715 + }, + { + "epoch": 2.0, + "grad_norm": 9.913809776306152, + "learning_rate": 2.3932960893854747e-06, + "loss": 0.7812, + "step": 716 + }, + { + "epoch": 2.0027932960893855, + "grad_norm": 15.437994003295898, + "learning_rate": 2.3966480446927373e-06, + "loss": 0.4453, + "step": 717 + }, + { + "epoch": 2.005586592178771, + "grad_norm": 1.2124661207199097, + "learning_rate": 2.4000000000000003e-06, + "loss": 0.4363, + "step": 718 + }, + { + "epoch": 2.0083798882681565, + "grad_norm": 1.5966084003448486, + "learning_rate": 2.403351955307263e-06, + "loss": 0.4042, + "step": 719 + }, + { + "epoch": 2.011173184357542, + "grad_norm": 3.0947916507720947, + "learning_rate": 2.406703910614525e-06, + "loss": 0.4487, + "step": 720 + }, + { + "epoch": 2.0139664804469275, + "grad_norm": 2.3459312915802, + "learning_rate": 2.4100558659217877e-06, + "loss": 0.4341, + "step": 721 + }, + { + "epoch": 2.016759776536313, + "grad_norm": 2.1044816970825195, + "learning_rate": 2.4134078212290504e-06, + "loss": 0.4676, + "step": 722 + }, + { + "epoch": 2.0195530726256985, + "grad_norm": 2.0787620544433594, + "learning_rate": 2.416759776536313e-06, + "loss": 0.4385, + "step": 723 + }, + { + "epoch": 2.022346368715084, + "grad_norm": 1.6578210592269897, + "learning_rate": 2.4201117318435756e-06, + "loss": 0.4226, + "step": 724 + }, + { + "epoch": 2.0251396648044695, + "grad_norm": 1.6172212362289429, + "learning_rate": 2.423463687150838e-06, + "loss": 0.4717, + "step": 725 + }, + { + "epoch": 2.0279329608938546, + "grad_norm": 2.0647099018096924, + "learning_rate": 2.4268156424581004e-06, + "loss": 0.4767, + "step": 726 + }, + { + "epoch": 2.03072625698324, + "grad_norm": 2.4714243412017822, + "learning_rate": 2.430167597765363e-06, + "loss": 0.4926, + "step": 727 + }, + { + "epoch": 2.0335195530726256, + "grad_norm": 1.9684044122695923, + "learning_rate": 2.4335195530726256e-06, + "loss": 0.5186, + "step": 728 + }, + { + "epoch": 2.036312849162011, + "grad_norm": 2.5510470867156982, + "learning_rate": 2.4368715083798882e-06, + "loss": 0.4715, + "step": 729 + }, + { + "epoch": 2.0391061452513966, + "grad_norm": 2.3160576820373535, + "learning_rate": 2.4402234636871513e-06, + "loss": 0.4791, + "step": 730 + }, + { + "epoch": 2.041899441340782, + "grad_norm": 1.8655046224594116, + "learning_rate": 2.4435754189944135e-06, + "loss": 0.5066, + "step": 731 + }, + { + "epoch": 2.0446927374301676, + "grad_norm": 1.763697862625122, + "learning_rate": 2.446927374301676e-06, + "loss": 0.5246, + "step": 732 + }, + { + "epoch": 2.047486033519553, + "grad_norm": 1.7792388200759888, + "learning_rate": 2.4502793296089387e-06, + "loss": 0.5229, + "step": 733 + }, + { + "epoch": 2.0502793296089385, + "grad_norm": 2.926753520965576, + "learning_rate": 2.4536312849162013e-06, + "loss": 0.5689, + "step": 734 + }, + { + "epoch": 2.053072625698324, + "grad_norm": 3.082181215286255, + "learning_rate": 2.456983240223464e-06, + "loss": 0.5305, + "step": 735 + }, + { + "epoch": 2.0558659217877095, + "grad_norm": 2.553727388381958, + "learning_rate": 2.460335195530726e-06, + "loss": 0.5691, + "step": 736 + }, + { + "epoch": 2.058659217877095, + "grad_norm": 3.3252177238464355, + "learning_rate": 2.4636871508379887e-06, + "loss": 0.5996, + "step": 737 + }, + { + "epoch": 2.0614525139664805, + "grad_norm": 3.4098668098449707, + "learning_rate": 2.4670391061452514e-06, + "loss": 0.5735, + "step": 738 + }, + { + "epoch": 2.064245810055866, + "grad_norm": 3.405611276626587, + "learning_rate": 2.470391061452514e-06, + "loss": 0.6769, + "step": 739 + }, + { + "epoch": 2.0670391061452515, + "grad_norm": 4.697253227233887, + "learning_rate": 2.4737430167597766e-06, + "loss": 0.7136, + "step": 740 + }, + { + "epoch": 2.069832402234637, + "grad_norm": 7.252675533294678, + "learning_rate": 2.4770949720670388e-06, + "loss": 0.8532, + "step": 741 + }, + { + "epoch": 2.0726256983240225, + "grad_norm": 2.5180931091308594, + "learning_rate": 2.480446927374302e-06, + "loss": 0.4504, + "step": 742 + }, + { + "epoch": 2.0754189944134076, + "grad_norm": 1.6111748218536377, + "learning_rate": 2.4837988826815644e-06, + "loss": 0.4058, + "step": 743 + }, + { + "epoch": 2.078212290502793, + "grad_norm": 2.2142696380615234, + "learning_rate": 2.487150837988827e-06, + "loss": 0.4682, + "step": 744 + }, + { + "epoch": 2.0810055865921786, + "grad_norm": 1.4057976007461548, + "learning_rate": 2.4905027932960897e-06, + "loss": 0.4264, + "step": 745 + }, + { + "epoch": 2.083798882681564, + "grad_norm": 1.3559712171554565, + "learning_rate": 2.4938547486033523e-06, + "loss": 0.4214, + "step": 746 + }, + { + "epoch": 2.0865921787709496, + "grad_norm": 3.0040721893310547, + "learning_rate": 2.4972067039106145e-06, + "loss": 0.4442, + "step": 747 + }, + { + "epoch": 2.089385474860335, + "grad_norm": 2.2289819717407227, + "learning_rate": 2.500558659217877e-06, + "loss": 0.4042, + "step": 748 + }, + { + "epoch": 2.0921787709497206, + "grad_norm": 2.196295976638794, + "learning_rate": 2.5039106145251397e-06, + "loss": 0.4596, + "step": 749 + }, + { + "epoch": 2.094972067039106, + "grad_norm": 2.419774293899536, + "learning_rate": 2.5072625698324023e-06, + "loss": 0.4368, + "step": 750 + }, + { + "epoch": 2.0977653631284916, + "grad_norm": 2.263261079788208, + "learning_rate": 2.510614525139665e-06, + "loss": 0.4925, + "step": 751 + }, + { + "epoch": 2.100558659217877, + "grad_norm": 2.1463863849639893, + "learning_rate": 2.513966480446927e-06, + "loss": 0.4606, + "step": 752 + }, + { + "epoch": 2.1033519553072626, + "grad_norm": 3.1462438106536865, + "learning_rate": 2.5173184357541898e-06, + "loss": 0.4747, + "step": 753 + }, + { + "epoch": 2.106145251396648, + "grad_norm": 2.922149896621704, + "learning_rate": 2.520670391061453e-06, + "loss": 0.4243, + "step": 754 + }, + { + "epoch": 2.1089385474860336, + "grad_norm": 3.0261402130126953, + "learning_rate": 2.5240223463687154e-06, + "loss": 0.4807, + "step": 755 + }, + { + "epoch": 2.111731843575419, + "grad_norm": 7.793262481689453, + "learning_rate": 2.527374301675978e-06, + "loss": 0.4593, + "step": 756 + }, + { + "epoch": 2.1145251396648046, + "grad_norm": 3.9324088096618652, + "learning_rate": 2.5307262569832402e-06, + "loss": 0.4452, + "step": 757 + }, + { + "epoch": 2.11731843575419, + "grad_norm": 3.364536762237549, + "learning_rate": 2.534078212290503e-06, + "loss": 0.5169, + "step": 758 + }, + { + "epoch": 2.1201117318435756, + "grad_norm": 2.2618565559387207, + "learning_rate": 2.5374301675977654e-06, + "loss": 0.5025, + "step": 759 + }, + { + "epoch": 2.122905027932961, + "grad_norm": 2.6475393772125244, + "learning_rate": 2.540782122905028e-06, + "loss": 0.5506, + "step": 760 + }, + { + "epoch": 2.1256983240223466, + "grad_norm": 2.7108049392700195, + "learning_rate": 2.5441340782122907e-06, + "loss": 0.5279, + "step": 761 + }, + { + "epoch": 2.1284916201117317, + "grad_norm": 10.060585021972656, + "learning_rate": 2.547486033519553e-06, + "loss": 0.5084, + "step": 762 + }, + { + "epoch": 2.131284916201117, + "grad_norm": 3.3217718601226807, + "learning_rate": 2.5508379888268155e-06, + "loss": 0.6148, + "step": 763 + }, + { + "epoch": 2.1340782122905027, + "grad_norm": 2.789785861968994, + "learning_rate": 2.554189944134078e-06, + "loss": 0.6386, + "step": 764 + }, + { + "epoch": 2.136871508379888, + "grad_norm": 3.6776909828186035, + "learning_rate": 2.5575418994413407e-06, + "loss": 0.7094, + "step": 765 + }, + { + "epoch": 2.1396648044692737, + "grad_norm": 13.651623725891113, + "learning_rate": 2.5608938547486038e-06, + "loss": 0.9567, + "step": 766 + }, + { + "epoch": 2.142458100558659, + "grad_norm": 1.9742310047149658, + "learning_rate": 2.564245810055866e-06, + "loss": 0.4837, + "step": 767 + }, + { + "epoch": 2.1452513966480447, + "grad_norm": 1.4257214069366455, + "learning_rate": 2.5675977653631286e-06, + "loss": 0.4157, + "step": 768 + }, + { + "epoch": 2.14804469273743, + "grad_norm": 2.3614726066589355, + "learning_rate": 2.570949720670391e-06, + "loss": 0.423, + "step": 769 + }, + { + "epoch": 2.1508379888268156, + "grad_norm": 1.6082231998443604, + "learning_rate": 2.574301675977654e-06, + "loss": 0.4441, + "step": 770 + }, + { + "epoch": 2.153631284916201, + "grad_norm": 1.6669729948043823, + "learning_rate": 2.5776536312849164e-06, + "loss": 0.3983, + "step": 771 + }, + { + "epoch": 2.1564245810055866, + "grad_norm": 1.710905909538269, + "learning_rate": 2.581005586592179e-06, + "loss": 0.4477, + "step": 772 + }, + { + "epoch": 2.159217877094972, + "grad_norm": 1.6602952480316162, + "learning_rate": 2.5843575418994412e-06, + "loss": 0.4363, + "step": 773 + }, + { + "epoch": 2.1620111731843576, + "grad_norm": 4.774791717529297, + "learning_rate": 2.587709497206704e-06, + "loss": 0.5104, + "step": 774 + }, + { + "epoch": 2.164804469273743, + "grad_norm": 2.1032819747924805, + "learning_rate": 2.5910614525139665e-06, + "loss": 0.4032, + "step": 775 + }, + { + "epoch": 2.1675977653631286, + "grad_norm": 3.6077799797058105, + "learning_rate": 2.594413407821229e-06, + "loss": 0.4328, + "step": 776 + }, + { + "epoch": 2.170391061452514, + "grad_norm": 1.5143489837646484, + "learning_rate": 2.5977653631284917e-06, + "loss": 0.4141, + "step": 777 + }, + { + "epoch": 2.1731843575418996, + "grad_norm": 1.584580898284912, + "learning_rate": 2.6011173184357543e-06, + "loss": 0.4139, + "step": 778 + }, + { + "epoch": 2.1759776536312847, + "grad_norm": 1.9167660474777222, + "learning_rate": 2.604469273743017e-06, + "loss": 0.4331, + "step": 779 + }, + { + "epoch": 2.17877094972067, + "grad_norm": 3.6957788467407227, + "learning_rate": 2.6078212290502795e-06, + "loss": 0.5401, + "step": 780 + }, + { + "epoch": 2.1815642458100557, + "grad_norm": 2.664004325866699, + "learning_rate": 2.611173184357542e-06, + "loss": 0.571, + "step": 781 + }, + { + "epoch": 2.184357541899441, + "grad_norm": 1.6747379302978516, + "learning_rate": 2.6145251396648048e-06, + "loss": 0.4936, + "step": 782 + }, + { + "epoch": 2.1871508379888267, + "grad_norm": 2.568706512451172, + "learning_rate": 2.617877094972067e-06, + "loss": 0.5281, + "step": 783 + }, + { + "epoch": 2.189944134078212, + "grad_norm": 4.233485221862793, + "learning_rate": 2.6212290502793296e-06, + "loss": 0.4915, + "step": 784 + }, + { + "epoch": 2.1927374301675977, + "grad_norm": 22.31960678100586, + "learning_rate": 2.624581005586592e-06, + "loss": 0.5257, + "step": 785 + }, + { + "epoch": 2.195530726256983, + "grad_norm": 5.974844455718994, + "learning_rate": 2.627932960893855e-06, + "loss": 0.5078, + "step": 786 + }, + { + "epoch": 2.1983240223463687, + "grad_norm": 4.961651802062988, + "learning_rate": 2.6312849162011174e-06, + "loss": 0.5926, + "step": 787 + }, + { + "epoch": 2.201117318435754, + "grad_norm": 8.470893859863281, + "learning_rate": 2.6346368715083796e-06, + "loss": 0.6049, + "step": 788 + }, + { + "epoch": 2.2039106145251397, + "grad_norm": 3.0153331756591797, + "learning_rate": 2.6379888268156422e-06, + "loss": 0.6014, + "step": 789 + }, + { + "epoch": 2.206703910614525, + "grad_norm": 3.172135829925537, + "learning_rate": 2.6413407821229053e-06, + "loss": 0.7079, + "step": 790 + }, + { + "epoch": 2.2094972067039107, + "grad_norm": 13.050090789794922, + "learning_rate": 2.644692737430168e-06, + "loss": 0.8322, + "step": 791 + }, + { + "epoch": 2.212290502793296, + "grad_norm": 2.283143997192383, + "learning_rate": 2.6480446927374305e-06, + "loss": 0.4765, + "step": 792 + }, + { + "epoch": 2.2150837988826817, + "grad_norm": 1.6502782106399536, + "learning_rate": 2.651396648044693e-06, + "loss": 0.4607, + "step": 793 + }, + { + "epoch": 2.217877094972067, + "grad_norm": 2.3054442405700684, + "learning_rate": 2.6547486033519553e-06, + "loss": 0.3865, + "step": 794 + }, + { + "epoch": 2.2206703910614527, + "grad_norm": 1.4920026063919067, + "learning_rate": 2.658100558659218e-06, + "loss": 0.3907, + "step": 795 + }, + { + "epoch": 2.223463687150838, + "grad_norm": 8.335897445678711, + "learning_rate": 2.6614525139664805e-06, + "loss": 0.4453, + "step": 796 + }, + { + "epoch": 2.2262569832402237, + "grad_norm": 1.3395084142684937, + "learning_rate": 2.664804469273743e-06, + "loss": 0.4355, + "step": 797 + }, + { + "epoch": 2.2290502793296088, + "grad_norm": 1.9491828680038452, + "learning_rate": 2.6681564245810058e-06, + "loss": 0.4929, + "step": 798 + }, + { + "epoch": 2.2318435754189943, + "grad_norm": 1.4206448793411255, + "learning_rate": 2.671508379888268e-06, + "loss": 0.4192, + "step": 799 + }, + { + "epoch": 2.2346368715083798, + "grad_norm": 3.3810160160064697, + "learning_rate": 2.6748603351955306e-06, + "loss": 0.4015, + "step": 800 + }, + { + "epoch": 2.2374301675977653, + "grad_norm": 1.6651445627212524, + "learning_rate": 2.678212290502793e-06, + "loss": 0.4435, + "step": 801 + }, + { + "epoch": 2.2402234636871508, + "grad_norm": 10.584975242614746, + "learning_rate": 2.681564245810056e-06, + "loss": 0.4096, + "step": 802 + }, + { + "epoch": 2.2430167597765363, + "grad_norm": 1.9810220003128052, + "learning_rate": 2.684916201117319e-06, + "loss": 0.4364, + "step": 803 + }, + { + "epoch": 2.2458100558659218, + "grad_norm": 2.507446050643921, + "learning_rate": 2.688268156424581e-06, + "loss": 0.4107, + "step": 804 + }, + { + "epoch": 2.2486033519553073, + "grad_norm": 3.678799629211426, + "learning_rate": 2.6916201117318437e-06, + "loss": 0.427, + "step": 805 + }, + { + "epoch": 2.2513966480446927, + "grad_norm": 8.231515884399414, + "learning_rate": 2.6949720670391063e-06, + "loss": 0.4946, + "step": 806 + }, + { + "epoch": 2.2541899441340782, + "grad_norm": 2.0500879287719727, + "learning_rate": 2.698324022346369e-06, + "loss": 0.4166, + "step": 807 + }, + { + "epoch": 2.2569832402234637, + "grad_norm": 2.2149972915649414, + "learning_rate": 2.7016759776536315e-06, + "loss": 0.5155, + "step": 808 + }, + { + "epoch": 2.2597765363128492, + "grad_norm": 2.256974220275879, + "learning_rate": 2.7050279329608937e-06, + "loss": 0.4712, + "step": 809 + }, + { + "epoch": 2.2625698324022347, + "grad_norm": 2.7403576374053955, + "learning_rate": 2.7083798882681563e-06, + "loss": 0.5768, + "step": 810 + }, + { + "epoch": 2.2653631284916202, + "grad_norm": 15.78232479095459, + "learning_rate": 2.711731843575419e-06, + "loss": 0.5702, + "step": 811 + }, + { + "epoch": 2.2681564245810057, + "grad_norm": 7.546295642852783, + "learning_rate": 2.7150837988826815e-06, + "loss": 0.4902, + "step": 812 + }, + { + "epoch": 2.2709497206703912, + "grad_norm": 3.6919734477996826, + "learning_rate": 2.718435754189944e-06, + "loss": 0.52, + "step": 813 + }, + { + "epoch": 2.2737430167597763, + "grad_norm": 4.238996505737305, + "learning_rate": 2.7217877094972068e-06, + "loss": 0.6123, + "step": 814 + }, + { + "epoch": 2.276536312849162, + "grad_norm": 6.941390514373779, + "learning_rate": 2.7251396648044694e-06, + "loss": 0.6482, + "step": 815 + }, + { + "epoch": 2.2793296089385473, + "grad_norm": 3.498861074447632, + "learning_rate": 2.728491620111732e-06, + "loss": 0.8026, + "step": 816 + }, + { + "epoch": 2.282122905027933, + "grad_norm": 1.6110559701919556, + "learning_rate": 2.7318435754189946e-06, + "loss": 0.3946, + "step": 817 + }, + { + "epoch": 2.2849162011173183, + "grad_norm": 2.9712185859680176, + "learning_rate": 2.7351955307262572e-06, + "loss": 0.404, + "step": 818 + }, + { + "epoch": 2.287709497206704, + "grad_norm": 1.5154914855957031, + "learning_rate": 2.73854748603352e-06, + "loss": 0.3424, + "step": 819 + }, + { + "epoch": 2.2905027932960893, + "grad_norm": 1.6249377727508545, + "learning_rate": 2.741899441340782e-06, + "loss": 0.4019, + "step": 820 + }, + { + "epoch": 2.293296089385475, + "grad_norm": 2.099909543991089, + "learning_rate": 2.7452513966480447e-06, + "loss": 0.4333, + "step": 821 + }, + { + "epoch": 2.2960893854748603, + "grad_norm": 1.52187979221344, + "learning_rate": 2.7486033519553073e-06, + "loss": 0.4498, + "step": 822 + }, + { + "epoch": 2.298882681564246, + "grad_norm": 1.6290833950042725, + "learning_rate": 2.75195530726257e-06, + "loss": 0.4104, + "step": 823 + }, + { + "epoch": 2.3016759776536313, + "grad_norm": 1.6507773399353027, + "learning_rate": 2.7553072625698325e-06, + "loss": 0.4557, + "step": 824 + }, + { + "epoch": 2.304469273743017, + "grad_norm": 2.0170226097106934, + "learning_rate": 2.7586592178770947e-06, + "loss": 0.403, + "step": 825 + }, + { + "epoch": 2.3072625698324023, + "grad_norm": 2.750206232070923, + "learning_rate": 2.7620111731843577e-06, + "loss": 0.4, + "step": 826 + }, + { + "epoch": 2.310055865921788, + "grad_norm": 2.36327862739563, + "learning_rate": 2.7653631284916204e-06, + "loss": 0.4371, + "step": 827 + }, + { + "epoch": 2.3128491620111733, + "grad_norm": 1.8770337104797363, + "learning_rate": 2.768715083798883e-06, + "loss": 0.4281, + "step": 828 + }, + { + "epoch": 2.315642458100559, + "grad_norm": 1.5466649532318115, + "learning_rate": 2.7720670391061456e-06, + "loss": 0.4709, + "step": 829 + }, + { + "epoch": 2.3184357541899443, + "grad_norm": 2.2570924758911133, + "learning_rate": 2.7754189944134078e-06, + "loss": 0.5141, + "step": 830 + }, + { + "epoch": 2.32122905027933, + "grad_norm": 1.8356703519821167, + "learning_rate": 2.7787709497206704e-06, + "loss": 0.4163, + "step": 831 + }, + { + "epoch": 2.3240223463687153, + "grad_norm": 1.9509092569351196, + "learning_rate": 2.782122905027933e-06, + "loss": 0.4929, + "step": 832 + }, + { + "epoch": 2.326815642458101, + "grad_norm": 2.727311372756958, + "learning_rate": 2.7854748603351956e-06, + "loss": 0.4734, + "step": 833 + }, + { + "epoch": 2.329608938547486, + "grad_norm": 2.3327295780181885, + "learning_rate": 2.7888268156424582e-06, + "loss": 0.5129, + "step": 834 + }, + { + "epoch": 2.3324022346368714, + "grad_norm": 3.8173024654388428, + "learning_rate": 2.7921787709497204e-06, + "loss": 0.4769, + "step": 835 + }, + { + "epoch": 2.335195530726257, + "grad_norm": 3.366360902786255, + "learning_rate": 2.795530726256983e-06, + "loss": 0.5417, + "step": 836 + }, + { + "epoch": 2.3379888268156424, + "grad_norm": 3.2680160999298096, + "learning_rate": 2.7988826815642457e-06, + "loss": 0.5086, + "step": 837 + }, + { + "epoch": 2.340782122905028, + "grad_norm": 3.4435629844665527, + "learning_rate": 2.8022346368715083e-06, + "loss": 0.5275, + "step": 838 + }, + { + "epoch": 2.3435754189944134, + "grad_norm": 2.170234441757202, + "learning_rate": 2.8055865921787713e-06, + "loss": 0.5373, + "step": 839 + }, + { + "epoch": 2.346368715083799, + "grad_norm": 3.72652268409729, + "learning_rate": 2.808938547486034e-06, + "loss": 0.5684, + "step": 840 + }, + { + "epoch": 2.3491620111731844, + "grad_norm": 3.534627914428711, + "learning_rate": 2.812290502793296e-06, + "loss": 0.8594, + "step": 841 + }, + { + "epoch": 2.35195530726257, + "grad_norm": 1.487287163734436, + "learning_rate": 2.8156424581005587e-06, + "loss": 0.4308, + "step": 842 + }, + { + "epoch": 2.3547486033519553, + "grad_norm": 1.233093023300171, + "learning_rate": 2.8189944134078214e-06, + "loss": 0.366, + "step": 843 + }, + { + "epoch": 2.357541899441341, + "grad_norm": 1.3019872903823853, + "learning_rate": 2.822346368715084e-06, + "loss": 0.3527, + "step": 844 + }, + { + "epoch": 2.3603351955307263, + "grad_norm": 1.265965223312378, + "learning_rate": 2.8256983240223466e-06, + "loss": 0.4128, + "step": 845 + }, + { + "epoch": 2.363128491620112, + "grad_norm": 1.8386411666870117, + "learning_rate": 2.8290502793296088e-06, + "loss": 0.4197, + "step": 846 + }, + { + "epoch": 2.3659217877094973, + "grad_norm": 1.689615249633789, + "learning_rate": 2.8324022346368714e-06, + "loss": 0.3925, + "step": 847 + }, + { + "epoch": 2.368715083798883, + "grad_norm": 2.5115861892700195, + "learning_rate": 2.835754189944134e-06, + "loss": 0.4478, + "step": 848 + }, + { + "epoch": 2.3715083798882683, + "grad_norm": 2.2385082244873047, + "learning_rate": 2.8391061452513966e-06, + "loss": 0.3999, + "step": 849 + }, + { + "epoch": 2.3743016759776534, + "grad_norm": 1.51487135887146, + "learning_rate": 2.8424581005586593e-06, + "loss": 0.3841, + "step": 850 + }, + { + "epoch": 2.377094972067039, + "grad_norm": 2.061770439147949, + "learning_rate": 2.845810055865922e-06, + "loss": 0.3673, + "step": 851 + }, + { + "epoch": 2.3798882681564244, + "grad_norm": 1.6525633335113525, + "learning_rate": 2.8491620111731845e-06, + "loss": 0.4338, + "step": 852 + }, + { + "epoch": 2.38268156424581, + "grad_norm": 2.2444138526916504, + "learning_rate": 2.852513966480447e-06, + "loss": 0.4038, + "step": 853 + }, + { + "epoch": 2.3854748603351954, + "grad_norm": 18.109643936157227, + "learning_rate": 2.8558659217877097e-06, + "loss": 0.4882, + "step": 854 + }, + { + "epoch": 2.388268156424581, + "grad_norm": 1.8350059986114502, + "learning_rate": 2.8592178770949723e-06, + "loss": 0.4353, + "step": 855 + }, + { + "epoch": 2.3910614525139664, + "grad_norm": 2.5742499828338623, + "learning_rate": 2.8625698324022345e-06, + "loss": 0.4198, + "step": 856 + }, + { + "epoch": 2.393854748603352, + "grad_norm": 1.7839794158935547, + "learning_rate": 2.865921787709497e-06, + "loss": 0.4408, + "step": 857 + }, + { + "epoch": 2.3966480446927374, + "grad_norm": 2.289353609085083, + "learning_rate": 2.8692737430167598e-06, + "loss": 0.464, + "step": 858 + }, + { + "epoch": 2.399441340782123, + "grad_norm": 2.520132541656494, + "learning_rate": 2.8726256983240224e-06, + "loss": 0.4344, + "step": 859 + }, + { + "epoch": 2.4022346368715084, + "grad_norm": 4.595797061920166, + "learning_rate": 2.875977653631285e-06, + "loss": 0.4629, + "step": 860 + }, + { + "epoch": 2.405027932960894, + "grad_norm": 3.2993080615997314, + "learning_rate": 2.879329608938547e-06, + "loss": 0.507, + "step": 861 + }, + { + "epoch": 2.4078212290502794, + "grad_norm": 2.290559768676758, + "learning_rate": 2.88268156424581e-06, + "loss": 0.488, + "step": 862 + }, + { + "epoch": 2.410614525139665, + "grad_norm": 2.3599212169647217, + "learning_rate": 2.886033519553073e-06, + "loss": 0.5718, + "step": 863 + }, + { + "epoch": 2.4134078212290504, + "grad_norm": 6.1311936378479, + "learning_rate": 2.8893854748603354e-06, + "loss": 0.5243, + "step": 864 + }, + { + "epoch": 2.416201117318436, + "grad_norm": 4.448702335357666, + "learning_rate": 2.892737430167598e-06, + "loss": 0.6898, + "step": 865 + }, + { + "epoch": 2.4189944134078214, + "grad_norm": 4.68412971496582, + "learning_rate": 2.8960893854748607e-06, + "loss": 0.6835, + "step": 866 + }, + { + "epoch": 2.421787709497207, + "grad_norm": 1.317013144493103, + "learning_rate": 2.899441340782123e-06, + "loss": 0.3956, + "step": 867 + }, + { + "epoch": 2.4245810055865924, + "grad_norm": 1.2126437425613403, + "learning_rate": 2.9027932960893855e-06, + "loss": 0.3722, + "step": 868 + }, + { + "epoch": 2.427374301675978, + "grad_norm": 2.039630174636841, + "learning_rate": 2.906145251396648e-06, + "loss": 0.3632, + "step": 869 + }, + { + "epoch": 2.430167597765363, + "grad_norm": 1.4606226682662964, + "learning_rate": 2.9094972067039107e-06, + "loss": 0.3894, + "step": 870 + }, + { + "epoch": 2.4329608938547485, + "grad_norm": 1.3030271530151367, + "learning_rate": 2.9128491620111733e-06, + "loss": 0.4288, + "step": 871 + }, + { + "epoch": 2.435754189944134, + "grad_norm": 1.8077303171157837, + "learning_rate": 2.9162011173184355e-06, + "loss": 0.359, + "step": 872 + }, + { + "epoch": 2.4385474860335195, + "grad_norm": 1.1585514545440674, + "learning_rate": 2.919553072625698e-06, + "loss": 0.3725, + "step": 873 + }, + { + "epoch": 2.441340782122905, + "grad_norm": 1.3698378801345825, + "learning_rate": 2.9229050279329608e-06, + "loss": 0.4356, + "step": 874 + }, + { + "epoch": 2.4441340782122905, + "grad_norm": 1.5453447103500366, + "learning_rate": 2.926256983240224e-06, + "loss": 0.4278, + "step": 875 + }, + { + "epoch": 2.446927374301676, + "grad_norm": 1.5376777648925781, + "learning_rate": 2.9296089385474864e-06, + "loss": 0.4218, + "step": 876 + }, + { + "epoch": 2.4497206703910615, + "grad_norm": 2.9291067123413086, + "learning_rate": 2.9329608938547486e-06, + "loss": 0.4322, + "step": 877 + }, + { + "epoch": 2.452513966480447, + "grad_norm": 1.9900550842285156, + "learning_rate": 2.9363128491620112e-06, + "loss": 0.4269, + "step": 878 + }, + { + "epoch": 2.4553072625698324, + "grad_norm": 2.0277698040008545, + "learning_rate": 2.939664804469274e-06, + "loss": 0.3665, + "step": 879 + }, + { + "epoch": 2.458100558659218, + "grad_norm": 1.9600462913513184, + "learning_rate": 2.9430167597765365e-06, + "loss": 0.4405, + "step": 880 + }, + { + "epoch": 2.4608938547486034, + "grad_norm": 4.011704444885254, + "learning_rate": 2.946368715083799e-06, + "loss": 0.411, + "step": 881 + }, + { + "epoch": 2.463687150837989, + "grad_norm": 1.9997632503509521, + "learning_rate": 2.9497206703910613e-06, + "loss": 0.4521, + "step": 882 + }, + { + "epoch": 2.4664804469273744, + "grad_norm": 10.953627586364746, + "learning_rate": 2.953072625698324e-06, + "loss": 0.45, + "step": 883 + }, + { + "epoch": 2.46927374301676, + "grad_norm": 2.124859571456909, + "learning_rate": 2.9564245810055865e-06, + "loss": 0.4494, + "step": 884 + }, + { + "epoch": 2.472067039106145, + "grad_norm": 2.119060516357422, + "learning_rate": 2.959776536312849e-06, + "loss": 0.4573, + "step": 885 + }, + { + "epoch": 2.4748603351955305, + "grad_norm": 1.8804141283035278, + "learning_rate": 2.9631284916201117e-06, + "loss": 0.4878, + "step": 886 + }, + { + "epoch": 2.477653631284916, + "grad_norm": 2.4910428524017334, + "learning_rate": 2.9664804469273748e-06, + "loss": 0.5197, + "step": 887 + }, + { + "epoch": 2.4804469273743015, + "grad_norm": 5.647790431976318, + "learning_rate": 2.969832402234637e-06, + "loss": 0.4983, + "step": 888 + }, + { + "epoch": 2.483240223463687, + "grad_norm": 5.405727386474609, + "learning_rate": 2.9731843575418996e-06, + "loss": 0.5653, + "step": 889 + }, + { + "epoch": 2.4860335195530725, + "grad_norm": 6.041434288024902, + "learning_rate": 2.976536312849162e-06, + "loss": 0.6194, + "step": 890 + }, + { + "epoch": 2.488826815642458, + "grad_norm": 3.7605721950531006, + "learning_rate": 2.979888268156425e-06, + "loss": 0.6714, + "step": 891 + }, + { + "epoch": 2.4916201117318435, + "grad_norm": 1.7148029804229736, + "learning_rate": 2.9832402234636874e-06, + "loss": 0.406, + "step": 892 + }, + { + "epoch": 2.494413407821229, + "grad_norm": 1.2828824520111084, + "learning_rate": 2.9865921787709496e-06, + "loss": 0.3644, + "step": 893 + }, + { + "epoch": 2.4972067039106145, + "grad_norm": 1.7568423748016357, + "learning_rate": 2.9899441340782122e-06, + "loss": 0.3923, + "step": 894 + }, + { + "epoch": 2.5, + "grad_norm": 2.0014357566833496, + "learning_rate": 2.993296089385475e-06, + "loss": 0.4067, + "step": 895 + }, + { + "epoch": 2.5027932960893855, + "grad_norm": 1.9997479915618896, + "learning_rate": 2.9966480446927375e-06, + "loss": 0.4237, + "step": 896 + }, + { + "epoch": 2.505586592178771, + "grad_norm": 2.058027982711792, + "learning_rate": 3e-06, + "loss": 0.3675, + "step": 897 + }, + { + "epoch": 2.5083798882681565, + "grad_norm": 1.6324915885925293, + "learning_rate": 3.0033519553072623e-06, + "loss": 0.4556, + "step": 898 + }, + { + "epoch": 2.511173184357542, + "grad_norm": 1.4901829957962036, + "learning_rate": 3.0067039106145253e-06, + "loss": 0.3793, + "step": 899 + }, + { + "epoch": 2.5139664804469275, + "grad_norm": 1.6047043800354004, + "learning_rate": 3.010055865921788e-06, + "loss": 0.4014, + "step": 900 + }, + { + "epoch": 2.516759776536313, + "grad_norm": 3.613027811050415, + "learning_rate": 3.0134078212290505e-06, + "loss": 0.4372, + "step": 901 + }, + { + "epoch": 2.5195530726256985, + "grad_norm": 1.9516769647598267, + "learning_rate": 3.016759776536313e-06, + "loss": 0.3813, + "step": 902 + }, + { + "epoch": 2.522346368715084, + "grad_norm": 1.7330446243286133, + "learning_rate": 3.0201117318435753e-06, + "loss": 0.4212, + "step": 903 + }, + { + "epoch": 2.5251396648044695, + "grad_norm": 1.918790340423584, + "learning_rate": 3.023463687150838e-06, + "loss": 0.3954, + "step": 904 + }, + { + "epoch": 2.527932960893855, + "grad_norm": 2.006517171859741, + "learning_rate": 3.0268156424581006e-06, + "loss": 0.4561, + "step": 905 + }, + { + "epoch": 2.5307262569832405, + "grad_norm": 2.7694389820098877, + "learning_rate": 3.030167597765363e-06, + "loss": 0.421, + "step": 906 + }, + { + "epoch": 2.5335195530726256, + "grad_norm": 12.75838851928711, + "learning_rate": 3.033519553072626e-06, + "loss": 0.4036, + "step": 907 + }, + { + "epoch": 2.536312849162011, + "grad_norm": 3.0600078105926514, + "learning_rate": 3.036871508379888e-06, + "loss": 0.4141, + "step": 908 + }, + { + "epoch": 2.5391061452513966, + "grad_norm": 2.6760451793670654, + "learning_rate": 3.0402234636871506e-06, + "loss": 0.4216, + "step": 909 + }, + { + "epoch": 2.541899441340782, + "grad_norm": 4.51766300201416, + "learning_rate": 3.0435754189944132e-06, + "loss": 0.52, + "step": 910 + }, + { + "epoch": 2.5446927374301676, + "grad_norm": 2.1919474601745605, + "learning_rate": 3.0469273743016763e-06, + "loss": 0.4874, + "step": 911 + }, + { + "epoch": 2.547486033519553, + "grad_norm": 3.025441884994507, + "learning_rate": 3.050279329608939e-06, + "loss": 0.4535, + "step": 912 + }, + { + "epoch": 2.5502793296089385, + "grad_norm": 2.2521986961364746, + "learning_rate": 3.0536312849162015e-06, + "loss": 0.4564, + "step": 913 + }, + { + "epoch": 2.553072625698324, + "grad_norm": 2.2791748046875, + "learning_rate": 3.0569832402234637e-06, + "loss": 0.553, + "step": 914 + }, + { + "epoch": 2.5558659217877095, + "grad_norm": 2.909482479095459, + "learning_rate": 3.0603351955307263e-06, + "loss": 0.5794, + "step": 915 + }, + { + "epoch": 2.558659217877095, + "grad_norm": 4.36000919342041, + "learning_rate": 3.063687150837989e-06, + "loss": 0.8173, + "step": 916 + }, + { + "epoch": 2.5614525139664805, + "grad_norm": 3.5961947441101074, + "learning_rate": 3.0670391061452515e-06, + "loss": 0.4211, + "step": 917 + }, + { + "epoch": 2.564245810055866, + "grad_norm": 1.292075276374817, + "learning_rate": 3.070391061452514e-06, + "loss": 0.3872, + "step": 918 + }, + { + "epoch": 2.5670391061452515, + "grad_norm": 1.0457656383514404, + "learning_rate": 3.0737430167597764e-06, + "loss": 0.3753, + "step": 919 + }, + { + "epoch": 2.5698324022346366, + "grad_norm": 1.402512550354004, + "learning_rate": 3.077094972067039e-06, + "loss": 0.3453, + "step": 920 + }, + { + "epoch": 2.572625698324022, + "grad_norm": 1.5532149076461792, + "learning_rate": 3.0804469273743016e-06, + "loss": 0.3966, + "step": 921 + }, + { + "epoch": 2.5754189944134076, + "grad_norm": 1.9059978723526, + "learning_rate": 3.083798882681564e-06, + "loss": 0.4078, + "step": 922 + }, + { + "epoch": 2.578212290502793, + "grad_norm": 1.3203620910644531, + "learning_rate": 3.0871508379888272e-06, + "loss": 0.3793, + "step": 923 + }, + { + "epoch": 2.5810055865921786, + "grad_norm": 1.2588735818862915, + "learning_rate": 3.0905027932960894e-06, + "loss": 0.4074, + "step": 924 + }, + { + "epoch": 2.583798882681564, + "grad_norm": 2.2567412853240967, + "learning_rate": 3.093854748603352e-06, + "loss": 0.3751, + "step": 925 + }, + { + "epoch": 2.5865921787709496, + "grad_norm": 1.5239313840866089, + "learning_rate": 3.0972067039106147e-06, + "loss": 0.4471, + "step": 926 + }, + { + "epoch": 2.589385474860335, + "grad_norm": 1.819555401802063, + "learning_rate": 3.1005586592178773e-06, + "loss": 0.3659, + "step": 927 + }, + { + "epoch": 2.5921787709497206, + "grad_norm": 2.1956536769866943, + "learning_rate": 3.10391061452514e-06, + "loss": 0.3956, + "step": 928 + }, + { + "epoch": 2.594972067039106, + "grad_norm": 1.6065884828567505, + "learning_rate": 3.107262569832402e-06, + "loss": 0.3836, + "step": 929 + }, + { + "epoch": 2.5977653631284916, + "grad_norm": 1.707672357559204, + "learning_rate": 3.1106145251396647e-06, + "loss": 0.4031, + "step": 930 + }, + { + "epoch": 2.600558659217877, + "grad_norm": 3.4139113426208496, + "learning_rate": 3.1139664804469273e-06, + "loss": 0.4224, + "step": 931 + }, + { + "epoch": 2.6033519553072626, + "grad_norm": 4.622681140899658, + "learning_rate": 3.11731843575419e-06, + "loss": 0.4359, + "step": 932 + }, + { + "epoch": 2.606145251396648, + "grad_norm": 2.0437448024749756, + "learning_rate": 3.1206703910614526e-06, + "loss": 0.43, + "step": 933 + }, + { + "epoch": 2.6089385474860336, + "grad_norm": 2.0084331035614014, + "learning_rate": 3.124022346368715e-06, + "loss": 0.4505, + "step": 934 + }, + { + "epoch": 2.611731843575419, + "grad_norm": 2.3053860664367676, + "learning_rate": 3.1273743016759778e-06, + "loss": 0.4501, + "step": 935 + }, + { + "epoch": 2.6145251396648046, + "grad_norm": 4.153201580047607, + "learning_rate": 3.1307262569832404e-06, + "loss": 0.418, + "step": 936 + }, + { + "epoch": 2.61731843575419, + "grad_norm": 1.7330175638198853, + "learning_rate": 3.134078212290503e-06, + "loss": 0.4686, + "step": 937 + }, + { + "epoch": 2.6201117318435756, + "grad_norm": 5.173537731170654, + "learning_rate": 3.1374301675977656e-06, + "loss": 0.5721, + "step": 938 + }, + { + "epoch": 2.622905027932961, + "grad_norm": 2.7377355098724365, + "learning_rate": 3.1407821229050282e-06, + "loss": 0.5272, + "step": 939 + }, + { + "epoch": 2.6256983240223466, + "grad_norm": 2.94636607170105, + "learning_rate": 3.1441340782122904e-06, + "loss": 0.5702, + "step": 940 + }, + { + "epoch": 2.628491620111732, + "grad_norm": 6.146542072296143, + "learning_rate": 3.147486033519553e-06, + "loss": 0.7821, + "step": 941 + }, + { + "epoch": 2.631284916201117, + "grad_norm": 3.6064701080322266, + "learning_rate": 3.1508379888268157e-06, + "loss": 0.4171, + "step": 942 + }, + { + "epoch": 2.6340782122905027, + "grad_norm": 1.4715392589569092, + "learning_rate": 3.1541899441340783e-06, + "loss": 0.3442, + "step": 943 + }, + { + "epoch": 2.636871508379888, + "grad_norm": 2.9554827213287354, + "learning_rate": 3.157541899441341e-06, + "loss": 0.375, + "step": 944 + }, + { + "epoch": 2.6396648044692737, + "grad_norm": 5.253285884857178, + "learning_rate": 3.160893854748603e-06, + "loss": 0.3553, + "step": 945 + }, + { + "epoch": 2.642458100558659, + "grad_norm": 1.995908260345459, + "learning_rate": 3.1642458100558657e-06, + "loss": 0.3384, + "step": 946 + }, + { + "epoch": 2.6452513966480447, + "grad_norm": 1.5907999277114868, + "learning_rate": 3.1675977653631287e-06, + "loss": 0.3435, + "step": 947 + }, + { + "epoch": 2.64804469273743, + "grad_norm": 1.5436711311340332, + "learning_rate": 3.1709497206703914e-06, + "loss": 0.421, + "step": 948 + }, + { + "epoch": 2.6508379888268156, + "grad_norm": 1.3006367683410645, + "learning_rate": 3.174301675977654e-06, + "loss": 0.3628, + "step": 949 + }, + { + "epoch": 2.653631284916201, + "grad_norm": 1.660723328590393, + "learning_rate": 3.177653631284916e-06, + "loss": 0.3668, + "step": 950 + }, + { + "epoch": 2.6564245810055866, + "grad_norm": 1.6349189281463623, + "learning_rate": 3.181005586592179e-06, + "loss": 0.3781, + "step": 951 + }, + { + "epoch": 2.659217877094972, + "grad_norm": 3.4873783588409424, + "learning_rate": 3.1843575418994414e-06, + "loss": 0.3897, + "step": 952 + }, + { + "epoch": 2.6620111731843576, + "grad_norm": 1.4385789632797241, + "learning_rate": 3.187709497206704e-06, + "loss": 0.4159, + "step": 953 + }, + { + "epoch": 2.664804469273743, + "grad_norm": 1.7710334062576294, + "learning_rate": 3.1910614525139666e-06, + "loss": 0.4032, + "step": 954 + }, + { + "epoch": 2.6675977653631286, + "grad_norm": 3.0238184928894043, + "learning_rate": 3.194413407821229e-06, + "loss": 0.3979, + "step": 955 + }, + { + "epoch": 2.6703910614525137, + "grad_norm": 2.936286449432373, + "learning_rate": 3.1977653631284914e-06, + "loss": 0.4237, + "step": 956 + }, + { + "epoch": 2.673184357541899, + "grad_norm": 1.7814221382141113, + "learning_rate": 3.201117318435754e-06, + "loss": 0.3962, + "step": 957 + }, + { + "epoch": 2.6759776536312847, + "grad_norm": 1.9644231796264648, + "learning_rate": 3.2044692737430167e-06, + "loss": 0.3782, + "step": 958 + }, + { + "epoch": 2.67877094972067, + "grad_norm": 1.9036448001861572, + "learning_rate": 3.2078212290502797e-06, + "loss": 0.4137, + "step": 959 + }, + { + "epoch": 2.6815642458100557, + "grad_norm": 1.7850133180618286, + "learning_rate": 3.2111731843575423e-06, + "loss": 0.4244, + "step": 960 + }, + { + "epoch": 2.684357541899441, + "grad_norm": 2.2023653984069824, + "learning_rate": 3.2145251396648045e-06, + "loss": 0.513, + "step": 961 + }, + { + "epoch": 2.6871508379888267, + "grad_norm": 2.0695505142211914, + "learning_rate": 3.217877094972067e-06, + "loss": 0.5612, + "step": 962 + }, + { + "epoch": 2.689944134078212, + "grad_norm": 3.0341434478759766, + "learning_rate": 3.2212290502793298e-06, + "loss": 0.4419, + "step": 963 + }, + { + "epoch": 2.6927374301675977, + "grad_norm": 2.3564138412475586, + "learning_rate": 3.2245810055865924e-06, + "loss": 0.4988, + "step": 964 + }, + { + "epoch": 2.695530726256983, + "grad_norm": 2.4233615398406982, + "learning_rate": 3.227932960893855e-06, + "loss": 0.5149, + "step": 965 + }, + { + "epoch": 2.6983240223463687, + "grad_norm": 3.9186410903930664, + "learning_rate": 3.231284916201117e-06, + "loss": 0.6469, + "step": 966 + }, + { + "epoch": 2.701117318435754, + "grad_norm": 1.626834750175476, + "learning_rate": 3.23463687150838e-06, + "loss": 0.3781, + "step": 967 + }, + { + "epoch": 2.7039106145251397, + "grad_norm": 1.1665780544281006, + "learning_rate": 3.2379888268156424e-06, + "loss": 0.3564, + "step": 968 + }, + { + "epoch": 2.706703910614525, + "grad_norm": 1.3284854888916016, + "learning_rate": 3.241340782122905e-06, + "loss": 0.3494, + "step": 969 + }, + { + "epoch": 2.7094972067039107, + "grad_norm": 1.5736455917358398, + "learning_rate": 3.2446927374301676e-06, + "loss": 0.351, + "step": 970 + }, + { + "epoch": 2.712290502793296, + "grad_norm": 1.4029985666275024, + "learning_rate": 3.2480446927374303e-06, + "loss": 0.3393, + "step": 971 + }, + { + "epoch": 2.7150837988826817, + "grad_norm": 1.621883511543274, + "learning_rate": 3.251396648044693e-06, + "loss": 0.3837, + "step": 972 + }, + { + "epoch": 2.717877094972067, + "grad_norm": 2.1428682804107666, + "learning_rate": 3.2547486033519555e-06, + "loss": 0.3322, + "step": 973 + }, + { + "epoch": 2.7206703910614527, + "grad_norm": 2.122093915939331, + "learning_rate": 3.258100558659218e-06, + "loss": 0.3641, + "step": 974 + }, + { + "epoch": 2.723463687150838, + "grad_norm": 1.7705615758895874, + "learning_rate": 3.2614525139664807e-06, + "loss": 0.3373, + "step": 975 + }, + { + "epoch": 2.7262569832402237, + "grad_norm": 1.4305355548858643, + "learning_rate": 3.264804469273743e-06, + "loss": 0.3856, + "step": 976 + }, + { + "epoch": 2.729050279329609, + "grad_norm": 1.760508418083191, + "learning_rate": 3.2681564245810055e-06, + "loss": 0.4187, + "step": 977 + }, + { + "epoch": 2.7318435754189943, + "grad_norm": 2.4990127086639404, + "learning_rate": 3.271508379888268e-06, + "loss": 0.3602, + "step": 978 + }, + { + "epoch": 2.7346368715083798, + "grad_norm": 1.5644584894180298, + "learning_rate": 3.2748603351955308e-06, + "loss": 0.4019, + "step": 979 + }, + { + "epoch": 2.7374301675977653, + "grad_norm": 3.966329574584961, + "learning_rate": 3.2782122905027934e-06, + "loss": 0.4018, + "step": 980 + }, + { + "epoch": 2.7402234636871508, + "grad_norm": 2.273045301437378, + "learning_rate": 3.281564245810056e-06, + "loss": 0.4329, + "step": 981 + }, + { + "epoch": 2.7430167597765363, + "grad_norm": 2.4368057250976562, + "learning_rate": 3.284916201117318e-06, + "loss": 0.3893, + "step": 982 + }, + { + "epoch": 2.7458100558659218, + "grad_norm": 3.9151909351348877, + "learning_rate": 3.2882681564245812e-06, + "loss": 0.3732, + "step": 983 + }, + { + "epoch": 2.7486033519553073, + "grad_norm": 3.034959316253662, + "learning_rate": 3.291620111731844e-06, + "loss": 0.4327, + "step": 984 + }, + { + "epoch": 2.7513966480446927, + "grad_norm": 2.18833327293396, + "learning_rate": 3.2949720670391065e-06, + "loss": 0.4247, + "step": 985 + }, + { + "epoch": 2.7541899441340782, + "grad_norm": 1.9602417945861816, + "learning_rate": 3.298324022346369e-06, + "loss": 0.45, + "step": 986 + }, + { + "epoch": 2.7569832402234637, + "grad_norm": 3.081732988357544, + "learning_rate": 3.3016759776536313e-06, + "loss": 0.4764, + "step": 987 + }, + { + "epoch": 2.7597765363128492, + "grad_norm": 2.4676642417907715, + "learning_rate": 3.305027932960894e-06, + "loss": 0.5186, + "step": 988 + }, + { + "epoch": 2.7625698324022347, + "grad_norm": 3.330019474029541, + "learning_rate": 3.3083798882681565e-06, + "loss": 0.5029, + "step": 989 + }, + { + "epoch": 2.7653631284916202, + "grad_norm": 3.71659517288208, + "learning_rate": 3.311731843575419e-06, + "loss": 0.5862, + "step": 990 + }, + { + "epoch": 2.7681564245810057, + "grad_norm": 3.900618314743042, + "learning_rate": 3.3150837988826817e-06, + "loss": 0.7893, + "step": 991 + }, + { + "epoch": 2.770949720670391, + "grad_norm": 1.118536114692688, + "learning_rate": 3.318435754189944e-06, + "loss": 0.3567, + "step": 992 + }, + { + "epoch": 2.7737430167597763, + "grad_norm": 0.9696893095970154, + "learning_rate": 3.3217877094972065e-06, + "loss": 0.3868, + "step": 993 + }, + { + "epoch": 2.776536312849162, + "grad_norm": 0.9937577247619629, + "learning_rate": 3.325139664804469e-06, + "loss": 0.3607, + "step": 994 + }, + { + "epoch": 2.7793296089385473, + "grad_norm": 2.161597967147827, + "learning_rate": 3.328491620111732e-06, + "loss": 0.3616, + "step": 995 + }, + { + "epoch": 2.782122905027933, + "grad_norm": 1.3333648443222046, + "learning_rate": 3.331843575418995e-06, + "loss": 0.4025, + "step": 996 + }, + { + "epoch": 2.7849162011173183, + "grad_norm": 4.907354831695557, + "learning_rate": 3.335195530726257e-06, + "loss": 0.3615, + "step": 997 + }, + { + "epoch": 2.787709497206704, + "grad_norm": 2.0579609870910645, + "learning_rate": 3.3385474860335196e-06, + "loss": 0.4128, + "step": 998 + }, + { + "epoch": 2.7905027932960893, + "grad_norm": 1.808306097984314, + "learning_rate": 3.3418994413407822e-06, + "loss": 0.3502, + "step": 999 + }, + { + "epoch": 2.793296089385475, + "grad_norm": 1.167189359664917, + "learning_rate": 3.345251396648045e-06, + "loss": 0.346, + "step": 1000 + }, + { + "epoch": 2.793296089385475, + "eval_cer": 0.21314086827490644, + "eval_loss": 0.4366828203201294, + "eval_runtime": 23.164, + "eval_samples_per_second": 195.907, + "eval_steps_per_second": 0.648, + "eval_wer": 0.4197089466774671, + "step": 1000 + }, + { + "epoch": 2.7960893854748603, + "grad_norm": 3.5702593326568604, + "learning_rate": 3.3486033519553075e-06, + "loss": 0.3866, + "step": 1001 + }, + { + "epoch": 2.798882681564246, + "grad_norm": 1.5501301288604736, + "learning_rate": 3.3519553072625697e-06, + "loss": 0.3772, + "step": 1002 + }, + { + "epoch": 2.8016759776536313, + "grad_norm": 2.2063379287719727, + "learning_rate": 3.3553072625698323e-06, + "loss": 0.4225, + "step": 1003 + }, + { + "epoch": 2.804469273743017, + "grad_norm": 1.3987849950790405, + "learning_rate": 3.358659217877095e-06, + "loss": 0.4006, + "step": 1004 + }, + { + "epoch": 2.8072625698324023, + "grad_norm": 1.5713845491409302, + "learning_rate": 3.3620111731843575e-06, + "loss": 0.4123, + "step": 1005 + }, + { + "epoch": 2.810055865921788, + "grad_norm": 1.2640436887741089, + "learning_rate": 3.36536312849162e-06, + "loss": 0.3681, + "step": 1006 + }, + { + "epoch": 2.8128491620111733, + "grad_norm": 2.3267722129821777, + "learning_rate": 3.368715083798883e-06, + "loss": 0.4189, + "step": 1007 + }, + { + "epoch": 2.815642458100559, + "grad_norm": 1.55404794216156, + "learning_rate": 3.3720670391061453e-06, + "loss": 0.4076, + "step": 1008 + }, + { + "epoch": 2.8184357541899443, + "grad_norm": 1.764844298362732, + "learning_rate": 3.375418994413408e-06, + "loss": 0.4581, + "step": 1009 + }, + { + "epoch": 2.82122905027933, + "grad_norm": 1.6694835424423218, + "learning_rate": 3.3787709497206706e-06, + "loss": 0.4208, + "step": 1010 + }, + { + "epoch": 2.8240223463687153, + "grad_norm": 1.9249577522277832, + "learning_rate": 3.382122905027933e-06, + "loss": 0.4143, + "step": 1011 + }, + { + "epoch": 2.826815642458101, + "grad_norm": 3.644536256790161, + "learning_rate": 3.385474860335196e-06, + "loss": 0.4458, + "step": 1012 + }, + { + "epoch": 2.8296089385474863, + "grad_norm": 3.30688738822937, + "learning_rate": 3.388826815642458e-06, + "loss": 0.4858, + "step": 1013 + }, + { + "epoch": 2.8324022346368714, + "grad_norm": 3.1424355506896973, + "learning_rate": 3.3921787709497206e-06, + "loss": 0.5046, + "step": 1014 + }, + { + "epoch": 2.835195530726257, + "grad_norm": 3.5599446296691895, + "learning_rate": 3.3955307262569832e-06, + "loss": 0.4762, + "step": 1015 + }, + { + "epoch": 2.8379888268156424, + "grad_norm": 5.154666423797607, + "learning_rate": 3.398882681564246e-06, + "loss": 0.6363, + "step": 1016 + }, + { + "epoch": 2.840782122905028, + "grad_norm": 1.4966647624969482, + "learning_rate": 3.4022346368715085e-06, + "loss": 0.3659, + "step": 1017 + }, + { + "epoch": 2.8435754189944134, + "grad_norm": 1.048509120941162, + "learning_rate": 3.4055865921787707e-06, + "loss": 0.3452, + "step": 1018 + }, + { + "epoch": 2.846368715083799, + "grad_norm": 1.1709436178207397, + "learning_rate": 3.4089385474860337e-06, + "loss": 0.3732, + "step": 1019 + }, + { + "epoch": 2.8491620111731844, + "grad_norm": 1.2480288743972778, + "learning_rate": 3.4122905027932963e-06, + "loss": 0.3534, + "step": 1020 + }, + { + "epoch": 2.85195530726257, + "grad_norm": 1.2058454751968384, + "learning_rate": 3.415642458100559e-06, + "loss": 0.3532, + "step": 1021 + }, + { + "epoch": 2.8547486033519553, + "grad_norm": 1.554110050201416, + "learning_rate": 3.4189944134078215e-06, + "loss": 0.3577, + "step": 1022 + }, + { + "epoch": 2.857541899441341, + "grad_norm": 3.5964910984039307, + "learning_rate": 3.4223463687150837e-06, + "loss": 0.385, + "step": 1023 + }, + { + "epoch": 2.8603351955307263, + "grad_norm": 1.5678935050964355, + "learning_rate": 3.4256983240223464e-06, + "loss": 0.3459, + "step": 1024 + }, + { + "epoch": 2.863128491620112, + "grad_norm": 1.3333802223205566, + "learning_rate": 3.429050279329609e-06, + "loss": 0.3219, + "step": 1025 + }, + { + "epoch": 2.8659217877094973, + "grad_norm": 1.2826341390609741, + "learning_rate": 3.4324022346368716e-06, + "loss": 0.3902, + "step": 1026 + }, + { + "epoch": 2.868715083798883, + "grad_norm": 1.950175404548645, + "learning_rate": 3.435754189944134e-06, + "loss": 0.3578, + "step": 1027 + }, + { + "epoch": 2.871508379888268, + "grad_norm": 31.124675750732422, + "learning_rate": 3.439106145251397e-06, + "loss": 0.4101, + "step": 1028 + }, + { + "epoch": 2.8743016759776534, + "grad_norm": 1.3327090740203857, + "learning_rate": 3.442458100558659e-06, + "loss": 0.3719, + "step": 1029 + }, + { + "epoch": 2.877094972067039, + "grad_norm": 1.4331103563308716, + "learning_rate": 3.4458100558659216e-06, + "loss": 0.3687, + "step": 1030 + }, + { + "epoch": 2.8798882681564244, + "grad_norm": 2.450512647628784, + "learning_rate": 3.4491620111731847e-06, + "loss": 0.3272, + "step": 1031 + }, + { + "epoch": 2.88268156424581, + "grad_norm": 1.997200846672058, + "learning_rate": 3.4525139664804473e-06, + "loss": 0.4216, + "step": 1032 + }, + { + "epoch": 2.8854748603351954, + "grad_norm": 2.1695902347564697, + "learning_rate": 3.45586592178771e-06, + "loss": 0.4125, + "step": 1033 + }, + { + "epoch": 2.888268156424581, + "grad_norm": 2.1260063648223877, + "learning_rate": 3.459217877094972e-06, + "loss": 0.3622, + "step": 1034 + }, + { + "epoch": 2.8910614525139664, + "grad_norm": 3.3801791667938232, + "learning_rate": 3.4625698324022347e-06, + "loss": 0.4046, + "step": 1035 + }, + { + "epoch": 2.893854748603352, + "grad_norm": 2.175487518310547, + "learning_rate": 3.4659217877094973e-06, + "loss": 0.3938, + "step": 1036 + }, + { + "epoch": 2.8966480446927374, + "grad_norm": 2.0540173053741455, + "learning_rate": 3.46927374301676e-06, + "loss": 0.465, + "step": 1037 + }, + { + "epoch": 2.899441340782123, + "grad_norm": 2.349916934967041, + "learning_rate": 3.4726256983240226e-06, + "loss": 0.4464, + "step": 1038 + }, + { + "epoch": 2.9022346368715084, + "grad_norm": 2.6910674571990967, + "learning_rate": 3.4759776536312847e-06, + "loss": 0.4994, + "step": 1039 + }, + { + "epoch": 2.905027932960894, + "grad_norm": 2.8544795513153076, + "learning_rate": 3.4793296089385474e-06, + "loss": 0.5191, + "step": 1040 + }, + { + "epoch": 2.9078212290502794, + "grad_norm": 8.905110359191895, + "learning_rate": 3.48268156424581e-06, + "loss": 0.6439, + "step": 1041 + }, + { + "epoch": 2.910614525139665, + "grad_norm": 2.171736001968384, + "learning_rate": 3.4860335195530726e-06, + "loss": 0.3589, + "step": 1042 + }, + { + "epoch": 2.9134078212290504, + "grad_norm": 1.3070627450942993, + "learning_rate": 3.4893854748603356e-06, + "loss": 0.3355, + "step": 1043 + }, + { + "epoch": 2.916201117318436, + "grad_norm": 1.1059659719467163, + "learning_rate": 3.492737430167598e-06, + "loss": 0.3399, + "step": 1044 + }, + { + "epoch": 2.9189944134078214, + "grad_norm": 1.034622311592102, + "learning_rate": 3.4960893854748604e-06, + "loss": 0.3678, + "step": 1045 + }, + { + "epoch": 2.921787709497207, + "grad_norm": 1.5194140672683716, + "learning_rate": 3.499441340782123e-06, + "loss": 0.3283, + "step": 1046 + }, + { + "epoch": 2.9245810055865924, + "grad_norm": 1.5572935342788696, + "learning_rate": 3.5027932960893857e-06, + "loss": 0.3847, + "step": 1047 + }, + { + "epoch": 2.927374301675978, + "grad_norm": 1.174438714981079, + "learning_rate": 3.5061452513966483e-06, + "loss": 0.3198, + "step": 1048 + }, + { + "epoch": 2.9301675977653634, + "grad_norm": 1.3590843677520752, + "learning_rate": 3.5094972067039105e-06, + "loss": 0.3291, + "step": 1049 + }, + { + "epoch": 2.9329608938547485, + "grad_norm": 1.0661094188690186, + "learning_rate": 3.512849162011173e-06, + "loss": 0.3338, + "step": 1050 + }, + { + "epoch": 2.935754189944134, + "grad_norm": 1.4307024478912354, + "learning_rate": 3.5162011173184357e-06, + "loss": 0.3522, + "step": 1051 + }, + { + "epoch": 2.9385474860335195, + "grad_norm": 1.8169820308685303, + "learning_rate": 3.5195530726256983e-06, + "loss": 0.3946, + "step": 1052 + }, + { + "epoch": 2.941340782122905, + "grad_norm": 3.09793758392334, + "learning_rate": 3.522905027932961e-06, + "loss": 0.3915, + "step": 1053 + }, + { + "epoch": 2.9441340782122905, + "grad_norm": 1.4726282358169556, + "learning_rate": 3.5262569832402236e-06, + "loss": 0.3604, + "step": 1054 + }, + { + "epoch": 2.946927374301676, + "grad_norm": 3.8493258953094482, + "learning_rate": 3.529608938547486e-06, + "loss": 0.377, + "step": 1055 + }, + { + "epoch": 2.9497206703910615, + "grad_norm": 1.5569509267807007, + "learning_rate": 3.532960893854749e-06, + "loss": 0.3694, + "step": 1056 + }, + { + "epoch": 2.952513966480447, + "grad_norm": 3.3749868869781494, + "learning_rate": 3.5363128491620114e-06, + "loss": 0.3916, + "step": 1057 + }, + { + "epoch": 2.9553072625698324, + "grad_norm": 1.5222928524017334, + "learning_rate": 3.539664804469274e-06, + "loss": 0.3849, + "step": 1058 + }, + { + "epoch": 2.958100558659218, + "grad_norm": 5.629571914672852, + "learning_rate": 3.5430167597765366e-06, + "loss": 0.3419, + "step": 1059 + }, + { + "epoch": 2.9608938547486034, + "grad_norm": 4.964062213897705, + "learning_rate": 3.546368715083799e-06, + "loss": 0.4152, + "step": 1060 + }, + { + "epoch": 2.963687150837989, + "grad_norm": 1.917149543762207, + "learning_rate": 3.5497206703910614e-06, + "loss": 0.4686, + "step": 1061 + }, + { + "epoch": 2.9664804469273744, + "grad_norm": 5.488239765167236, + "learning_rate": 3.553072625698324e-06, + "loss": 0.4223, + "step": 1062 + }, + { + "epoch": 2.9692737430167595, + "grad_norm": 1.889108419418335, + "learning_rate": 3.5564245810055867e-06, + "loss": 0.4472, + "step": 1063 + }, + { + "epoch": 2.972067039106145, + "grad_norm": 3.355389356613159, + "learning_rate": 3.5597765363128493e-06, + "loss": 0.4476, + "step": 1064 + }, + { + "epoch": 2.9748603351955305, + "grad_norm": 2.454160451889038, + "learning_rate": 3.5631284916201115e-06, + "loss": 0.5084, + "step": 1065 + }, + { + "epoch": 2.977653631284916, + "grad_norm": 2.8285460472106934, + "learning_rate": 3.566480446927374e-06, + "loss": 0.7239, + "step": 1066 + }, + { + "epoch": 2.9804469273743015, + "grad_norm": 1.3528121709823608, + "learning_rate": 3.569832402234637e-06, + "loss": 0.3691, + "step": 1067 + }, + { + "epoch": 2.983240223463687, + "grad_norm": 5.557601451873779, + "learning_rate": 3.5731843575418998e-06, + "loss": 0.358, + "step": 1068 + }, + { + "epoch": 2.9860335195530725, + "grad_norm": 2.047055959701538, + "learning_rate": 3.5765363128491624e-06, + "loss": 0.3204, + "step": 1069 + }, + { + "epoch": 2.988826815642458, + "grad_norm": 1.5284423828125, + "learning_rate": 3.5798882681564246e-06, + "loss": 0.3705, + "step": 1070 + }, + { + "epoch": 2.9916201117318435, + "grad_norm": 1.858002781867981, + "learning_rate": 3.583240223463687e-06, + "loss": 0.3311, + "step": 1071 + }, + { + "epoch": 2.994413407821229, + "grad_norm": 2.381492853164673, + "learning_rate": 3.58659217877095e-06, + "loss": 0.3669, + "step": 1072 + }, + { + "epoch": 2.9972067039106145, + "grad_norm": 1.9589487314224243, + "learning_rate": 3.5899441340782124e-06, + "loss": 0.4612, + "step": 1073 + }, + { + "epoch": 3.0, + "grad_norm": 3.3665831089019775, + "learning_rate": 3.593296089385475e-06, + "loss": 0.5913, + "step": 1074 + }, + { + "epoch": 3.0027932960893855, + "grad_norm": 1.4088499546051025, + "learning_rate": 3.5966480446927376e-06, + "loss": 0.3368, + "step": 1075 + }, + { + "epoch": 3.005586592178771, + "grad_norm": 1.1289281845092773, + "learning_rate": 3.6e-06, + "loss": 0.3441, + "step": 1076 + }, + { + "epoch": 3.0083798882681565, + "grad_norm": 1.1794946193695068, + "learning_rate": 3.6033519553072625e-06, + "loss": 0.3435, + "step": 1077 + }, + { + "epoch": 3.011173184357542, + "grad_norm": 1.3832049369812012, + "learning_rate": 3.606703910614525e-06, + "loss": 0.3204, + "step": 1078 + }, + { + "epoch": 3.0139664804469275, + "grad_norm": 1.7648067474365234, + "learning_rate": 3.610055865921788e-06, + "loss": 0.3686, + "step": 1079 + }, + { + "epoch": 3.016759776536313, + "grad_norm": 1.54685640335083, + "learning_rate": 3.6134078212290507e-06, + "loss": 0.3717, + "step": 1080 + }, + { + "epoch": 3.0195530726256985, + "grad_norm": 1.331088900566101, + "learning_rate": 3.616759776536313e-06, + "loss": 0.3589, + "step": 1081 + }, + { + "epoch": 3.022346368715084, + "grad_norm": 1.3855265378952026, + "learning_rate": 3.6201117318435755e-06, + "loss": 0.3606, + "step": 1082 + }, + { + "epoch": 3.0251396648044695, + "grad_norm": 1.206202745437622, + "learning_rate": 3.623463687150838e-06, + "loss": 0.3116, + "step": 1083 + }, + { + "epoch": 3.0279329608938546, + "grad_norm": 1.3088760375976562, + "learning_rate": 3.6268156424581008e-06, + "loss": 0.3459, + "step": 1084 + }, + { + "epoch": 3.03072625698324, + "grad_norm": 1.5826489925384521, + "learning_rate": 3.6301675977653634e-06, + "loss": 0.4077, + "step": 1085 + }, + { + "epoch": 3.0335195530726256, + "grad_norm": 2.547030210494995, + "learning_rate": 3.6335195530726256e-06, + "loss": 0.3763, + "step": 1086 + }, + { + "epoch": 3.036312849162011, + "grad_norm": 1.4516581296920776, + "learning_rate": 3.636871508379888e-06, + "loss": 0.3779, + "step": 1087 + }, + { + "epoch": 3.0391061452513966, + "grad_norm": 3.4305832386016846, + "learning_rate": 3.640223463687151e-06, + "loss": 0.4088, + "step": 1088 + }, + { + "epoch": 3.041899441340782, + "grad_norm": 1.9310340881347656, + "learning_rate": 3.6435754189944134e-06, + "loss": 0.3391, + "step": 1089 + }, + { + "epoch": 3.0446927374301676, + "grad_norm": 5.487971305847168, + "learning_rate": 3.646927374301676e-06, + "loss": 0.322, + "step": 1090 + }, + { + "epoch": 3.047486033519553, + "grad_norm": 5.023561477661133, + "learning_rate": 3.6502793296089387e-06, + "loss": 0.3217, + "step": 1091 + }, + { + "epoch": 3.0502793296089385, + "grad_norm": 3.0015006065368652, + "learning_rate": 3.6536312849162013e-06, + "loss": 0.4098, + "step": 1092 + }, + { + "epoch": 3.053072625698324, + "grad_norm": 2.449122428894043, + "learning_rate": 3.656983240223464e-06, + "loss": 0.3776, + "step": 1093 + }, + { + "epoch": 3.0558659217877095, + "grad_norm": 4.40966796875, + "learning_rate": 3.6603351955307265e-06, + "loss": 0.452, + "step": 1094 + }, + { + "epoch": 3.058659217877095, + "grad_norm": 2.3427793979644775, + "learning_rate": 3.663687150837989e-06, + "loss": 0.4344, + "step": 1095 + }, + { + "epoch": 3.0614525139664805, + "grad_norm": 2.9425125122070312, + "learning_rate": 3.6670391061452513e-06, + "loss": 0.3964, + "step": 1096 + }, + { + "epoch": 3.064245810055866, + "grad_norm": 2.8161933422088623, + "learning_rate": 3.670391061452514e-06, + "loss": 0.4214, + "step": 1097 + }, + { + "epoch": 3.0670391061452515, + "grad_norm": 7.407336235046387, + "learning_rate": 3.6737430167597765e-06, + "loss": 0.5607, + "step": 1098 + }, + { + "epoch": 3.069832402234637, + "grad_norm": 3.5610036849975586, + "learning_rate": 3.677094972067039e-06, + "loss": 0.717, + "step": 1099 + }, + { + "epoch": 3.0726256983240225, + "grad_norm": 1.5323798656463623, + "learning_rate": 3.6804469273743018e-06, + "loss": 0.35, + "step": 1100 + }, + { + "epoch": 3.0754189944134076, + "grad_norm": 1.0983535051345825, + "learning_rate": 3.6837988826815644e-06, + "loss": 0.3436, + "step": 1101 + }, + { + "epoch": 3.078212290502793, + "grad_norm": 0.9438749551773071, + "learning_rate": 3.6871508379888266e-06, + "loss": 0.3106, + "step": 1102 + }, + { + "epoch": 3.0810055865921786, + "grad_norm": 1.8678596019744873, + "learning_rate": 3.6905027932960896e-06, + "loss": 0.354, + "step": 1103 + }, + { + "epoch": 3.083798882681564, + "grad_norm": 4.479010105133057, + "learning_rate": 3.6938547486033522e-06, + "loss": 0.3403, + "step": 1104 + }, + { + "epoch": 3.0865921787709496, + "grad_norm": 1.6001774072647095, + "learning_rate": 3.697206703910615e-06, + "loss": 0.4414, + "step": 1105 + }, + { + "epoch": 3.089385474860335, + "grad_norm": 1.868389368057251, + "learning_rate": 3.7005586592178775e-06, + "loss": 0.3645, + "step": 1106 + }, + { + "epoch": 3.0921787709497206, + "grad_norm": 2.8334648609161377, + "learning_rate": 3.7039106145251397e-06, + "loss": 0.3446, + "step": 1107 + }, + { + "epoch": 3.094972067039106, + "grad_norm": 1.948793649673462, + "learning_rate": 3.7072625698324023e-06, + "loss": 0.3608, + "step": 1108 + }, + { + "epoch": 3.0977653631284916, + "grad_norm": 1.7837179899215698, + "learning_rate": 3.710614525139665e-06, + "loss": 0.353, + "step": 1109 + }, + { + "epoch": 3.100558659217877, + "grad_norm": 1.262573480606079, + "learning_rate": 3.7139664804469275e-06, + "loss": 0.3794, + "step": 1110 + }, + { + "epoch": 3.1033519553072626, + "grad_norm": 1.3717130422592163, + "learning_rate": 3.71731843575419e-06, + "loss": 0.3525, + "step": 1111 + }, + { + "epoch": 3.106145251396648, + "grad_norm": 1.570822834968567, + "learning_rate": 3.7206703910614523e-06, + "loss": 0.3583, + "step": 1112 + }, + { + "epoch": 3.1089385474860336, + "grad_norm": 1.551196575164795, + "learning_rate": 3.724022346368715e-06, + "loss": 0.4333, + "step": 1113 + }, + { + "epoch": 3.111731843575419, + "grad_norm": 2.600823163986206, + "learning_rate": 3.7273743016759775e-06, + "loss": 0.3633, + "step": 1114 + }, + { + "epoch": 3.1145251396648046, + "grad_norm": 1.8274368047714233, + "learning_rate": 3.7307262569832406e-06, + "loss": 0.4138, + "step": 1115 + }, + { + "epoch": 3.11731843575419, + "grad_norm": 1.6556206941604614, + "learning_rate": 3.734078212290503e-06, + "loss": 0.4025, + "step": 1116 + }, + { + "epoch": 3.1201117318435756, + "grad_norm": 2.323518753051758, + "learning_rate": 3.7374301675977654e-06, + "loss": 0.426, + "step": 1117 + }, + { + "epoch": 3.122905027932961, + "grad_norm": 1.6889681816101074, + "learning_rate": 3.740782122905028e-06, + "loss": 0.4116, + "step": 1118 + }, + { + "epoch": 3.1256983240223466, + "grad_norm": 2.101938009262085, + "learning_rate": 3.7441340782122906e-06, + "loss": 0.4144, + "step": 1119 + }, + { + "epoch": 3.1284916201117317, + "grad_norm": 1.969752550125122, + "learning_rate": 3.7474860335195532e-06, + "loss": 0.4204, + "step": 1120 + }, + { + "epoch": 3.131284916201117, + "grad_norm": 2.230369806289673, + "learning_rate": 3.7508379888268154e-06, + "loss": 0.3999, + "step": 1121 + }, + { + "epoch": 3.1340782122905027, + "grad_norm": 2.7185142040252686, + "learning_rate": 3.7541899441340785e-06, + "loss": 0.4288, + "step": 1122 + }, + { + "epoch": 3.136871508379888, + "grad_norm": 9.903284072875977, + "learning_rate": 3.7575418994413407e-06, + "loss": 0.4987, + "step": 1123 + }, + { + "epoch": 3.1396648044692737, + "grad_norm": 3.898953437805176, + "learning_rate": 3.7608938547486037e-06, + "loss": 0.7116, + "step": 1124 + }, + { + "epoch": 3.142458100558659, + "grad_norm": 1.2296949625015259, + "learning_rate": 3.764245810055866e-06, + "loss": 0.3265, + "step": 1125 + }, + { + "epoch": 3.1452513966480447, + "grad_norm": 0.8335631489753723, + "learning_rate": 3.767597765363128e-06, + "loss": 0.3129, + "step": 1126 + }, + { + "epoch": 3.14804469273743, + "grad_norm": 1.0560312271118164, + "learning_rate": 3.7709497206703915e-06, + "loss": 0.3302, + "step": 1127 + }, + { + "epoch": 3.1508379888268156, + "grad_norm": 1.4834799766540527, + "learning_rate": 3.7743016759776537e-06, + "loss": 0.3055, + "step": 1128 + }, + { + "epoch": 3.153631284916201, + "grad_norm": 1.5117125511169434, + "learning_rate": 3.7776536312849168e-06, + "loss": 0.3548, + "step": 1129 + }, + { + "epoch": 3.1564245810055866, + "grad_norm": 1.2366617918014526, + "learning_rate": 3.781005586592179e-06, + "loss": 0.3241, + "step": 1130 + }, + { + "epoch": 3.159217877094972, + "grad_norm": 2.311779737472534, + "learning_rate": 3.784357541899441e-06, + "loss": 0.3426, + "step": 1131 + }, + { + "epoch": 3.1620111731843576, + "grad_norm": 1.3708757162094116, + "learning_rate": 3.787709497206704e-06, + "loss": 0.3231, + "step": 1132 + }, + { + "epoch": 3.164804469273743, + "grad_norm": 1.3573986291885376, + "learning_rate": 3.7910614525139664e-06, + "loss": 0.3175, + "step": 1133 + }, + { + "epoch": 3.1675977653631286, + "grad_norm": 1.7755740880966187, + "learning_rate": 3.7944134078212294e-06, + "loss": 0.3902, + "step": 1134 + }, + { + "epoch": 3.170391061452514, + "grad_norm": 2.745845317840576, + "learning_rate": 3.7977653631284916e-06, + "loss": 0.3345, + "step": 1135 + }, + { + "epoch": 3.1731843575418996, + "grad_norm": 2.376347541809082, + "learning_rate": 3.801117318435754e-06, + "loss": 0.3269, + "step": 1136 + }, + { + "epoch": 3.1759776536312847, + "grad_norm": 2.0311834812164307, + "learning_rate": 3.804469273743017e-06, + "loss": 0.3749, + "step": 1137 + }, + { + "epoch": 3.17877094972067, + "grad_norm": 1.365647792816162, + "learning_rate": 3.807821229050279e-06, + "loss": 0.3659, + "step": 1138 + }, + { + "epoch": 3.1815642458100557, + "grad_norm": 1.4635369777679443, + "learning_rate": 3.8111731843575425e-06, + "loss": 0.3436, + "step": 1139 + }, + { + "epoch": 3.184357541899441, + "grad_norm": 3.2827043533325195, + "learning_rate": 3.8145251396648047e-06, + "loss": 0.3464, + "step": 1140 + }, + { + "epoch": 3.1871508379888267, + "grad_norm": 2.3158135414123535, + "learning_rate": 3.8178770949720665e-06, + "loss": 0.3644, + "step": 1141 + }, + { + "epoch": 3.189944134078212, + "grad_norm": 1.9942140579223633, + "learning_rate": 3.8212290502793295e-06, + "loss": 0.3572, + "step": 1142 + }, + { + "epoch": 3.1927374301675977, + "grad_norm": 2.0316874980926514, + "learning_rate": 3.824581005586592e-06, + "loss": 0.3873, + "step": 1143 + }, + { + "epoch": 3.195530726256983, + "grad_norm": 4.500475883483887, + "learning_rate": 3.827932960893855e-06, + "loss": 0.4036, + "step": 1144 + }, + { + "epoch": 3.1983240223463687, + "grad_norm": 9.19706916809082, + "learning_rate": 3.831284916201118e-06, + "loss": 0.4073, + "step": 1145 + }, + { + "epoch": 3.201117318435754, + "grad_norm": 2.7144596576690674, + "learning_rate": 3.83463687150838e-06, + "loss": 0.3936, + "step": 1146 + }, + { + "epoch": 3.2039106145251397, + "grad_norm": 2.6496899127960205, + "learning_rate": 3.837988826815643e-06, + "loss": 0.4623, + "step": 1147 + }, + { + "epoch": 3.206703910614525, + "grad_norm": 5.261826515197754, + "learning_rate": 3.841340782122905e-06, + "loss": 0.4878, + "step": 1148 + }, + { + "epoch": 3.2094972067039107, + "grad_norm": 16.117843627929688, + "learning_rate": 3.844692737430168e-06, + "loss": 0.6016, + "step": 1149 + }, + { + "epoch": 3.212290502793296, + "grad_norm": 1.2679048776626587, + "learning_rate": 3.8480446927374304e-06, + "loss": 0.3271, + "step": 1150 + }, + { + "epoch": 3.2150837988826817, + "grad_norm": 0.8199213743209839, + "learning_rate": 3.851396648044693e-06, + "loss": 0.3003, + "step": 1151 + }, + { + "epoch": 3.217877094972067, + "grad_norm": 0.8534806370735168, + "learning_rate": 3.854748603351956e-06, + "loss": 0.3368, + "step": 1152 + }, + { + "epoch": 3.2206703910614527, + "grad_norm": 1.4127224683761597, + "learning_rate": 3.858100558659218e-06, + "loss": 0.3438, + "step": 1153 + }, + { + "epoch": 3.223463687150838, + "grad_norm": 1.0002996921539307, + "learning_rate": 3.861452513966481e-06, + "loss": 0.311, + "step": 1154 + }, + { + "epoch": 3.2262569832402237, + "grad_norm": 1.2609870433807373, + "learning_rate": 3.864804469273743e-06, + "loss": 0.3301, + "step": 1155 + }, + { + "epoch": 3.2290502793296088, + "grad_norm": 1.8136091232299805, + "learning_rate": 3.868156424581005e-06, + "loss": 0.3625, + "step": 1156 + }, + { + "epoch": 3.2318435754189943, + "grad_norm": 1.4170751571655273, + "learning_rate": 3.871508379888268e-06, + "loss": 0.3431, + "step": 1157 + }, + { + "epoch": 3.2346368715083798, + "grad_norm": 1.3396705389022827, + "learning_rate": 3.8748603351955305e-06, + "loss": 0.3113, + "step": 1158 + }, + { + "epoch": 3.2374301675977653, + "grad_norm": 1.278328537940979, + "learning_rate": 3.8782122905027936e-06, + "loss": 0.3655, + "step": 1159 + }, + { + "epoch": 3.2402234636871508, + "grad_norm": 1.652985692024231, + "learning_rate": 3.881564245810056e-06, + "loss": 0.315, + "step": 1160 + }, + { + "epoch": 3.2430167597765363, + "grad_norm": 1.591238260269165, + "learning_rate": 3.884916201117319e-06, + "loss": 0.3546, + "step": 1161 + }, + { + "epoch": 3.2458100558659218, + "grad_norm": 1.30949866771698, + "learning_rate": 3.888268156424581e-06, + "loss": 0.3604, + "step": 1162 + }, + { + "epoch": 3.2486033519553073, + "grad_norm": 1.5504283905029297, + "learning_rate": 3.891620111731843e-06, + "loss": 0.3424, + "step": 1163 + }, + { + "epoch": 3.2513966480446927, + "grad_norm": 1.6199322938919067, + "learning_rate": 3.894972067039106e-06, + "loss": 0.3506, + "step": 1164 + }, + { + "epoch": 3.2541899441340782, + "grad_norm": 1.8205665349960327, + "learning_rate": 3.898324022346368e-06, + "loss": 0.3633, + "step": 1165 + }, + { + "epoch": 3.2569832402234637, + "grad_norm": 3.1932666301727295, + "learning_rate": 3.9016759776536314e-06, + "loss": 0.3802, + "step": 1166 + }, + { + "epoch": 3.2597765363128492, + "grad_norm": 1.7619726657867432, + "learning_rate": 3.905027932960894e-06, + "loss": 0.3305, + "step": 1167 + }, + { + "epoch": 3.2625698324022347, + "grad_norm": 2.681483507156372, + "learning_rate": 3.908379888268156e-06, + "loss": 0.3865, + "step": 1168 + }, + { + "epoch": 3.2653631284916202, + "grad_norm": 1.676092267036438, + "learning_rate": 3.91173184357542e-06, + "loss": 0.4004, + "step": 1169 + }, + { + "epoch": 3.2681564245810057, + "grad_norm": 2.2918808460235596, + "learning_rate": 3.915083798882682e-06, + "loss": 0.4123, + "step": 1170 + }, + { + "epoch": 3.2709497206703912, + "grad_norm": 2.2460927963256836, + "learning_rate": 3.918435754189945e-06, + "loss": 0.4298, + "step": 1171 + }, + { + "epoch": 3.2737430167597763, + "grad_norm": 3.2791333198547363, + "learning_rate": 3.921787709497207e-06, + "loss": 0.4445, + "step": 1172 + }, + { + "epoch": 3.276536312849162, + "grad_norm": 3.20371675491333, + "learning_rate": 3.925139664804469e-06, + "loss": 0.5013, + "step": 1173 + }, + { + "epoch": 3.2793296089385473, + "grad_norm": 3.692323923110962, + "learning_rate": 3.928491620111732e-06, + "loss": 0.5411, + "step": 1174 + }, + { + "epoch": 3.282122905027933, + "grad_norm": 2.9045214653015137, + "learning_rate": 3.9318435754189946e-06, + "loss": 0.3856, + "step": 1175 + }, + { + "epoch": 3.2849162011173183, + "grad_norm": 1.0197888612747192, + "learning_rate": 3.935195530726258e-06, + "loss": 0.3172, + "step": 1176 + }, + { + "epoch": 3.287709497206704, + "grad_norm": 1.2208348512649536, + "learning_rate": 3.93854748603352e-06, + "loss": 0.324, + "step": 1177 + }, + { + "epoch": 3.2905027932960893, + "grad_norm": 1.6910191774368286, + "learning_rate": 3.941899441340782e-06, + "loss": 0.3185, + "step": 1178 + }, + { + "epoch": 3.293296089385475, + "grad_norm": 1.2190288305282593, + "learning_rate": 3.945251396648045e-06, + "loss": 0.3285, + "step": 1179 + }, + { + "epoch": 3.2960893854748603, + "grad_norm": 1.2894783020019531, + "learning_rate": 3.948603351955307e-06, + "loss": 0.3124, + "step": 1180 + }, + { + "epoch": 3.298882681564246, + "grad_norm": 1.3836427927017212, + "learning_rate": 3.95195530726257e-06, + "loss": 0.3444, + "step": 1181 + }, + { + "epoch": 3.3016759776536313, + "grad_norm": 1.3779702186584473, + "learning_rate": 3.9553072625698325e-06, + "loss": 0.3228, + "step": 1182 + }, + { + "epoch": 3.304469273743017, + "grad_norm": 1.1909891366958618, + "learning_rate": 3.958659217877095e-06, + "loss": 0.3173, + "step": 1183 + }, + { + "epoch": 3.3072625698324023, + "grad_norm": 2.021533250808716, + "learning_rate": 3.962011173184358e-06, + "loss": 0.3221, + "step": 1184 + }, + { + "epoch": 3.310055865921788, + "grad_norm": 1.494223713874817, + "learning_rate": 3.96536312849162e-06, + "loss": 0.4184, + "step": 1185 + }, + { + "epoch": 3.3128491620111733, + "grad_norm": 6.690242767333984, + "learning_rate": 3.968715083798883e-06, + "loss": 0.3304, + "step": 1186 + }, + { + "epoch": 3.315642458100559, + "grad_norm": 1.384992241859436, + "learning_rate": 3.972067039106145e-06, + "loss": 0.3641, + "step": 1187 + }, + { + "epoch": 3.3184357541899443, + "grad_norm": 1.506320595741272, + "learning_rate": 3.975418994413407e-06, + "loss": 0.3621, + "step": 1188 + }, + { + "epoch": 3.32122905027933, + "grad_norm": 1.4999268054962158, + "learning_rate": 3.97877094972067e-06, + "loss": 0.3228, + "step": 1189 + }, + { + "epoch": 3.3240223463687153, + "grad_norm": 1.496852993965149, + "learning_rate": 3.9821229050279325e-06, + "loss": 0.362, + "step": 1190 + }, + { + "epoch": 3.326815642458101, + "grad_norm": 1.4983439445495605, + "learning_rate": 3.9854748603351956e-06, + "loss": 0.3703, + "step": 1191 + }, + { + "epoch": 3.329608938547486, + "grad_norm": 1.9659347534179688, + "learning_rate": 3.988826815642458e-06, + "loss": 0.3991, + "step": 1192 + }, + { + "epoch": 3.3324022346368714, + "grad_norm": 1.6705873012542725, + "learning_rate": 3.992178770949721e-06, + "loss": 0.3789, + "step": 1193 + }, + { + "epoch": 3.335195530726257, + "grad_norm": 19.2963924407959, + "learning_rate": 3.995530726256984e-06, + "loss": 0.4024, + "step": 1194 + }, + { + "epoch": 3.3379888268156424, + "grad_norm": 2.930727243423462, + "learning_rate": 3.998882681564246e-06, + "loss": 0.4608, + "step": 1195 + }, + { + "epoch": 3.340782122905028, + "grad_norm": 7.114203453063965, + "learning_rate": 4.002234636871509e-06, + "loss": 0.4177, + "step": 1196 + }, + { + "epoch": 3.3435754189944134, + "grad_norm": 2.4212806224823, + "learning_rate": 4.005586592178771e-06, + "loss": 0.4511, + "step": 1197 + }, + { + "epoch": 3.346368715083799, + "grad_norm": 5.972248077392578, + "learning_rate": 4.0089385474860335e-06, + "loss": 0.4565, + "step": 1198 + }, + { + "epoch": 3.3491620111731844, + "grad_norm": 4.7472243309021, + "learning_rate": 4.0122905027932965e-06, + "loss": 0.6003, + "step": 1199 + }, + { + "epoch": 3.35195530726257, + "grad_norm": 1.5404484272003174, + "learning_rate": 4.015642458100559e-06, + "loss": 0.3856, + "step": 1200 + }, + { + "epoch": 3.3547486033519553, + "grad_norm": 1.4620275497436523, + "learning_rate": 4.018994413407822e-06, + "loss": 0.3458, + "step": 1201 + }, + { + "epoch": 3.357541899441341, + "grad_norm": 1.4137049913406372, + "learning_rate": 4.022346368715084e-06, + "loss": 0.3291, + "step": 1202 + }, + { + "epoch": 3.3603351955307263, + "grad_norm": 1.1721950769424438, + "learning_rate": 4.025698324022346e-06, + "loss": 0.3109, + "step": 1203 + }, + { + "epoch": 3.363128491620112, + "grad_norm": 1.1389591693878174, + "learning_rate": 4.029050279329609e-06, + "loss": 0.3006, + "step": 1204 + }, + { + "epoch": 3.3659217877094973, + "grad_norm": 1.271691918373108, + "learning_rate": 4.032402234636871e-06, + "loss": 0.2988, + "step": 1205 + }, + { + "epoch": 3.368715083798883, + "grad_norm": 1.1986027956008911, + "learning_rate": 4.035754189944134e-06, + "loss": 0.3197, + "step": 1206 + }, + { + "epoch": 3.3715083798882683, + "grad_norm": 1.6436811685562134, + "learning_rate": 4.039106145251397e-06, + "loss": 0.3018, + "step": 1207 + }, + { + "epoch": 3.3743016759776534, + "grad_norm": 1.1320551633834839, + "learning_rate": 4.04245810055866e-06, + "loss": 0.2937, + "step": 1208 + }, + { + "epoch": 3.377094972067039, + "grad_norm": 2.2733938694000244, + "learning_rate": 4.045810055865922e-06, + "loss": 0.336, + "step": 1209 + }, + { + "epoch": 3.3798882681564244, + "grad_norm": 1.1354655027389526, + "learning_rate": 4.049162011173184e-06, + "loss": 0.3426, + "step": 1210 + }, + { + "epoch": 3.38268156424581, + "grad_norm": 1.4168121814727783, + "learning_rate": 4.052513966480447e-06, + "loss": 0.3171, + "step": 1211 + }, + { + "epoch": 3.3854748603351954, + "grad_norm": 5.084190368652344, + "learning_rate": 4.055865921787709e-06, + "loss": 0.3168, + "step": 1212 + }, + { + "epoch": 3.388268156424581, + "grad_norm": 2.158146619796753, + "learning_rate": 4.059217877094972e-06, + "loss": 0.3603, + "step": 1213 + }, + { + "epoch": 3.3910614525139664, + "grad_norm": 1.4581831693649292, + "learning_rate": 4.0625698324022345e-06, + "loss": 0.3307, + "step": 1214 + }, + { + "epoch": 3.393854748603352, + "grad_norm": 3.828110694885254, + "learning_rate": 4.065921787709497e-06, + "loss": 0.3386, + "step": 1215 + }, + { + "epoch": 3.3966480446927374, + "grad_norm": 2.5315611362457275, + "learning_rate": 4.06927374301676e-06, + "loss": 0.3515, + "step": 1216 + }, + { + "epoch": 3.399441340782123, + "grad_norm": 2.607835054397583, + "learning_rate": 4.072625698324023e-06, + "loss": 0.3798, + "step": 1217 + }, + { + "epoch": 3.4022346368715084, + "grad_norm": 5.791173934936523, + "learning_rate": 4.075977653631286e-06, + "loss": 0.3843, + "step": 1218 + }, + { + "epoch": 3.405027932960894, + "grad_norm": 2.78159499168396, + "learning_rate": 4.079329608938548e-06, + "loss": 0.4342, + "step": 1219 + }, + { + "epoch": 3.4078212290502794, + "grad_norm": 3.2733118534088135, + "learning_rate": 4.08268156424581e-06, + "loss": 0.4747, + "step": 1220 + }, + { + "epoch": 3.410614525139665, + "grad_norm": 2.1844592094421387, + "learning_rate": 4.086033519553073e-06, + "loss": 0.3955, + "step": 1221 + }, + { + "epoch": 3.4134078212290504, + "grad_norm": 3.520319938659668, + "learning_rate": 4.089385474860335e-06, + "loss": 0.5267, + "step": 1222 + }, + { + "epoch": 3.416201117318436, + "grad_norm": 3.160015344619751, + "learning_rate": 4.0927374301675984e-06, + "loss": 0.4214, + "step": 1223 + }, + { + "epoch": 3.4189944134078214, + "grad_norm": Infinity, + "learning_rate": 4.0927374301675984e-06, + "loss": 0.6036, + "step": 1224 + }, + { + "epoch": 3.421787709497207, + "grad_norm": 1.0105278491973877, + "learning_rate": 4.096089385474861e-06, + "loss": 0.3341, + "step": 1225 + }, + { + "epoch": 3.4245810055865924, + "grad_norm": 1.0004974603652954, + "learning_rate": 4.099441340782123e-06, + "loss": 0.3322, + "step": 1226 + }, + { + "epoch": 3.427374301675978, + "grad_norm": 0.8637301921844482, + "learning_rate": 4.102793296089386e-06, + "loss": 0.2901, + "step": 1227 + }, + { + "epoch": 3.430167597765363, + "grad_norm": 0.9320588707923889, + "learning_rate": 4.106145251396648e-06, + "loss": 0.3139, + "step": 1228 + }, + { + "epoch": 3.4329608938547485, + "grad_norm": 1.3901019096374512, + "learning_rate": 4.109497206703911e-06, + "loss": 0.3093, + "step": 1229 + }, + { + "epoch": 3.435754189944134, + "grad_norm": 1.0396521091461182, + "learning_rate": 4.112849162011173e-06, + "loss": 0.3252, + "step": 1230 + }, + { + "epoch": 3.4385474860335195, + "grad_norm": 1.0676325559616089, + "learning_rate": 4.1162011173184355e-06, + "loss": 0.3212, + "step": 1231 + }, + { + "epoch": 3.441340782122905, + "grad_norm": 2.5555248260498047, + "learning_rate": 4.1195530726256985e-06, + "loss": 0.3423, + "step": 1232 + }, + { + "epoch": 3.4441340782122905, + "grad_norm": 1.2287794351577759, + "learning_rate": 4.122905027932961e-06, + "loss": 0.2822, + "step": 1233 + }, + { + "epoch": 3.446927374301676, + "grad_norm": 1.551735520362854, + "learning_rate": 4.126256983240224e-06, + "loss": 0.3271, + "step": 1234 + }, + { + "epoch": 3.4497206703910615, + "grad_norm": 1.3163206577301025, + "learning_rate": 4.129608938547486e-06, + "loss": 0.2979, + "step": 1235 + }, + { + "epoch": 3.452513966480447, + "grad_norm": 1.4434260129928589, + "learning_rate": 4.132960893854748e-06, + "loss": 0.3284, + "step": 1236 + }, + { + "epoch": 3.4553072625698324, + "grad_norm": 1.5730010271072388, + "learning_rate": 4.136312849162011e-06, + "loss": 0.3203, + "step": 1237 + }, + { + "epoch": 3.458100558659218, + "grad_norm": 2.531080722808838, + "learning_rate": 4.139664804469273e-06, + "loss": 0.3913, + "step": 1238 + }, + { + "epoch": 3.4608938547486034, + "grad_norm": 1.2969385385513306, + "learning_rate": 4.143016759776536e-06, + "loss": 0.3601, + "step": 1239 + }, + { + "epoch": 3.463687150837989, + "grad_norm": 1.6375553607940674, + "learning_rate": 4.146368715083799e-06, + "loss": 0.3538, + "step": 1240 + }, + { + "epoch": 3.4664804469273744, + "grad_norm": 2.251586675643921, + "learning_rate": 4.149720670391061e-06, + "loss": 0.3787, + "step": 1241 + }, + { + "epoch": 3.46927374301676, + "grad_norm": 2.0671019554138184, + "learning_rate": 4.153072625698324e-06, + "loss": 0.3631, + "step": 1242 + }, + { + "epoch": 3.472067039106145, + "grad_norm": 2.127230405807495, + "learning_rate": 4.156424581005587e-06, + "loss": 0.362, + "step": 1243 + }, + { + "epoch": 3.4748603351955305, + "grad_norm": 2.292429208755493, + "learning_rate": 4.15977653631285e-06, + "loss": 0.3751, + "step": 1244 + }, + { + "epoch": 3.477653631284916, + "grad_norm": 2.55267333984375, + "learning_rate": 4.163128491620112e-06, + "loss": 0.3652, + "step": 1245 + }, + { + "epoch": 3.4804469273743015, + "grad_norm": 4.776055335998535, + "learning_rate": 4.166480446927374e-06, + "loss": 0.4016, + "step": 1246 + }, + { + "epoch": 3.483240223463687, + "grad_norm": 3.2314183712005615, + "learning_rate": 4.169832402234637e-06, + "loss": 0.4349, + "step": 1247 + }, + { + "epoch": 3.4860335195530725, + "grad_norm": 5.541086196899414, + "learning_rate": 4.1731843575418995e-06, + "loss": 0.495, + "step": 1248 + }, + { + "epoch": 3.488826815642458, + "grad_norm": 3.618764638900757, + "learning_rate": 4.1765363128491626e-06, + "loss": 0.6371, + "step": 1249 + }, + { + "epoch": 3.4916201117318435, + "grad_norm": 0.9980988502502441, + "learning_rate": 4.179888268156425e-06, + "loss": 0.3137, + "step": 1250 + }, + { + "epoch": 3.494413407821229, + "grad_norm": 0.8711819648742676, + "learning_rate": 4.183240223463687e-06, + "loss": 0.2963, + "step": 1251 + }, + { + "epoch": 3.4972067039106145, + "grad_norm": 1.4326032400131226, + "learning_rate": 4.18659217877095e-06, + "loss": 0.2871, + "step": 1252 + }, + { + "epoch": 3.5, + "grad_norm": 1.9187122583389282, + "learning_rate": 4.189944134078212e-06, + "loss": 0.3973, + "step": 1253 + }, + { + "epoch": 3.5027932960893855, + "grad_norm": 1.0688508749008179, + "learning_rate": 4.193296089385475e-06, + "loss": 0.3211, + "step": 1254 + }, + { + "epoch": 3.505586592178771, + "grad_norm": 1.3896992206573486, + "learning_rate": 4.196648044692737e-06, + "loss": 0.3122, + "step": 1255 + }, + { + "epoch": 3.5083798882681565, + "grad_norm": 1.759398102760315, + "learning_rate": 4.2000000000000004e-06, + "loss": 0.3102, + "step": 1256 + }, + { + "epoch": 3.511173184357542, + "grad_norm": 1.4857362508773804, + "learning_rate": 4.203351955307263e-06, + "loss": 0.3387, + "step": 1257 + }, + { + "epoch": 3.5139664804469275, + "grad_norm": 1.5543674230575562, + "learning_rate": 4.206703910614525e-06, + "loss": 0.3318, + "step": 1258 + }, + { + "epoch": 3.516759776536313, + "grad_norm": 1.4485273361206055, + "learning_rate": 4.210055865921788e-06, + "loss": 0.3597, + "step": 1259 + }, + { + "epoch": 3.5195530726256985, + "grad_norm": 1.8954466581344604, + "learning_rate": 4.21340782122905e-06, + "loss": 0.3232, + "step": 1260 + }, + { + "epoch": 3.522346368715084, + "grad_norm": 1.5319074392318726, + "learning_rate": 4.216759776536313e-06, + "loss": 0.3554, + "step": 1261 + }, + { + "epoch": 3.5251396648044695, + "grad_norm": 1.4043940305709839, + "learning_rate": 4.220111731843575e-06, + "loss": 0.3295, + "step": 1262 + }, + { + "epoch": 3.527932960893855, + "grad_norm": 1.88729989528656, + "learning_rate": 4.2234636871508375e-06, + "loss": 0.3569, + "step": 1263 + }, + { + "epoch": 3.5307262569832405, + "grad_norm": 1.31914222240448, + "learning_rate": 4.2268156424581005e-06, + "loss": 0.342, + "step": 1264 + }, + { + "epoch": 3.5335195530726256, + "grad_norm": 1.4942641258239746, + "learning_rate": 4.230167597765363e-06, + "loss": 0.3465, + "step": 1265 + }, + { + "epoch": 3.536312849162011, + "grad_norm": 1.6203111410140991, + "learning_rate": 4.233519553072626e-06, + "loss": 0.3687, + "step": 1266 + }, + { + "epoch": 3.5391061452513966, + "grad_norm": 1.4986796379089355, + "learning_rate": 4.236871508379889e-06, + "loss": 0.3604, + "step": 1267 + }, + { + "epoch": 3.541899441340782, + "grad_norm": 2.2067713737487793, + "learning_rate": 4.240223463687151e-06, + "loss": 0.3559, + "step": 1268 + }, + { + "epoch": 3.5446927374301676, + "grad_norm": 1.3800028562545776, + "learning_rate": 4.243575418994414e-06, + "loss": 0.3592, + "step": 1269 + }, + { + "epoch": 3.547486033519553, + "grad_norm": 1.8495538234710693, + "learning_rate": 4.246927374301676e-06, + "loss": 0.4062, + "step": 1270 + }, + { + "epoch": 3.5502793296089385, + "grad_norm": 3.7579400539398193, + "learning_rate": 4.250279329608939e-06, + "loss": 0.4399, + "step": 1271 + }, + { + "epoch": 3.553072625698324, + "grad_norm": 2.481085777282715, + "learning_rate": 4.2536312849162015e-06, + "loss": 0.378, + "step": 1272 + }, + { + "epoch": 3.5558659217877095, + "grad_norm": 2.969104766845703, + "learning_rate": 4.256983240223464e-06, + "loss": 0.4488, + "step": 1273 + }, + { + "epoch": 3.558659217877095, + "grad_norm": 4.857105255126953, + "learning_rate": 4.260335195530727e-06, + "loss": 0.6223, + "step": 1274 + }, + { + "epoch": 3.5614525139664805, + "grad_norm": 0.742992639541626, + "learning_rate": 4.263687150837989e-06, + "loss": 0.295, + "step": 1275 + }, + { + "epoch": 3.564245810055866, + "grad_norm": 1.4554110765457153, + "learning_rate": 4.267039106145252e-06, + "loss": 0.2772, + "step": 1276 + }, + { + "epoch": 3.5670391061452515, + "grad_norm": 0.8429902791976929, + "learning_rate": 4.270391061452514e-06, + "loss": 0.3075, + "step": 1277 + }, + { + "epoch": 3.5698324022346366, + "grad_norm": 0.8672071695327759, + "learning_rate": 4.273743016759776e-06, + "loss": 0.2908, + "step": 1278 + }, + { + "epoch": 3.572625698324022, + "grad_norm": 1.5792009830474854, + "learning_rate": 4.277094972067039e-06, + "loss": 0.2915, + "step": 1279 + }, + { + "epoch": 3.5754189944134076, + "grad_norm": 1.3984836339950562, + "learning_rate": 4.2804469273743015e-06, + "loss": 0.327, + "step": 1280 + }, + { + "epoch": 3.578212290502793, + "grad_norm": 1.220801591873169, + "learning_rate": 4.2837988826815646e-06, + "loss": 0.3007, + "step": 1281 + }, + { + "epoch": 3.5810055865921786, + "grad_norm": 1.139129877090454, + "learning_rate": 4.287150837988827e-06, + "loss": 0.3479, + "step": 1282 + }, + { + "epoch": 3.583798882681564, + "grad_norm": 1.615267276763916, + "learning_rate": 4.290502793296089e-06, + "loss": 0.3037, + "step": 1283 + }, + { + "epoch": 3.5865921787709496, + "grad_norm": 2.7950356006622314, + "learning_rate": 4.293854748603352e-06, + "loss": 0.3026, + "step": 1284 + }, + { + "epoch": 3.589385474860335, + "grad_norm": 1.9023959636688232, + "learning_rate": 4.297206703910614e-06, + "loss": 0.3474, + "step": 1285 + }, + { + "epoch": 3.5921787709497206, + "grad_norm": 1.4593322277069092, + "learning_rate": 4.300558659217877e-06, + "loss": 0.3245, + "step": 1286 + }, + { + "epoch": 3.594972067039106, + "grad_norm": 1.489857792854309, + "learning_rate": 4.303910614525139e-06, + "loss": 0.3394, + "step": 1287 + }, + { + "epoch": 3.5977653631284916, + "grad_norm": 2.714019536972046, + "learning_rate": 4.307262569832402e-06, + "loss": 0.345, + "step": 1288 + }, + { + "epoch": 3.600558659217877, + "grad_norm": 1.4554060697555542, + "learning_rate": 4.310614525139665e-06, + "loss": 0.3332, + "step": 1289 + }, + { + "epoch": 3.6033519553072626, + "grad_norm": 1.4200193881988525, + "learning_rate": 4.313966480446927e-06, + "loss": 0.3079, + "step": 1290 + }, + { + "epoch": 3.606145251396648, + "grad_norm": 1.7599602937698364, + "learning_rate": 4.317318435754191e-06, + "loss": 0.322, + "step": 1291 + }, + { + "epoch": 3.6089385474860336, + "grad_norm": 1.626774549484253, + "learning_rate": 4.320670391061453e-06, + "loss": 0.3323, + "step": 1292 + }, + { + "epoch": 3.611731843575419, + "grad_norm": 3.5924267768859863, + "learning_rate": 4.324022346368715e-06, + "loss": 0.3574, + "step": 1293 + }, + { + "epoch": 3.6145251396648046, + "grad_norm": 1.6880981922149658, + "learning_rate": 4.327374301675978e-06, + "loss": 0.4084, + "step": 1294 + }, + { + "epoch": 3.61731843575419, + "grad_norm": 2.328871250152588, + "learning_rate": 4.33072625698324e-06, + "loss": 0.3676, + "step": 1295 + }, + { + "epoch": 3.6201117318435756, + "grad_norm": 3.6545512676239014, + "learning_rate": 4.334078212290503e-06, + "loss": 0.4233, + "step": 1296 + }, + { + "epoch": 3.622905027932961, + "grad_norm": 2.4232983589172363, + "learning_rate": 4.3374301675977656e-06, + "loss": 0.4269, + "step": 1297 + }, + { + "epoch": 3.6256983240223466, + "grad_norm": 4.286740303039551, + "learning_rate": 4.340782122905028e-06, + "loss": 0.4684, + "step": 1298 + }, + { + "epoch": 3.628491620111732, + "grad_norm": 7.273785591125488, + "learning_rate": 4.344134078212291e-06, + "loss": 0.5436, + "step": 1299 + }, + { + "epoch": 3.631284916201117, + "grad_norm": 1.3584160804748535, + "learning_rate": 4.347486033519553e-06, + "loss": 0.2918, + "step": 1300 + }, + { + "epoch": 3.6340782122905027, + "grad_norm": 1.1079648733139038, + "learning_rate": 4.350837988826816e-06, + "loss": 0.2836, + "step": 1301 + }, + { + "epoch": 3.636871508379888, + "grad_norm": 1.1159429550170898, + "learning_rate": 4.354189944134078e-06, + "loss": 0.3148, + "step": 1302 + }, + { + "epoch": 3.6396648044692737, + "grad_norm": 1.1226056814193726, + "learning_rate": 4.357541899441341e-06, + "loss": 0.3129, + "step": 1303 + }, + { + "epoch": 3.642458100558659, + "grad_norm": 1.036586046218872, + "learning_rate": 4.3608938547486035e-06, + "loss": 0.2778, + "step": 1304 + }, + { + "epoch": 3.6452513966480447, + "grad_norm": 1.2653192281723022, + "learning_rate": 4.364245810055866e-06, + "loss": 0.3379, + "step": 1305 + }, + { + "epoch": 3.64804469273743, + "grad_norm": 1.8584184646606445, + "learning_rate": 4.367597765363129e-06, + "loss": 0.3043, + "step": 1306 + }, + { + "epoch": 3.6508379888268156, + "grad_norm": 1.4460487365722656, + "learning_rate": 4.370949720670391e-06, + "loss": 0.3312, + "step": 1307 + }, + { + "epoch": 3.653631284916201, + "grad_norm": 1.544199824333191, + "learning_rate": 4.374301675977654e-06, + "loss": 0.3208, + "step": 1308 + }, + { + "epoch": 3.6564245810055866, + "grad_norm": 1.0816829204559326, + "learning_rate": 4.377653631284916e-06, + "loss": 0.3312, + "step": 1309 + }, + { + "epoch": 3.659217877094972, + "grad_norm": 1.0235631465911865, + "learning_rate": 4.381005586592178e-06, + "loss": 0.3123, + "step": 1310 + }, + { + "epoch": 3.6620111731843576, + "grad_norm": 1.4816781282424927, + "learning_rate": 4.384357541899441e-06, + "loss": 0.3288, + "step": 1311 + }, + { + "epoch": 3.664804469273743, + "grad_norm": 1.3982703685760498, + "learning_rate": 4.3877094972067035e-06, + "loss": 0.292, + "step": 1312 + }, + { + "epoch": 3.6675977653631286, + "grad_norm": 1.7121195793151855, + "learning_rate": 4.391061452513967e-06, + "loss": 0.2755, + "step": 1313 + }, + { + "epoch": 3.6703910614525137, + "grad_norm": 2.871760606765747, + "learning_rate": 4.394413407821229e-06, + "loss": 0.3719, + "step": 1314 + }, + { + "epoch": 3.673184357541899, + "grad_norm": 1.3116278648376465, + "learning_rate": 4.397765363128492e-06, + "loss": 0.3948, + "step": 1315 + }, + { + "epoch": 3.6759776536312847, + "grad_norm": 1.8066879510879517, + "learning_rate": 4.401117318435755e-06, + "loss": 0.3522, + "step": 1316 + }, + { + "epoch": 3.67877094972067, + "grad_norm": 1.9206733703613281, + "learning_rate": 4.404469273743017e-06, + "loss": 0.3332, + "step": 1317 + }, + { + "epoch": 3.6815642458100557, + "grad_norm": 2.8793301582336426, + "learning_rate": 4.40782122905028e-06, + "loss": 0.4462, + "step": 1318 + }, + { + "epoch": 3.684357541899441, + "grad_norm": 2.256497383117676, + "learning_rate": 4.411173184357542e-06, + "loss": 0.3854, + "step": 1319 + }, + { + "epoch": 3.6871508379888267, + "grad_norm": 1.956012487411499, + "learning_rate": 4.4145251396648045e-06, + "loss": 0.3895, + "step": 1320 + }, + { + "epoch": 3.689944134078212, + "grad_norm": 2.5333101749420166, + "learning_rate": 4.4178770949720675e-06, + "loss": 0.4069, + "step": 1321 + }, + { + "epoch": 3.6927374301675977, + "grad_norm": 2.829958915710449, + "learning_rate": 4.42122905027933e-06, + "loss": 0.402, + "step": 1322 + }, + { + "epoch": 3.695530726256983, + "grad_norm": 10.57265853881836, + "learning_rate": 4.424581005586593e-06, + "loss": 0.4581, + "step": 1323 + }, + { + "epoch": 3.6983240223463687, + "grad_norm": 4.017953872680664, + "learning_rate": 4.427932960893855e-06, + "loss": 0.6406, + "step": 1324 + }, + { + "epoch": 3.701117318435754, + "grad_norm": 1.3446170091629028, + "learning_rate": 4.431284916201117e-06, + "loss": 0.3533, + "step": 1325 + }, + { + "epoch": 3.7039106145251397, + "grad_norm": 1.1496012210845947, + "learning_rate": 4.43463687150838e-06, + "loss": 0.3011, + "step": 1326 + }, + { + "epoch": 3.706703910614525, + "grad_norm": 1.2443937063217163, + "learning_rate": 4.437988826815642e-06, + "loss": 0.272, + "step": 1327 + }, + { + "epoch": 3.7094972067039107, + "grad_norm": 1.3341974020004272, + "learning_rate": 4.441340782122905e-06, + "loss": 0.3193, + "step": 1328 + }, + { + "epoch": 3.712290502793296, + "grad_norm": 4.2614874839782715, + "learning_rate": 4.444692737430168e-06, + "loss": 0.3554, + "step": 1329 + }, + { + "epoch": 3.7150837988826817, + "grad_norm": 2.271589517593384, + "learning_rate": 4.44804469273743e-06, + "loss": 0.3231, + "step": 1330 + }, + { + "epoch": 3.717877094972067, + "grad_norm": 2.1845242977142334, + "learning_rate": 4.451396648044693e-06, + "loss": 0.3121, + "step": 1331 + }, + { + "epoch": 3.7206703910614527, + "grad_norm": 1.1860986948013306, + "learning_rate": 4.454748603351955e-06, + "loss": 0.2835, + "step": 1332 + }, + { + "epoch": 3.723463687150838, + "grad_norm": 1.4681758880615234, + "learning_rate": 4.458100558659218e-06, + "loss": 0.3077, + "step": 1333 + }, + { + "epoch": 3.7262569832402237, + "grad_norm": 1.1743826866149902, + "learning_rate": 4.46145251396648e-06, + "loss": 0.3558, + "step": 1334 + }, + { + "epoch": 3.729050279329609, + "grad_norm": 1.6545650959014893, + "learning_rate": 4.4648044692737424e-06, + "loss": 0.3176, + "step": 1335 + }, + { + "epoch": 3.7318435754189943, + "grad_norm": 1.597456693649292, + "learning_rate": 4.4681564245810055e-06, + "loss": 0.3053, + "step": 1336 + }, + { + "epoch": 3.7346368715083798, + "grad_norm": 1.2069505453109741, + "learning_rate": 4.471508379888268e-06, + "loss": 0.3373, + "step": 1337 + }, + { + "epoch": 3.7374301675977653, + "grad_norm": 1.3008424043655396, + "learning_rate": 4.474860335195531e-06, + "loss": 0.2854, + "step": 1338 + }, + { + "epoch": 3.7402234636871508, + "grad_norm": 1.5795600414276123, + "learning_rate": 4.478212290502794e-06, + "loss": 0.328, + "step": 1339 + }, + { + "epoch": 3.7430167597765363, + "grad_norm": 2.0557682514190674, + "learning_rate": 4.481564245810056e-06, + "loss": 0.3827, + "step": 1340 + }, + { + "epoch": 3.7458100558659218, + "grad_norm": 1.6960523128509521, + "learning_rate": 4.484916201117319e-06, + "loss": 0.3637, + "step": 1341 + }, + { + "epoch": 3.7486033519553073, + "grad_norm": 1.7345499992370605, + "learning_rate": 4.488268156424581e-06, + "loss": 0.3534, + "step": 1342 + }, + { + "epoch": 3.7513966480446927, + "grad_norm": 2.679267406463623, + "learning_rate": 4.491620111731844e-06, + "loss": 0.3603, + "step": 1343 + }, + { + "epoch": 3.7541899441340782, + "grad_norm": 2.601855993270874, + "learning_rate": 4.494972067039106e-06, + "loss": 0.3578, + "step": 1344 + }, + { + "epoch": 3.7569832402234637, + "grad_norm": 1.9907031059265137, + "learning_rate": 4.498324022346369e-06, + "loss": 0.3188, + "step": 1345 + }, + { + "epoch": 3.7597765363128492, + "grad_norm": 2.021497964859009, + "learning_rate": 4.501675977653632e-06, + "loss": 0.4182, + "step": 1346 + }, + { + "epoch": 3.7625698324022347, + "grad_norm": 2.519223690032959, + "learning_rate": 4.505027932960894e-06, + "loss": 0.425, + "step": 1347 + }, + { + "epoch": 3.7653631284916202, + "grad_norm": 3.9739372730255127, + "learning_rate": 4.508379888268157e-06, + "loss": 0.506, + "step": 1348 + }, + { + "epoch": 3.7681564245810057, + "grad_norm": 3.612534999847412, + "learning_rate": 4.511731843575419e-06, + "loss": 0.5864, + "step": 1349 + }, + { + "epoch": 3.770949720670391, + "grad_norm": 0.8302329778671265, + "learning_rate": 4.515083798882682e-06, + "loss": 0.309, + "step": 1350 + }, + { + "epoch": 3.7737430167597763, + "grad_norm": 0.8852688670158386, + "learning_rate": 4.518435754189944e-06, + "loss": 0.3008, + "step": 1351 + }, + { + "epoch": 3.776536312849162, + "grad_norm": 1.0472099781036377, + "learning_rate": 4.5217877094972065e-06, + "loss": 0.3066, + "step": 1352 + }, + { + "epoch": 3.7793296089385473, + "grad_norm": 1.0146552324295044, + "learning_rate": 4.5251396648044695e-06, + "loss": 0.3165, + "step": 1353 + }, + { + "epoch": 3.782122905027933, + "grad_norm": 0.8513569831848145, + "learning_rate": 4.528491620111732e-06, + "loss": 0.3055, + "step": 1354 + }, + { + "epoch": 3.7849162011173183, + "grad_norm": 1.5159393548965454, + "learning_rate": 4.531843575418995e-06, + "loss": 0.32, + "step": 1355 + }, + { + "epoch": 3.787709497206704, + "grad_norm": 1.3136472702026367, + "learning_rate": 4.535195530726257e-06, + "loss": 0.3458, + "step": 1356 + }, + { + "epoch": 3.7905027932960893, + "grad_norm": 2.299961805343628, + "learning_rate": 4.538547486033519e-06, + "loss": 0.3033, + "step": 1357 + }, + { + "epoch": 3.793296089385475, + "grad_norm": 1.6148791313171387, + "learning_rate": 4.541899441340782e-06, + "loss": 0.3034, + "step": 1358 + }, + { + "epoch": 3.7960893854748603, + "grad_norm": 1.7344541549682617, + "learning_rate": 4.545251396648044e-06, + "loss": 0.3054, + "step": 1359 + }, + { + "epoch": 3.798882681564246, + "grad_norm": 1.2550467252731323, + "learning_rate": 4.548603351955307e-06, + "loss": 0.3479, + "step": 1360 + }, + { + "epoch": 3.8016759776536313, + "grad_norm": 1.1956249475479126, + "learning_rate": 4.55195530726257e-06, + "loss": 0.3389, + "step": 1361 + }, + { + "epoch": 3.804469273743017, + "grad_norm": 1.4012634754180908, + "learning_rate": 4.555307262569832e-06, + "loss": 0.2966, + "step": 1362 + }, + { + "epoch": 3.8072625698324023, + "grad_norm": 2.992264747619629, + "learning_rate": 4.558659217877096e-06, + "loss": 0.3825, + "step": 1363 + }, + { + "epoch": 3.810055865921788, + "grad_norm": 2.46866774559021, + "learning_rate": 4.562011173184358e-06, + "loss": 0.3272, + "step": 1364 + }, + { + "epoch": 3.8128491620111733, + "grad_norm": 1.9699687957763672, + "learning_rate": 4.565363128491621e-06, + "loss": 0.3357, + "step": 1365 + }, + { + "epoch": 3.815642458100559, + "grad_norm": 1.4382697343826294, + "learning_rate": 4.568715083798883e-06, + "loss": 0.3418, + "step": 1366 + }, + { + "epoch": 3.8184357541899443, + "grad_norm": 1.4273732900619507, + "learning_rate": 4.572067039106145e-06, + "loss": 0.3264, + "step": 1367 + }, + { + "epoch": 3.82122905027933, + "grad_norm": 1.4358539581298828, + "learning_rate": 4.575418994413408e-06, + "loss": 0.328, + "step": 1368 + }, + { + "epoch": 3.8240223463687153, + "grad_norm": 2.1353566646575928, + "learning_rate": 4.5787709497206705e-06, + "loss": 0.3306, + "step": 1369 + }, + { + "epoch": 3.826815642458101, + "grad_norm": 2.1310083866119385, + "learning_rate": 4.5821229050279336e-06, + "loss": 0.4416, + "step": 1370 + }, + { + "epoch": 3.8296089385474863, + "grad_norm": 1.7869678735733032, + "learning_rate": 4.585474860335196e-06, + "loss": 0.3898, + "step": 1371 + }, + { + "epoch": 3.8324022346368714, + "grad_norm": 2.1416068077087402, + "learning_rate": 4.588826815642458e-06, + "loss": 0.4516, + "step": 1372 + }, + { + "epoch": 3.835195530726257, + "grad_norm": 3.639678716659546, + "learning_rate": 4.592178770949721e-06, + "loss": 0.4521, + "step": 1373 + }, + { + "epoch": 3.8379888268156424, + "grad_norm": 3.8462321758270264, + "learning_rate": 4.595530726256983e-06, + "loss": 0.5412, + "step": 1374 + }, + { + "epoch": 3.840782122905028, + "grad_norm": 1.4868900775909424, + "learning_rate": 4.598882681564246e-06, + "loss": 0.3332, + "step": 1375 + }, + { + "epoch": 3.8435754189944134, + "grad_norm": 1.484424114227295, + "learning_rate": 4.602234636871508e-06, + "loss": 0.3185, + "step": 1376 + }, + { + "epoch": 3.846368715083799, + "grad_norm": 1.0440224409103394, + "learning_rate": 4.605586592178771e-06, + "loss": 0.3223, + "step": 1377 + }, + { + "epoch": 3.8491620111731844, + "grad_norm": 1.0802586078643799, + "learning_rate": 4.608938547486034e-06, + "loss": 0.3223, + "step": 1378 + }, + { + "epoch": 3.85195530726257, + "grad_norm": 0.8876691460609436, + "learning_rate": 4.612290502793296e-06, + "loss": 0.2889, + "step": 1379 + }, + { + "epoch": 3.8547486033519553, + "grad_norm": 1.035894751548767, + "learning_rate": 4.615642458100559e-06, + "loss": 0.2843, + "step": 1380 + }, + { + "epoch": 3.857541899441341, + "grad_norm": 1.3959912061691284, + "learning_rate": 4.618994413407821e-06, + "loss": 0.2943, + "step": 1381 + }, + { + "epoch": 3.8603351955307263, + "grad_norm": 1.419358730316162, + "learning_rate": 4.622346368715083e-06, + "loss": 0.3507, + "step": 1382 + }, + { + "epoch": 3.863128491620112, + "grad_norm": 1.020471453666687, + "learning_rate": 4.625698324022346e-06, + "loss": 0.2928, + "step": 1383 + }, + { + "epoch": 3.8659217877094973, + "grad_norm": 1.0672205686569214, + "learning_rate": 4.6290502793296085e-06, + "loss": 0.3251, + "step": 1384 + }, + { + "epoch": 3.868715083798883, + "grad_norm": 1.4566621780395508, + "learning_rate": 4.6324022346368715e-06, + "loss": 0.3123, + "step": 1385 + }, + { + "epoch": 3.871508379888268, + "grad_norm": 1.8961660861968994, + "learning_rate": 4.635754189944134e-06, + "loss": 0.3423, + "step": 1386 + }, + { + "epoch": 3.8743016759776534, + "grad_norm": 3.8542118072509766, + "learning_rate": 4.639106145251397e-06, + "loss": 0.3081, + "step": 1387 + }, + { + "epoch": 3.877094972067039, + "grad_norm": 1.4976840019226074, + "learning_rate": 4.64245810055866e-06, + "loss": 0.3147, + "step": 1388 + }, + { + "epoch": 3.8798882681564244, + "grad_norm": 1.6414885520935059, + "learning_rate": 4.645810055865922e-06, + "loss": 0.3285, + "step": 1389 + }, + { + "epoch": 3.88268156424581, + "grad_norm": 2.6283016204833984, + "learning_rate": 4.649162011173185e-06, + "loss": 0.3035, + "step": 1390 + }, + { + "epoch": 3.8854748603351954, + "grad_norm": 1.5289987325668335, + "learning_rate": 4.652513966480447e-06, + "loss": 0.3422, + "step": 1391 + }, + { + "epoch": 3.888268156424581, + "grad_norm": 1.8149363994598389, + "learning_rate": 4.655865921787709e-06, + "loss": 0.3574, + "step": 1392 + }, + { + "epoch": 3.8910614525139664, + "grad_norm": 2.0374598503112793, + "learning_rate": 4.6592178770949725e-06, + "loss": 0.3891, + "step": 1393 + }, + { + "epoch": 3.893854748603352, + "grad_norm": 1.675881028175354, + "learning_rate": 4.662569832402235e-06, + "loss": 0.3765, + "step": 1394 + }, + { + "epoch": 3.8966480446927374, + "grad_norm": 7.783751010894775, + "learning_rate": 4.665921787709498e-06, + "loss": 0.3855, + "step": 1395 + }, + { + "epoch": 3.899441340782123, + "grad_norm": 2.330472230911255, + "learning_rate": 4.66927374301676e-06, + "loss": 0.3987, + "step": 1396 + }, + { + "epoch": 3.9022346368715084, + "grad_norm": 2.929588794708252, + "learning_rate": 4.672625698324023e-06, + "loss": 0.4205, + "step": 1397 + }, + { + "epoch": 3.905027932960894, + "grad_norm": 3.2448537349700928, + "learning_rate": 4.675977653631285e-06, + "loss": 0.5134, + "step": 1398 + }, + { + "epoch": 3.9078212290502794, + "grad_norm": 5.120497226715088, + "learning_rate": 4.679329608938547e-06, + "loss": 0.6171, + "step": 1399 + }, + { + "epoch": 3.910614525139665, + "grad_norm": 1.7573543787002563, + "learning_rate": 4.68268156424581e-06, + "loss": 0.3112, + "step": 1400 + }, + { + "epoch": 3.9134078212290504, + "grad_norm": 0.9426363706588745, + "learning_rate": 4.6860335195530725e-06, + "loss": 0.2923, + "step": 1401 + }, + { + "epoch": 3.916201117318436, + "grad_norm": 0.847605288028717, + "learning_rate": 4.689385474860336e-06, + "loss": 0.2864, + "step": 1402 + }, + { + "epoch": 3.9189944134078214, + "grad_norm": 0.7622266411781311, + "learning_rate": 4.692737430167598e-06, + "loss": 0.2776, + "step": 1403 + }, + { + "epoch": 3.921787709497207, + "grad_norm": 1.0128124952316284, + "learning_rate": 4.69608938547486e-06, + "loss": 0.3358, + "step": 1404 + }, + { + "epoch": 3.9245810055865924, + "grad_norm": 1.4241243600845337, + "learning_rate": 4.699441340782123e-06, + "loss": 0.2721, + "step": 1405 + }, + { + "epoch": 3.927374301675978, + "grad_norm": 1.0260761976242065, + "learning_rate": 4.702793296089385e-06, + "loss": 0.3149, + "step": 1406 + }, + { + "epoch": 3.9301675977653634, + "grad_norm": 0.8978831171989441, + "learning_rate": 4.706145251396648e-06, + "loss": 0.3025, + "step": 1407 + }, + { + "epoch": 3.9329608938547485, + "grad_norm": 1.0879201889038086, + "learning_rate": 4.7094972067039104e-06, + "loss": 0.2993, + "step": 1408 + }, + { + "epoch": 3.935754189944134, + "grad_norm": 1.5749925374984741, + "learning_rate": 4.712849162011173e-06, + "loss": 0.3235, + "step": 1409 + }, + { + "epoch": 3.9385474860335195, + "grad_norm": 1.2913912534713745, + "learning_rate": 4.716201117318436e-06, + "loss": 0.3081, + "step": 1410 + }, + { + "epoch": 3.941340782122905, + "grad_norm": 1.5889540910720825, + "learning_rate": 4.719553072625699e-06, + "loss": 0.3258, + "step": 1411 + }, + { + "epoch": 3.9441340782122905, + "grad_norm": 1.8847867250442505, + "learning_rate": 4.722905027932962e-06, + "loss": 0.3008, + "step": 1412 + }, + { + "epoch": 3.946927374301676, + "grad_norm": 1.4849687814712524, + "learning_rate": 4.726256983240224e-06, + "loss": 0.3218, + "step": 1413 + }, + { + "epoch": 3.9497206703910615, + "grad_norm": 1.1775243282318115, + "learning_rate": 4.729608938547486e-06, + "loss": 0.2975, + "step": 1414 + }, + { + "epoch": 3.952513966480447, + "grad_norm": 1.1609933376312256, + "learning_rate": 4.732960893854749e-06, + "loss": 0.3264, + "step": 1415 + }, + { + "epoch": 3.9553072625698324, + "grad_norm": 1.2967239618301392, + "learning_rate": 4.736312849162011e-06, + "loss": 0.3574, + "step": 1416 + }, + { + "epoch": 3.958100558659218, + "grad_norm": 2.7718393802642822, + "learning_rate": 4.739664804469274e-06, + "loss": 0.3356, + "step": 1417 + }, + { + "epoch": 3.9608938547486034, + "grad_norm": 1.4305980205535889, + "learning_rate": 4.743016759776537e-06, + "loss": 0.3245, + "step": 1418 + }, + { + "epoch": 3.963687150837989, + "grad_norm": 2.132070779800415, + "learning_rate": 4.746368715083799e-06, + "loss": 0.3555, + "step": 1419 + }, + { + "epoch": 3.9664804469273744, + "grad_norm": 2.5774402618408203, + "learning_rate": 4.749720670391062e-06, + "loss": 0.3487, + "step": 1420 + }, + { + "epoch": 3.9692737430167595, + "grad_norm": 1.8792834281921387, + "learning_rate": 4.753072625698324e-06, + "loss": 0.3864, + "step": 1421 + }, + { + "epoch": 3.972067039106145, + "grad_norm": 3.632439613342285, + "learning_rate": 4.756424581005587e-06, + "loss": 0.4399, + "step": 1422 + }, + { + "epoch": 3.9748603351955305, + "grad_norm": 2.122959852218628, + "learning_rate": 4.759776536312849e-06, + "loss": 0.4063, + "step": 1423 + }, + { + "epoch": 3.977653631284916, + "grad_norm": 4.096036911010742, + "learning_rate": 4.7631284916201114e-06, + "loss": 0.5628, + "step": 1424 + }, + { + "epoch": 3.9804469273743015, + "grad_norm": 1.2448925971984863, + "learning_rate": 4.7664804469273745e-06, + "loss": 0.3142, + "step": 1425 + }, + { + "epoch": 3.983240223463687, + "grad_norm": 2.2381834983825684, + "learning_rate": 4.769832402234637e-06, + "loss": 0.3131, + "step": 1426 + }, + { + "epoch": 3.9860335195530725, + "grad_norm": 1.2145166397094727, + "learning_rate": 4.7731843575419e-06, + "loss": 0.2769, + "step": 1427 + }, + { + "epoch": 3.988826815642458, + "grad_norm": 2.5460033416748047, + "learning_rate": 4.776536312849162e-06, + "loss": 0.3001, + "step": 1428 + }, + { + "epoch": 3.9916201117318435, + "grad_norm": 1.36093008518219, + "learning_rate": 4.779888268156424e-06, + "loss": 0.3191, + "step": 1429 + }, + { + "epoch": 3.994413407821229, + "grad_norm": 3.8495054244995117, + "learning_rate": 4.783240223463687e-06, + "loss": 0.3738, + "step": 1430 + }, + { + "epoch": 3.9972067039106145, + "grad_norm": 1.8115493059158325, + "learning_rate": 4.786592178770949e-06, + "loss": 0.3427, + "step": 1431 + }, + { + "epoch": 4.0, + "grad_norm": 3.456529378890991, + "learning_rate": 4.789944134078212e-06, + "loss": 0.4379, + "step": 1432 + }, + { + "epoch": 4.0027932960893855, + "grad_norm": 1.2401498556137085, + "learning_rate": 4.7932960893854746e-06, + "loss": 0.3311, + "step": 1433 + }, + { + "epoch": 4.005586592178771, + "grad_norm": 0.731011688709259, + "learning_rate": 4.796648044692737e-06, + "loss": 0.2668, + "step": 1434 + }, + { + "epoch": 4.0083798882681565, + "grad_norm": 1.2079449892044067, + "learning_rate": 4.800000000000001e-06, + "loss": 0.2477, + "step": 1435 + }, + { + "epoch": 4.011173184357542, + "grad_norm": 1.0756217241287231, + "learning_rate": 4.803351955307263e-06, + "loss": 0.3046, + "step": 1436 + }, + { + "epoch": 4.0139664804469275, + "grad_norm": 1.318786859512329, + "learning_rate": 4.806703910614526e-06, + "loss": 0.3183, + "step": 1437 + }, + { + "epoch": 4.016759776536313, + "grad_norm": 1.2232187986373901, + "learning_rate": 4.810055865921788e-06, + "loss": 0.3109, + "step": 1438 + }, + { + "epoch": 4.0195530726256985, + "grad_norm": 1.079311728477478, + "learning_rate": 4.81340782122905e-06, + "loss": 0.2901, + "step": 1439 + }, + { + "epoch": 4.022346368715084, + "grad_norm": 0.8263248205184937, + "learning_rate": 4.816759776536313e-06, + "loss": 0.2788, + "step": 1440 + }, + { + "epoch": 4.0251396648044695, + "grad_norm": 1.3990592956542969, + "learning_rate": 4.8201117318435755e-06, + "loss": 0.2958, + "step": 1441 + }, + { + "epoch": 4.027932960893855, + "grad_norm": 1.7201941013336182, + "learning_rate": 4.8234636871508385e-06, + "loss": 0.2874, + "step": 1442 + }, + { + "epoch": 4.0307262569832405, + "grad_norm": 1.1468440294265747, + "learning_rate": 4.826815642458101e-06, + "loss": 0.2808, + "step": 1443 + }, + { + "epoch": 4.033519553072626, + "grad_norm": 1.1514559984207153, + "learning_rate": 4.830167597765364e-06, + "loss": 0.3541, + "step": 1444 + }, + { + "epoch": 4.0363128491620115, + "grad_norm": 1.6899693012237549, + "learning_rate": 4.833519553072626e-06, + "loss": 0.2822, + "step": 1445 + }, + { + "epoch": 4.039106145251397, + "grad_norm": 1.2757807970046997, + "learning_rate": 4.836871508379888e-06, + "loss": 0.3409, + "step": 1446 + }, + { + "epoch": 4.0418994413407825, + "grad_norm": 1.4153623580932617, + "learning_rate": 4.840223463687151e-06, + "loss": 0.336, + "step": 1447 + }, + { + "epoch": 4.044692737430168, + "grad_norm": 1.1445424556732178, + "learning_rate": 4.843575418994413e-06, + "loss": 0.2627, + "step": 1448 + }, + { + "epoch": 4.0474860335195535, + "grad_norm": 1.4836316108703613, + "learning_rate": 4.846927374301676e-06, + "loss": 0.3532, + "step": 1449 + }, + { + "epoch": 4.050279329608939, + "grad_norm": 1.5797406435012817, + "learning_rate": 4.850279329608939e-06, + "loss": 0.3111, + "step": 1450 + }, + { + "epoch": 4.053072625698324, + "grad_norm": 1.5097262859344482, + "learning_rate": 4.853631284916201e-06, + "loss": 0.3437, + "step": 1451 + }, + { + "epoch": 4.055865921787709, + "grad_norm": 1.7183111906051636, + "learning_rate": 4.856983240223464e-06, + "loss": 0.3515, + "step": 1452 + }, + { + "epoch": 4.058659217877095, + "grad_norm": 1.976593255996704, + "learning_rate": 4.860335195530726e-06, + "loss": 0.3841, + "step": 1453 + }, + { + "epoch": 4.06145251396648, + "grad_norm": 2.150451898574829, + "learning_rate": 4.863687150837989e-06, + "loss": 0.3765, + "step": 1454 + }, + { + "epoch": 4.064245810055866, + "grad_norm": 2.2433841228485107, + "learning_rate": 4.867039106145251e-06, + "loss": 0.4174, + "step": 1455 + }, + { + "epoch": 4.067039106145251, + "grad_norm": 1.7732762098312378, + "learning_rate": 4.8703910614525134e-06, + "loss": 0.3746, + "step": 1456 + }, + { + "epoch": 4.069832402234637, + "grad_norm": 4.701582908630371, + "learning_rate": 4.8737430167597765e-06, + "loss": 0.5597, + "step": 1457 + }, + { + "epoch": 4.072625698324022, + "grad_norm": 1.64872407913208, + "learning_rate": 4.877094972067039e-06, + "loss": 0.2856, + "step": 1458 + }, + { + "epoch": 4.075418994413408, + "grad_norm": 0.7819265127182007, + "learning_rate": 4.8804469273743026e-06, + "loss": 0.2479, + "step": 1459 + }, + { + "epoch": 4.078212290502793, + "grad_norm": 1.087489128112793, + "learning_rate": 4.883798882681565e-06, + "loss": 0.2947, + "step": 1460 + }, + { + "epoch": 4.081005586592179, + "grad_norm": 1.2870094776153564, + "learning_rate": 4.887150837988827e-06, + "loss": 0.3739, + "step": 1461 + }, + { + "epoch": 4.083798882681564, + "grad_norm": 2.6875267028808594, + "learning_rate": 4.89050279329609e-06, + "loss": 0.3395, + "step": 1462 + }, + { + "epoch": 4.08659217877095, + "grad_norm": 1.4719398021697998, + "learning_rate": 4.893854748603352e-06, + "loss": 0.3128, + "step": 1463 + }, + { + "epoch": 4.089385474860335, + "grad_norm": 1.4132143259048462, + "learning_rate": 4.897206703910615e-06, + "loss": 0.3265, + "step": 1464 + }, + { + "epoch": 4.092178770949721, + "grad_norm": 0.9505769610404968, + "learning_rate": 4.900558659217877e-06, + "loss": 0.3259, + "step": 1465 + }, + { + "epoch": 4.094972067039106, + "grad_norm": 0.9185716509819031, + "learning_rate": 4.90391061452514e-06, + "loss": 0.2889, + "step": 1466 + }, + { + "epoch": 4.097765363128492, + "grad_norm": 1.3268781900405884, + "learning_rate": 4.907262569832403e-06, + "loss": 0.3326, + "step": 1467 + }, + { + "epoch": 4.100558659217877, + "grad_norm": 1.1900060176849365, + "learning_rate": 4.910614525139665e-06, + "loss": 0.2809, + "step": 1468 + }, + { + "epoch": 4.103351955307263, + "grad_norm": 1.3500521183013916, + "learning_rate": 4.913966480446928e-06, + "loss": 0.2866, + "step": 1469 + }, + { + "epoch": 4.106145251396648, + "grad_norm": 1.199096918106079, + "learning_rate": 4.91731843575419e-06, + "loss": 0.2592, + "step": 1470 + }, + { + "epoch": 4.108938547486034, + "grad_norm": 2.6659438610076904, + "learning_rate": 4.920670391061452e-06, + "loss": 0.3008, + "step": 1471 + }, + { + "epoch": 4.111731843575419, + "grad_norm": 1.5160025358200073, + "learning_rate": 4.924022346368715e-06, + "loss": 0.3951, + "step": 1472 + }, + { + "epoch": 4.114525139664805, + "grad_norm": 2.1113078594207764, + "learning_rate": 4.9273743016759775e-06, + "loss": 0.3211, + "step": 1473 + }, + { + "epoch": 4.11731843575419, + "grad_norm": 1.964053750038147, + "learning_rate": 4.9307262569832405e-06, + "loss": 0.3125, + "step": 1474 + }, + { + "epoch": 4.120111731843576, + "grad_norm": 1.3445509672164917, + "learning_rate": 4.934078212290503e-06, + "loss": 0.3399, + "step": 1475 + }, + { + "epoch": 4.122905027932961, + "grad_norm": 3.57043194770813, + "learning_rate": 4.937430167597765e-06, + "loss": 0.3183, + "step": 1476 + }, + { + "epoch": 4.125698324022347, + "grad_norm": 2.00640869140625, + "learning_rate": 4.940782122905028e-06, + "loss": 0.3601, + "step": 1477 + }, + { + "epoch": 4.128491620111732, + "grad_norm": 1.4489836692810059, + "learning_rate": 4.94413407821229e-06, + "loss": 0.325, + "step": 1478 + }, + { + "epoch": 4.131284916201118, + "grad_norm": 1.6279054880142212, + "learning_rate": 4.947486033519553e-06, + "loss": 0.3692, + "step": 1479 + }, + { + "epoch": 4.134078212290503, + "grad_norm": 2.017972707748413, + "learning_rate": 4.950837988826815e-06, + "loss": 0.3778, + "step": 1480 + }, + { + "epoch": 4.136871508379889, + "grad_norm": 3.9486801624298096, + "learning_rate": 4.9541899441340776e-06, + "loss": 0.4082, + "step": 1481 + }, + { + "epoch": 4.139664804469274, + "grad_norm": 8.270044326782227, + "learning_rate": 4.957541899441341e-06, + "loss": 0.4107, + "step": 1482 + }, + { + "epoch": 4.14245810055866, + "grad_norm": 1.5810019969940186, + "learning_rate": 4.960893854748604e-06, + "loss": 0.3515, + "step": 1483 + }, + { + "epoch": 4.145251396648045, + "grad_norm": 1.126284122467041, + "learning_rate": 4.964245810055867e-06, + "loss": 0.3515, + "step": 1484 + }, + { + "epoch": 4.148044692737431, + "grad_norm": 0.9932015538215637, + "learning_rate": 4.967597765363129e-06, + "loss": 0.2929, + "step": 1485 + }, + { + "epoch": 4.150837988826815, + "grad_norm": 0.9566678404808044, + "learning_rate": 4.970949720670391e-06, + "loss": 0.2812, + "step": 1486 + }, + { + "epoch": 4.153631284916201, + "grad_norm": 1.0570974349975586, + "learning_rate": 4.974301675977654e-06, + "loss": 0.2498, + "step": 1487 + }, + { + "epoch": 4.156424581005586, + "grad_norm": 0.9708537459373474, + "learning_rate": 4.977653631284916e-06, + "loss": 0.3087, + "step": 1488 + }, + { + "epoch": 4.159217877094972, + "grad_norm": 1.0620113611221313, + "learning_rate": 4.981005586592179e-06, + "loss": 0.3143, + "step": 1489 + }, + { + "epoch": 4.162011173184357, + "grad_norm": 1.5860751867294312, + "learning_rate": 4.9843575418994415e-06, + "loss": 0.3005, + "step": 1490 + }, + { + "epoch": 4.164804469273743, + "grad_norm": 0.9743533730506897, + "learning_rate": 4.9877094972067046e-06, + "loss": 0.2765, + "step": 1491 + }, + { + "epoch": 4.167597765363128, + "grad_norm": 1.30978262424469, + "learning_rate": 4.991061452513967e-06, + "loss": 0.3383, + "step": 1492 + }, + { + "epoch": 4.170391061452514, + "grad_norm": 1.3617440462112427, + "learning_rate": 4.994413407821229e-06, + "loss": 0.2999, + "step": 1493 + }, + { + "epoch": 4.173184357541899, + "grad_norm": 1.3727173805236816, + "learning_rate": 4.997765363128492e-06, + "loss": 0.2727, + "step": 1494 + }, + { + "epoch": 4.175977653631285, + "grad_norm": 1.0447728633880615, + "learning_rate": 5.001117318435754e-06, + "loss": 0.3211, + "step": 1495 + }, + { + "epoch": 4.17877094972067, + "grad_norm": 1.192089557647705, + "learning_rate": 5.004469273743017e-06, + "loss": 0.2816, + "step": 1496 + }, + { + "epoch": 4.181564245810056, + "grad_norm": 1.6479718685150146, + "learning_rate": 5.007821229050279e-06, + "loss": 0.2994, + "step": 1497 + }, + { + "epoch": 4.184357541899441, + "grad_norm": 1.486892580986023, + "learning_rate": 5.011173184357542e-06, + "loss": 0.3166, + "step": 1498 + }, + { + "epoch": 4.187150837988827, + "grad_norm": 1.4377686977386475, + "learning_rate": 5.014525139664805e-06, + "loss": 0.2741, + "step": 1499 + }, + { + "epoch": 4.189944134078212, + "grad_norm": 2.5752599239349365, + "learning_rate": 5.017877094972067e-06, + "loss": 0.2795, + "step": 1500 + }, + { + "epoch": 4.192737430167598, + "grad_norm": 3.276172399520874, + "learning_rate": 5.02122905027933e-06, + "loss": 0.3185, + "step": 1501 + }, + { + "epoch": 4.195530726256983, + "grad_norm": 2.945248603820801, + "learning_rate": 5.024581005586592e-06, + "loss": 0.331, + "step": 1502 + }, + { + "epoch": 4.198324022346369, + "grad_norm": 2.7255618572235107, + "learning_rate": 5.027932960893854e-06, + "loss": 0.3696, + "step": 1503 + }, + { + "epoch": 4.201117318435754, + "grad_norm": 3.2913920879364014, + "learning_rate": 5.031284916201117e-06, + "loss": 0.3814, + "step": 1504 + }, + { + "epoch": 4.20391061452514, + "grad_norm": 2.363116502761841, + "learning_rate": 5.0346368715083795e-06, + "loss": 0.3922, + "step": 1505 + }, + { + "epoch": 4.206703910614525, + "grad_norm": 5.622950077056885, + "learning_rate": 5.0379888268156425e-06, + "loss": 0.4881, + "step": 1506 + }, + { + "epoch": 4.209497206703911, + "grad_norm": 9.848299980163574, + "learning_rate": 5.041340782122906e-06, + "loss": 0.5709, + "step": 1507 + }, + { + "epoch": 4.212290502793296, + "grad_norm": 1.1636101007461548, + "learning_rate": 5.044692737430168e-06, + "loss": 0.2819, + "step": 1508 + }, + { + "epoch": 4.215083798882682, + "grad_norm": 1.4857945442199707, + "learning_rate": 5.048044692737431e-06, + "loss": 0.3358, + "step": 1509 + }, + { + "epoch": 4.217877094972067, + "grad_norm": 0.9207989573478699, + "learning_rate": 5.051396648044693e-06, + "loss": 0.2569, + "step": 1510 + }, + { + "epoch": 4.220670391061453, + "grad_norm": 1.3626958131790161, + "learning_rate": 5.054748603351956e-06, + "loss": 0.3051, + "step": 1511 + }, + { + "epoch": 4.223463687150838, + "grad_norm": 1.2928950786590576, + "learning_rate": 5.058100558659218e-06, + "loss": 0.315, + "step": 1512 + }, + { + "epoch": 4.226256983240224, + "grad_norm": 1.5114504098892212, + "learning_rate": 5.0614525139664804e-06, + "loss": 0.3196, + "step": 1513 + }, + { + "epoch": 4.229050279329609, + "grad_norm": 0.9563342332839966, + "learning_rate": 5.0648044692737435e-06, + "loss": 0.2801, + "step": 1514 + }, + { + "epoch": 4.231843575418995, + "grad_norm": 1.3455272912979126, + "learning_rate": 5.068156424581006e-06, + "loss": 0.2706, + "step": 1515 + }, + { + "epoch": 4.23463687150838, + "grad_norm": 2.2796568870544434, + "learning_rate": 5.071508379888269e-06, + "loss": 0.2699, + "step": 1516 + }, + { + "epoch": 4.237430167597766, + "grad_norm": 1.032025933265686, + "learning_rate": 5.074860335195531e-06, + "loss": 0.3002, + "step": 1517 + }, + { + "epoch": 4.240223463687151, + "grad_norm": 1.3637992143630981, + "learning_rate": 5.078212290502793e-06, + "loss": 0.3074, + "step": 1518 + }, + { + "epoch": 4.243016759776537, + "grad_norm": 1.054673671722412, + "learning_rate": 5.081564245810056e-06, + "loss": 0.3327, + "step": 1519 + }, + { + "epoch": 4.245810055865922, + "grad_norm": 5.084787368774414, + "learning_rate": 5.084916201117318e-06, + "loss": 0.2639, + "step": 1520 + }, + { + "epoch": 4.248603351955307, + "grad_norm": 1.5533266067504883, + "learning_rate": 5.088268156424581e-06, + "loss": 0.3381, + "step": 1521 + }, + { + "epoch": 4.251396648044693, + "grad_norm": 1.450925350189209, + "learning_rate": 5.0916201117318435e-06, + "loss": 0.2971, + "step": 1522 + }, + { + "epoch": 4.254189944134078, + "grad_norm": 1.5946859121322632, + "learning_rate": 5.094972067039106e-06, + "loss": 0.3208, + "step": 1523 + }, + { + "epoch": 4.256983240223463, + "grad_norm": 1.505895733833313, + "learning_rate": 5.098324022346369e-06, + "loss": 0.3498, + "step": 1524 + }, + { + "epoch": 4.259776536312849, + "grad_norm": 1.4467741250991821, + "learning_rate": 5.101675977653631e-06, + "loss": 0.3403, + "step": 1525 + }, + { + "epoch": 4.262569832402234, + "grad_norm": 8.925929069519043, + "learning_rate": 5.105027932960894e-06, + "loss": 0.2846, + "step": 1526 + }, + { + "epoch": 4.26536312849162, + "grad_norm": 1.339359998703003, + "learning_rate": 5.108379888268156e-06, + "loss": 0.3237, + "step": 1527 + }, + { + "epoch": 4.268156424581005, + "grad_norm": 2.28544282913208, + "learning_rate": 5.111731843575418e-06, + "loss": 0.3694, + "step": 1528 + }, + { + "epoch": 4.270949720670391, + "grad_norm": 4.504724502563477, + "learning_rate": 5.1150837988826814e-06, + "loss": 0.4471, + "step": 1529 + }, + { + "epoch": 4.273743016759776, + "grad_norm": 3.083725929260254, + "learning_rate": 5.118435754189944e-06, + "loss": 0.3536, + "step": 1530 + }, + { + "epoch": 4.276536312849162, + "grad_norm": 3.9613888263702393, + "learning_rate": 5.1217877094972075e-06, + "loss": 0.4729, + "step": 1531 + }, + { + "epoch": 4.279329608938547, + "grad_norm": 4.515844345092773, + "learning_rate": 5.12513966480447e-06, + "loss": 0.5014, + "step": 1532 + }, + { + "epoch": 4.282122905027933, + "grad_norm": 0.7878180742263794, + "learning_rate": 5.128491620111732e-06, + "loss": 0.304, + "step": 1533 + }, + { + "epoch": 4.284916201117318, + "grad_norm": 0.8707952499389648, + "learning_rate": 5.131843575418995e-06, + "loss": 0.2948, + "step": 1534 + }, + { + "epoch": 4.287709497206704, + "grad_norm": 1.2841767072677612, + "learning_rate": 5.135195530726257e-06, + "loss": 0.2833, + "step": 1535 + }, + { + "epoch": 4.290502793296089, + "grad_norm": 0.8695642948150635, + "learning_rate": 5.13854748603352e-06, + "loss": 0.2751, + "step": 1536 + }, + { + "epoch": 4.293296089385475, + "grad_norm": 1.3415955305099487, + "learning_rate": 5.141899441340782e-06, + "loss": 0.2753, + "step": 1537 + }, + { + "epoch": 4.29608938547486, + "grad_norm": 0.8643090128898621, + "learning_rate": 5.145251396648045e-06, + "loss": 0.2649, + "step": 1538 + }, + { + "epoch": 4.298882681564246, + "grad_norm": 0.9092831611633301, + "learning_rate": 5.148603351955308e-06, + "loss": 0.3086, + "step": 1539 + }, + { + "epoch": 4.301675977653631, + "grad_norm": 1.3636720180511475, + "learning_rate": 5.15195530726257e-06, + "loss": 0.2795, + "step": 1540 + }, + { + "epoch": 4.304469273743017, + "grad_norm": 1.4960674047470093, + "learning_rate": 5.155307262569833e-06, + "loss": 0.2895, + "step": 1541 + }, + { + "epoch": 4.307262569832402, + "grad_norm": 1.5171529054641724, + "learning_rate": 5.158659217877095e-06, + "loss": 0.3154, + "step": 1542 + }, + { + "epoch": 4.310055865921788, + "grad_norm": 1.3178282976150513, + "learning_rate": 5.162011173184358e-06, + "loss": 0.2843, + "step": 1543 + }, + { + "epoch": 4.312849162011173, + "grad_norm": 1.2845525741577148, + "learning_rate": 5.16536312849162e-06, + "loss": 0.2865, + "step": 1544 + }, + { + "epoch": 4.315642458100559, + "grad_norm": 1.256445050239563, + "learning_rate": 5.1687150837988824e-06, + "loss": 0.3245, + "step": 1545 + }, + { + "epoch": 4.318435754189944, + "grad_norm": 1.407440423965454, + "learning_rate": 5.1720670391061455e-06, + "loss": 0.2912, + "step": 1546 + }, + { + "epoch": 4.32122905027933, + "grad_norm": 1.281035304069519, + "learning_rate": 5.175418994413408e-06, + "loss": 0.3107, + "step": 1547 + }, + { + "epoch": 4.324022346368715, + "grad_norm": 1.9041616916656494, + "learning_rate": 5.178770949720671e-06, + "loss": 0.3095, + "step": 1548 + }, + { + "epoch": 4.326815642458101, + "grad_norm": 1.5764857530593872, + "learning_rate": 5.182122905027933e-06, + "loss": 0.3139, + "step": 1549 + }, + { + "epoch": 4.329608938547486, + "grad_norm": 4.560598850250244, + "learning_rate": 5.185474860335195e-06, + "loss": 0.3787, + "step": 1550 + }, + { + "epoch": 4.332402234636872, + "grad_norm": 3.5645463466644287, + "learning_rate": 5.188826815642458e-06, + "loss": 0.3577, + "step": 1551 + }, + { + "epoch": 4.335195530726257, + "grad_norm": 3.247612953186035, + "learning_rate": 5.19217877094972e-06, + "loss": 0.3474, + "step": 1552 + }, + { + "epoch": 4.337988826815643, + "grad_norm": 1.5145725011825562, + "learning_rate": 5.195530726256983e-06, + "loss": 0.3865, + "step": 1553 + }, + { + "epoch": 4.340782122905028, + "grad_norm": 1.6970195770263672, + "learning_rate": 5.1988826815642456e-06, + "loss": 0.3397, + "step": 1554 + }, + { + "epoch": 4.343575418994414, + "grad_norm": 3.8149614334106445, + "learning_rate": 5.202234636871509e-06, + "loss": 0.372, + "step": 1555 + }, + { + "epoch": 4.346368715083799, + "grad_norm": 5.783229351043701, + "learning_rate": 5.205586592178772e-06, + "loss": 0.4728, + "step": 1556 + }, + { + "epoch": 4.349162011173185, + "grad_norm": 4.281739711761475, + "learning_rate": 5.208938547486034e-06, + "loss": 0.5925, + "step": 1557 + }, + { + "epoch": 4.351955307262569, + "grad_norm": 1.0336763858795166, + "learning_rate": 5.212290502793297e-06, + "loss": 0.3126, + "step": 1558 + }, + { + "epoch": 4.354748603351955, + "grad_norm": 0.8853062391281128, + "learning_rate": 5.215642458100559e-06, + "loss": 0.2993, + "step": 1559 + }, + { + "epoch": 4.35754189944134, + "grad_norm": 0.9578348398208618, + "learning_rate": 5.218994413407821e-06, + "loss": 0.2504, + "step": 1560 + }, + { + "epoch": 4.360335195530726, + "grad_norm": 1.4860813617706299, + "learning_rate": 5.222346368715084e-06, + "loss": 0.2725, + "step": 1561 + }, + { + "epoch": 4.363128491620111, + "grad_norm": 1.0964237451553345, + "learning_rate": 5.2256983240223465e-06, + "loss": 0.328, + "step": 1562 + }, + { + "epoch": 4.365921787709497, + "grad_norm": 1.0044019222259521, + "learning_rate": 5.2290502793296095e-06, + "loss": 0.3088, + "step": 1563 + }, + { + "epoch": 4.368715083798882, + "grad_norm": 0.7566937804222107, + "learning_rate": 5.232402234636872e-06, + "loss": 0.2493, + "step": 1564 + }, + { + "epoch": 4.371508379888268, + "grad_norm": 0.8541187047958374, + "learning_rate": 5.235754189944134e-06, + "loss": 0.2502, + "step": 1565 + }, + { + "epoch": 4.374301675977653, + "grad_norm": 0.9311847686767578, + "learning_rate": 5.239106145251397e-06, + "loss": 0.2405, + "step": 1566 + }, + { + "epoch": 4.377094972067039, + "grad_norm": 1.022479772567749, + "learning_rate": 5.242458100558659e-06, + "loss": 0.2819, + "step": 1567 + }, + { + "epoch": 4.379888268156424, + "grad_norm": 1.0698456764221191, + "learning_rate": 5.245810055865922e-06, + "loss": 0.2728, + "step": 1568 + }, + { + "epoch": 4.38268156424581, + "grad_norm": 1.1081345081329346, + "learning_rate": 5.249162011173184e-06, + "loss": 0.3167, + "step": 1569 + }, + { + "epoch": 4.385474860335195, + "grad_norm": 1.3194606304168701, + "learning_rate": 5.2525139664804466e-06, + "loss": 0.2625, + "step": 1570 + }, + { + "epoch": 4.388268156424581, + "grad_norm": 1.3161207437515259, + "learning_rate": 5.25586592178771e-06, + "loss": 0.293, + "step": 1571 + }, + { + "epoch": 4.391061452513966, + "grad_norm": 1.5326688289642334, + "learning_rate": 5.259217877094972e-06, + "loss": 0.2908, + "step": 1572 + }, + { + "epoch": 4.393854748603352, + "grad_norm": 1.2458175420761108, + "learning_rate": 5.262569832402235e-06, + "loss": 0.332, + "step": 1573 + }, + { + "epoch": 4.396648044692737, + "grad_norm": 1.6765296459197998, + "learning_rate": 5.265921787709497e-06, + "loss": 0.2899, + "step": 1574 + }, + { + "epoch": 4.399441340782123, + "grad_norm": 1.7160992622375488, + "learning_rate": 5.269273743016759e-06, + "loss": 0.3716, + "step": 1575 + }, + { + "epoch": 4.402234636871508, + "grad_norm": 2.1460771560668945, + "learning_rate": 5.272625698324022e-06, + "loss": 0.3094, + "step": 1576 + }, + { + "epoch": 4.405027932960894, + "grad_norm": 1.4962223768234253, + "learning_rate": 5.2759776536312845e-06, + "loss": 0.3293, + "step": 1577 + }, + { + "epoch": 4.407821229050279, + "grad_norm": 1.6034629344940186, + "learning_rate": 5.2793296089385475e-06, + "loss": 0.3545, + "step": 1578 + }, + { + "epoch": 4.410614525139665, + "grad_norm": 2.717482805252075, + "learning_rate": 5.2826815642458105e-06, + "loss": 0.2985, + "step": 1579 + }, + { + "epoch": 4.41340782122905, + "grad_norm": 2.791872978210449, + "learning_rate": 5.286033519553073e-06, + "loss": 0.3696, + "step": 1580 + }, + { + "epoch": 4.416201117318436, + "grad_norm": 3.677290439605713, + "learning_rate": 5.289385474860336e-06, + "loss": 0.4232, + "step": 1581 + }, + { + "epoch": 4.418994413407821, + "grad_norm": 3.354666233062744, + "learning_rate": 5.292737430167598e-06, + "loss": 0.5212, + "step": 1582 + }, + { + "epoch": 4.421787709497207, + "grad_norm": 1.468185544013977, + "learning_rate": 5.296089385474861e-06, + "loss": 0.262, + "step": 1583 + }, + { + "epoch": 4.424581005586592, + "grad_norm": 1.0890363454818726, + "learning_rate": 5.299441340782123e-06, + "loss": 0.3026, + "step": 1584 + }, + { + "epoch": 4.427374301675978, + "grad_norm": 1.4341472387313843, + "learning_rate": 5.302793296089386e-06, + "loss": 0.3124, + "step": 1585 + }, + { + "epoch": 4.430167597765363, + "grad_norm": 1.0492998361587524, + "learning_rate": 5.306145251396648e-06, + "loss": 0.2781, + "step": 1586 + }, + { + "epoch": 4.432960893854749, + "grad_norm": 1.210907220840454, + "learning_rate": 5.309497206703911e-06, + "loss": 0.2921, + "step": 1587 + }, + { + "epoch": 4.435754189944134, + "grad_norm": 1.474865436553955, + "learning_rate": 5.312849162011174e-06, + "loss": 0.258, + "step": 1588 + }, + { + "epoch": 4.43854748603352, + "grad_norm": 1.3780678510665894, + "learning_rate": 5.316201117318436e-06, + "loss": 0.2491, + "step": 1589 + }, + { + "epoch": 4.441340782122905, + "grad_norm": 1.1540414094924927, + "learning_rate": 5.319553072625699e-06, + "loss": 0.2857, + "step": 1590 + }, + { + "epoch": 4.444134078212291, + "grad_norm": 1.573499083518982, + "learning_rate": 5.322905027932961e-06, + "loss": 0.2737, + "step": 1591 + }, + { + "epoch": 4.446927374301676, + "grad_norm": 1.1156741380691528, + "learning_rate": 5.326256983240223e-06, + "loss": 0.2956, + "step": 1592 + }, + { + "epoch": 4.449720670391061, + "grad_norm": 1.3208931684494019, + "learning_rate": 5.329608938547486e-06, + "loss": 0.2859, + "step": 1593 + }, + { + "epoch": 4.452513966480447, + "grad_norm": 1.0765372514724731, + "learning_rate": 5.3329608938547485e-06, + "loss": 0.2613, + "step": 1594 + }, + { + "epoch": 4.455307262569832, + "grad_norm": 1.2362526655197144, + "learning_rate": 5.3363128491620115e-06, + "loss": 0.3109, + "step": 1595 + }, + { + "epoch": 4.4581005586592175, + "grad_norm": 1.3804892301559448, + "learning_rate": 5.339664804469274e-06, + "loss": 0.3711, + "step": 1596 + }, + { + "epoch": 4.460893854748603, + "grad_norm": 1.1731045246124268, + "learning_rate": 5.343016759776536e-06, + "loss": 0.2692, + "step": 1597 + }, + { + "epoch": 4.4636871508379885, + "grad_norm": 1.4707483053207397, + "learning_rate": 5.346368715083799e-06, + "loss": 0.3249, + "step": 1598 + }, + { + "epoch": 4.466480446927374, + "grad_norm": 2.924112319946289, + "learning_rate": 5.349720670391061e-06, + "loss": 0.2715, + "step": 1599 + }, + { + "epoch": 4.4692737430167595, + "grad_norm": 1.5153597593307495, + "learning_rate": 5.353072625698324e-06, + "loss": 0.3376, + "step": 1600 + }, + { + "epoch": 4.472067039106145, + "grad_norm": 1.8851146697998047, + "learning_rate": 5.356424581005586e-06, + "loss": 0.3178, + "step": 1601 + }, + { + "epoch": 4.4748603351955305, + "grad_norm": 1.5888110399246216, + "learning_rate": 5.359776536312849e-06, + "loss": 0.3363, + "step": 1602 + }, + { + "epoch": 4.477653631284916, + "grad_norm": 2.179880380630493, + "learning_rate": 5.363128491620112e-06, + "loss": 0.3642, + "step": 1603 + }, + { + "epoch": 4.4804469273743015, + "grad_norm": 1.5621927976608276, + "learning_rate": 5.366480446927375e-06, + "loss": 0.3253, + "step": 1604 + }, + { + "epoch": 4.483240223463687, + "grad_norm": 2.7325499057769775, + "learning_rate": 5.369832402234638e-06, + "loss": 0.3608, + "step": 1605 + }, + { + "epoch": 4.4860335195530725, + "grad_norm": 4.6276774406433105, + "learning_rate": 5.3731843575419e-06, + "loss": 0.4207, + "step": 1606 + }, + { + "epoch": 4.488826815642458, + "grad_norm": 2.9533889293670654, + "learning_rate": 5.376536312849162e-06, + "loss": 0.4772, + "step": 1607 + }, + { + "epoch": 4.4916201117318435, + "grad_norm": 0.857501745223999, + "learning_rate": 5.379888268156425e-06, + "loss": 0.2975, + "step": 1608 + }, + { + "epoch": 4.494413407821229, + "grad_norm": 0.9046820402145386, + "learning_rate": 5.383240223463687e-06, + "loss": 0.2819, + "step": 1609 + }, + { + "epoch": 4.4972067039106145, + "grad_norm": 0.7863813042640686, + "learning_rate": 5.38659217877095e-06, + "loss": 0.2813, + "step": 1610 + }, + { + "epoch": 4.5, + "grad_norm": 8.41663646697998, + "learning_rate": 5.3899441340782125e-06, + "loss": 0.3115, + "step": 1611 + }, + { + "epoch": 4.5027932960893855, + "grad_norm": 0.9606806039810181, + "learning_rate": 5.393296089385475e-06, + "loss": 0.2702, + "step": 1612 + }, + { + "epoch": 4.505586592178771, + "grad_norm": 0.9988352060317993, + "learning_rate": 5.396648044692738e-06, + "loss": 0.2842, + "step": 1613 + }, + { + "epoch": 4.5083798882681565, + "grad_norm": 1.2612104415893555, + "learning_rate": 5.4e-06, + "loss": 0.2958, + "step": 1614 + }, + { + "epoch": 4.511173184357542, + "grad_norm": 0.978295624256134, + "learning_rate": 5.403351955307263e-06, + "loss": 0.3038, + "step": 1615 + }, + { + "epoch": 4.5139664804469275, + "grad_norm": 1.2410566806793213, + "learning_rate": 5.406703910614525e-06, + "loss": 0.2743, + "step": 1616 + }, + { + "epoch": 4.516759776536313, + "grad_norm": 1.0406793355941772, + "learning_rate": 5.410055865921787e-06, + "loss": 0.2884, + "step": 1617 + }, + { + "epoch": 4.5195530726256985, + "grad_norm": 1.1533211469650269, + "learning_rate": 5.4134078212290504e-06, + "loss": 0.3323, + "step": 1618 + }, + { + "epoch": 4.522346368715084, + "grad_norm": 1.148693561553955, + "learning_rate": 5.416759776536313e-06, + "loss": 0.2798, + "step": 1619 + }, + { + "epoch": 4.5251396648044695, + "grad_norm": 2.592644214630127, + "learning_rate": 5.420111731843576e-06, + "loss": 0.3138, + "step": 1620 + }, + { + "epoch": 4.527932960893855, + "grad_norm": 1.7396408319473267, + "learning_rate": 5.423463687150838e-06, + "loss": 0.3445, + "step": 1621 + }, + { + "epoch": 4.5307262569832405, + "grad_norm": 1.4195936918258667, + "learning_rate": 5.4268156424581e-06, + "loss": 0.2829, + "step": 1622 + }, + { + "epoch": 4.533519553072626, + "grad_norm": 1.9928443431854248, + "learning_rate": 5.430167597765363e-06, + "loss": 0.2605, + "step": 1623 + }, + { + "epoch": 4.5363128491620115, + "grad_norm": 7.171860218048096, + "learning_rate": 5.433519553072625e-06, + "loss": 0.3154, + "step": 1624 + }, + { + "epoch": 4.539106145251397, + "grad_norm": 4.764532089233398, + "learning_rate": 5.436871508379888e-06, + "loss": 0.3188, + "step": 1625 + }, + { + "epoch": 4.5418994413407825, + "grad_norm": 2.70971941947937, + "learning_rate": 5.4402234636871505e-06, + "loss": 0.3515, + "step": 1626 + }, + { + "epoch": 4.544692737430168, + "grad_norm": 3.047136068344116, + "learning_rate": 5.4435754189944135e-06, + "loss": 0.3397, + "step": 1627 + }, + { + "epoch": 4.547486033519553, + "grad_norm": 1.9845876693725586, + "learning_rate": 5.446927374301677e-06, + "loss": 0.3041, + "step": 1628 + }, + { + "epoch": 4.550279329608939, + "grad_norm": 2.8338406085968018, + "learning_rate": 5.450279329608939e-06, + "loss": 0.3528, + "step": 1629 + }, + { + "epoch": 4.553072625698324, + "grad_norm": 2.6183621883392334, + "learning_rate": 5.453631284916202e-06, + "loss": 0.4293, + "step": 1630 + }, + { + "epoch": 4.55586592178771, + "grad_norm": 4.303993225097656, + "learning_rate": 5.456983240223464e-06, + "loss": 0.4418, + "step": 1631 + }, + { + "epoch": 4.558659217877095, + "grad_norm": 3.164017915725708, + "learning_rate": 5.460335195530727e-06, + "loss": 0.5525, + "step": 1632 + }, + { + "epoch": 4.56145251396648, + "grad_norm": 1.0821175575256348, + "learning_rate": 5.463687150837989e-06, + "loss": 0.3091, + "step": 1633 + }, + { + "epoch": 4.564245810055866, + "grad_norm": 0.7574735283851624, + "learning_rate": 5.4670391061452514e-06, + "loss": 0.2572, + "step": 1634 + }, + { + "epoch": 4.567039106145251, + "grad_norm": 0.9690719246864319, + "learning_rate": 5.4703910614525145e-06, + "loss": 0.2532, + "step": 1635 + }, + { + "epoch": 4.569832402234637, + "grad_norm": 0.8309781551361084, + "learning_rate": 5.473743016759777e-06, + "loss": 0.277, + "step": 1636 + }, + { + "epoch": 4.572625698324022, + "grad_norm": 1.895836353302002, + "learning_rate": 5.47709497206704e-06, + "loss": 0.2934, + "step": 1637 + }, + { + "epoch": 4.575418994413408, + "grad_norm": 1.1264020204544067, + "learning_rate": 5.480446927374302e-06, + "loss": 0.289, + "step": 1638 + }, + { + "epoch": 4.578212290502793, + "grad_norm": 1.6840568780899048, + "learning_rate": 5.483798882681564e-06, + "loss": 0.2861, + "step": 1639 + }, + { + "epoch": 4.581005586592179, + "grad_norm": 1.4028780460357666, + "learning_rate": 5.487150837988827e-06, + "loss": 0.2837, + "step": 1640 + }, + { + "epoch": 4.583798882681564, + "grad_norm": 0.9129101037979126, + "learning_rate": 5.490502793296089e-06, + "loss": 0.2731, + "step": 1641 + }, + { + "epoch": 4.58659217877095, + "grad_norm": 1.094743013381958, + "learning_rate": 5.493854748603352e-06, + "loss": 0.3192, + "step": 1642 + }, + { + "epoch": 4.589385474860335, + "grad_norm": 1.05113685131073, + "learning_rate": 5.4972067039106146e-06, + "loss": 0.3095, + "step": 1643 + }, + { + "epoch": 4.592178770949721, + "grad_norm": 0.9239780902862549, + "learning_rate": 5.500558659217877e-06, + "loss": 0.2915, + "step": 1644 + }, + { + "epoch": 4.594972067039106, + "grad_norm": 2.0118825435638428, + "learning_rate": 5.50391061452514e-06, + "loss": 0.3196, + "step": 1645 + }, + { + "epoch": 4.597765363128492, + "grad_norm": 1.6237080097198486, + "learning_rate": 5.507262569832402e-06, + "loss": 0.312, + "step": 1646 + }, + { + "epoch": 4.600558659217877, + "grad_norm": 1.2120535373687744, + "learning_rate": 5.510614525139665e-06, + "loss": 0.2984, + "step": 1647 + }, + { + "epoch": 4.603351955307263, + "grad_norm": 1.2273308038711548, + "learning_rate": 5.513966480446927e-06, + "loss": 0.3048, + "step": 1648 + }, + { + "epoch": 4.606145251396648, + "grad_norm": 2.7366268634796143, + "learning_rate": 5.517318435754189e-06, + "loss": 0.3112, + "step": 1649 + }, + { + "epoch": 4.608938547486034, + "grad_norm": 1.8701125383377075, + "learning_rate": 5.5206703910614524e-06, + "loss": 0.2988, + "step": 1650 + }, + { + "epoch": 4.611731843575419, + "grad_norm": 1.19278085231781, + "learning_rate": 5.5240223463687155e-06, + "loss": 0.328, + "step": 1651 + }, + { + "epoch": 4.614525139664805, + "grad_norm": 21.988941192626953, + "learning_rate": 5.5273743016759785e-06, + "loss": 0.3415, + "step": 1652 + }, + { + "epoch": 4.61731843575419, + "grad_norm": 1.9260237216949463, + "learning_rate": 5.530726256983241e-06, + "loss": 0.3527, + "step": 1653 + }, + { + "epoch": 4.620111731843576, + "grad_norm": 1.4782971143722534, + "learning_rate": 5.534078212290503e-06, + "loss": 0.3477, + "step": 1654 + }, + { + "epoch": 4.622905027932961, + "grad_norm": 3.425361394882202, + "learning_rate": 5.537430167597766e-06, + "loss": 0.3865, + "step": 1655 + }, + { + "epoch": 4.625698324022347, + "grad_norm": 2.766385793685913, + "learning_rate": 5.540782122905028e-06, + "loss": 0.3308, + "step": 1656 + }, + { + "epoch": 4.628491620111732, + "grad_norm": 3.199098825454712, + "learning_rate": 5.544134078212291e-06, + "loss": 0.5018, + "step": 1657 + }, + { + "epoch": 4.631284916201118, + "grad_norm": 0.8875945806503296, + "learning_rate": 5.547486033519553e-06, + "loss": 0.2867, + "step": 1658 + }, + { + "epoch": 4.634078212290503, + "grad_norm": 0.7914174199104309, + "learning_rate": 5.5508379888268156e-06, + "loss": 0.2871, + "step": 1659 + }, + { + "epoch": 4.636871508379889, + "grad_norm": 0.922480583190918, + "learning_rate": 5.554189944134079e-06, + "loss": 0.2674, + "step": 1660 + }, + { + "epoch": 4.639664804469274, + "grad_norm": 0.8805686235427856, + "learning_rate": 5.557541899441341e-06, + "loss": 0.3006, + "step": 1661 + }, + { + "epoch": 4.64245810055866, + "grad_norm": 0.9327067136764526, + "learning_rate": 5.560893854748604e-06, + "loss": 0.2592, + "step": 1662 + }, + { + "epoch": 4.645251396648044, + "grad_norm": 3.246739625930786, + "learning_rate": 5.564245810055866e-06, + "loss": 0.283, + "step": 1663 + }, + { + "epoch": 4.648044692737431, + "grad_norm": 1.0465214252471924, + "learning_rate": 5.567597765363128e-06, + "loss": 0.2565, + "step": 1664 + }, + { + "epoch": 4.650837988826815, + "grad_norm": 0.9519761204719543, + "learning_rate": 5.570949720670391e-06, + "loss": 0.3246, + "step": 1665 + }, + { + "epoch": 4.653631284916202, + "grad_norm": 1.0139083862304688, + "learning_rate": 5.5743016759776534e-06, + "loss": 0.3017, + "step": 1666 + }, + { + "epoch": 4.656424581005586, + "grad_norm": 1.680413842201233, + "learning_rate": 5.5776536312849165e-06, + "loss": 0.3106, + "step": 1667 + }, + { + "epoch": 4.659217877094972, + "grad_norm": 1.0927317142486572, + "learning_rate": 5.581005586592179e-06, + "loss": 0.3416, + "step": 1668 + }, + { + "epoch": 4.662011173184357, + "grad_norm": 1.409177303314209, + "learning_rate": 5.584357541899441e-06, + "loss": 0.3156, + "step": 1669 + }, + { + "epoch": 4.664804469273743, + "grad_norm": 1.9920543432235718, + "learning_rate": 5.587709497206704e-06, + "loss": 0.2881, + "step": 1670 + }, + { + "epoch": 4.667597765363128, + "grad_norm": 1.2798773050308228, + "learning_rate": 5.591061452513966e-06, + "loss": 0.2963, + "step": 1671 + }, + { + "epoch": 4.670391061452514, + "grad_norm": 1.5273224115371704, + "learning_rate": 5.594413407821229e-06, + "loss": 0.2996, + "step": 1672 + }, + { + "epoch": 4.673184357541899, + "grad_norm": 1.7455084323883057, + "learning_rate": 5.597765363128491e-06, + "loss": 0.3594, + "step": 1673 + }, + { + "epoch": 4.675977653631285, + "grad_norm": 1.4745991230010986, + "learning_rate": 5.6011173184357535e-06, + "loss": 0.3221, + "step": 1674 + }, + { + "epoch": 4.67877094972067, + "grad_norm": 3.5801167488098145, + "learning_rate": 5.6044692737430166e-06, + "loss": 0.3414, + "step": 1675 + }, + { + "epoch": 4.681564245810056, + "grad_norm": 1.4924659729003906, + "learning_rate": 5.60782122905028e-06, + "loss": 0.3124, + "step": 1676 + }, + { + "epoch": 4.684357541899441, + "grad_norm": 1.3395344018936157, + "learning_rate": 5.611173184357543e-06, + "loss": 0.3215, + "step": 1677 + }, + { + "epoch": 4.687150837988827, + "grad_norm": 1.837404489517212, + "learning_rate": 5.614525139664805e-06, + "loss": 0.3186, + "step": 1678 + }, + { + "epoch": 4.689944134078212, + "grad_norm": 1.6547943353652954, + "learning_rate": 5.617877094972068e-06, + "loss": 0.3238, + "step": 1679 + }, + { + "epoch": 4.692737430167598, + "grad_norm": 1.941409945487976, + "learning_rate": 5.62122905027933e-06, + "loss": 0.3213, + "step": 1680 + }, + { + "epoch": 4.695530726256983, + "grad_norm": 2.4112207889556885, + "learning_rate": 5.624581005586592e-06, + "loss": 0.4041, + "step": 1681 + }, + { + "epoch": 4.698324022346369, + "grad_norm": 2.9815597534179688, + "learning_rate": 5.627932960893855e-06, + "loss": 0.5663, + "step": 1682 + }, + { + "epoch": 4.701117318435754, + "grad_norm": 0.6422210335731506, + "learning_rate": 5.6312849162011175e-06, + "loss": 0.2649, + "step": 1683 + }, + { + "epoch": 4.70391061452514, + "grad_norm": 2.1800665855407715, + "learning_rate": 5.6346368715083805e-06, + "loss": 0.2623, + "step": 1684 + }, + { + "epoch": 4.706703910614525, + "grad_norm": 1.052063226699829, + "learning_rate": 5.637988826815643e-06, + "loss": 0.2864, + "step": 1685 + }, + { + "epoch": 4.709497206703911, + "grad_norm": 0.8951178789138794, + "learning_rate": 5.641340782122905e-06, + "loss": 0.25, + "step": 1686 + }, + { + "epoch": 4.712290502793296, + "grad_norm": 1.1276708841323853, + "learning_rate": 5.644692737430168e-06, + "loss": 0.2814, + "step": 1687 + }, + { + "epoch": 4.715083798882682, + "grad_norm": 1.1869267225265503, + "learning_rate": 5.64804469273743e-06, + "loss": 0.3223, + "step": 1688 + }, + { + "epoch": 4.717877094972067, + "grad_norm": 0.8003553748130798, + "learning_rate": 5.651396648044693e-06, + "loss": 0.313, + "step": 1689 + }, + { + "epoch": 4.720670391061453, + "grad_norm": 1.2638916969299316, + "learning_rate": 5.654748603351955e-06, + "loss": 0.2703, + "step": 1690 + }, + { + "epoch": 4.723463687150838, + "grad_norm": 1.6387133598327637, + "learning_rate": 5.6581005586592176e-06, + "loss": 0.3059, + "step": 1691 + }, + { + "epoch": 4.726256983240224, + "grad_norm": 1.5128118991851807, + "learning_rate": 5.661452513966481e-06, + "loss": 0.3147, + "step": 1692 + }, + { + "epoch": 4.729050279329609, + "grad_norm": 1.1124974489212036, + "learning_rate": 5.664804469273743e-06, + "loss": 0.2528, + "step": 1693 + }, + { + "epoch": 4.731843575418995, + "grad_norm": 1.7912098169326782, + "learning_rate": 5.668156424581006e-06, + "loss": 0.3394, + "step": 1694 + }, + { + "epoch": 4.73463687150838, + "grad_norm": 2.289757013320923, + "learning_rate": 5.671508379888268e-06, + "loss": 0.2556, + "step": 1695 + }, + { + "epoch": 4.737430167597766, + "grad_norm": 1.2847394943237305, + "learning_rate": 5.67486033519553e-06, + "loss": 0.3521, + "step": 1696 + }, + { + "epoch": 4.740223463687151, + "grad_norm": 1.217038631439209, + "learning_rate": 5.678212290502793e-06, + "loss": 0.2631, + "step": 1697 + }, + { + "epoch": 4.743016759776537, + "grad_norm": 1.273397445678711, + "learning_rate": 5.6815642458100555e-06, + "loss": 0.3291, + "step": 1698 + }, + { + "epoch": 4.745810055865922, + "grad_norm": 1.1548024415969849, + "learning_rate": 5.6849162011173185e-06, + "loss": 0.2681, + "step": 1699 + }, + { + "epoch": 4.748603351955307, + "grad_norm": 1.596364974975586, + "learning_rate": 5.6882681564245815e-06, + "loss": 0.3225, + "step": 1700 + }, + { + "epoch": 4.751396648044693, + "grad_norm": 1.8989055156707764, + "learning_rate": 5.691620111731844e-06, + "loss": 0.3521, + "step": 1701 + }, + { + "epoch": 4.754189944134078, + "grad_norm": 1.7660237550735474, + "learning_rate": 5.694972067039107e-06, + "loss": 0.3656, + "step": 1702 + }, + { + "epoch": 4.756983240223463, + "grad_norm": 1.6747233867645264, + "learning_rate": 5.698324022346369e-06, + "loss": 0.3102, + "step": 1703 + }, + { + "epoch": 4.759776536312849, + "grad_norm": 2.096008062362671, + "learning_rate": 5.701675977653632e-06, + "loss": 0.3262, + "step": 1704 + }, + { + "epoch": 4.762569832402234, + "grad_norm": 1.9695038795471191, + "learning_rate": 5.705027932960894e-06, + "loss": 0.3631, + "step": 1705 + }, + { + "epoch": 4.76536312849162, + "grad_norm": 1.7403218746185303, + "learning_rate": 5.708379888268156e-06, + "loss": 0.3738, + "step": 1706 + }, + { + "epoch": 4.768156424581005, + "grad_norm": 4.977904796600342, + "learning_rate": 5.7117318435754194e-06, + "loss": 0.5622, + "step": 1707 + }, + { + "epoch": 4.770949720670391, + "grad_norm": 1.0284661054611206, + "learning_rate": 5.715083798882682e-06, + "loss": 0.3309, + "step": 1708 + }, + { + "epoch": 4.773743016759776, + "grad_norm": 1.6932798624038696, + "learning_rate": 5.718435754189945e-06, + "loss": 0.2603, + "step": 1709 + }, + { + "epoch": 4.776536312849162, + "grad_norm": 1.7542182207107544, + "learning_rate": 5.721787709497207e-06, + "loss": 0.3003, + "step": 1710 + }, + { + "epoch": 4.779329608938547, + "grad_norm": 1.2171169519424438, + "learning_rate": 5.725139664804469e-06, + "loss": 0.2638, + "step": 1711 + }, + { + "epoch": 4.782122905027933, + "grad_norm": 1.9779964685440063, + "learning_rate": 5.728491620111732e-06, + "loss": 0.2477, + "step": 1712 + }, + { + "epoch": 4.784916201117318, + "grad_norm": 1.023616909980774, + "learning_rate": 5.731843575418994e-06, + "loss": 0.2777, + "step": 1713 + }, + { + "epoch": 4.787709497206704, + "grad_norm": 0.9917799234390259, + "learning_rate": 5.735195530726257e-06, + "loss": 0.3007, + "step": 1714 + }, + { + "epoch": 4.790502793296089, + "grad_norm": 0.7936597466468811, + "learning_rate": 5.7385474860335195e-06, + "loss": 0.2702, + "step": 1715 + }, + { + "epoch": 4.793296089385475, + "grad_norm": 1.1395992040634155, + "learning_rate": 5.741899441340782e-06, + "loss": 0.2918, + "step": 1716 + }, + { + "epoch": 4.79608938547486, + "grad_norm": 1.5889990329742432, + "learning_rate": 5.745251396648045e-06, + "loss": 0.2631, + "step": 1717 + }, + { + "epoch": 4.798882681564246, + "grad_norm": 1.457538366317749, + "learning_rate": 5.748603351955307e-06, + "loss": 0.3238, + "step": 1718 + }, + { + "epoch": 4.801675977653631, + "grad_norm": 1.4057542085647583, + "learning_rate": 5.75195530726257e-06, + "loss": 0.3094, + "step": 1719 + }, + { + "epoch": 4.804469273743017, + "grad_norm": 1.1749536991119385, + "learning_rate": 5.755307262569832e-06, + "loss": 0.3024, + "step": 1720 + }, + { + "epoch": 4.807262569832402, + "grad_norm": 5.724274635314941, + "learning_rate": 5.758659217877094e-06, + "loss": 0.2812, + "step": 1721 + }, + { + "epoch": 4.810055865921788, + "grad_norm": 1.1219087839126587, + "learning_rate": 5.762011173184357e-06, + "loss": 0.2801, + "step": 1722 + }, + { + "epoch": 4.812849162011173, + "grad_norm": 1.4101289510726929, + "learning_rate": 5.76536312849162e-06, + "loss": 0.2996, + "step": 1723 + }, + { + "epoch": 4.815642458100559, + "grad_norm": 1.1659393310546875, + "learning_rate": 5.7687150837988835e-06, + "loss": 0.3147, + "step": 1724 + }, + { + "epoch": 4.818435754189944, + "grad_norm": 1.2521016597747803, + "learning_rate": 5.772067039106146e-06, + "loss": 0.2679, + "step": 1725 + }, + { + "epoch": 4.82122905027933, + "grad_norm": 5.823552131652832, + "learning_rate": 5.775418994413409e-06, + "loss": 0.3143, + "step": 1726 + }, + { + "epoch": 4.824022346368715, + "grad_norm": 1.599360704421997, + "learning_rate": 5.778770949720671e-06, + "loss": 0.4052, + "step": 1727 + }, + { + "epoch": 4.826815642458101, + "grad_norm": 2.6810944080352783, + "learning_rate": 5.782122905027933e-06, + "loss": 0.3409, + "step": 1728 + }, + { + "epoch": 4.829608938547486, + "grad_norm": 1.5446301698684692, + "learning_rate": 5.785474860335196e-06, + "loss": 0.3604, + "step": 1729 + }, + { + "epoch": 4.832402234636872, + "grad_norm": 10.702712059020996, + "learning_rate": 5.788826815642458e-06, + "loss": 0.3989, + "step": 1730 + }, + { + "epoch": 4.835195530726257, + "grad_norm": 2.136373519897461, + "learning_rate": 5.792178770949721e-06, + "loss": 0.3921, + "step": 1731 + }, + { + "epoch": 4.837988826815643, + "grad_norm": 3.0789315700531006, + "learning_rate": 5.7955307262569836e-06, + "loss": 0.5398, + "step": 1732 + }, + { + "epoch": 4.840782122905028, + "grad_norm": 1.1233313083648682, + "learning_rate": 5.798882681564246e-06, + "loss": 0.2937, + "step": 1733 + }, + { + "epoch": 4.843575418994414, + "grad_norm": 0.8784751296043396, + "learning_rate": 5.802234636871509e-06, + "loss": 0.2578, + "step": 1734 + }, + { + "epoch": 4.846368715083798, + "grad_norm": 0.6654068827629089, + "learning_rate": 5.805586592178771e-06, + "loss": 0.2447, + "step": 1735 + }, + { + "epoch": 4.849162011173185, + "grad_norm": 1.6538870334625244, + "learning_rate": 5.808938547486034e-06, + "loss": 0.2556, + "step": 1736 + }, + { + "epoch": 4.851955307262569, + "grad_norm": 0.9368259906768799, + "learning_rate": 5.812290502793296e-06, + "loss": 0.2881, + "step": 1737 + }, + { + "epoch": 4.854748603351956, + "grad_norm": 1.725964903831482, + "learning_rate": 5.815642458100558e-06, + "loss": 0.2971, + "step": 1738 + }, + { + "epoch": 4.85754189944134, + "grad_norm": 5.4336137771606445, + "learning_rate": 5.8189944134078214e-06, + "loss": 0.2668, + "step": 1739 + }, + { + "epoch": 4.860335195530726, + "grad_norm": 0.8479229211807251, + "learning_rate": 5.822346368715084e-06, + "loss": 0.2853, + "step": 1740 + }, + { + "epoch": 4.863128491620111, + "grad_norm": 1.2538812160491943, + "learning_rate": 5.825698324022347e-06, + "loss": 0.2615, + "step": 1741 + }, + { + "epoch": 4.865921787709497, + "grad_norm": 3.66921067237854, + "learning_rate": 5.829050279329609e-06, + "loss": 0.2767, + "step": 1742 + }, + { + "epoch": 4.868715083798882, + "grad_norm": 1.1578377485275269, + "learning_rate": 5.832402234636871e-06, + "loss": 0.2882, + "step": 1743 + }, + { + "epoch": 4.871508379888268, + "grad_norm": 2.2936933040618896, + "learning_rate": 5.835754189944134e-06, + "loss": 0.2603, + "step": 1744 + }, + { + "epoch": 4.874301675977653, + "grad_norm": 1.2245551347732544, + "learning_rate": 5.839106145251396e-06, + "loss": 0.2558, + "step": 1745 + }, + { + "epoch": 4.877094972067039, + "grad_norm": 2.2031617164611816, + "learning_rate": 5.842458100558659e-06, + "loss": 0.307, + "step": 1746 + }, + { + "epoch": 4.879888268156424, + "grad_norm": 2.274758815765381, + "learning_rate": 5.8458100558659215e-06, + "loss": 0.2893, + "step": 1747 + }, + { + "epoch": 4.88268156424581, + "grad_norm": 1.0624945163726807, + "learning_rate": 5.8491620111731846e-06, + "loss": 0.2809, + "step": 1748 + }, + { + "epoch": 4.885474860335195, + "grad_norm": 1.8155237436294556, + "learning_rate": 5.852513966480448e-06, + "loss": 0.3617, + "step": 1749 + }, + { + "epoch": 4.888268156424581, + "grad_norm": 1.9903337955474854, + "learning_rate": 5.85586592178771e-06, + "loss": 0.3026, + "step": 1750 + }, + { + "epoch": 4.891061452513966, + "grad_norm": 1.5544184446334839, + "learning_rate": 5.859217877094973e-06, + "loss": 0.3589, + "step": 1751 + }, + { + "epoch": 4.893854748603352, + "grad_norm": 2.0307083129882812, + "learning_rate": 5.862569832402235e-06, + "loss": 0.326, + "step": 1752 + }, + { + "epoch": 4.896648044692737, + "grad_norm": 1.7254600524902344, + "learning_rate": 5.865921787709497e-06, + "loss": 0.3595, + "step": 1753 + }, + { + "epoch": 4.899441340782123, + "grad_norm": 2.2822396755218506, + "learning_rate": 5.86927374301676e-06, + "loss": 0.352, + "step": 1754 + }, + { + "epoch": 4.902234636871508, + "grad_norm": 1.8167017698287964, + "learning_rate": 5.8726256983240224e-06, + "loss": 0.3547, + "step": 1755 + }, + { + "epoch": 4.905027932960894, + "grad_norm": 2.465867280960083, + "learning_rate": 5.8759776536312855e-06, + "loss": 0.4391, + "step": 1756 + }, + { + "epoch": 4.907821229050279, + "grad_norm": 2.6938037872314453, + "learning_rate": 5.879329608938548e-06, + "loss": 0.4993, + "step": 1757 + }, + { + "epoch": 4.910614525139665, + "grad_norm": 1.0118259191513062, + "learning_rate": 5.88268156424581e-06, + "loss": 0.2509, + "step": 1758 + }, + { + "epoch": 4.91340782122905, + "grad_norm": 0.9897668957710266, + "learning_rate": 5.886033519553073e-06, + "loss": 0.283, + "step": 1759 + }, + { + "epoch": 4.916201117318436, + "grad_norm": 1.288469672203064, + "learning_rate": 5.889385474860335e-06, + "loss": 0.2991, + "step": 1760 + }, + { + "epoch": 4.918994413407821, + "grad_norm": 1.0939676761627197, + "learning_rate": 5.892737430167598e-06, + "loss": 0.2713, + "step": 1761 + }, + { + "epoch": 4.921787709497207, + "grad_norm": 1.3626909255981445, + "learning_rate": 5.89608938547486e-06, + "loss": 0.3298, + "step": 1762 + }, + { + "epoch": 4.924581005586592, + "grad_norm": 0.7375345826148987, + "learning_rate": 5.8994413407821225e-06, + "loss": 0.279, + "step": 1763 + }, + { + "epoch": 4.927374301675978, + "grad_norm": 1.1015418767929077, + "learning_rate": 5.9027932960893856e-06, + "loss": 0.2925, + "step": 1764 + }, + { + "epoch": 4.930167597765363, + "grad_norm": 0.9700621962547302, + "learning_rate": 5.906145251396648e-06, + "loss": 0.3146, + "step": 1765 + }, + { + "epoch": 4.932960893854749, + "grad_norm": 0.9800953269004822, + "learning_rate": 5.909497206703911e-06, + "loss": 0.2622, + "step": 1766 + }, + { + "epoch": 4.935754189944134, + "grad_norm": 1.1149142980575562, + "learning_rate": 5.912849162011173e-06, + "loss": 0.2908, + "step": 1767 + }, + { + "epoch": 4.93854748603352, + "grad_norm": 1.248996615409851, + "learning_rate": 5.916201117318435e-06, + "loss": 0.3304, + "step": 1768 + }, + { + "epoch": 4.941340782122905, + "grad_norm": 1.439240574836731, + "learning_rate": 5.919553072625698e-06, + "loss": 0.2826, + "step": 1769 + }, + { + "epoch": 4.94413407821229, + "grad_norm": 1.1775747537612915, + "learning_rate": 5.92290502793296e-06, + "loss": 0.2927, + "step": 1770 + }, + { + "epoch": 4.946927374301676, + "grad_norm": 1.345774531364441, + "learning_rate": 5.9262569832402235e-06, + "loss": 0.3084, + "step": 1771 + }, + { + "epoch": 4.949720670391061, + "grad_norm": 1.1216381788253784, + "learning_rate": 5.9296089385474865e-06, + "loss": 0.2876, + "step": 1772 + }, + { + "epoch": 4.952513966480447, + "grad_norm": 1.0626263618469238, + "learning_rate": 5.9329608938547495e-06, + "loss": 0.3001, + "step": 1773 + }, + { + "epoch": 4.955307262569832, + "grad_norm": 1.5037040710449219, + "learning_rate": 5.936312849162012e-06, + "loss": 0.3557, + "step": 1774 + }, + { + "epoch": 4.9581005586592175, + "grad_norm": 1.5268263816833496, + "learning_rate": 5.939664804469274e-06, + "loss": 0.309, + "step": 1775 + }, + { + "epoch": 4.960893854748603, + "grad_norm": 1.465101718902588, + "learning_rate": 5.943016759776537e-06, + "loss": 0.3181, + "step": 1776 + }, + { + "epoch": 4.9636871508379885, + "grad_norm": 1.3725422620773315, + "learning_rate": 5.946368715083799e-06, + "loss": 0.31, + "step": 1777 + }, + { + "epoch": 4.966480446927374, + "grad_norm": 1.2763099670410156, + "learning_rate": 5.949720670391062e-06, + "loss": 0.3324, + "step": 1778 + }, + { + "epoch": 4.9692737430167595, + "grad_norm": 3.9610350131988525, + "learning_rate": 5.953072625698324e-06, + "loss": 0.3549, + "step": 1779 + }, + { + "epoch": 4.972067039106145, + "grad_norm": 3.469139575958252, + "learning_rate": 5.9564245810055866e-06, + "loss": 0.3537, + "step": 1780 + }, + { + "epoch": 4.9748603351955305, + "grad_norm": 2.3919594287872314, + "learning_rate": 5.95977653631285e-06, + "loss": 0.4372, + "step": 1781 + }, + { + "epoch": 4.977653631284916, + "grad_norm": 4.204289436340332, + "learning_rate": 5.963128491620112e-06, + "loss": 0.4941, + "step": 1782 + }, + { + "epoch": 4.9804469273743015, + "grad_norm": 0.7141598463058472, + "learning_rate": 5.966480446927375e-06, + "loss": 0.2702, + "step": 1783 + }, + { + "epoch": 4.983240223463687, + "grad_norm": 1.0259298086166382, + "learning_rate": 5.969832402234637e-06, + "loss": 0.3296, + "step": 1784 + }, + { + "epoch": 4.9860335195530725, + "grad_norm": 0.7820119261741638, + "learning_rate": 5.973184357541899e-06, + "loss": 0.2556, + "step": 1785 + }, + { + "epoch": 4.988826815642458, + "grad_norm": 1.4178752899169922, + "learning_rate": 5.976536312849162e-06, + "loss": 0.3503, + "step": 1786 + }, + { + "epoch": 4.9916201117318435, + "grad_norm": 1.2120929956436157, + "learning_rate": 5.9798882681564245e-06, + "loss": 0.2539, + "step": 1787 + }, + { + "epoch": 4.994413407821229, + "grad_norm": 1.3807432651519775, + "learning_rate": 5.9832402234636875e-06, + "loss": 0.2591, + "step": 1788 + }, + { + "epoch": 4.9972067039106145, + "grad_norm": 1.454654335975647, + "learning_rate": 5.98659217877095e-06, + "loss": 0.3339, + "step": 1789 + }, + { + "epoch": 5.0, + "grad_norm": 2.088444471359253, + "learning_rate": 5.989944134078212e-06, + "loss": 0.4254, + "step": 1790 + }, + { + "epoch": 5.0027932960893855, + "grad_norm": 1.2672319412231445, + "learning_rate": 5.993296089385475e-06, + "loss": 0.2961, + "step": 1791 + }, + { + "epoch": 5.005586592178771, + "grad_norm": 1.2128766775131226, + "learning_rate": 5.996648044692737e-06, + "loss": 0.3071, + "step": 1792 + }, + { + "epoch": 5.0083798882681565, + "grad_norm": 0.7507612109184265, + "learning_rate": 6e-06, + "loss": 0.2974, + "step": 1793 + }, + { + "epoch": 5.011173184357542, + "grad_norm": 0.8452123999595642, + "learning_rate": 6.003351955307262e-06, + "loss": 0.2501, + "step": 1794 + }, + { + "epoch": 5.0139664804469275, + "grad_norm": 0.9172201752662659, + "learning_rate": 6.0067039106145245e-06, + "loss": 0.2794, + "step": 1795 + }, + { + "epoch": 5.016759776536313, + "grad_norm": 0.7636535167694092, + "learning_rate": 6.010055865921788e-06, + "loss": 0.3021, + "step": 1796 + }, + { + "epoch": 5.0195530726256985, + "grad_norm": 0.7818847298622131, + "learning_rate": 6.013407821229051e-06, + "loss": 0.2655, + "step": 1797 + }, + { + "epoch": 5.022346368715084, + "grad_norm": 0.9756902456283569, + "learning_rate": 6.016759776536314e-06, + "loss": 0.3119, + "step": 1798 + }, + { + "epoch": 5.0251396648044695, + "grad_norm": 1.1357636451721191, + "learning_rate": 6.020111731843576e-06, + "loss": 0.2611, + "step": 1799 + }, + { + "epoch": 5.027932960893855, + "grad_norm": 1.320099115371704, + "learning_rate": 6.023463687150838e-06, + "loss": 0.2743, + "step": 1800 + }, + { + "epoch": 5.0307262569832405, + "grad_norm": 1.0840080976486206, + "learning_rate": 6.026815642458101e-06, + "loss": 0.3098, + "step": 1801 + }, + { + "epoch": 5.033519553072626, + "grad_norm": 1.2945151329040527, + "learning_rate": 6.030167597765363e-06, + "loss": 0.2831, + "step": 1802 + }, + { + "epoch": 5.0363128491620115, + "grad_norm": 0.9378376603126526, + "learning_rate": 6.033519553072626e-06, + "loss": 0.2786, + "step": 1803 + }, + { + "epoch": 5.039106145251397, + "grad_norm": 1.179083228111267, + "learning_rate": 6.0368715083798885e-06, + "loss": 0.2834, + "step": 1804 + }, + { + "epoch": 5.0418994413407825, + "grad_norm": 1.059796690940857, + "learning_rate": 6.040223463687151e-06, + "loss": 0.2699, + "step": 1805 + }, + { + "epoch": 5.044692737430168, + "grad_norm": 1.3870470523834229, + "learning_rate": 6.043575418994414e-06, + "loss": 0.3201, + "step": 1806 + }, + { + "epoch": 5.0474860335195535, + "grad_norm": 4.20770788192749, + "learning_rate": 6.046927374301676e-06, + "loss": 0.2772, + "step": 1807 + }, + { + "epoch": 5.050279329608939, + "grad_norm": 1.4773064851760864, + "learning_rate": 6.050279329608939e-06, + "loss": 0.314, + "step": 1808 + }, + { + "epoch": 5.053072625698324, + "grad_norm": 5.713160991668701, + "learning_rate": 6.053631284916201e-06, + "loss": 0.3337, + "step": 1809 + }, + { + "epoch": 5.055865921787709, + "grad_norm": 1.4888461828231812, + "learning_rate": 6.056983240223463e-06, + "loss": 0.3241, + "step": 1810 + }, + { + "epoch": 5.058659217877095, + "grad_norm": 1.6337289810180664, + "learning_rate": 6.060335195530726e-06, + "loss": 0.2889, + "step": 1811 + }, + { + "epoch": 5.06145251396648, + "grad_norm": 7.11538553237915, + "learning_rate": 6.063687150837989e-06, + "loss": 0.2978, + "step": 1812 + }, + { + "epoch": 5.064245810055866, + "grad_norm": 1.7709788084030151, + "learning_rate": 6.067039106145252e-06, + "loss": 0.3756, + "step": 1813 + }, + { + "epoch": 5.067039106145251, + "grad_norm": 3.513477087020874, + "learning_rate": 6.070391061452514e-06, + "loss": 0.3836, + "step": 1814 + }, + { + "epoch": 5.069832402234637, + "grad_norm": 4.283865928649902, + "learning_rate": 6.073743016759776e-06, + "loss": 0.4868, + "step": 1815 + }, + { + "epoch": 5.072625698324022, + "grad_norm": 0.9538938403129578, + "learning_rate": 6.077094972067039e-06, + "loss": 0.2753, + "step": 1816 + }, + { + "epoch": 5.075418994413408, + "grad_norm": 0.6690722703933716, + "learning_rate": 6.080446927374301e-06, + "loss": 0.2322, + "step": 1817 + }, + { + "epoch": 5.078212290502793, + "grad_norm": 0.7294425368309021, + "learning_rate": 6.083798882681564e-06, + "loss": 0.2662, + "step": 1818 + }, + { + "epoch": 5.081005586592179, + "grad_norm": 1.1045304536819458, + "learning_rate": 6.0871508379888265e-06, + "loss": 0.3075, + "step": 1819 + }, + { + "epoch": 5.083798882681564, + "grad_norm": 0.9229918718338013, + "learning_rate": 6.09050279329609e-06, + "loss": 0.2832, + "step": 1820 + }, + { + "epoch": 5.08659217877095, + "grad_norm": 0.7271966934204102, + "learning_rate": 6.0938547486033525e-06, + "loss": 0.3115, + "step": 1821 + }, + { + "epoch": 5.089385474860335, + "grad_norm": 0.786655843257904, + "learning_rate": 6.097206703910615e-06, + "loss": 0.2503, + "step": 1822 + }, + { + "epoch": 5.092178770949721, + "grad_norm": 0.9051518440246582, + "learning_rate": 6.100558659217878e-06, + "loss": 0.2543, + "step": 1823 + }, + { + "epoch": 5.094972067039106, + "grad_norm": 0.8098540306091309, + "learning_rate": 6.10391061452514e-06, + "loss": 0.2678, + "step": 1824 + }, + { + "epoch": 5.097765363128492, + "grad_norm": 0.9535631537437439, + "learning_rate": 6.107262569832403e-06, + "loss": 0.27, + "step": 1825 + }, + { + "epoch": 5.100558659217877, + "grad_norm": 1.050157070159912, + "learning_rate": 6.110614525139665e-06, + "loss": 0.2545, + "step": 1826 + }, + { + "epoch": 5.103351955307263, + "grad_norm": 1.0207644701004028, + "learning_rate": 6.113966480446927e-06, + "loss": 0.2878, + "step": 1827 + }, + { + "epoch": 5.106145251396648, + "grad_norm": 0.8761366605758667, + "learning_rate": 6.1173184357541904e-06, + "loss": 0.2449, + "step": 1828 + }, + { + "epoch": 5.108938547486034, + "grad_norm": 2.2719826698303223, + "learning_rate": 6.120670391061453e-06, + "loss": 0.3209, + "step": 1829 + }, + { + "epoch": 5.111731843575419, + "grad_norm": 1.2294787168502808, + "learning_rate": 6.124022346368716e-06, + "loss": 0.2937, + "step": 1830 + }, + { + "epoch": 5.114525139664805, + "grad_norm": 2.0678517818450928, + "learning_rate": 6.127374301675978e-06, + "loss": 0.3323, + "step": 1831 + }, + { + "epoch": 5.11731843575419, + "grad_norm": 1.9384267330169678, + "learning_rate": 6.13072625698324e-06, + "loss": 0.2921, + "step": 1832 + }, + { + "epoch": 5.120111731843576, + "grad_norm": 1.202404260635376, + "learning_rate": 6.134078212290503e-06, + "loss": 0.2783, + "step": 1833 + }, + { + "epoch": 5.122905027932961, + "grad_norm": 3.0719265937805176, + "learning_rate": 6.137430167597765e-06, + "loss": 0.3317, + "step": 1834 + }, + { + "epoch": 5.125698324022347, + "grad_norm": 1.1859261989593506, + "learning_rate": 6.140782122905028e-06, + "loss": 0.3351, + "step": 1835 + }, + { + "epoch": 5.128491620111732, + "grad_norm": 1.4899580478668213, + "learning_rate": 6.1441340782122905e-06, + "loss": 0.3156, + "step": 1836 + }, + { + "epoch": 5.131284916201118, + "grad_norm": 1.806089162826538, + "learning_rate": 6.147486033519553e-06, + "loss": 0.3717, + "step": 1837 + }, + { + "epoch": 5.134078212290503, + "grad_norm": 2.575892448425293, + "learning_rate": 6.150837988826816e-06, + "loss": 0.3716, + "step": 1838 + }, + { + "epoch": 5.136871508379889, + "grad_norm": 7.089779853820801, + "learning_rate": 6.154189944134078e-06, + "loss": 0.3787, + "step": 1839 + }, + { + "epoch": 5.139664804469274, + "grad_norm": 3.901911497116089, + "learning_rate": 6.157541899441341e-06, + "loss": 0.4939, + "step": 1840 + }, + { + "epoch": 5.14245810055866, + "grad_norm": 1.0321241617202759, + "learning_rate": 6.160893854748603e-06, + "loss": 0.2727, + "step": 1841 + }, + { + "epoch": 5.145251396648045, + "grad_norm": 0.6617045402526855, + "learning_rate": 6.164245810055865e-06, + "loss": 0.2297, + "step": 1842 + }, + { + "epoch": 5.148044692737431, + "grad_norm": 2.140003204345703, + "learning_rate": 6.167597765363128e-06, + "loss": 0.2613, + "step": 1843 + }, + { + "epoch": 5.150837988826815, + "grad_norm": 0.8014543056488037, + "learning_rate": 6.1709497206703914e-06, + "loss": 0.2817, + "step": 1844 + }, + { + "epoch": 5.153631284916201, + "grad_norm": 0.8606362342834473, + "learning_rate": 6.1743016759776545e-06, + "loss": 0.2605, + "step": 1845 + }, + { + "epoch": 5.156424581005586, + "grad_norm": 0.8060360550880432, + "learning_rate": 6.177653631284917e-06, + "loss": 0.2525, + "step": 1846 + }, + { + "epoch": 5.159217877094972, + "grad_norm": 1.4615522623062134, + "learning_rate": 6.181005586592179e-06, + "loss": 0.267, + "step": 1847 + }, + { + "epoch": 5.162011173184357, + "grad_norm": 1.1299731731414795, + "learning_rate": 6.184357541899442e-06, + "loss": 0.2656, + "step": 1848 + }, + { + "epoch": 5.164804469273743, + "grad_norm": 1.0976790189743042, + "learning_rate": 6.187709497206704e-06, + "loss": 0.2624, + "step": 1849 + }, + { + "epoch": 5.167597765363128, + "grad_norm": 0.8851279020309448, + "learning_rate": 6.191061452513967e-06, + "loss": 0.277, + "step": 1850 + }, + { + "epoch": 5.170391061452514, + "grad_norm": 1.1434439420700073, + "learning_rate": 6.194413407821229e-06, + "loss": 0.3086, + "step": 1851 + }, + { + "epoch": 5.173184357541899, + "grad_norm": 1.2450363636016846, + "learning_rate": 6.1977653631284915e-06, + "loss": 0.2806, + "step": 1852 + }, + { + "epoch": 5.175977653631285, + "grad_norm": 1.1080976724624634, + "learning_rate": 6.2011173184357546e-06, + "loss": 0.2772, + "step": 1853 + }, + { + "epoch": 5.17877094972067, + "grad_norm": 1.5471917390823364, + "learning_rate": 6.204469273743017e-06, + "loss": 0.2594, + "step": 1854 + }, + { + "epoch": 5.181564245810056, + "grad_norm": 1.0797456502914429, + "learning_rate": 6.20782122905028e-06, + "loss": 0.2781, + "step": 1855 + }, + { + "epoch": 5.184357541899441, + "grad_norm": 0.9587633013725281, + "learning_rate": 6.211173184357542e-06, + "loss": 0.277, + "step": 1856 + }, + { + "epoch": 5.187150837988827, + "grad_norm": 1.4839032888412476, + "learning_rate": 6.214525139664804e-06, + "loss": 0.2933, + "step": 1857 + }, + { + "epoch": 5.189944134078212, + "grad_norm": 2.5825908184051514, + "learning_rate": 6.217877094972067e-06, + "loss": 0.3077, + "step": 1858 + }, + { + "epoch": 5.192737430167598, + "grad_norm": 1.3689639568328857, + "learning_rate": 6.221229050279329e-06, + "loss": 0.311, + "step": 1859 + }, + { + "epoch": 5.195530726256983, + "grad_norm": 1.3146567344665527, + "learning_rate": 6.2245810055865924e-06, + "loss": 0.3014, + "step": 1860 + }, + { + "epoch": 5.198324022346369, + "grad_norm": 1.3786956071853638, + "learning_rate": 6.227932960893855e-06, + "loss": 0.3175, + "step": 1861 + }, + { + "epoch": 5.201117318435754, + "grad_norm": 1.6835644245147705, + "learning_rate": 6.231284916201117e-06, + "loss": 0.3603, + "step": 1862 + }, + { + "epoch": 5.20391061452514, + "grad_norm": 2.0575127601623535, + "learning_rate": 6.23463687150838e-06, + "loss": 0.3218, + "step": 1863 + }, + { + "epoch": 5.206703910614525, + "grad_norm": 2.647763252258301, + "learning_rate": 6.237988826815642e-06, + "loss": 0.3619, + "step": 1864 + }, + { + "epoch": 5.209497206703911, + "grad_norm": 3.532555103302002, + "learning_rate": 6.241340782122905e-06, + "loss": 0.4459, + "step": 1865 + }, + { + "epoch": 5.212290502793296, + "grad_norm": 0.8198737502098083, + "learning_rate": 6.244692737430167e-06, + "loss": 0.3143, + "step": 1866 + }, + { + "epoch": 5.215083798882682, + "grad_norm": 0.878190815448761, + "learning_rate": 6.24804469273743e-06, + "loss": 0.2641, + "step": 1867 + }, + { + "epoch": 5.217877094972067, + "grad_norm": 0.6988477110862732, + "learning_rate": 6.251396648044693e-06, + "loss": 0.2556, + "step": 1868 + }, + { + "epoch": 5.220670391061453, + "grad_norm": 0.9250651001930237, + "learning_rate": 6.2547486033519556e-06, + "loss": 0.3013, + "step": 1869 + }, + { + "epoch": 5.223463687150838, + "grad_norm": 1.2071925401687622, + "learning_rate": 6.258100558659219e-06, + "loss": 0.2367, + "step": 1870 + }, + { + "epoch": 5.226256983240224, + "grad_norm": 0.9094240665435791, + "learning_rate": 6.261452513966481e-06, + "loss": 0.2588, + "step": 1871 + }, + { + "epoch": 5.229050279329609, + "grad_norm": 1.033064365386963, + "learning_rate": 6.264804469273744e-06, + "loss": 0.2912, + "step": 1872 + }, + { + "epoch": 5.231843575418995, + "grad_norm": 0.7117978930473328, + "learning_rate": 6.268156424581006e-06, + "loss": 0.2723, + "step": 1873 + }, + { + "epoch": 5.23463687150838, + "grad_norm": 0.8705118894577026, + "learning_rate": 6.271508379888268e-06, + "loss": 0.2409, + "step": 1874 + }, + { + "epoch": 5.237430167597766, + "grad_norm": 1.34247887134552, + "learning_rate": 6.274860335195531e-06, + "loss": 0.2574, + "step": 1875 + }, + { + "epoch": 5.240223463687151, + "grad_norm": 1.0620986223220825, + "learning_rate": 6.2782122905027935e-06, + "loss": 0.351, + "step": 1876 + }, + { + "epoch": 5.243016759776537, + "grad_norm": 1.0671035051345825, + "learning_rate": 6.2815642458100565e-06, + "loss": 0.2707, + "step": 1877 + }, + { + "epoch": 5.245810055865922, + "grad_norm": 2.140720844268799, + "learning_rate": 6.284916201117319e-06, + "loss": 0.2969, + "step": 1878 + }, + { + "epoch": 5.248603351955307, + "grad_norm": 1.4124557971954346, + "learning_rate": 6.288268156424581e-06, + "loss": 0.303, + "step": 1879 + }, + { + "epoch": 5.251396648044693, + "grad_norm": 1.2751896381378174, + "learning_rate": 6.291620111731844e-06, + "loss": 0.2945, + "step": 1880 + }, + { + "epoch": 5.254189944134078, + "grad_norm": 0.9082738757133484, + "learning_rate": 6.294972067039106e-06, + "loss": 0.2644, + "step": 1881 + }, + { + "epoch": 5.256983240223463, + "grad_norm": 0.9649515748023987, + "learning_rate": 6.298324022346369e-06, + "loss": 0.3076, + "step": 1882 + }, + { + "epoch": 5.259776536312849, + "grad_norm": 1.3101574182510376, + "learning_rate": 6.301675977653631e-06, + "loss": 0.256, + "step": 1883 + }, + { + "epoch": 5.262569832402234, + "grad_norm": 1.78885817527771, + "learning_rate": 6.3050279329608935e-06, + "loss": 0.3374, + "step": 1884 + }, + { + "epoch": 5.26536312849162, + "grad_norm": 1.4074186086654663, + "learning_rate": 6.3083798882681566e-06, + "loss": 0.319, + "step": 1885 + }, + { + "epoch": 5.268156424581005, + "grad_norm": 1.564381718635559, + "learning_rate": 6.311731843575419e-06, + "loss": 0.3119, + "step": 1886 + }, + { + "epoch": 5.270949720670391, + "grad_norm": 2.002225637435913, + "learning_rate": 6.315083798882682e-06, + "loss": 0.3284, + "step": 1887 + }, + { + "epoch": 5.273743016759776, + "grad_norm": 1.8313493728637695, + "learning_rate": 6.318435754189944e-06, + "loss": 0.3259, + "step": 1888 + }, + { + "epoch": 5.276536312849162, + "grad_norm": 2.0545575618743896, + "learning_rate": 6.321787709497206e-06, + "loss": 0.3745, + "step": 1889 + }, + { + "epoch": 5.279329608938547, + "grad_norm": 5.590494632720947, + "learning_rate": 6.325139664804469e-06, + "loss": 0.4327, + "step": 1890 + }, + { + "epoch": 5.282122905027933, + "grad_norm": 0.7241694927215576, + "learning_rate": 6.328491620111731e-06, + "loss": 0.2879, + "step": 1891 + }, + { + "epoch": 5.284916201117318, + "grad_norm": 0.9245277643203735, + "learning_rate": 6.331843575418995e-06, + "loss": 0.2646, + "step": 1892 + }, + { + "epoch": 5.287709497206704, + "grad_norm": 1.2119693756103516, + "learning_rate": 6.3351955307262575e-06, + "loss": 0.2828, + "step": 1893 + }, + { + "epoch": 5.290502793296089, + "grad_norm": 0.9561535120010376, + "learning_rate": 6.33854748603352e-06, + "loss": 0.2482, + "step": 1894 + }, + { + "epoch": 5.293296089385475, + "grad_norm": 2.1700072288513184, + "learning_rate": 6.341899441340783e-06, + "loss": 0.2583, + "step": 1895 + }, + { + "epoch": 5.29608938547486, + "grad_norm": 1.0299097299575806, + "learning_rate": 6.345251396648045e-06, + "loss": 0.27, + "step": 1896 + }, + { + "epoch": 5.298882681564246, + "grad_norm": 0.8930662274360657, + "learning_rate": 6.348603351955308e-06, + "loss": 0.2334, + "step": 1897 + }, + { + "epoch": 5.301675977653631, + "grad_norm": 0.9146484136581421, + "learning_rate": 6.35195530726257e-06, + "loss": 0.2714, + "step": 1898 + }, + { + "epoch": 5.304469273743017, + "grad_norm": 1.1597564220428467, + "learning_rate": 6.355307262569832e-06, + "loss": 0.2444, + "step": 1899 + }, + { + "epoch": 5.307262569832402, + "grad_norm": 1.2910884618759155, + "learning_rate": 6.358659217877095e-06, + "loss": 0.2881, + "step": 1900 + }, + { + "epoch": 5.310055865921788, + "grad_norm": 1.233657717704773, + "learning_rate": 6.362011173184358e-06, + "loss": 0.2615, + "step": 1901 + }, + { + "epoch": 5.312849162011173, + "grad_norm": 1.4418509006500244, + "learning_rate": 6.365363128491621e-06, + "loss": 0.293, + "step": 1902 + }, + { + "epoch": 5.315642458100559, + "grad_norm": 1.1399755477905273, + "learning_rate": 6.368715083798883e-06, + "loss": 0.2415, + "step": 1903 + }, + { + "epoch": 5.318435754189944, + "grad_norm": 1.1861231327056885, + "learning_rate": 6.372067039106145e-06, + "loss": 0.2872, + "step": 1904 + }, + { + "epoch": 5.32122905027933, + "grad_norm": 1.1408406496047974, + "learning_rate": 6.375418994413408e-06, + "loss": 0.2466, + "step": 1905 + }, + { + "epoch": 5.324022346368715, + "grad_norm": 1.1017338037490845, + "learning_rate": 6.37877094972067e-06, + "loss": 0.295, + "step": 1906 + }, + { + "epoch": 5.326815642458101, + "grad_norm": 1.0633206367492676, + "learning_rate": 6.382122905027933e-06, + "loss": 0.2295, + "step": 1907 + }, + { + "epoch": 5.329608938547486, + "grad_norm": 1.1643683910369873, + "learning_rate": 6.3854748603351955e-06, + "loss": 0.27, + "step": 1908 + }, + { + "epoch": 5.332402234636872, + "grad_norm": 1.4966927766799927, + "learning_rate": 6.388826815642458e-06, + "loss": 0.277, + "step": 1909 + }, + { + "epoch": 5.335195530726257, + "grad_norm": 1.605310082435608, + "learning_rate": 6.392178770949721e-06, + "loss": 0.3031, + "step": 1910 + }, + { + "epoch": 5.337988826815643, + "grad_norm": 2.071439504623413, + "learning_rate": 6.395530726256983e-06, + "loss": 0.3495, + "step": 1911 + }, + { + "epoch": 5.340782122905028, + "grad_norm": 1.651806354522705, + "learning_rate": 6.398882681564246e-06, + "loss": 0.3318, + "step": 1912 + }, + { + "epoch": 5.343575418994414, + "grad_norm": 2.214076042175293, + "learning_rate": 6.402234636871508e-06, + "loss": 0.3341, + "step": 1913 + }, + { + "epoch": 5.346368715083799, + "grad_norm": 3.251127004623413, + "learning_rate": 6.405586592178771e-06, + "loss": 0.4076, + "step": 1914 + }, + { + "epoch": 5.349162011173185, + "grad_norm": 2.584912061691284, + "learning_rate": 6.408938547486033e-06, + "loss": 0.4619, + "step": 1915 + }, + { + "epoch": 5.351955307262569, + "grad_norm": 1.704502820968628, + "learning_rate": 6.412290502793296e-06, + "loss": 0.2503, + "step": 1916 + }, + { + "epoch": 5.354748603351955, + "grad_norm": 0.6236934065818787, + "learning_rate": 6.4156424581005594e-06, + "loss": 0.2316, + "step": 1917 + }, + { + "epoch": 5.35754189944134, + "grad_norm": 1.3246259689331055, + "learning_rate": 6.418994413407822e-06, + "loss": 0.273, + "step": 1918 + }, + { + "epoch": 5.360335195530726, + "grad_norm": 0.6451730728149414, + "learning_rate": 6.422346368715085e-06, + "loss": 0.2346, + "step": 1919 + }, + { + "epoch": 5.363128491620111, + "grad_norm": 0.9821768403053284, + "learning_rate": 6.425698324022347e-06, + "loss": 0.3323, + "step": 1920 + }, + { + "epoch": 5.365921787709497, + "grad_norm": 0.8737133741378784, + "learning_rate": 6.429050279329609e-06, + "loss": 0.2283, + "step": 1921 + }, + { + "epoch": 5.368715083798882, + "grad_norm": 0.6160896420478821, + "learning_rate": 6.432402234636872e-06, + "loss": 0.2786, + "step": 1922 + }, + { + "epoch": 5.371508379888268, + "grad_norm": 0.9305797815322876, + "learning_rate": 6.435754189944134e-06, + "loss": 0.2434, + "step": 1923 + }, + { + "epoch": 5.374301675977653, + "grad_norm": 0.9042353630065918, + "learning_rate": 6.439106145251397e-06, + "loss": 0.246, + "step": 1924 + }, + { + "epoch": 5.377094972067039, + "grad_norm": 0.9169382452964783, + "learning_rate": 6.4424581005586595e-06, + "loss": 0.2569, + "step": 1925 + }, + { + "epoch": 5.379888268156424, + "grad_norm": 1.1165077686309814, + "learning_rate": 6.445810055865922e-06, + "loss": 0.2976, + "step": 1926 + }, + { + "epoch": 5.38268156424581, + "grad_norm": 0.9713110327720642, + "learning_rate": 6.449162011173185e-06, + "loss": 0.2677, + "step": 1927 + }, + { + "epoch": 5.385474860335195, + "grad_norm": 1.2221273183822632, + "learning_rate": 6.452513966480447e-06, + "loss": 0.2947, + "step": 1928 + }, + { + "epoch": 5.388268156424581, + "grad_norm": 1.9063212871551514, + "learning_rate": 6.45586592178771e-06, + "loss": 0.2847, + "step": 1929 + }, + { + "epoch": 5.391061452513966, + "grad_norm": 1.387919545173645, + "learning_rate": 6.459217877094972e-06, + "loss": 0.3072, + "step": 1930 + }, + { + "epoch": 5.393854748603352, + "grad_norm": 1.0844817161560059, + "learning_rate": 6.462569832402234e-06, + "loss": 0.2913, + "step": 1931 + }, + { + "epoch": 5.396648044692737, + "grad_norm": 1.5822914838790894, + "learning_rate": 6.465921787709497e-06, + "loss": 0.2804, + "step": 1932 + }, + { + "epoch": 5.399441340782123, + "grad_norm": 1.4664349555969238, + "learning_rate": 6.46927374301676e-06, + "loss": 0.2642, + "step": 1933 + }, + { + "epoch": 5.402234636871508, + "grad_norm": 1.6619762182235718, + "learning_rate": 6.472625698324023e-06, + "loss": 0.298, + "step": 1934 + }, + { + "epoch": 5.405027932960894, + "grad_norm": 1.4820046424865723, + "learning_rate": 6.475977653631285e-06, + "loss": 0.374, + "step": 1935 + }, + { + "epoch": 5.407821229050279, + "grad_norm": 1.9781765937805176, + "learning_rate": 6.479329608938547e-06, + "loss": 0.3133, + "step": 1936 + }, + { + "epoch": 5.410614525139665, + "grad_norm": 1.7736289501190186, + "learning_rate": 6.48268156424581e-06, + "loss": 0.3278, + "step": 1937 + }, + { + "epoch": 5.41340782122905, + "grad_norm": 4.863185882568359, + "learning_rate": 6.486033519553072e-06, + "loss": 0.3176, + "step": 1938 + }, + { + "epoch": 5.416201117318436, + "grad_norm": 2.272547483444214, + "learning_rate": 6.489385474860335e-06, + "loss": 0.3383, + "step": 1939 + }, + { + "epoch": 5.418994413407821, + "grad_norm": 3.433352470397949, + "learning_rate": 6.492737430167598e-06, + "loss": 0.5664, + "step": 1940 + }, + { + "epoch": 5.421787709497207, + "grad_norm": 0.7983115315437317, + "learning_rate": 6.4960893854748605e-06, + "loss": 0.3163, + "step": 1941 + }, + { + "epoch": 5.424581005586592, + "grad_norm": 0.9780985116958618, + "learning_rate": 6.4994413407821236e-06, + "loss": 0.2512, + "step": 1942 + }, + { + "epoch": 5.427374301675978, + "grad_norm": 0.7865659594535828, + "learning_rate": 6.502793296089386e-06, + "loss": 0.2776, + "step": 1943 + }, + { + "epoch": 5.430167597765363, + "grad_norm": 0.8958929777145386, + "learning_rate": 6.506145251396649e-06, + "loss": 0.2651, + "step": 1944 + }, + { + "epoch": 5.432960893854749, + "grad_norm": 1.2326228618621826, + "learning_rate": 6.509497206703911e-06, + "loss": 0.2742, + "step": 1945 + }, + { + "epoch": 5.435754189944134, + "grad_norm": 0.8631238341331482, + "learning_rate": 6.512849162011173e-06, + "loss": 0.2947, + "step": 1946 + }, + { + "epoch": 5.43854748603352, + "grad_norm": 1.3510725498199463, + "learning_rate": 6.516201117318436e-06, + "loss": 0.2846, + "step": 1947 + }, + { + "epoch": 5.441340782122905, + "grad_norm": 1.1109312772750854, + "learning_rate": 6.519553072625698e-06, + "loss": 0.2503, + "step": 1948 + }, + { + "epoch": 5.444134078212291, + "grad_norm": 0.9137476682662964, + "learning_rate": 6.5229050279329614e-06, + "loss": 0.2506, + "step": 1949 + }, + { + "epoch": 5.446927374301676, + "grad_norm": 6.344702243804932, + "learning_rate": 6.526256983240224e-06, + "loss": 0.2715, + "step": 1950 + }, + { + "epoch": 5.449720670391061, + "grad_norm": 30.595067977905273, + "learning_rate": 6.529608938547486e-06, + "loss": 0.2883, + "step": 1951 + }, + { + "epoch": 5.452513966480447, + "grad_norm": 1.171781063079834, + "learning_rate": 6.532960893854749e-06, + "loss": 0.257, + "step": 1952 + }, + { + "epoch": 5.455307262569832, + "grad_norm": 1.084414005279541, + "learning_rate": 6.536312849162011e-06, + "loss": 0.2464, + "step": 1953 + }, + { + "epoch": 5.4581005586592175, + "grad_norm": 1.6217060089111328, + "learning_rate": 6.539664804469274e-06, + "loss": 0.2878, + "step": 1954 + }, + { + "epoch": 5.460893854748603, + "grad_norm": 1.8531297445297241, + "learning_rate": 6.543016759776536e-06, + "loss": 0.2478, + "step": 1955 + }, + { + "epoch": 5.4636871508379885, + "grad_norm": 1.1444119215011597, + "learning_rate": 6.5463687150837985e-06, + "loss": 0.2437, + "step": 1956 + }, + { + "epoch": 5.466480446927374, + "grad_norm": 1.9606194496154785, + "learning_rate": 6.5497206703910615e-06, + "loss": 0.2765, + "step": 1957 + }, + { + "epoch": 5.4692737430167595, + "grad_norm": 2.263803482055664, + "learning_rate": 6.553072625698324e-06, + "loss": 0.3311, + "step": 1958 + }, + { + "epoch": 5.472067039106145, + "grad_norm": 2.379793167114258, + "learning_rate": 6.556424581005587e-06, + "loss": 0.2984, + "step": 1959 + }, + { + "epoch": 5.4748603351955305, + "grad_norm": 1.4399487972259521, + "learning_rate": 6.559776536312849e-06, + "loss": 0.2575, + "step": 1960 + }, + { + "epoch": 5.477653631284916, + "grad_norm": 2.0233500003814697, + "learning_rate": 6.563128491620112e-06, + "loss": 0.3136, + "step": 1961 + }, + { + "epoch": 5.4804469273743015, + "grad_norm": 2.4556782245635986, + "learning_rate": 6.566480446927374e-06, + "loss": 0.3411, + "step": 1962 + }, + { + "epoch": 5.483240223463687, + "grad_norm": 2.2601141929626465, + "learning_rate": 6.569832402234636e-06, + "loss": 0.3335, + "step": 1963 + }, + { + "epoch": 5.4860335195530725, + "grad_norm": 2.1010568141937256, + "learning_rate": 6.5731843575419e-06, + "loss": 0.3824, + "step": 1964 + }, + { + "epoch": 5.488826815642458, + "grad_norm": 7.153081893920898, + "learning_rate": 6.5765363128491624e-06, + "loss": 0.4843, + "step": 1965 + }, + { + "epoch": 5.4916201117318435, + "grad_norm": 0.931185245513916, + "learning_rate": 6.5798882681564255e-06, + "loss": 0.2698, + "step": 1966 + }, + { + "epoch": 5.494413407821229, + "grad_norm": 0.7334734201431274, + "learning_rate": 6.583240223463688e-06, + "loss": 0.2676, + "step": 1967 + }, + { + "epoch": 5.4972067039106145, + "grad_norm": 0.8505213856697083, + "learning_rate": 6.58659217877095e-06, + "loss": 0.2311, + "step": 1968 + }, + { + "epoch": 5.5, + "grad_norm": 1.0038577318191528, + "learning_rate": 6.589944134078213e-06, + "loss": 0.2704, + "step": 1969 + }, + { + "epoch": 5.5027932960893855, + "grad_norm": 0.684473991394043, + "learning_rate": 6.593296089385475e-06, + "loss": 0.2682, + "step": 1970 + }, + { + "epoch": 5.505586592178771, + "grad_norm": 0.9912981986999512, + "learning_rate": 6.596648044692738e-06, + "loss": 0.2609, + "step": 1971 + }, + { + "epoch": 5.5083798882681565, + "grad_norm": 1.5605647563934326, + "learning_rate": 6.6e-06, + "loss": 0.2887, + "step": 1972 + }, + { + "epoch": 5.511173184357542, + "grad_norm": 1.1440966129302979, + "learning_rate": 6.6033519553072625e-06, + "loss": 0.2846, + "step": 1973 + }, + { + "epoch": 5.5139664804469275, + "grad_norm": 1.014465093612671, + "learning_rate": 6.6067039106145256e-06, + "loss": 0.2632, + "step": 1974 + }, + { + "epoch": 5.516759776536313, + "grad_norm": 2.4433534145355225, + "learning_rate": 6.610055865921788e-06, + "loss": 0.3009, + "step": 1975 + }, + { + "epoch": 5.5195530726256985, + "grad_norm": 0.9188408255577087, + "learning_rate": 6.613407821229051e-06, + "loss": 0.2736, + "step": 1976 + }, + { + "epoch": 5.522346368715084, + "grad_norm": 1.1153956651687622, + "learning_rate": 6.616759776536313e-06, + "loss": 0.2758, + "step": 1977 + }, + { + "epoch": 5.5251396648044695, + "grad_norm": 1.096919298171997, + "learning_rate": 6.620111731843575e-06, + "loss": 0.2687, + "step": 1978 + }, + { + "epoch": 5.527932960893855, + "grad_norm": 1.0518863201141357, + "learning_rate": 6.623463687150838e-06, + "loss": 0.2582, + "step": 1979 + }, + { + "epoch": 5.5307262569832405, + "grad_norm": 1.4609042406082153, + "learning_rate": 6.6268156424581e-06, + "loss": 0.2542, + "step": 1980 + }, + { + "epoch": 5.533519553072626, + "grad_norm": 1.3449817895889282, + "learning_rate": 6.6301675977653635e-06, + "loss": 0.2876, + "step": 1981 + }, + { + "epoch": 5.5363128491620115, + "grad_norm": 1.7738093137741089, + "learning_rate": 6.633519553072626e-06, + "loss": 0.3277, + "step": 1982 + }, + { + "epoch": 5.539106145251397, + "grad_norm": 1.1897963285446167, + "learning_rate": 6.636871508379888e-06, + "loss": 0.2933, + "step": 1983 + }, + { + "epoch": 5.5418994413407825, + "grad_norm": 1.4338127374649048, + "learning_rate": 6.640223463687151e-06, + "loss": 0.2833, + "step": 1984 + }, + { + "epoch": 5.544692737430168, + "grad_norm": 2.3423495292663574, + "learning_rate": 6.643575418994413e-06, + "loss": 0.2912, + "step": 1985 + }, + { + "epoch": 5.547486033519553, + "grad_norm": 2.0876171588897705, + "learning_rate": 6.646927374301676e-06, + "loss": 0.2891, + "step": 1986 + }, + { + "epoch": 5.550279329608939, + "grad_norm": 3.223083734512329, + "learning_rate": 6.650279329608938e-06, + "loss": 0.3121, + "step": 1987 + }, + { + "epoch": 5.553072625698324, + "grad_norm": 2.073512315750122, + "learning_rate": 6.653631284916201e-06, + "loss": 0.3358, + "step": 1988 + }, + { + "epoch": 5.55586592178771, + "grad_norm": 1.805420994758606, + "learning_rate": 6.656983240223464e-06, + "loss": 0.3697, + "step": 1989 + }, + { + "epoch": 5.558659217877095, + "grad_norm": 3.182495594024658, + "learning_rate": 6.6603351955307266e-06, + "loss": 0.5011, + "step": 1990 + }, + { + "epoch": 5.56145251396648, + "grad_norm": 0.6142599582672119, + "learning_rate": 6.66368715083799e-06, + "loss": 0.2252, + "step": 1991 + }, + { + "epoch": 5.564245810055866, + "grad_norm": 0.7567723989486694, + "learning_rate": 6.667039106145252e-06, + "loss": 0.2628, + "step": 1992 + }, + { + "epoch": 5.567039106145251, + "grad_norm": 0.7566982507705688, + "learning_rate": 6.670391061452514e-06, + "loss": 0.2517, + "step": 1993 + }, + { + "epoch": 5.569832402234637, + "grad_norm": 1.0873514413833618, + "learning_rate": 6.673743016759777e-06, + "loss": 0.2622, + "step": 1994 + }, + { + "epoch": 5.572625698324022, + "grad_norm": 0.7600097060203552, + "learning_rate": 6.677094972067039e-06, + "loss": 0.2495, + "step": 1995 + }, + { + "epoch": 5.575418994413408, + "grad_norm": 1.0502632856369019, + "learning_rate": 6.680446927374302e-06, + "loss": 0.2689, + "step": 1996 + }, + { + "epoch": 5.578212290502793, + "grad_norm": 1.287634015083313, + "learning_rate": 6.6837988826815645e-06, + "loss": 0.2613, + "step": 1997 + }, + { + "epoch": 5.581005586592179, + "grad_norm": 0.6878490447998047, + "learning_rate": 6.687150837988827e-06, + "loss": 0.2733, + "step": 1998 + }, + { + "epoch": 5.583798882681564, + "grad_norm": 0.9757861495018005, + "learning_rate": 6.69050279329609e-06, + "loss": 0.269, + "step": 1999 + }, + { + "epoch": 5.58659217877095, + "grad_norm": 1.3301657438278198, + "learning_rate": 6.693854748603352e-06, + "loss": 0.2972, + "step": 2000 + }, + { + "epoch": 5.58659217877095, + "eval_cer": 0.19975341778228753, + "eval_loss": 0.3149757981300354, + "eval_runtime": 23.3517, + "eval_samples_per_second": 194.333, + "eval_steps_per_second": 0.642, + "eval_wer": 0.3690958353118628, + "step": 2000 + }, + { + "epoch": 5.589385474860335, + "grad_norm": 1.8227416276931763, + "learning_rate": 6.697206703910615e-06, + "loss": 0.2958, + "step": 2001 + }, + { + "epoch": 5.592178770949721, + "grad_norm": 1.4708274602890015, + "learning_rate": 6.700558659217877e-06, + "loss": 0.2539, + "step": 2002 + }, + { + "epoch": 5.594972067039106, + "grad_norm": 1.2759531736373901, + "learning_rate": 6.703910614525139e-06, + "loss": 0.2468, + "step": 2003 + }, + { + "epoch": 5.597765363128492, + "grad_norm": 3.0070583820343018, + "learning_rate": 6.707262569832402e-06, + "loss": 0.2836, + "step": 2004 + }, + { + "epoch": 5.600558659217877, + "grad_norm": 0.9663261771202087, + "learning_rate": 6.7106145251396645e-06, + "loss": 0.3013, + "step": 2005 + }, + { + "epoch": 5.603351955307263, + "grad_norm": 1.1295732259750366, + "learning_rate": 6.713966480446928e-06, + "loss": 0.3279, + "step": 2006 + }, + { + "epoch": 5.606145251396648, + "grad_norm": 1.4885385036468506, + "learning_rate": 6.71731843575419e-06, + "loss": 0.2904, + "step": 2007 + }, + { + "epoch": 5.608938547486034, + "grad_norm": 2.010406732559204, + "learning_rate": 6.720670391061453e-06, + "loss": 0.2545, + "step": 2008 + }, + { + "epoch": 5.611731843575419, + "grad_norm": 1.2382876873016357, + "learning_rate": 6.724022346368715e-06, + "loss": 0.2781, + "step": 2009 + }, + { + "epoch": 5.614525139664805, + "grad_norm": 1.2483317852020264, + "learning_rate": 6.727374301675977e-06, + "loss": 0.2978, + "step": 2010 + }, + { + "epoch": 5.61731843575419, + "grad_norm": 2.8138298988342285, + "learning_rate": 6.73072625698324e-06, + "loss": 0.2727, + "step": 2011 + }, + { + "epoch": 5.620111731843576, + "grad_norm": 1.701602578163147, + "learning_rate": 6.734078212290503e-06, + "loss": 0.3398, + "step": 2012 + }, + { + "epoch": 5.622905027932961, + "grad_norm": 3.7735376358032227, + "learning_rate": 6.737430167597766e-06, + "loss": 0.3521, + "step": 2013 + }, + { + "epoch": 5.625698324022347, + "grad_norm": 2.0885562896728516, + "learning_rate": 6.7407821229050285e-06, + "loss": 0.3499, + "step": 2014 + }, + { + "epoch": 5.628491620111732, + "grad_norm": 3.3985230922698975, + "learning_rate": 6.744134078212291e-06, + "loss": 0.4734, + "step": 2015 + }, + { + "epoch": 5.631284916201118, + "grad_norm": 0.6720626950263977, + "learning_rate": 6.747486033519554e-06, + "loss": 0.264, + "step": 2016 + }, + { + "epoch": 5.634078212290503, + "grad_norm": 0.5850576758384705, + "learning_rate": 6.750837988826816e-06, + "loss": 0.2633, + "step": 2017 + }, + { + "epoch": 5.636871508379889, + "grad_norm": 0.6633307933807373, + "learning_rate": 6.754189944134079e-06, + "loss": 0.2387, + "step": 2018 + }, + { + "epoch": 5.639664804469274, + "grad_norm": 0.826386034488678, + "learning_rate": 6.757541899441341e-06, + "loss": 0.2502, + "step": 2019 + }, + { + "epoch": 5.64245810055866, + "grad_norm": 0.6703612208366394, + "learning_rate": 6.760893854748603e-06, + "loss": 0.2625, + "step": 2020 + }, + { + "epoch": 5.645251396648044, + "grad_norm": 0.8407956957817078, + "learning_rate": 6.764245810055866e-06, + "loss": 0.2549, + "step": 2021 + }, + { + "epoch": 5.648044692737431, + "grad_norm": 0.7338083386421204, + "learning_rate": 6.767597765363129e-06, + "loss": 0.2775, + "step": 2022 + }, + { + "epoch": 5.650837988826815, + "grad_norm": 1.3895729780197144, + "learning_rate": 6.770949720670392e-06, + "loss": 0.2609, + "step": 2023 + }, + { + "epoch": 5.653631284916202, + "grad_norm": 1.4501115083694458, + "learning_rate": 6.774301675977654e-06, + "loss": 0.251, + "step": 2024 + }, + { + "epoch": 5.656424581005586, + "grad_norm": 1.1225801706314087, + "learning_rate": 6.777653631284916e-06, + "loss": 0.2433, + "step": 2025 + }, + { + "epoch": 5.659217877094972, + "grad_norm": 1.0761078596115112, + "learning_rate": 6.781005586592179e-06, + "loss": 0.296, + "step": 2026 + }, + { + "epoch": 5.662011173184357, + "grad_norm": 1.1168713569641113, + "learning_rate": 6.784357541899441e-06, + "loss": 0.2946, + "step": 2027 + }, + { + "epoch": 5.664804469273743, + "grad_norm": 1.24300217628479, + "learning_rate": 6.787709497206704e-06, + "loss": 0.2999, + "step": 2028 + }, + { + "epoch": 5.667597765363128, + "grad_norm": 0.9698011875152588, + "learning_rate": 6.7910614525139665e-06, + "loss": 0.2677, + "step": 2029 + }, + { + "epoch": 5.670391061452514, + "grad_norm": 1.0524662733078003, + "learning_rate": 6.794413407821229e-06, + "loss": 0.2606, + "step": 2030 + }, + { + "epoch": 5.673184357541899, + "grad_norm": 1.0513242483139038, + "learning_rate": 6.797765363128492e-06, + "loss": 0.2486, + "step": 2031 + }, + { + "epoch": 5.675977653631285, + "grad_norm": 1.1049373149871826, + "learning_rate": 6.801117318435754e-06, + "loss": 0.2799, + "step": 2032 + }, + { + "epoch": 5.67877094972067, + "grad_norm": 1.2116186618804932, + "learning_rate": 6.804469273743017e-06, + "loss": 0.2827, + "step": 2033 + }, + { + "epoch": 5.681564245810056, + "grad_norm": 1.7134239673614502, + "learning_rate": 6.807821229050279e-06, + "loss": 0.3106, + "step": 2034 + }, + { + "epoch": 5.684357541899441, + "grad_norm": 2.010478973388672, + "learning_rate": 6.811173184357541e-06, + "loss": 0.3173, + "step": 2035 + }, + { + "epoch": 5.687150837988827, + "grad_norm": 1.6770652532577515, + "learning_rate": 6.814525139664804e-06, + "loss": 0.2874, + "step": 2036 + }, + { + "epoch": 5.689944134078212, + "grad_norm": 1.3954970836639404, + "learning_rate": 6.817877094972067e-06, + "loss": 0.2944, + "step": 2037 + }, + { + "epoch": 5.692737430167598, + "grad_norm": 2.0753307342529297, + "learning_rate": 6.8212290502793304e-06, + "loss": 0.345, + "step": 2038 + }, + { + "epoch": 5.695530726256983, + "grad_norm": 2.4927005767822266, + "learning_rate": 6.824581005586593e-06, + "loss": 0.4013, + "step": 2039 + }, + { + "epoch": 5.698324022346369, + "grad_norm": 5.110206604003906, + "learning_rate": 6.827932960893855e-06, + "loss": 0.436, + "step": 2040 + }, + { + "epoch": 5.701117318435754, + "grad_norm": 0.6868772506713867, + "learning_rate": 6.831284916201118e-06, + "loss": 0.2418, + "step": 2041 + }, + { + "epoch": 5.70391061452514, + "grad_norm": 0.5611029863357544, + "learning_rate": 6.83463687150838e-06, + "loss": 0.2306, + "step": 2042 + }, + { + "epoch": 5.706703910614525, + "grad_norm": 0.7051751017570496, + "learning_rate": 6.837988826815643e-06, + "loss": 0.2173, + "step": 2043 + }, + { + "epoch": 5.709497206703911, + "grad_norm": 0.9078261852264404, + "learning_rate": 6.841340782122905e-06, + "loss": 0.2625, + "step": 2044 + }, + { + "epoch": 5.712290502793296, + "grad_norm": 0.744576632976532, + "learning_rate": 6.8446927374301675e-06, + "loss": 0.2828, + "step": 2045 + }, + { + "epoch": 5.715083798882682, + "grad_norm": 0.9236341714859009, + "learning_rate": 6.8480446927374305e-06, + "loss": 0.2681, + "step": 2046 + }, + { + "epoch": 5.717877094972067, + "grad_norm": 2.7060635089874268, + "learning_rate": 6.851396648044693e-06, + "loss": 0.2369, + "step": 2047 + }, + { + "epoch": 5.720670391061453, + "grad_norm": 0.9518362879753113, + "learning_rate": 6.854748603351956e-06, + "loss": 0.2602, + "step": 2048 + }, + { + "epoch": 5.723463687150838, + "grad_norm": 1.6614538431167603, + "learning_rate": 6.858100558659218e-06, + "loss": 0.2624, + "step": 2049 + }, + { + "epoch": 5.726256983240224, + "grad_norm": 1.1964311599731445, + "learning_rate": 6.86145251396648e-06, + "loss": 0.2935, + "step": 2050 + }, + { + "epoch": 5.729050279329609, + "grad_norm": 0.9713432788848877, + "learning_rate": 6.864804469273743e-06, + "loss": 0.2694, + "step": 2051 + }, + { + "epoch": 5.731843575418995, + "grad_norm": 1.1331589221954346, + "learning_rate": 6.868156424581005e-06, + "loss": 0.3024, + "step": 2052 + }, + { + "epoch": 5.73463687150838, + "grad_norm": 0.8688535094261169, + "learning_rate": 6.871508379888268e-06, + "loss": 0.2637, + "step": 2053 + }, + { + "epoch": 5.737430167597766, + "grad_norm": 0.9312444925308228, + "learning_rate": 6.874860335195531e-06, + "loss": 0.2917, + "step": 2054 + }, + { + "epoch": 5.740223463687151, + "grad_norm": 1.364878535270691, + "learning_rate": 6.878212290502794e-06, + "loss": 0.2556, + "step": 2055 + }, + { + "epoch": 5.743016759776537, + "grad_norm": 1.040281891822815, + "learning_rate": 6.881564245810056e-06, + "loss": 0.2564, + "step": 2056 + }, + { + "epoch": 5.745810055865922, + "grad_norm": 1.1204071044921875, + "learning_rate": 6.884916201117318e-06, + "loss": 0.2961, + "step": 2057 + }, + { + "epoch": 5.748603351955307, + "grad_norm": 1.1044206619262695, + "learning_rate": 6.888268156424581e-06, + "loss": 0.2635, + "step": 2058 + }, + { + "epoch": 5.751396648044693, + "grad_norm": 1.3795826435089111, + "learning_rate": 6.891620111731843e-06, + "loss": 0.3039, + "step": 2059 + }, + { + "epoch": 5.754189944134078, + "grad_norm": 1.735266923904419, + "learning_rate": 6.894972067039106e-06, + "loss": 0.2673, + "step": 2060 + }, + { + "epoch": 5.756983240223463, + "grad_norm": 2.28765606880188, + "learning_rate": 6.898324022346369e-06, + "loss": 0.3109, + "step": 2061 + }, + { + "epoch": 5.759776536312849, + "grad_norm": 2.5249814987182617, + "learning_rate": 6.9016759776536315e-06, + "loss": 0.2887, + "step": 2062 + }, + { + "epoch": 5.762569832402234, + "grad_norm": 3.8563451766967773, + "learning_rate": 6.9050279329608946e-06, + "loss": 0.3171, + "step": 2063 + }, + { + "epoch": 5.76536312849162, + "grad_norm": 2.6447842121124268, + "learning_rate": 6.908379888268157e-06, + "loss": 0.3598, + "step": 2064 + }, + { + "epoch": 5.768156424581005, + "grad_norm": 4.488938331604004, + "learning_rate": 6.91173184357542e-06, + "loss": 0.4862, + "step": 2065 + }, + { + "epoch": 5.770949720670391, + "grad_norm": 2.5192551612854004, + "learning_rate": 6.915083798882682e-06, + "loss": 0.3045, + "step": 2066 + }, + { + "epoch": 5.773743016759776, + "grad_norm": 0.738484263420105, + "learning_rate": 6.918435754189944e-06, + "loss": 0.2819, + "step": 2067 + }, + { + "epoch": 5.776536312849162, + "grad_norm": 0.8400179743766785, + "learning_rate": 6.921787709497207e-06, + "loss": 0.2431, + "step": 2068 + }, + { + "epoch": 5.779329608938547, + "grad_norm": 0.848404049873352, + "learning_rate": 6.925139664804469e-06, + "loss": 0.2502, + "step": 2069 + }, + { + "epoch": 5.782122905027933, + "grad_norm": 1.1590913534164429, + "learning_rate": 6.9284916201117325e-06, + "loss": 0.274, + "step": 2070 + }, + { + "epoch": 5.784916201117318, + "grad_norm": 1.0684713125228882, + "learning_rate": 6.931843575418995e-06, + "loss": 0.2432, + "step": 2071 + }, + { + "epoch": 5.787709497206704, + "grad_norm": 0.9052929282188416, + "learning_rate": 6.935195530726257e-06, + "loss": 0.2568, + "step": 2072 + }, + { + "epoch": 5.790502793296089, + "grad_norm": 0.9717419147491455, + "learning_rate": 6.93854748603352e-06, + "loss": 0.3103, + "step": 2073 + }, + { + "epoch": 5.793296089385475, + "grad_norm": 0.732385516166687, + "learning_rate": 6.941899441340782e-06, + "loss": 0.232, + "step": 2074 + }, + { + "epoch": 5.79608938547486, + "grad_norm": 0.927345335483551, + "learning_rate": 6.945251396648045e-06, + "loss": 0.2591, + "step": 2075 + }, + { + "epoch": 5.798882681564246, + "grad_norm": 1.27933931350708, + "learning_rate": 6.948603351955307e-06, + "loss": 0.2517, + "step": 2076 + }, + { + "epoch": 5.801675977653631, + "grad_norm": 2.439098834991455, + "learning_rate": 6.9519553072625695e-06, + "loss": 0.2455, + "step": 2077 + }, + { + "epoch": 5.804469273743017, + "grad_norm": 1.1038702726364136, + "learning_rate": 6.9553072625698325e-06, + "loss": 0.2602, + "step": 2078 + }, + { + "epoch": 5.807262569832402, + "grad_norm": 1.0192159414291382, + "learning_rate": 6.958659217877095e-06, + "loss": 0.2927, + "step": 2079 + }, + { + "epoch": 5.810055865921788, + "grad_norm": 1.4004619121551514, + "learning_rate": 6.962011173184358e-06, + "loss": 0.2683, + "step": 2080 + }, + { + "epoch": 5.812849162011173, + "grad_norm": 1.2071666717529297, + "learning_rate": 6.96536312849162e-06, + "loss": 0.2935, + "step": 2081 + }, + { + "epoch": 5.815642458100559, + "grad_norm": 1.5014749765396118, + "learning_rate": 6.968715083798882e-06, + "loss": 0.2766, + "step": 2082 + }, + { + "epoch": 5.818435754189944, + "grad_norm": 1.6254545450210571, + "learning_rate": 6.972067039106145e-06, + "loss": 0.281, + "step": 2083 + }, + { + "epoch": 5.82122905027933, + "grad_norm": 1.2197872400283813, + "learning_rate": 6.975418994413407e-06, + "loss": 0.2784, + "step": 2084 + }, + { + "epoch": 5.824022346368715, + "grad_norm": 1.5583257675170898, + "learning_rate": 6.978770949720671e-06, + "loss": 0.2869, + "step": 2085 + }, + { + "epoch": 5.826815642458101, + "grad_norm": 2.3515028953552246, + "learning_rate": 6.9821229050279335e-06, + "loss": 0.2857, + "step": 2086 + }, + { + "epoch": 5.829608938547486, + "grad_norm": 6.81102991104126, + "learning_rate": 6.985474860335196e-06, + "loss": 0.3271, + "step": 2087 + }, + { + "epoch": 5.832402234636872, + "grad_norm": 2.3442203998565674, + "learning_rate": 6.988826815642459e-06, + "loss": 0.2982, + "step": 2088 + }, + { + "epoch": 5.835195530726257, + "grad_norm": 2.1208465099334717, + "learning_rate": 6.992178770949721e-06, + "loss": 0.3272, + "step": 2089 + }, + { + "epoch": 5.837988826815643, + "grad_norm": 4.333054542541504, + "learning_rate": 6.995530726256984e-06, + "loss": 0.5245, + "step": 2090 + }, + { + "epoch": 5.840782122905028, + "grad_norm": 0.8545054197311401, + "learning_rate": 6.998882681564246e-06, + "loss": 0.2397, + "step": 2091 + }, + { + "epoch": 5.843575418994414, + "grad_norm": 0.7762572169303894, + "learning_rate": 7.002234636871508e-06, + "loss": 0.2472, + "step": 2092 + }, + { + "epoch": 5.846368715083798, + "grad_norm": 0.6536039710044861, + "learning_rate": 7.005586592178771e-06, + "loss": 0.2544, + "step": 2093 + }, + { + "epoch": 5.849162011173185, + "grad_norm": 0.6678842902183533, + "learning_rate": 7.0089385474860335e-06, + "loss": 0.2597, + "step": 2094 + }, + { + "epoch": 5.851955307262569, + "grad_norm": 0.7839565873146057, + "learning_rate": 7.012290502793297e-06, + "loss": 0.2812, + "step": 2095 + }, + { + "epoch": 5.854748603351956, + "grad_norm": 1.0155751705169678, + "learning_rate": 7.015642458100559e-06, + "loss": 0.2847, + "step": 2096 + }, + { + "epoch": 5.85754189944134, + "grad_norm": 0.8711690902709961, + "learning_rate": 7.018994413407821e-06, + "loss": 0.2954, + "step": 2097 + }, + { + "epoch": 5.860335195530726, + "grad_norm": 1.1747411489486694, + "learning_rate": 7.022346368715084e-06, + "loss": 0.2595, + "step": 2098 + }, + { + "epoch": 5.863128491620111, + "grad_norm": 1.075469970703125, + "learning_rate": 7.025698324022346e-06, + "loss": 0.2589, + "step": 2099 + }, + { + "epoch": 5.865921787709497, + "grad_norm": 1.3710185289382935, + "learning_rate": 7.029050279329609e-06, + "loss": 0.2669, + "step": 2100 + }, + { + "epoch": 5.868715083798882, + "grad_norm": 0.8869255781173706, + "learning_rate": 7.0324022346368714e-06, + "loss": 0.2826, + "step": 2101 + }, + { + "epoch": 5.871508379888268, + "grad_norm": 1.3674373626708984, + "learning_rate": 7.0357541899441345e-06, + "loss": 0.2879, + "step": 2102 + }, + { + "epoch": 5.874301675977653, + "grad_norm": 1.3655052185058594, + "learning_rate": 7.039106145251397e-06, + "loss": 0.2962, + "step": 2103 + }, + { + "epoch": 5.877094972067039, + "grad_norm": 1.1838172674179077, + "learning_rate": 7.042458100558659e-06, + "loss": 0.2681, + "step": 2104 + }, + { + "epoch": 5.879888268156424, + "grad_norm": 1.0533887147903442, + "learning_rate": 7.045810055865922e-06, + "loss": 0.3204, + "step": 2105 + }, + { + "epoch": 5.88268156424581, + "grad_norm": 1.4076406955718994, + "learning_rate": 7.049162011173184e-06, + "loss": 0.2705, + "step": 2106 + }, + { + "epoch": 5.885474860335195, + "grad_norm": 1.7825850248336792, + "learning_rate": 7.052513966480447e-06, + "loss": 0.2741, + "step": 2107 + }, + { + "epoch": 5.888268156424581, + "grad_norm": 1.5789629220962524, + "learning_rate": 7.055865921787709e-06, + "loss": 0.2928, + "step": 2108 + }, + { + "epoch": 5.891061452513966, + "grad_norm": 1.201353669166565, + "learning_rate": 7.059217877094972e-06, + "loss": 0.2848, + "step": 2109 + }, + { + "epoch": 5.893854748603352, + "grad_norm": 1.4042176008224487, + "learning_rate": 7.062569832402235e-06, + "loss": 0.3385, + "step": 2110 + }, + { + "epoch": 5.896648044692737, + "grad_norm": 1.335472583770752, + "learning_rate": 7.065921787709498e-06, + "loss": 0.2933, + "step": 2111 + }, + { + "epoch": 5.899441340782123, + "grad_norm": 1.7052851915359497, + "learning_rate": 7.069273743016761e-06, + "loss": 0.322, + "step": 2112 + }, + { + "epoch": 5.902234636871508, + "grad_norm": 2.5387299060821533, + "learning_rate": 7.072625698324023e-06, + "loss": 0.3339, + "step": 2113 + }, + { + "epoch": 5.905027932960894, + "grad_norm": 5.469228267669678, + "learning_rate": 7.075977653631285e-06, + "loss": 0.3758, + "step": 2114 + }, + { + "epoch": 5.907821229050279, + "grad_norm": 2.2529990673065186, + "learning_rate": 7.079329608938548e-06, + "loss": 0.4084, + "step": 2115 + }, + { + "epoch": 5.910614525139665, + "grad_norm": 0.6878659129142761, + "learning_rate": 7.08268156424581e-06, + "loss": 0.2666, + "step": 2116 + }, + { + "epoch": 5.91340782122905, + "grad_norm": 0.679978609085083, + "learning_rate": 7.086033519553073e-06, + "loss": 0.2734, + "step": 2117 + }, + { + "epoch": 5.916201117318436, + "grad_norm": 0.7094521522521973, + "learning_rate": 7.0893854748603355e-06, + "loss": 0.2625, + "step": 2118 + }, + { + "epoch": 5.918994413407821, + "grad_norm": 1.1350759267807007, + "learning_rate": 7.092737430167598e-06, + "loss": 0.2712, + "step": 2119 + }, + { + "epoch": 5.921787709497207, + "grad_norm": 1.8431992530822754, + "learning_rate": 7.096089385474861e-06, + "loss": 0.275, + "step": 2120 + }, + { + "epoch": 5.924581005586592, + "grad_norm": 1.011098861694336, + "learning_rate": 7.099441340782123e-06, + "loss": 0.2501, + "step": 2121 + }, + { + "epoch": 5.927374301675978, + "grad_norm": 0.9488484263420105, + "learning_rate": 7.102793296089386e-06, + "loss": 0.261, + "step": 2122 + }, + { + "epoch": 5.930167597765363, + "grad_norm": 0.8140328526496887, + "learning_rate": 7.106145251396648e-06, + "loss": 0.2257, + "step": 2123 + }, + { + "epoch": 5.932960893854749, + "grad_norm": 0.7470210194587708, + "learning_rate": 7.10949720670391e-06, + "loss": 0.2646, + "step": 2124 + }, + { + "epoch": 5.935754189944134, + "grad_norm": 0.9226143956184387, + "learning_rate": 7.112849162011173e-06, + "loss": 0.232, + "step": 2125 + }, + { + "epoch": 5.93854748603352, + "grad_norm": 0.7682480812072754, + "learning_rate": 7.1162011173184355e-06, + "loss": 0.2317, + "step": 2126 + }, + { + "epoch": 5.941340782122905, + "grad_norm": 1.13335120677948, + "learning_rate": 7.119553072625699e-06, + "loss": 0.2595, + "step": 2127 + }, + { + "epoch": 5.94413407821229, + "grad_norm": 0.9354434609413147, + "learning_rate": 7.122905027932961e-06, + "loss": 0.2438, + "step": 2128 + }, + { + "epoch": 5.946927374301676, + "grad_norm": 1.469827651977539, + "learning_rate": 7.126256983240223e-06, + "loss": 0.2311, + "step": 2129 + }, + { + "epoch": 5.949720670391061, + "grad_norm": 1.4485294818878174, + "learning_rate": 7.129608938547486e-06, + "loss": 0.2545, + "step": 2130 + }, + { + "epoch": 5.952513966480447, + "grad_norm": 1.9421591758728027, + "learning_rate": 7.132960893854748e-06, + "loss": 0.2664, + "step": 2131 + }, + { + "epoch": 5.955307262569832, + "grad_norm": 1.1769770383834839, + "learning_rate": 7.136312849162011e-06, + "loss": 0.2632, + "step": 2132 + }, + { + "epoch": 5.9581005586592175, + "grad_norm": 1.1700462102890015, + "learning_rate": 7.139664804469274e-06, + "loss": 0.2978, + "step": 2133 + }, + { + "epoch": 5.960893854748603, + "grad_norm": 3.313950300216675, + "learning_rate": 7.1430167597765365e-06, + "loss": 0.329, + "step": 2134 + }, + { + "epoch": 5.9636871508379885, + "grad_norm": 1.7780276536941528, + "learning_rate": 7.1463687150837995e-06, + "loss": 0.3294, + "step": 2135 + }, + { + "epoch": 5.966480446927374, + "grad_norm": 1.6255534887313843, + "learning_rate": 7.149720670391062e-06, + "loss": 0.3505, + "step": 2136 + }, + { + "epoch": 5.9692737430167595, + "grad_norm": 1.9375156164169312, + "learning_rate": 7.153072625698325e-06, + "loss": 0.3263, + "step": 2137 + }, + { + "epoch": 5.972067039106145, + "grad_norm": 2.1164588928222656, + "learning_rate": 7.156424581005587e-06, + "loss": 0.3463, + "step": 2138 + }, + { + "epoch": 5.9748603351955305, + "grad_norm": 1.6237902641296387, + "learning_rate": 7.159776536312849e-06, + "loss": 0.3904, + "step": 2139 + }, + { + "epoch": 5.977653631284916, + "grad_norm": 3.2170326709747314, + "learning_rate": 7.163128491620112e-06, + "loss": 0.4162, + "step": 2140 + }, + { + "epoch": 5.9804469273743015, + "grad_norm": 1.1411993503570557, + "learning_rate": 7.166480446927374e-06, + "loss": 0.3088, + "step": 2141 + }, + { + "epoch": 5.983240223463687, + "grad_norm": 0.8384201526641846, + "learning_rate": 7.169832402234637e-06, + "loss": 0.2533, + "step": 2142 + }, + { + "epoch": 5.9860335195530725, + "grad_norm": 0.7205768823623657, + "learning_rate": 7.1731843575419e-06, + "loss": 0.2742, + "step": 2143 + }, + { + "epoch": 5.988826815642458, + "grad_norm": 1.0364054441452026, + "learning_rate": 7.176536312849162e-06, + "loss": 0.242, + "step": 2144 + }, + { + "epoch": 5.9916201117318435, + "grad_norm": 1.1900904178619385, + "learning_rate": 7.179888268156425e-06, + "loss": 0.2857, + "step": 2145 + }, + { + "epoch": 5.994413407821229, + "grad_norm": 1.067307949066162, + "learning_rate": 7.183240223463687e-06, + "loss": 0.2685, + "step": 2146 + }, + { + "epoch": 5.9972067039106145, + "grad_norm": 2.051083564758301, + "learning_rate": 7.18659217877095e-06, + "loss": 0.3221, + "step": 2147 + }, + { + "epoch": 6.0, + "grad_norm": 2.0260848999023438, + "learning_rate": 7.189944134078212e-06, + "loss": 0.4065, + "step": 2148 + }, + { + "epoch": 6.0027932960893855, + "grad_norm": 0.6674203872680664, + "learning_rate": 7.193296089385475e-06, + "loss": 0.2151, + "step": 2149 + }, + { + "epoch": 6.005586592178771, + "grad_norm": 0.7570977210998535, + "learning_rate": 7.1966480446927375e-06, + "loss": 0.2325, + "step": 2150 + }, + { + "epoch": 6.0083798882681565, + "grad_norm": 0.7511481046676636, + "learning_rate": 7.2e-06, + "loss": 0.2576, + "step": 2151 + }, + { + "epoch": 6.011173184357542, + "grad_norm": 0.5580072999000549, + "learning_rate": 7.203351955307263e-06, + "loss": 0.2217, + "step": 2152 + }, + { + "epoch": 6.0139664804469275, + "grad_norm": 0.7397994995117188, + "learning_rate": 7.206703910614525e-06, + "loss": 0.2446, + "step": 2153 + }, + { + "epoch": 6.016759776536313, + "grad_norm": 0.7719267010688782, + "learning_rate": 7.210055865921788e-06, + "loss": 0.2267, + "step": 2154 + }, + { + "epoch": 6.0195530726256985, + "grad_norm": 0.8569240570068359, + "learning_rate": 7.21340782122905e-06, + "loss": 0.25, + "step": 2155 + }, + { + "epoch": 6.022346368715084, + "grad_norm": 0.7992863655090332, + "learning_rate": 7.216759776536312e-06, + "loss": 0.2573, + "step": 2156 + }, + { + "epoch": 6.0251396648044695, + "grad_norm": 0.8160316944122314, + "learning_rate": 7.220111731843576e-06, + "loss": 0.2129, + "step": 2157 + }, + { + "epoch": 6.027932960893855, + "grad_norm": 0.8069496750831604, + "learning_rate": 7.223463687150838e-06, + "loss": 0.2766, + "step": 2158 + }, + { + "epoch": 6.0307262569832405, + "grad_norm": 0.8967238664627075, + "learning_rate": 7.2268156424581014e-06, + "loss": 0.2544, + "step": 2159 + }, + { + "epoch": 6.033519553072626, + "grad_norm": 2.4012157917022705, + "learning_rate": 7.230167597765364e-06, + "loss": 0.2748, + "step": 2160 + }, + { + "epoch": 6.0363128491620115, + "grad_norm": 0.9489363431930542, + "learning_rate": 7.233519553072626e-06, + "loss": 0.2654, + "step": 2161 + }, + { + "epoch": 6.039106145251397, + "grad_norm": 0.9613003730773926, + "learning_rate": 7.236871508379889e-06, + "loss": 0.2734, + "step": 2162 + }, + { + "epoch": 6.0418994413407825, + "grad_norm": 0.9921791553497314, + "learning_rate": 7.240223463687151e-06, + "loss": 0.2806, + "step": 2163 + }, + { + "epoch": 6.044692737430168, + "grad_norm": 1.0521907806396484, + "learning_rate": 7.243575418994414e-06, + "loss": 0.2829, + "step": 2164 + }, + { + "epoch": 6.0474860335195535, + "grad_norm": 1.749269723892212, + "learning_rate": 7.246927374301676e-06, + "loss": 0.2496, + "step": 2165 + }, + { + "epoch": 6.050279329608939, + "grad_norm": 1.5117545127868652, + "learning_rate": 7.2502793296089385e-06, + "loss": 0.2854, + "step": 2166 + }, + { + "epoch": 6.053072625698324, + "grad_norm": 1.4560297727584839, + "learning_rate": 7.2536312849162015e-06, + "loss": 0.3084, + "step": 2167 + }, + { + "epoch": 6.055865921787709, + "grad_norm": 1.2727333307266235, + "learning_rate": 7.256983240223464e-06, + "loss": 0.2612, + "step": 2168 + }, + { + "epoch": 6.058659217877095, + "grad_norm": 1.3676636219024658, + "learning_rate": 7.260335195530727e-06, + "loss": 0.2948, + "step": 2169 + }, + { + "epoch": 6.06145251396648, + "grad_norm": 1.3633805513381958, + "learning_rate": 7.263687150837989e-06, + "loss": 0.2659, + "step": 2170 + }, + { + "epoch": 6.064245810055866, + "grad_norm": 2.4130022525787354, + "learning_rate": 7.267039106145251e-06, + "loss": 0.339, + "step": 2171 + }, + { + "epoch": 6.067039106145251, + "grad_norm": 1.7509381771087646, + "learning_rate": 7.270391061452514e-06, + "loss": 0.3818, + "step": 2172 + }, + { + "epoch": 6.069832402234637, + "grad_norm": 3.350858211517334, + "learning_rate": 7.273743016759776e-06, + "loss": 0.4546, + "step": 2173 + }, + { + "epoch": 6.072625698324022, + "grad_norm": 0.9328500628471375, + "learning_rate": 7.277094972067039e-06, + "loss": 0.2605, + "step": 2174 + }, + { + "epoch": 6.075418994413408, + "grad_norm": 1.144251823425293, + "learning_rate": 7.280446927374302e-06, + "loss": 0.2535, + "step": 2175 + }, + { + "epoch": 6.078212290502793, + "grad_norm": 1.0755751132965088, + "learning_rate": 7.283798882681564e-06, + "loss": 0.2403, + "step": 2176 + }, + { + "epoch": 6.081005586592179, + "grad_norm": 0.7465524077415466, + "learning_rate": 7.287150837988827e-06, + "loss": 0.2592, + "step": 2177 + }, + { + "epoch": 6.083798882681564, + "grad_norm": 1.3771581649780273, + "learning_rate": 7.290502793296089e-06, + "loss": 0.2364, + "step": 2178 + }, + { + "epoch": 6.08659217877095, + "grad_norm": 0.8130316734313965, + "learning_rate": 7.293854748603352e-06, + "loss": 0.2546, + "step": 2179 + }, + { + "epoch": 6.089385474860335, + "grad_norm": 0.9218255281448364, + "learning_rate": 7.297206703910614e-06, + "loss": 0.2502, + "step": 2180 + }, + { + "epoch": 6.092178770949721, + "grad_norm": 1.0040212869644165, + "learning_rate": 7.300558659217877e-06, + "loss": 0.2736, + "step": 2181 + }, + { + "epoch": 6.094972067039106, + "grad_norm": 0.8594334125518799, + "learning_rate": 7.30391061452514e-06, + "loss": 0.2804, + "step": 2182 + }, + { + "epoch": 6.097765363128492, + "grad_norm": 0.8601797223091125, + "learning_rate": 7.3072625698324025e-06, + "loss": 0.2736, + "step": 2183 + }, + { + "epoch": 6.100558659217877, + "grad_norm": 1.077185034751892, + "learning_rate": 7.3106145251396656e-06, + "loss": 0.2795, + "step": 2184 + }, + { + "epoch": 6.103351955307263, + "grad_norm": 1.2119171619415283, + "learning_rate": 7.313966480446928e-06, + "loss": 0.2638, + "step": 2185 + }, + { + "epoch": 6.106145251396648, + "grad_norm": 0.9618368744850159, + "learning_rate": 7.31731843575419e-06, + "loss": 0.2504, + "step": 2186 + }, + { + "epoch": 6.108938547486034, + "grad_norm": 0.9121984839439392, + "learning_rate": 7.320670391061453e-06, + "loss": 0.279, + "step": 2187 + }, + { + "epoch": 6.111731843575419, + "grad_norm": 0.8849733471870422, + "learning_rate": 7.324022346368715e-06, + "loss": 0.1993, + "step": 2188 + }, + { + "epoch": 6.114525139664805, + "grad_norm": 1.1839087009429932, + "learning_rate": 7.327374301675978e-06, + "loss": 0.2628, + "step": 2189 + }, + { + "epoch": 6.11731843575419, + "grad_norm": 1.58015775680542, + "learning_rate": 7.33072625698324e-06, + "loss": 0.2726, + "step": 2190 + }, + { + "epoch": 6.120111731843576, + "grad_norm": 1.0803951025009155, + "learning_rate": 7.334078212290503e-06, + "loss": 0.2448, + "step": 2191 + }, + { + "epoch": 6.122905027932961, + "grad_norm": 6.596610069274902, + "learning_rate": 7.337430167597766e-06, + "loss": 0.3224, + "step": 2192 + }, + { + "epoch": 6.125698324022347, + "grad_norm": 1.174576997756958, + "learning_rate": 7.340782122905028e-06, + "loss": 0.2742, + "step": 2193 + }, + { + "epoch": 6.128491620111732, + "grad_norm": 1.8159934282302856, + "learning_rate": 7.344134078212291e-06, + "loss": 0.3337, + "step": 2194 + }, + { + "epoch": 6.131284916201118, + "grad_norm": 1.8854289054870605, + "learning_rate": 7.347486033519553e-06, + "loss": 0.2987, + "step": 2195 + }, + { + "epoch": 6.134078212290503, + "grad_norm": 1.6136109828948975, + "learning_rate": 7.350837988826816e-06, + "loss": 0.3095, + "step": 2196 + }, + { + "epoch": 6.136871508379889, + "grad_norm": 1.7766263484954834, + "learning_rate": 7.354189944134078e-06, + "loss": 0.3226, + "step": 2197 + }, + { + "epoch": 6.139664804469274, + "grad_norm": 3.249023914337158, + "learning_rate": 7.3575418994413405e-06, + "loss": 0.4942, + "step": 2198 + }, + { + "epoch": 6.14245810055866, + "grad_norm": 0.7257541418075562, + "learning_rate": 7.3608938547486035e-06, + "loss": 0.2589, + "step": 2199 + }, + { + "epoch": 6.145251396648045, + "grad_norm": 0.6266637444496155, + "learning_rate": 7.364245810055866e-06, + "loss": 0.2307, + "step": 2200 + }, + { + "epoch": 6.148044692737431, + "grad_norm": 0.7693330645561218, + "learning_rate": 7.367597765363129e-06, + "loss": 0.2633, + "step": 2201 + }, + { + "epoch": 6.150837988826815, + "grad_norm": 0.8122349977493286, + "learning_rate": 7.370949720670391e-06, + "loss": 0.3035, + "step": 2202 + }, + { + "epoch": 6.153631284916201, + "grad_norm": 47.300254821777344, + "learning_rate": 7.374301675977653e-06, + "loss": 0.3059, + "step": 2203 + }, + { + "epoch": 6.156424581005586, + "grad_norm": 0.798048734664917, + "learning_rate": 7.377653631284916e-06, + "loss": 0.2256, + "step": 2204 + }, + { + "epoch": 6.159217877094972, + "grad_norm": 1.2339509725570679, + "learning_rate": 7.381005586592179e-06, + "loss": 0.2638, + "step": 2205 + }, + { + "epoch": 6.162011173184357, + "grad_norm": 0.6825451850891113, + "learning_rate": 7.384357541899442e-06, + "loss": 0.246, + "step": 2206 + }, + { + "epoch": 6.164804469273743, + "grad_norm": 1.00334632396698, + "learning_rate": 7.3877094972067045e-06, + "loss": 0.2459, + "step": 2207 + }, + { + "epoch": 6.167597765363128, + "grad_norm": 2.6108391284942627, + "learning_rate": 7.391061452513967e-06, + "loss": 0.2738, + "step": 2208 + }, + { + "epoch": 6.170391061452514, + "grad_norm": 0.8892171382904053, + "learning_rate": 7.39441340782123e-06, + "loss": 0.2825, + "step": 2209 + }, + { + "epoch": 6.173184357541899, + "grad_norm": 1.0082368850708008, + "learning_rate": 7.397765363128492e-06, + "loss": 0.2437, + "step": 2210 + }, + { + "epoch": 6.175977653631285, + "grad_norm": 0.9910778403282166, + "learning_rate": 7.401117318435755e-06, + "loss": 0.2778, + "step": 2211 + }, + { + "epoch": 6.17877094972067, + "grad_norm": 1.423555612564087, + "learning_rate": 7.404469273743017e-06, + "loss": 0.2554, + "step": 2212 + }, + { + "epoch": 6.181564245810056, + "grad_norm": 1.2521388530731201, + "learning_rate": 7.407821229050279e-06, + "loss": 0.2157, + "step": 2213 + }, + { + "epoch": 6.184357541899441, + "grad_norm": 1.1600899696350098, + "learning_rate": 7.411173184357542e-06, + "loss": 0.2595, + "step": 2214 + }, + { + "epoch": 6.187150837988827, + "grad_norm": 1.1315069198608398, + "learning_rate": 7.4145251396648045e-06, + "loss": 0.2611, + "step": 2215 + }, + { + "epoch": 6.189944134078212, + "grad_norm": 1.6112828254699707, + "learning_rate": 7.417877094972068e-06, + "loss": 0.3003, + "step": 2216 + }, + { + "epoch": 6.192737430167598, + "grad_norm": 1.9910656213760376, + "learning_rate": 7.42122905027933e-06, + "loss": 0.2681, + "step": 2217 + }, + { + "epoch": 6.195530726256983, + "grad_norm": 1.1366863250732422, + "learning_rate": 7.424581005586592e-06, + "loss": 0.2942, + "step": 2218 + }, + { + "epoch": 6.198324022346369, + "grad_norm": 1.9309693574905396, + "learning_rate": 7.427932960893855e-06, + "loss": 0.3235, + "step": 2219 + }, + { + "epoch": 6.201117318435754, + "grad_norm": 1.2944386005401611, + "learning_rate": 7.431284916201117e-06, + "loss": 0.3194, + "step": 2220 + }, + { + "epoch": 6.20391061452514, + "grad_norm": 2.621053457260132, + "learning_rate": 7.43463687150838e-06, + "loss": 0.3, + "step": 2221 + }, + { + "epoch": 6.206703910614525, + "grad_norm": 1.8166555166244507, + "learning_rate": 7.4379888268156424e-06, + "loss": 0.2828, + "step": 2222 + }, + { + "epoch": 6.209497206703911, + "grad_norm": 3.4187123775482178, + "learning_rate": 7.441340782122905e-06, + "loss": 0.4321, + "step": 2223 + }, + { + "epoch": 6.212290502793296, + "grad_norm": 0.7675819993019104, + "learning_rate": 7.444692737430168e-06, + "loss": 0.2494, + "step": 2224 + }, + { + "epoch": 6.215083798882682, + "grad_norm": 0.6779239177703857, + "learning_rate": 7.44804469273743e-06, + "loss": 0.3102, + "step": 2225 + }, + { + "epoch": 6.217877094972067, + "grad_norm": 0.8509693145751953, + "learning_rate": 7.451396648044693e-06, + "loss": 0.2225, + "step": 2226 + }, + { + "epoch": 6.220670391061453, + "grad_norm": 0.7323559522628784, + "learning_rate": 7.454748603351955e-06, + "loss": 0.2571, + "step": 2227 + }, + { + "epoch": 6.223463687150838, + "grad_norm": 0.6218514442443848, + "learning_rate": 7.458100558659217e-06, + "loss": 0.2435, + "step": 2228 + }, + { + "epoch": 6.226256983240224, + "grad_norm": 0.7344107627868652, + "learning_rate": 7.461452513966481e-06, + "loss": 0.3008, + "step": 2229 + }, + { + "epoch": 6.229050279329609, + "grad_norm": 0.7879393696784973, + "learning_rate": 7.464804469273743e-06, + "loss": 0.3076, + "step": 2230 + }, + { + "epoch": 6.231843575418995, + "grad_norm": 0.9306378960609436, + "learning_rate": 7.468156424581006e-06, + "loss": 0.2575, + "step": 2231 + }, + { + "epoch": 6.23463687150838, + "grad_norm": 0.8269739747047424, + "learning_rate": 7.471508379888269e-06, + "loss": 0.2375, + "step": 2232 + }, + { + "epoch": 6.237430167597766, + "grad_norm": 1.805976152420044, + "learning_rate": 7.474860335195531e-06, + "loss": 0.2479, + "step": 2233 + }, + { + "epoch": 6.240223463687151, + "grad_norm": 0.7366883754730225, + "learning_rate": 7.478212290502794e-06, + "loss": 0.2529, + "step": 2234 + }, + { + "epoch": 6.243016759776537, + "grad_norm": 1.9558982849121094, + "learning_rate": 7.481564245810056e-06, + "loss": 0.2212, + "step": 2235 + }, + { + "epoch": 6.245810055865922, + "grad_norm": 0.9245476722717285, + "learning_rate": 7.484916201117319e-06, + "loss": 0.2349, + "step": 2236 + }, + { + "epoch": 6.248603351955307, + "grad_norm": 0.908221960067749, + "learning_rate": 7.488268156424581e-06, + "loss": 0.2494, + "step": 2237 + }, + { + "epoch": 6.251396648044693, + "grad_norm": 1.3782377243041992, + "learning_rate": 7.4916201117318434e-06, + "loss": 0.2976, + "step": 2238 + }, + { + "epoch": 6.254189944134078, + "grad_norm": 0.9188522696495056, + "learning_rate": 7.4949720670391065e-06, + "loss": 0.2331, + "step": 2239 + }, + { + "epoch": 6.256983240223463, + "grad_norm": 0.9659701585769653, + "learning_rate": 7.498324022346369e-06, + "loss": 0.2662, + "step": 2240 + }, + { + "epoch": 6.259776536312849, + "grad_norm": 1.2501978874206543, + "learning_rate": 7.501675977653631e-06, + "loss": 0.2962, + "step": 2241 + }, + { + "epoch": 6.262569832402234, + "grad_norm": 2.7984468936920166, + "learning_rate": 7.505027932960894e-06, + "loss": 0.2842, + "step": 2242 + }, + { + "epoch": 6.26536312849162, + "grad_norm": 1.4299174547195435, + "learning_rate": 7.508379888268157e-06, + "loss": 0.3103, + "step": 2243 + }, + { + "epoch": 6.268156424581005, + "grad_norm": 1.6845965385437012, + "learning_rate": 7.511731843575418e-06, + "loss": 0.2851, + "step": 2244 + }, + { + "epoch": 6.270949720670391, + "grad_norm": 1.5310617685317993, + "learning_rate": 7.515083798882681e-06, + "loss": 0.3146, + "step": 2245 + }, + { + "epoch": 6.273743016759776, + "grad_norm": 2.0364766120910645, + "learning_rate": 7.518435754189944e-06, + "loss": 0.3228, + "step": 2246 + }, + { + "epoch": 6.276536312849162, + "grad_norm": 2.6157684326171875, + "learning_rate": 7.521787709497207e-06, + "loss": 0.3826, + "step": 2247 + }, + { + "epoch": 6.279329608938547, + "grad_norm": 2.6933958530426025, + "learning_rate": 7.525139664804469e-06, + "loss": 0.431, + "step": 2248 + }, + { + "epoch": 6.282122905027933, + "grad_norm": 1.8192555904388428, + "learning_rate": 7.528491620111732e-06, + "loss": 0.2915, + "step": 2249 + }, + { + "epoch": 6.284916201117318, + "grad_norm": 0.6415170431137085, + "learning_rate": 7.531843575418995e-06, + "loss": 0.2404, + "step": 2250 + }, + { + "epoch": 6.287709497206704, + "grad_norm": 0.674987256526947, + "learning_rate": 7.535195530726256e-06, + "loss": 0.2389, + "step": 2251 + }, + { + "epoch": 6.290502793296089, + "grad_norm": 0.8160474896430969, + "learning_rate": 7.538547486033519e-06, + "loss": 0.278, + "step": 2252 + }, + { + "epoch": 6.293296089385475, + "grad_norm": 0.9531134366989136, + "learning_rate": 7.541899441340783e-06, + "loss": 0.2773, + "step": 2253 + }, + { + "epoch": 6.29608938547486, + "grad_norm": 0.7926580309867859, + "learning_rate": 7.5452513966480444e-06, + "loss": 0.2461, + "step": 2254 + }, + { + "epoch": 6.298882681564246, + "grad_norm": 0.7576619982719421, + "learning_rate": 7.5486033519553075e-06, + "loss": 0.2895, + "step": 2255 + }, + { + "epoch": 6.301675977653631, + "grad_norm": 0.9751829504966736, + "learning_rate": 7.5519553072625705e-06, + "loss": 0.2679, + "step": 2256 + }, + { + "epoch": 6.304469273743017, + "grad_norm": 0.930585503578186, + "learning_rate": 7.5553072625698336e-06, + "loss": 0.2371, + "step": 2257 + }, + { + "epoch": 6.307262569832402, + "grad_norm": 1.4522671699523926, + "learning_rate": 7.558659217877095e-06, + "loss": 0.2785, + "step": 2258 + }, + { + "epoch": 6.310055865921788, + "grad_norm": 0.9191986322402954, + "learning_rate": 7.562011173184358e-06, + "loss": 0.3197, + "step": 2259 + }, + { + "epoch": 6.312849162011173, + "grad_norm": 1.0127182006835938, + "learning_rate": 7.565363128491621e-06, + "loss": 0.3055, + "step": 2260 + }, + { + "epoch": 6.315642458100559, + "grad_norm": 0.9640551805496216, + "learning_rate": 7.568715083798882e-06, + "loss": 0.2279, + "step": 2261 + }, + { + "epoch": 6.318435754189944, + "grad_norm": 0.9853895902633667, + "learning_rate": 7.572067039106145e-06, + "loss": 0.244, + "step": 2262 + }, + { + "epoch": 6.32122905027933, + "grad_norm": 0.9309731125831604, + "learning_rate": 7.575418994413408e-06, + "loss": 0.2264, + "step": 2263 + }, + { + "epoch": 6.324022346368715, + "grad_norm": 1.3780181407928467, + "learning_rate": 7.57877094972067e-06, + "loss": 0.2906, + "step": 2264 + }, + { + "epoch": 6.326815642458101, + "grad_norm": 1.0852978229522705, + "learning_rate": 7.582122905027933e-06, + "loss": 0.2498, + "step": 2265 + }, + { + "epoch": 6.329608938547486, + "grad_norm": 1.2889713048934937, + "learning_rate": 7.585474860335196e-06, + "loss": 0.2865, + "step": 2266 + }, + { + "epoch": 6.332402234636872, + "grad_norm": 1.2451090812683105, + "learning_rate": 7.588826815642459e-06, + "loss": 0.267, + "step": 2267 + }, + { + "epoch": 6.335195530726257, + "grad_norm": 1.0736894607543945, + "learning_rate": 7.59217877094972e-06, + "loss": 0.2737, + "step": 2268 + }, + { + "epoch": 6.337988826815643, + "grad_norm": 1.4693222045898438, + "learning_rate": 7.595530726256983e-06, + "loss": 0.2407, + "step": 2269 + }, + { + "epoch": 6.340782122905028, + "grad_norm": 1.2255269289016724, + "learning_rate": 7.598882681564246e-06, + "loss": 0.3037, + "step": 2270 + }, + { + "epoch": 6.343575418994414, + "grad_norm": 1.6936461925506592, + "learning_rate": 7.602234636871508e-06, + "loss": 0.26, + "step": 2271 + }, + { + "epoch": 6.346368715083799, + "grad_norm": 1.8275705575942993, + "learning_rate": 7.605586592178771e-06, + "loss": 0.383, + "step": 2272 + }, + { + "epoch": 6.349162011173185, + "grad_norm": 2.8930904865264893, + "learning_rate": 7.608938547486034e-06, + "loss": 0.477, + "step": 2273 + }, + { + "epoch": 6.351955307262569, + "grad_norm": 0.6241574287414551, + "learning_rate": 7.612290502793297e-06, + "loss": 0.2652, + "step": 2274 + }, + { + "epoch": 6.354748603351955, + "grad_norm": 0.6486414670944214, + "learning_rate": 7.615642458100558e-06, + "loss": 0.2198, + "step": 2275 + }, + { + "epoch": 6.35754189944134, + "grad_norm": 0.7639493346214294, + "learning_rate": 7.618994413407821e-06, + "loss": 0.2775, + "step": 2276 + }, + { + "epoch": 6.360335195530726, + "grad_norm": 0.7557418346405029, + "learning_rate": 7.622346368715085e-06, + "loss": 0.2467, + "step": 2277 + }, + { + "epoch": 6.363128491620111, + "grad_norm": 0.730185866355896, + "learning_rate": 7.625698324022346e-06, + "loss": 0.2516, + "step": 2278 + }, + { + "epoch": 6.365921787709497, + "grad_norm": 1.0833936929702759, + "learning_rate": 7.629050279329609e-06, + "loss": 0.2264, + "step": 2279 + }, + { + "epoch": 6.368715083798882, + "grad_norm": 0.9825930595397949, + "learning_rate": 7.632402234636872e-06, + "loss": 0.2657, + "step": 2280 + }, + { + "epoch": 6.371508379888268, + "grad_norm": 0.7073759436607361, + "learning_rate": 7.635754189944133e-06, + "loss": 0.2298, + "step": 2281 + }, + { + "epoch": 6.374301675977653, + "grad_norm": 0.8036351203918457, + "learning_rate": 7.639106145251396e-06, + "loss": 0.2234, + "step": 2282 + }, + { + "epoch": 6.377094972067039, + "grad_norm": 1.02799391746521, + "learning_rate": 7.642458100558659e-06, + "loss": 0.2642, + "step": 2283 + }, + { + "epoch": 6.379888268156424, + "grad_norm": 2.3131885528564453, + "learning_rate": 7.645810055865922e-06, + "loss": 0.2254, + "step": 2284 + }, + { + "epoch": 6.38268156424581, + "grad_norm": 0.8716560006141663, + "learning_rate": 7.649162011173183e-06, + "loss": 0.2438, + "step": 2285 + }, + { + "epoch": 6.385474860335195, + "grad_norm": 0.944170355796814, + "learning_rate": 7.652513966480446e-06, + "loss": 0.2008, + "step": 2286 + }, + { + "epoch": 6.388268156424581, + "grad_norm": 0.9214158654212952, + "learning_rate": 7.65586592178771e-06, + "loss": 0.2355, + "step": 2287 + }, + { + "epoch": 6.391061452513966, + "grad_norm": 1.422365427017212, + "learning_rate": 7.65921787709497e-06, + "loss": 0.2733, + "step": 2288 + }, + { + "epoch": 6.393854748603352, + "grad_norm": 1.0737468004226685, + "learning_rate": 7.662569832402236e-06, + "loss": 0.2549, + "step": 2289 + }, + { + "epoch": 6.396648044692737, + "grad_norm": 1.3597850799560547, + "learning_rate": 7.665921787709499e-06, + "loss": 0.2701, + "step": 2290 + }, + { + "epoch": 6.399441340782123, + "grad_norm": 1.5490623712539673, + "learning_rate": 7.66927374301676e-06, + "loss": 0.2841, + "step": 2291 + }, + { + "epoch": 6.402234636871508, + "grad_norm": 1.3039532899856567, + "learning_rate": 7.672625698324023e-06, + "loss": 0.2907, + "step": 2292 + }, + { + "epoch": 6.405027932960894, + "grad_norm": 1.7567405700683594, + "learning_rate": 7.675977653631286e-06, + "loss": 0.2817, + "step": 2293 + }, + { + "epoch": 6.407821229050279, + "grad_norm": 1.216231107711792, + "learning_rate": 7.679329608938549e-06, + "loss": 0.2922, + "step": 2294 + }, + { + "epoch": 6.410614525139665, + "grad_norm": 2.2196550369262695, + "learning_rate": 7.68268156424581e-06, + "loss": 0.3496, + "step": 2295 + }, + { + "epoch": 6.41340782122905, + "grad_norm": 1.5434997081756592, + "learning_rate": 7.686033519553073e-06, + "loss": 0.365, + "step": 2296 + }, + { + "epoch": 6.416201117318436, + "grad_norm": 6.480259418487549, + "learning_rate": 7.689385474860337e-06, + "loss": 0.3593, + "step": 2297 + }, + { + "epoch": 6.418994413407821, + "grad_norm": 4.260974407196045, + "learning_rate": 7.692737430167598e-06, + "loss": 0.4344, + "step": 2298 + }, + { + "epoch": 6.421787709497207, + "grad_norm": 0.6312860250473022, + "learning_rate": 7.696089385474861e-06, + "loss": 0.2624, + "step": 2299 + }, + { + "epoch": 6.424581005586592, + "grad_norm": 0.6236568689346313, + "learning_rate": 7.699441340782124e-06, + "loss": 0.2405, + "step": 2300 + }, + { + "epoch": 6.427374301675978, + "grad_norm": 0.5002043843269348, + "learning_rate": 7.702793296089385e-06, + "loss": 0.2125, + "step": 2301 + }, + { + "epoch": 6.430167597765363, + "grad_norm": 0.8223304748535156, + "learning_rate": 7.706145251396648e-06, + "loss": 0.2287, + "step": 2302 + }, + { + "epoch": 6.432960893854749, + "grad_norm": 1.051814079284668, + "learning_rate": 7.709497206703911e-06, + "loss": 0.2327, + "step": 2303 + }, + { + "epoch": 6.435754189944134, + "grad_norm": 0.6217700839042664, + "learning_rate": 7.712849162011174e-06, + "loss": 0.2221, + "step": 2304 + }, + { + "epoch": 6.43854748603352, + "grad_norm": 0.8439684510231018, + "learning_rate": 7.716201117318436e-06, + "loss": 0.2658, + "step": 2305 + }, + { + "epoch": 6.441340782122905, + "grad_norm": 0.7174456119537354, + "learning_rate": 7.719553072625699e-06, + "loss": 0.1983, + "step": 2306 + }, + { + "epoch": 6.444134078212291, + "grad_norm": 0.6004683971405029, + "learning_rate": 7.722905027932962e-06, + "loss": 0.2163, + "step": 2307 + }, + { + "epoch": 6.446927374301676, + "grad_norm": 0.704346776008606, + "learning_rate": 7.726256983240223e-06, + "loss": 0.2518, + "step": 2308 + }, + { + "epoch": 6.449720670391061, + "grad_norm": 0.8626233339309692, + "learning_rate": 7.729608938547486e-06, + "loss": 0.2346, + "step": 2309 + }, + { + "epoch": 6.452513966480447, + "grad_norm": 0.7971252202987671, + "learning_rate": 7.73296089385475e-06, + "loss": 0.2256, + "step": 2310 + }, + { + "epoch": 6.455307262569832, + "grad_norm": 0.78908771276474, + "learning_rate": 7.73631284916201e-06, + "loss": 0.2974, + "step": 2311 + }, + { + "epoch": 6.4581005586592175, + "grad_norm": 0.7469732165336609, + "learning_rate": 7.739664804469274e-06, + "loss": 0.2499, + "step": 2312 + }, + { + "epoch": 6.460893854748603, + "grad_norm": 1.4002691507339478, + "learning_rate": 7.743016759776537e-06, + "loss": 0.2941, + "step": 2313 + }, + { + "epoch": 6.4636871508379885, + "grad_norm": 1.1159323453903198, + "learning_rate": 7.7463687150838e-06, + "loss": 0.2557, + "step": 2314 + }, + { + "epoch": 6.466480446927374, + "grad_norm": 0.9297694563865662, + "learning_rate": 7.749720670391061e-06, + "loss": 0.2704, + "step": 2315 + }, + { + "epoch": 6.4692737430167595, + "grad_norm": 1.1559346914291382, + "learning_rate": 7.753072625698324e-06, + "loss": 0.2438, + "step": 2316 + }, + { + "epoch": 6.472067039106145, + "grad_norm": 1.516727089881897, + "learning_rate": 7.756424581005587e-06, + "loss": 0.24, + "step": 2317 + }, + { + "epoch": 6.4748603351955305, + "grad_norm": 1.0412923097610474, + "learning_rate": 7.759776536312848e-06, + "loss": 0.2586, + "step": 2318 + }, + { + "epoch": 6.477653631284916, + "grad_norm": 1.2830396890640259, + "learning_rate": 7.763128491620112e-06, + "loss": 0.2773, + "step": 2319 + }, + { + "epoch": 6.4804469273743015, + "grad_norm": 1.8285704851150513, + "learning_rate": 7.766480446927375e-06, + "loss": 0.2833, + "step": 2320 + }, + { + "epoch": 6.483240223463687, + "grad_norm": 3.1357343196868896, + "learning_rate": 7.769832402234638e-06, + "loss": 0.3176, + "step": 2321 + }, + { + "epoch": 6.4860335195530725, + "grad_norm": 1.644368052482605, + "learning_rate": 7.773184357541899e-06, + "loss": 0.3597, + "step": 2322 + }, + { + "epoch": 6.488826815642458, + "grad_norm": 3.4724059104919434, + "learning_rate": 7.776536312849162e-06, + "loss": 0.5015, + "step": 2323 + }, + { + "epoch": 6.4916201117318435, + "grad_norm": 0.881718099117279, + "learning_rate": 7.779888268156425e-06, + "loss": 0.2429, + "step": 2324 + }, + { + "epoch": 6.494413407821229, + "grad_norm": 1.8005470037460327, + "learning_rate": 7.783240223463686e-06, + "loss": 0.2113, + "step": 2325 + }, + { + "epoch": 6.4972067039106145, + "grad_norm": 0.5498993396759033, + "learning_rate": 7.78659217877095e-06, + "loss": 0.229, + "step": 2326 + }, + { + "epoch": 6.5, + "grad_norm": 0.8668805360794067, + "learning_rate": 7.789944134078212e-06, + "loss": 0.245, + "step": 2327 + }, + { + "epoch": 6.5027932960893855, + "grad_norm": 0.759845495223999, + "learning_rate": 7.793296089385474e-06, + "loss": 0.2385, + "step": 2328 + }, + { + "epoch": 6.505586592178771, + "grad_norm": 0.7176125049591064, + "learning_rate": 7.796648044692737e-06, + "loss": 0.248, + "step": 2329 + }, + { + "epoch": 6.5083798882681565, + "grad_norm": 0.8777843117713928, + "learning_rate": 7.8e-06, + "loss": 0.2529, + "step": 2330 + }, + { + "epoch": 6.511173184357542, + "grad_norm": 0.666235089302063, + "learning_rate": 7.803351955307263e-06, + "loss": 0.215, + "step": 2331 + }, + { + "epoch": 6.5139664804469275, + "grad_norm": 0.9951727390289307, + "learning_rate": 7.806703910614524e-06, + "loss": 0.254, + "step": 2332 + }, + { + "epoch": 6.516759776536313, + "grad_norm": 1.0236783027648926, + "learning_rate": 7.810055865921787e-06, + "loss": 0.2442, + "step": 2333 + }, + { + "epoch": 6.5195530726256985, + "grad_norm": 0.8996778726577759, + "learning_rate": 7.81340782122905e-06, + "loss": 0.2532, + "step": 2334 + }, + { + "epoch": 6.522346368715084, + "grad_norm": 0.8303968906402588, + "learning_rate": 7.816759776536312e-06, + "loss": 0.2455, + "step": 2335 + }, + { + "epoch": 6.5251396648044695, + "grad_norm": 1.092720866203308, + "learning_rate": 7.820111731843575e-06, + "loss": 0.2745, + "step": 2336 + }, + { + "epoch": 6.527932960893855, + "grad_norm": 0.8832992911338806, + "learning_rate": 7.82346368715084e-06, + "loss": 0.2521, + "step": 2337 + }, + { + "epoch": 6.5307262569832405, + "grad_norm": 0.9645662903785706, + "learning_rate": 7.8268156424581e-06, + "loss": 0.2414, + "step": 2338 + }, + { + "epoch": 6.533519553072626, + "grad_norm": 1.6709059476852417, + "learning_rate": 7.830167597765364e-06, + "loss": 0.2155, + "step": 2339 + }, + { + "epoch": 6.5363128491620115, + "grad_norm": 1.9959568977355957, + "learning_rate": 7.833519553072627e-06, + "loss": 0.2647, + "step": 2340 + }, + { + "epoch": 6.539106145251397, + "grad_norm": 1.1195966005325317, + "learning_rate": 7.83687150837989e-06, + "loss": 0.2702, + "step": 2341 + }, + { + "epoch": 6.5418994413407825, + "grad_norm": 1.2897202968597412, + "learning_rate": 7.840223463687151e-06, + "loss": 0.3075, + "step": 2342 + }, + { + "epoch": 6.544692737430168, + "grad_norm": 1.5463488101959229, + "learning_rate": 7.843575418994414e-06, + "loss": 0.3278, + "step": 2343 + }, + { + "epoch": 6.547486033519553, + "grad_norm": 1.2007213830947876, + "learning_rate": 7.846927374301677e-06, + "loss": 0.2714, + "step": 2344 + }, + { + "epoch": 6.550279329608939, + "grad_norm": 1.320127248764038, + "learning_rate": 7.850279329608939e-06, + "loss": 0.2657, + "step": 2345 + }, + { + "epoch": 6.553072625698324, + "grad_norm": 1.3413342237472534, + "learning_rate": 7.853631284916202e-06, + "loss": 0.3081, + "step": 2346 + }, + { + "epoch": 6.55586592178771, + "grad_norm": 1.4163066148757935, + "learning_rate": 7.856983240223465e-06, + "loss": 0.3643, + "step": 2347 + }, + { + "epoch": 6.558659217877095, + "grad_norm": 2.8471434116363525, + "learning_rate": 7.860335195530726e-06, + "loss": 0.4382, + "step": 2348 + }, + { + "epoch": 6.56145251396648, + "grad_norm": 0.6180393099784851, + "learning_rate": 7.863687150837989e-06, + "loss": 0.2321, + "step": 2349 + }, + { + "epoch": 6.564245810055866, + "grad_norm": 0.7030313611030579, + "learning_rate": 7.867039106145252e-06, + "loss": 0.2429, + "step": 2350 + }, + { + "epoch": 6.567039106145251, + "grad_norm": 0.6093187928199768, + "learning_rate": 7.870391061452515e-06, + "loss": 0.2132, + "step": 2351 + }, + { + "epoch": 6.569832402234637, + "grad_norm": 0.919064462184906, + "learning_rate": 7.873743016759777e-06, + "loss": 0.207, + "step": 2352 + }, + { + "epoch": 6.572625698324022, + "grad_norm": 0.5931017398834229, + "learning_rate": 7.87709497206704e-06, + "loss": 0.2502, + "step": 2353 + }, + { + "epoch": 6.575418994413408, + "grad_norm": 0.7466760873794556, + "learning_rate": 7.880446927374303e-06, + "loss": 0.2551, + "step": 2354 + }, + { + "epoch": 6.578212290502793, + "grad_norm": 0.9994369149208069, + "learning_rate": 7.883798882681564e-06, + "loss": 0.2979, + "step": 2355 + }, + { + "epoch": 6.581005586592179, + "grad_norm": 0.8313371539115906, + "learning_rate": 7.887150837988827e-06, + "loss": 0.2441, + "step": 2356 + }, + { + "epoch": 6.583798882681564, + "grad_norm": 0.7616198658943176, + "learning_rate": 7.89050279329609e-06, + "loss": 0.2475, + "step": 2357 + }, + { + "epoch": 6.58659217877095, + "grad_norm": 0.9754396080970764, + "learning_rate": 7.893854748603351e-06, + "loss": 0.2573, + "step": 2358 + }, + { + "epoch": 6.589385474860335, + "grad_norm": 1.2909948825836182, + "learning_rate": 7.897206703910614e-06, + "loss": 0.2678, + "step": 2359 + }, + { + "epoch": 6.592178770949721, + "grad_norm": 0.8542158007621765, + "learning_rate": 7.900558659217877e-06, + "loss": 0.2393, + "step": 2360 + }, + { + "epoch": 6.594972067039106, + "grad_norm": 1.0143964290618896, + "learning_rate": 7.90391061452514e-06, + "loss": 0.2467, + "step": 2361 + }, + { + "epoch": 6.597765363128492, + "grad_norm": 1.2067657709121704, + "learning_rate": 7.907262569832402e-06, + "loss": 0.2839, + "step": 2362 + }, + { + "epoch": 6.600558659217877, + "grad_norm": 1.1325100660324097, + "learning_rate": 7.910614525139665e-06, + "loss": 0.2456, + "step": 2363 + }, + { + "epoch": 6.603351955307263, + "grad_norm": 1.0284523963928223, + "learning_rate": 7.913966480446928e-06, + "loss": 0.2283, + "step": 2364 + }, + { + "epoch": 6.606145251396648, + "grad_norm": 1.6291004419326782, + "learning_rate": 7.91731843575419e-06, + "loss": 0.2348, + "step": 2365 + }, + { + "epoch": 6.608938547486034, + "grad_norm": 1.0554550886154175, + "learning_rate": 7.920670391061452e-06, + "loss": 0.2779, + "step": 2366 + }, + { + "epoch": 6.611731843575419, + "grad_norm": 1.1165153980255127, + "learning_rate": 7.924022346368715e-06, + "loss": 0.2645, + "step": 2367 + }, + { + "epoch": 6.614525139664805, + "grad_norm": 1.359704613685608, + "learning_rate": 7.927374301675978e-06, + "loss": 0.2805, + "step": 2368 + }, + { + "epoch": 6.61731843575419, + "grad_norm": 1.7631462812423706, + "learning_rate": 7.93072625698324e-06, + "loss": 0.3014, + "step": 2369 + }, + { + "epoch": 6.620111731843576, + "grad_norm": 1.5041688680648804, + "learning_rate": 7.934078212290503e-06, + "loss": 0.307, + "step": 2370 + }, + { + "epoch": 6.622905027932961, + "grad_norm": 1.4722208976745605, + "learning_rate": 7.937430167597766e-06, + "loss": 0.2664, + "step": 2371 + }, + { + "epoch": 6.625698324022347, + "grad_norm": 1.707302212715149, + "learning_rate": 7.940782122905027e-06, + "loss": 0.3047, + "step": 2372 + }, + { + "epoch": 6.628491620111732, + "grad_norm": 2.389885663986206, + "learning_rate": 7.94413407821229e-06, + "loss": 0.4273, + "step": 2373 + }, + { + "epoch": 6.631284916201118, + "grad_norm": 0.8264630436897278, + "learning_rate": 7.947486033519553e-06, + "loss": 0.2744, + "step": 2374 + }, + { + "epoch": 6.634078212290503, + "grad_norm": 1.0649847984313965, + "learning_rate": 7.950837988826815e-06, + "loss": 0.2141, + "step": 2375 + }, + { + "epoch": 6.636871508379889, + "grad_norm": 0.6062030792236328, + "learning_rate": 7.954189944134078e-06, + "loss": 0.2486, + "step": 2376 + }, + { + "epoch": 6.639664804469274, + "grad_norm": 0.62831711769104, + "learning_rate": 7.95754189944134e-06, + "loss": 0.2447, + "step": 2377 + }, + { + "epoch": 6.64245810055866, + "grad_norm": 0.6828584671020508, + "learning_rate": 7.960893854748604e-06, + "loss": 0.2527, + "step": 2378 + }, + { + "epoch": 6.645251396648044, + "grad_norm": 1.0552016496658325, + "learning_rate": 7.964245810055865e-06, + "loss": 0.2498, + "step": 2379 + }, + { + "epoch": 6.648044692737431, + "grad_norm": 0.7988876104354858, + "learning_rate": 7.967597765363128e-06, + "loss": 0.2363, + "step": 2380 + }, + { + "epoch": 6.650837988826815, + "grad_norm": 0.8158085942268372, + "learning_rate": 7.970949720670391e-06, + "loss": 0.2837, + "step": 2381 + }, + { + "epoch": 6.653631284916202, + "grad_norm": 0.7179872989654541, + "learning_rate": 7.974301675977652e-06, + "loss": 0.215, + "step": 2382 + }, + { + "epoch": 6.656424581005586, + "grad_norm": 1.2913224697113037, + "learning_rate": 7.977653631284916e-06, + "loss": 0.2692, + "step": 2383 + }, + { + "epoch": 6.659217877094972, + "grad_norm": 0.7526054382324219, + "learning_rate": 7.981005586592179e-06, + "loss": 0.2513, + "step": 2384 + }, + { + "epoch": 6.662011173184357, + "grad_norm": 1.0224648714065552, + "learning_rate": 7.984357541899442e-06, + "loss": 0.2155, + "step": 2385 + }, + { + "epoch": 6.664804469273743, + "grad_norm": 0.8891600966453552, + "learning_rate": 7.987709497206705e-06, + "loss": 0.244, + "step": 2386 + }, + { + "epoch": 6.667597765363128, + "grad_norm": 0.9150322079658508, + "learning_rate": 7.991061452513968e-06, + "loss": 0.2619, + "step": 2387 + }, + { + "epoch": 6.670391061452514, + "grad_norm": 1.3269736766815186, + "learning_rate": 7.99441340782123e-06, + "loss": 0.2611, + "step": 2388 + }, + { + "epoch": 6.673184357541899, + "grad_norm": 0.8710314631462097, + "learning_rate": 7.997765363128492e-06, + "loss": 0.2665, + "step": 2389 + }, + { + "epoch": 6.675977653631285, + "grad_norm": 3.151125431060791, + "learning_rate": 8.001117318435755e-06, + "loss": 0.249, + "step": 2390 + }, + { + "epoch": 6.67877094972067, + "grad_norm": 1.0233358144760132, + "learning_rate": 8.004469273743018e-06, + "loss": 0.2692, + "step": 2391 + }, + { + "epoch": 6.681564245810056, + "grad_norm": 1.297886610031128, + "learning_rate": 8.00782122905028e-06, + "loss": 0.2766, + "step": 2392 + }, + { + "epoch": 6.684357541899441, + "grad_norm": 2.2539267539978027, + "learning_rate": 8.011173184357543e-06, + "loss": 0.2964, + "step": 2393 + }, + { + "epoch": 6.687150837988827, + "grad_norm": 1.7841030359268188, + "learning_rate": 8.014525139664806e-06, + "loss": 0.297, + "step": 2394 + }, + { + "epoch": 6.689944134078212, + "grad_norm": 1.382910966873169, + "learning_rate": 8.017877094972067e-06, + "loss": 0.2689, + "step": 2395 + }, + { + "epoch": 6.692737430167598, + "grad_norm": 1.5241793394088745, + "learning_rate": 8.02122905027933e-06, + "loss": 0.2723, + "step": 2396 + }, + { + "epoch": 6.695530726256983, + "grad_norm": 4.375805377960205, + "learning_rate": 8.024581005586593e-06, + "loss": 0.3248, + "step": 2397 + }, + { + "epoch": 6.698324022346369, + "grad_norm": 2.8248586654663086, + "learning_rate": 8.027932960893856e-06, + "loss": 0.43, + "step": 2398 + }, + { + "epoch": 6.701117318435754, + "grad_norm": 0.8342903256416321, + "learning_rate": 8.031284916201117e-06, + "loss": 0.2615, + "step": 2399 + }, + { + "epoch": 6.70391061452514, + "grad_norm": 1.0374438762664795, + "learning_rate": 8.03463687150838e-06, + "loss": 0.2996, + "step": 2400 + }, + { + "epoch": 6.706703910614525, + "grad_norm": 3.9716293811798096, + "learning_rate": 8.037988826815643e-06, + "loss": 0.2495, + "step": 2401 + }, + { + "epoch": 6.709497206703911, + "grad_norm": 1.1983615159988403, + "learning_rate": 8.041340782122905e-06, + "loss": 0.2919, + "step": 2402 + }, + { + "epoch": 6.712290502793296, + "grad_norm": 1.1084867715835571, + "learning_rate": 8.044692737430168e-06, + "loss": 0.29, + "step": 2403 + }, + { + "epoch": 6.715083798882682, + "grad_norm": 0.8564739227294922, + "learning_rate": 8.048044692737431e-06, + "loss": 0.2254, + "step": 2404 + }, + { + "epoch": 6.717877094972067, + "grad_norm": 0.6860163807868958, + "learning_rate": 8.051396648044692e-06, + "loss": 0.2382, + "step": 2405 + }, + { + "epoch": 6.720670391061453, + "grad_norm": 0.707258939743042, + "learning_rate": 8.054748603351955e-06, + "loss": 0.2729, + "step": 2406 + }, + { + "epoch": 6.723463687150838, + "grad_norm": 0.6122349500656128, + "learning_rate": 8.058100558659218e-06, + "loss": 0.2379, + "step": 2407 + }, + { + "epoch": 6.726256983240224, + "grad_norm": 0.7053366899490356, + "learning_rate": 8.061452513966481e-06, + "loss": 0.2273, + "step": 2408 + }, + { + "epoch": 6.729050279329609, + "grad_norm": 0.8459175825119019, + "learning_rate": 8.064804469273743e-06, + "loss": 0.2312, + "step": 2409 + }, + { + "epoch": 6.731843575418995, + "grad_norm": 0.921342670917511, + "learning_rate": 8.068156424581006e-06, + "loss": 0.235, + "step": 2410 + }, + { + "epoch": 6.73463687150838, + "grad_norm": 1.1678828001022339, + "learning_rate": 8.071508379888269e-06, + "loss": 0.2626, + "step": 2411 + }, + { + "epoch": 6.737430167597766, + "grad_norm": 1.0094329118728638, + "learning_rate": 8.07486033519553e-06, + "loss": 0.2388, + "step": 2412 + }, + { + "epoch": 6.740223463687151, + "grad_norm": 0.9951021671295166, + "learning_rate": 8.078212290502793e-06, + "loss": 0.2547, + "step": 2413 + }, + { + "epoch": 6.743016759776537, + "grad_norm": 1.0590084791183472, + "learning_rate": 8.081564245810056e-06, + "loss": 0.2749, + "step": 2414 + }, + { + "epoch": 6.745810055865922, + "grad_norm": 1.1764403581619263, + "learning_rate": 8.08491620111732e-06, + "loss": 0.2521, + "step": 2415 + }, + { + "epoch": 6.748603351955307, + "grad_norm": 1.3071478605270386, + "learning_rate": 8.08826815642458e-06, + "loss": 0.237, + "step": 2416 + }, + { + "epoch": 6.751396648044693, + "grad_norm": 1.5600473880767822, + "learning_rate": 8.091620111731844e-06, + "loss": 0.2889, + "step": 2417 + }, + { + "epoch": 6.754189944134078, + "grad_norm": 1.2294938564300537, + "learning_rate": 8.094972067039107e-06, + "loss": 0.27, + "step": 2418 + }, + { + "epoch": 6.756983240223463, + "grad_norm": 1.067320466041565, + "learning_rate": 8.098324022346368e-06, + "loss": 0.2973, + "step": 2419 + }, + { + "epoch": 6.759776536312849, + "grad_norm": 1.3259230852127075, + "learning_rate": 8.101675977653631e-06, + "loss": 0.2836, + "step": 2420 + }, + { + "epoch": 6.762569832402234, + "grad_norm": 1.3542134761810303, + "learning_rate": 8.105027932960894e-06, + "loss": 0.2634, + "step": 2421 + }, + { + "epoch": 6.76536312849162, + "grad_norm": 1.7185776233673096, + "learning_rate": 8.108379888268155e-06, + "loss": 0.335, + "step": 2422 + }, + { + "epoch": 6.768156424581005, + "grad_norm": 2.1864089965820312, + "learning_rate": 8.111731843575418e-06, + "loss": 0.4137, + "step": 2423 + }, + { + "epoch": 6.770949720670391, + "grad_norm": 0.5752627849578857, + "learning_rate": 8.115083798882682e-06, + "loss": 0.2246, + "step": 2424 + }, + { + "epoch": 6.773743016759776, + "grad_norm": 0.6615093350410461, + "learning_rate": 8.118435754189945e-06, + "loss": 0.2542, + "step": 2425 + }, + { + "epoch": 6.776536312849162, + "grad_norm": 0.5666368007659912, + "learning_rate": 8.121787709497206e-06, + "loss": 0.2167, + "step": 2426 + }, + { + "epoch": 6.779329608938547, + "grad_norm": 1.4073940515518188, + "learning_rate": 8.125139664804469e-06, + "loss": 0.2237, + "step": 2427 + }, + { + "epoch": 6.782122905027933, + "grad_norm": 0.7296780347824097, + "learning_rate": 8.128491620111732e-06, + "loss": 0.2564, + "step": 2428 + }, + { + "epoch": 6.784916201117318, + "grad_norm": 0.6092986464500427, + "learning_rate": 8.131843575418993e-06, + "loss": 0.2379, + "step": 2429 + }, + { + "epoch": 6.787709497206704, + "grad_norm": 0.6894410252571106, + "learning_rate": 8.135195530726256e-06, + "loss": 0.2312, + "step": 2430 + }, + { + "epoch": 6.790502793296089, + "grad_norm": 0.7891924381256104, + "learning_rate": 8.13854748603352e-06, + "loss": 0.2472, + "step": 2431 + }, + { + "epoch": 6.793296089385475, + "grad_norm": 1.2291131019592285, + "learning_rate": 8.14189944134078e-06, + "loss": 0.2424, + "step": 2432 + }, + { + "epoch": 6.79608938547486, + "grad_norm": 1.1932799816131592, + "learning_rate": 8.145251396648045e-06, + "loss": 0.2573, + "step": 2433 + }, + { + "epoch": 6.798882681564246, + "grad_norm": 1.1688883304595947, + "learning_rate": 8.148603351955309e-06, + "loss": 0.2705, + "step": 2434 + }, + { + "epoch": 6.801675977653631, + "grad_norm": 1.0601025819778442, + "learning_rate": 8.151955307262572e-06, + "loss": 0.2943, + "step": 2435 + }, + { + "epoch": 6.804469273743017, + "grad_norm": 1.610451340675354, + "learning_rate": 8.155307262569833e-06, + "loss": 0.2305, + "step": 2436 + }, + { + "epoch": 6.807262569832402, + "grad_norm": 0.9473409652709961, + "learning_rate": 8.158659217877096e-06, + "loss": 0.2724, + "step": 2437 + }, + { + "epoch": 6.810055865921788, + "grad_norm": 1.0041630268096924, + "learning_rate": 8.162011173184359e-06, + "loss": 0.282, + "step": 2438 + }, + { + "epoch": 6.812849162011173, + "grad_norm": 1.9253342151641846, + "learning_rate": 8.16536312849162e-06, + "loss": 0.23, + "step": 2439 + }, + { + "epoch": 6.815642458100559, + "grad_norm": 0.9120733737945557, + "learning_rate": 8.168715083798883e-06, + "loss": 0.2598, + "step": 2440 + }, + { + "epoch": 6.818435754189944, + "grad_norm": 1.1191697120666504, + "learning_rate": 8.172067039106146e-06, + "loss": 0.2429, + "step": 2441 + }, + { + "epoch": 6.82122905027933, + "grad_norm": 1.2319564819335938, + "learning_rate": 8.175418994413408e-06, + "loss": 0.282, + "step": 2442 + }, + { + "epoch": 6.824022346368715, + "grad_norm": 1.0104602575302124, + "learning_rate": 8.17877094972067e-06, + "loss": 0.2331, + "step": 2443 + }, + { + "epoch": 6.826815642458101, + "grad_norm": 1.2411385774612427, + "learning_rate": 8.182122905027934e-06, + "loss": 0.2583, + "step": 2444 + }, + { + "epoch": 6.829608938547486, + "grad_norm": 8.61441421508789, + "learning_rate": 8.185474860335197e-06, + "loss": 0.2965, + "step": 2445 + }, + { + "epoch": 6.832402234636872, + "grad_norm": 1.7882484197616577, + "learning_rate": 8.188826815642458e-06, + "loss": 0.3355, + "step": 2446 + }, + { + "epoch": 6.835195530726257, + "grad_norm": 3.8582112789154053, + "learning_rate": 8.192178770949721e-06, + "loss": 0.3532, + "step": 2447 + }, + { + "epoch": 6.837988826815643, + "grad_norm": 1.9028856754302979, + "learning_rate": 8.195530726256984e-06, + "loss": 0.4295, + "step": 2448 + }, + { + "epoch": 6.840782122905028, + "grad_norm": 1.2784637212753296, + "learning_rate": 8.198882681564246e-06, + "loss": 0.2908, + "step": 2449 + }, + { + "epoch": 6.843575418994414, + "grad_norm": 0.5910930633544922, + "learning_rate": 8.202234636871509e-06, + "loss": 0.2531, + "step": 2450 + }, + { + "epoch": 6.846368715083798, + "grad_norm": 0.8645721077919006, + "learning_rate": 8.205586592178772e-06, + "loss": 0.2621, + "step": 2451 + }, + { + "epoch": 6.849162011173185, + "grad_norm": 0.9056130051612854, + "learning_rate": 8.208938547486033e-06, + "loss": 0.2408, + "step": 2452 + }, + { + "epoch": 6.851955307262569, + "grad_norm": 0.8525672554969788, + "learning_rate": 8.212290502793296e-06, + "loss": 0.3166, + "step": 2453 + }, + { + "epoch": 6.854748603351956, + "grad_norm": 0.8727729320526123, + "learning_rate": 8.215642458100559e-06, + "loss": 0.2449, + "step": 2454 + }, + { + "epoch": 6.85754189944134, + "grad_norm": 0.6567447185516357, + "learning_rate": 8.218994413407822e-06, + "loss": 0.2627, + "step": 2455 + }, + { + "epoch": 6.860335195530726, + "grad_norm": 1.2302520275115967, + "learning_rate": 8.222346368715084e-06, + "loss": 0.2368, + "step": 2456 + }, + { + "epoch": 6.863128491620111, + "grad_norm": 0.7202404737472534, + "learning_rate": 8.225698324022347e-06, + "loss": 0.2181, + "step": 2457 + }, + { + "epoch": 6.865921787709497, + "grad_norm": 0.8089212775230408, + "learning_rate": 8.22905027932961e-06, + "loss": 0.2527, + "step": 2458 + }, + { + "epoch": 6.868715083798882, + "grad_norm": 0.9817788004875183, + "learning_rate": 8.232402234636871e-06, + "loss": 0.3072, + "step": 2459 + }, + { + "epoch": 6.871508379888268, + "grad_norm": 1.1296138763427734, + "learning_rate": 8.235754189944134e-06, + "loss": 0.2833, + "step": 2460 + }, + { + "epoch": 6.874301675977653, + "grad_norm": 0.7771025896072388, + "learning_rate": 8.239106145251397e-06, + "loss": 0.2291, + "step": 2461 + }, + { + "epoch": 6.877094972067039, + "grad_norm": 1.1460479497909546, + "learning_rate": 8.24245810055866e-06, + "loss": 0.2608, + "step": 2462 + }, + { + "epoch": 6.879888268156424, + "grad_norm": 0.8554694652557373, + "learning_rate": 8.245810055865921e-06, + "loss": 0.2514, + "step": 2463 + }, + { + "epoch": 6.88268156424581, + "grad_norm": 1.2113595008850098, + "learning_rate": 8.249162011173184e-06, + "loss": 0.3186, + "step": 2464 + }, + { + "epoch": 6.885474860335195, + "grad_norm": 0.9225947856903076, + "learning_rate": 8.252513966480447e-06, + "loss": 0.263, + "step": 2465 + }, + { + "epoch": 6.888268156424581, + "grad_norm": 1.4984246492385864, + "learning_rate": 8.255865921787709e-06, + "loss": 0.2443, + "step": 2466 + }, + { + "epoch": 6.891061452513966, + "grad_norm": 3.7180302143096924, + "learning_rate": 8.259217877094972e-06, + "loss": 0.23, + "step": 2467 + }, + { + "epoch": 6.893854748603352, + "grad_norm": 1.4208378791809082, + "learning_rate": 8.262569832402235e-06, + "loss": 0.272, + "step": 2468 + }, + { + "epoch": 6.896648044692737, + "grad_norm": 1.7611887454986572, + "learning_rate": 8.265921787709496e-06, + "loss": 0.2693, + "step": 2469 + }, + { + "epoch": 6.899441340782123, + "grad_norm": 2.092797040939331, + "learning_rate": 8.26927374301676e-06, + "loss": 0.3203, + "step": 2470 + }, + { + "epoch": 6.902234636871508, + "grad_norm": 1.6150494813919067, + "learning_rate": 8.272625698324022e-06, + "loss": 0.2926, + "step": 2471 + }, + { + "epoch": 6.905027932960894, + "grad_norm": 1.6558969020843506, + "learning_rate": 8.275977653631285e-06, + "loss": 0.3298, + "step": 2472 + }, + { + "epoch": 6.907821229050279, + "grad_norm": 6.067139625549316, + "learning_rate": 8.279329608938547e-06, + "loss": 0.36, + "step": 2473 + }, + { + "epoch": 6.910614525139665, + "grad_norm": 0.7174433469772339, + "learning_rate": 8.28268156424581e-06, + "loss": 0.2542, + "step": 2474 + }, + { + "epoch": 6.91340782122905, + "grad_norm": 0.5978436470031738, + "learning_rate": 8.286033519553073e-06, + "loss": 0.2382, + "step": 2475 + }, + { + "epoch": 6.916201117318436, + "grad_norm": 0.7757250666618347, + "learning_rate": 8.289385474860334e-06, + "loss": 0.2431, + "step": 2476 + }, + { + "epoch": 6.918994413407821, + "grad_norm": 0.5992730259895325, + "learning_rate": 8.292737430167597e-06, + "loss": 0.2647, + "step": 2477 + }, + { + "epoch": 6.921787709497207, + "grad_norm": 0.6706070303916931, + "learning_rate": 8.29608938547486e-06, + "loss": 0.2676, + "step": 2478 + }, + { + "epoch": 6.924581005586592, + "grad_norm": 0.6442945599555969, + "learning_rate": 8.299441340782122e-06, + "loss": 0.2107, + "step": 2479 + }, + { + "epoch": 6.927374301675978, + "grad_norm": 0.843197226524353, + "learning_rate": 8.302793296089385e-06, + "loss": 0.2211, + "step": 2480 + }, + { + "epoch": 6.930167597765363, + "grad_norm": 0.5908825397491455, + "learning_rate": 8.306145251396648e-06, + "loss": 0.2294, + "step": 2481 + }, + { + "epoch": 6.932960893854749, + "grad_norm": 0.6788244247436523, + "learning_rate": 8.309497206703912e-06, + "loss": 0.2299, + "step": 2482 + }, + { + "epoch": 6.935754189944134, + "grad_norm": 1.076216459274292, + "learning_rate": 8.312849162011174e-06, + "loss": 0.2568, + "step": 2483 + }, + { + "epoch": 6.93854748603352, + "grad_norm": 0.6588031053543091, + "learning_rate": 8.316201117318437e-06, + "loss": 0.2129, + "step": 2484 + }, + { + "epoch": 6.941340782122905, + "grad_norm": 0.7600060701370239, + "learning_rate": 8.3195530726257e-06, + "loss": 0.2573, + "step": 2485 + }, + { + "epoch": 6.94413407821229, + "grad_norm": 0.9387103319168091, + "learning_rate": 8.322905027932961e-06, + "loss": 0.2266, + "step": 2486 + }, + { + "epoch": 6.946927374301676, + "grad_norm": 1.1759326457977295, + "learning_rate": 8.326256983240224e-06, + "loss": 0.2489, + "step": 2487 + }, + { + "epoch": 6.949720670391061, + "grad_norm": 1.2589986324310303, + "learning_rate": 8.329608938547487e-06, + "loss": 0.2687, + "step": 2488 + }, + { + "epoch": 6.952513966480447, + "grad_norm": 0.9400416612625122, + "learning_rate": 8.332960893854749e-06, + "loss": 0.254, + "step": 2489 + }, + { + "epoch": 6.955307262569832, + "grad_norm": 1.1208220720291138, + "learning_rate": 8.336312849162012e-06, + "loss": 0.3276, + "step": 2490 + }, + { + "epoch": 6.9581005586592175, + "grad_norm": 1.043319582939148, + "learning_rate": 8.339664804469275e-06, + "loss": 0.2347, + "step": 2491 + }, + { + "epoch": 6.960893854748603, + "grad_norm": 2.405142068862915, + "learning_rate": 8.343016759776538e-06, + "loss": 0.239, + "step": 2492 + }, + { + "epoch": 6.9636871508379885, + "grad_norm": 1.3740649223327637, + "learning_rate": 8.346368715083799e-06, + "loss": 0.3256, + "step": 2493 + }, + { + "epoch": 6.966480446927374, + "grad_norm": 2.1521716117858887, + "learning_rate": 8.349720670391062e-06, + "loss": 0.2896, + "step": 2494 + }, + { + "epoch": 6.9692737430167595, + "grad_norm": 1.6444002389907837, + "learning_rate": 8.353072625698325e-06, + "loss": 0.3146, + "step": 2495 + }, + { + "epoch": 6.972067039106145, + "grad_norm": 2.270613670349121, + "learning_rate": 8.356424581005586e-06, + "loss": 0.3219, + "step": 2496 + }, + { + "epoch": 6.9748603351955305, + "grad_norm": 2.471982717514038, + "learning_rate": 8.35977653631285e-06, + "loss": 0.3131, + "step": 2497 + }, + { + "epoch": 6.977653631284916, + "grad_norm": 5.335579872131348, + "learning_rate": 8.363128491620113e-06, + "loss": 0.4262, + "step": 2498 + }, + { + "epoch": 6.9804469273743015, + "grad_norm": 0.614124059677124, + "learning_rate": 8.366480446927374e-06, + "loss": 0.2257, + "step": 2499 + }, + { + "epoch": 6.983240223463687, + "grad_norm": 0.6187222003936768, + "learning_rate": 8.369832402234637e-06, + "loss": 0.2461, + "step": 2500 + }, + { + "epoch": 6.9860335195530725, + "grad_norm": 1.0464085340499878, + "learning_rate": 8.3731843575419e-06, + "loss": 0.2376, + "step": 2501 + }, + { + "epoch": 6.988826815642458, + "grad_norm": 1.903355598449707, + "learning_rate": 8.376536312849163e-06, + "loss": 0.2412, + "step": 2502 + }, + { + "epoch": 6.9916201117318435, + "grad_norm": 0.8759481310844421, + "learning_rate": 8.379888268156424e-06, + "loss": 0.2267, + "step": 2503 + }, + { + "epoch": 6.994413407821229, + "grad_norm": 0.8437309265136719, + "learning_rate": 8.383240223463687e-06, + "loss": 0.2282, + "step": 2504 + }, + { + "epoch": 6.9972067039106145, + "grad_norm": 1.1924951076507568, + "learning_rate": 8.38659217877095e-06, + "loss": 0.2686, + "step": 2505 + }, + { + "epoch": 7.0, + "grad_norm": 3.3172404766082764, + "learning_rate": 8.389944134078212e-06, + "loss": 0.3193, + "step": 2506 + }, + { + "epoch": 7.0027932960893855, + "grad_norm": 0.7812169790267944, + "learning_rate": 8.393296089385475e-06, + "loss": 0.244, + "step": 2507 + }, + { + "epoch": 7.005586592178771, + "grad_norm": 0.5318458080291748, + "learning_rate": 8.396648044692738e-06, + "loss": 0.2231, + "step": 2508 + }, + { + "epoch": 7.0083798882681565, + "grad_norm": 0.6655611395835876, + "learning_rate": 8.400000000000001e-06, + "loss": 0.248, + "step": 2509 + }, + { + "epoch": 7.011173184357542, + "grad_norm": 1.0428818464279175, + "learning_rate": 8.403351955307262e-06, + "loss": 0.234, + "step": 2510 + }, + { + "epoch": 7.0139664804469275, + "grad_norm": 0.6005908846855164, + "learning_rate": 8.406703910614525e-06, + "loss": 0.1926, + "step": 2511 + }, + { + "epoch": 7.016759776536313, + "grad_norm": 0.9278674125671387, + "learning_rate": 8.410055865921788e-06, + "loss": 0.2241, + "step": 2512 + }, + { + "epoch": 7.0195530726256985, + "grad_norm": 0.8069068193435669, + "learning_rate": 8.41340782122905e-06, + "loss": 0.2562, + "step": 2513 + }, + { + "epoch": 7.022346368715084, + "grad_norm": 1.0319709777832031, + "learning_rate": 8.416759776536313e-06, + "loss": 0.2535, + "step": 2514 + }, + { + "epoch": 7.0251396648044695, + "grad_norm": 0.7144967913627625, + "learning_rate": 8.420111731843576e-06, + "loss": 0.2523, + "step": 2515 + }, + { + "epoch": 7.027932960893855, + "grad_norm": 0.8145877122879028, + "learning_rate": 8.423463687150837e-06, + "loss": 0.2261, + "step": 2516 + }, + { + "epoch": 7.0307262569832405, + "grad_norm": 0.7589735984802246, + "learning_rate": 8.4268156424581e-06, + "loss": 0.2335, + "step": 2517 + }, + { + "epoch": 7.033519553072626, + "grad_norm": 0.8694164752960205, + "learning_rate": 8.430167597765363e-06, + "loss": 0.2334, + "step": 2518 + }, + { + "epoch": 7.0363128491620115, + "grad_norm": 1.2436352968215942, + "learning_rate": 8.433519553072626e-06, + "loss": 0.2391, + "step": 2519 + }, + { + "epoch": 7.039106145251397, + "grad_norm": 0.7446537017822266, + "learning_rate": 8.436871508379888e-06, + "loss": 0.2174, + "step": 2520 + }, + { + "epoch": 7.0418994413407825, + "grad_norm": 0.8652095794677734, + "learning_rate": 8.44022346368715e-06, + "loss": 0.2398, + "step": 2521 + }, + { + "epoch": 7.044692737430168, + "grad_norm": 0.9876266717910767, + "learning_rate": 8.443575418994414e-06, + "loss": 0.2347, + "step": 2522 + }, + { + "epoch": 7.0474860335195535, + "grad_norm": 0.8945322036743164, + "learning_rate": 8.446927374301675e-06, + "loss": 0.2414, + "step": 2523 + }, + { + "epoch": 7.050279329608939, + "grad_norm": 1.4168322086334229, + "learning_rate": 8.450279329608938e-06, + "loss": 0.23, + "step": 2524 + }, + { + "epoch": 7.053072625698324, + "grad_norm": 1.420703649520874, + "learning_rate": 8.453631284916201e-06, + "loss": 0.2586, + "step": 2525 + }, + { + "epoch": 7.055865921787709, + "grad_norm": 1.6616653203964233, + "learning_rate": 8.456983240223462e-06, + "loss": 0.2495, + "step": 2526 + }, + { + "epoch": 7.058659217877095, + "grad_norm": 2.1397900581359863, + "learning_rate": 8.460335195530725e-06, + "loss": 0.2333, + "step": 2527 + }, + { + "epoch": 7.06145251396648, + "grad_norm": 3.2026326656341553, + "learning_rate": 8.463687150837988e-06, + "loss": 0.3009, + "step": 2528 + }, + { + "epoch": 7.064245810055866, + "grad_norm": 4.070718288421631, + "learning_rate": 8.467039106145252e-06, + "loss": 0.3055, + "step": 2529 + }, + { + "epoch": 7.067039106145251, + "grad_norm": 2.1454920768737793, + "learning_rate": 8.470391061452515e-06, + "loss": 0.2866, + "step": 2530 + }, + { + "epoch": 7.069832402234637, + "grad_norm": 3.2344064712524414, + "learning_rate": 8.473743016759778e-06, + "loss": 0.5049, + "step": 2531 + }, + { + "epoch": 7.072625698324022, + "grad_norm": 0.9821122288703918, + "learning_rate": 8.47709497206704e-06, + "loss": 0.2396, + "step": 2532 + }, + { + "epoch": 7.075418994413408, + "grad_norm": 0.7460352778434753, + "learning_rate": 8.480446927374302e-06, + "loss": 0.231, + "step": 2533 + }, + { + "epoch": 7.078212290502793, + "grad_norm": 0.844428539276123, + "learning_rate": 8.483798882681565e-06, + "loss": 0.2437, + "step": 2534 + }, + { + "epoch": 7.081005586592179, + "grad_norm": 0.7356999516487122, + "learning_rate": 8.487150837988828e-06, + "loss": 0.2432, + "step": 2535 + }, + { + "epoch": 7.083798882681564, + "grad_norm": 0.6686387062072754, + "learning_rate": 8.49050279329609e-06, + "loss": 0.2421, + "step": 2536 + }, + { + "epoch": 7.08659217877095, + "grad_norm": 0.8969717025756836, + "learning_rate": 8.493854748603352e-06, + "loss": 0.2215, + "step": 2537 + }, + { + "epoch": 7.089385474860335, + "grad_norm": 0.7196751236915588, + "learning_rate": 8.497206703910615e-06, + "loss": 0.1988, + "step": 2538 + }, + { + "epoch": 7.092178770949721, + "grad_norm": 3.0435330867767334, + "learning_rate": 8.500558659217879e-06, + "loss": 0.2159, + "step": 2539 + }, + { + "epoch": 7.094972067039106, + "grad_norm": 0.8835919499397278, + "learning_rate": 8.50391061452514e-06, + "loss": 0.2142, + "step": 2540 + }, + { + "epoch": 7.097765363128492, + "grad_norm": 1.1821701526641846, + "learning_rate": 8.507262569832403e-06, + "loss": 0.2872, + "step": 2541 + }, + { + "epoch": 7.100558659217877, + "grad_norm": 0.8100203275680542, + "learning_rate": 8.510614525139666e-06, + "loss": 0.248, + "step": 2542 + }, + { + "epoch": 7.103351955307263, + "grad_norm": 0.872600793838501, + "learning_rate": 8.513966480446927e-06, + "loss": 0.2258, + "step": 2543 + }, + { + "epoch": 7.106145251396648, + "grad_norm": 0.9138769507408142, + "learning_rate": 8.51731843575419e-06, + "loss": 0.2511, + "step": 2544 + }, + { + "epoch": 7.108938547486034, + "grad_norm": 1.0851434469223022, + "learning_rate": 8.520670391061453e-06, + "loss": 0.2388, + "step": 2545 + }, + { + "epoch": 7.111731843575419, + "grad_norm": 1.049847960472107, + "learning_rate": 8.524022346368715e-06, + "loss": 0.2282, + "step": 2546 + }, + { + "epoch": 7.114525139664805, + "grad_norm": 1.232094407081604, + "learning_rate": 8.527374301675978e-06, + "loss": 0.2521, + "step": 2547 + }, + { + "epoch": 7.11731843575419, + "grad_norm": 1.2906361818313599, + "learning_rate": 8.53072625698324e-06, + "loss": 0.2482, + "step": 2548 + }, + { + "epoch": 7.120111731843576, + "grad_norm": 1.2293163537979126, + "learning_rate": 8.534078212290504e-06, + "loss": 0.2363, + "step": 2549 + }, + { + "epoch": 7.122905027932961, + "grad_norm": 1.2293190956115723, + "learning_rate": 8.537430167597765e-06, + "loss": 0.2719, + "step": 2550 + }, + { + "epoch": 7.125698324022347, + "grad_norm": 1.0016010999679565, + "learning_rate": 8.540782122905028e-06, + "loss": 0.2661, + "step": 2551 + }, + { + "epoch": 7.128491620111732, + "grad_norm": 1.2443692684173584, + "learning_rate": 8.544134078212291e-06, + "loss": 0.2693, + "step": 2552 + }, + { + "epoch": 7.131284916201118, + "grad_norm": 1.8228652477264404, + "learning_rate": 8.547486033519553e-06, + "loss": 0.316, + "step": 2553 + }, + { + "epoch": 7.134078212290503, + "grad_norm": 1.6360968351364136, + "learning_rate": 8.550837988826816e-06, + "loss": 0.3623, + "step": 2554 + }, + { + "epoch": 7.136871508379889, + "grad_norm": 2.1772830486297607, + "learning_rate": 8.554189944134079e-06, + "loss": 0.3246, + "step": 2555 + }, + { + "epoch": 7.139664804469274, + "grad_norm": 2.527968168258667, + "learning_rate": 8.557541899441342e-06, + "loss": 0.4254, + "step": 2556 + }, + { + "epoch": 7.14245810055866, + "grad_norm": 1.1121256351470947, + "learning_rate": 8.560893854748603e-06, + "loss": 0.2622, + "step": 2557 + }, + { + "epoch": 7.145251396648045, + "grad_norm": 0.7189643979072571, + "learning_rate": 8.564245810055866e-06, + "loss": 0.25, + "step": 2558 + }, + { + "epoch": 7.148044692737431, + "grad_norm": 0.8268055319786072, + "learning_rate": 8.567597765363129e-06, + "loss": 0.2293, + "step": 2559 + }, + { + "epoch": 7.150837988826815, + "grad_norm": 0.8735764622688293, + "learning_rate": 8.57094972067039e-06, + "loss": 0.2336, + "step": 2560 + }, + { + "epoch": 7.153631284916201, + "grad_norm": 2.6199584007263184, + "learning_rate": 8.574301675977654e-06, + "loss": 0.2779, + "step": 2561 + }, + { + "epoch": 7.156424581005586, + "grad_norm": 0.9971876740455627, + "learning_rate": 8.577653631284917e-06, + "loss": 0.2665, + "step": 2562 + }, + { + "epoch": 7.159217877094972, + "grad_norm": 0.7999356389045715, + "learning_rate": 8.581005586592178e-06, + "loss": 0.2391, + "step": 2563 + }, + { + "epoch": 7.162011173184357, + "grad_norm": 0.9085678458213806, + "learning_rate": 8.584357541899441e-06, + "loss": 0.2838, + "step": 2564 + }, + { + "epoch": 7.164804469273743, + "grad_norm": 0.8175155520439148, + "learning_rate": 8.587709497206704e-06, + "loss": 0.2432, + "step": 2565 + }, + { + "epoch": 7.167597765363128, + "grad_norm": 0.9841026067733765, + "learning_rate": 8.591061452513967e-06, + "loss": 0.2465, + "step": 2566 + }, + { + "epoch": 7.170391061452514, + "grad_norm": 0.9684368371963501, + "learning_rate": 8.594413407821228e-06, + "loss": 0.2191, + "step": 2567 + }, + { + "epoch": 7.173184357541899, + "grad_norm": 1.1561537981033325, + "learning_rate": 8.597765363128491e-06, + "loss": 0.2142, + "step": 2568 + }, + { + "epoch": 7.175977653631285, + "grad_norm": 1.2218389511108398, + "learning_rate": 8.601117318435754e-06, + "loss": 0.2319, + "step": 2569 + }, + { + "epoch": 7.17877094972067, + "grad_norm": 0.9846516251564026, + "learning_rate": 8.604469273743016e-06, + "loss": 0.2558, + "step": 2570 + }, + { + "epoch": 7.181564245810056, + "grad_norm": 1.303324818611145, + "learning_rate": 8.607821229050279e-06, + "loss": 0.2258, + "step": 2571 + }, + { + "epoch": 7.184357541899441, + "grad_norm": 0.8223280906677246, + "learning_rate": 8.611173184357542e-06, + "loss": 0.2333, + "step": 2572 + }, + { + "epoch": 7.187150837988827, + "grad_norm": 2.1640729904174805, + "learning_rate": 8.614525139664803e-06, + "loss": 0.2632, + "step": 2573 + }, + { + "epoch": 7.189944134078212, + "grad_norm": 1.0180349349975586, + "learning_rate": 8.617877094972066e-06, + "loss": 0.2844, + "step": 2574 + }, + { + "epoch": 7.192737430167598, + "grad_norm": 1.054057240486145, + "learning_rate": 8.62122905027933e-06, + "loss": 0.2438, + "step": 2575 + }, + { + "epoch": 7.195530726256983, + "grad_norm": 1.1661990880966187, + "learning_rate": 8.624581005586592e-06, + "loss": 0.2593, + "step": 2576 + }, + { + "epoch": 7.198324022346369, + "grad_norm": 3.754551649093628, + "learning_rate": 8.627932960893854e-06, + "loss": 0.2529, + "step": 2577 + }, + { + "epoch": 7.201117318435754, + "grad_norm": 2.3860392570495605, + "learning_rate": 8.631284916201118e-06, + "loss": 0.2644, + "step": 2578 + }, + { + "epoch": 7.20391061452514, + "grad_norm": 1.8407928943634033, + "learning_rate": 8.634636871508381e-06, + "loss": 0.3003, + "step": 2579 + }, + { + "epoch": 7.206703910614525, + "grad_norm": 4.817329406738281, + "learning_rate": 8.637988826815643e-06, + "loss": 0.2894, + "step": 2580 + }, + { + "epoch": 7.209497206703911, + "grad_norm": 7.567015171051025, + "learning_rate": 8.641340782122906e-06, + "loss": 0.3661, + "step": 2581 + }, + { + "epoch": 7.212290502793296, + "grad_norm": 1.8832502365112305, + "learning_rate": 8.644692737430169e-06, + "loss": 0.2084, + "step": 2582 + }, + { + "epoch": 7.215083798882682, + "grad_norm": 0.9256683588027954, + "learning_rate": 8.64804469273743e-06, + "loss": 0.2504, + "step": 2583 + }, + { + "epoch": 7.217877094972067, + "grad_norm": 0.5735813975334167, + "learning_rate": 8.651396648044693e-06, + "loss": 0.2473, + "step": 2584 + }, + { + "epoch": 7.220670391061453, + "grad_norm": 0.7028632164001465, + "learning_rate": 8.654748603351956e-06, + "loss": 0.236, + "step": 2585 + }, + { + "epoch": 7.223463687150838, + "grad_norm": 0.6966735124588013, + "learning_rate": 8.65810055865922e-06, + "loss": 0.2404, + "step": 2586 + }, + { + "epoch": 7.226256983240224, + "grad_norm": 0.6106359362602234, + "learning_rate": 8.66145251396648e-06, + "loss": 0.2204, + "step": 2587 + }, + { + "epoch": 7.229050279329609, + "grad_norm": 1.5831005573272705, + "learning_rate": 8.664804469273744e-06, + "loss": 0.2339, + "step": 2588 + }, + { + "epoch": 7.231843575418995, + "grad_norm": 0.9576526284217834, + "learning_rate": 8.668156424581007e-06, + "loss": 0.2485, + "step": 2589 + }, + { + "epoch": 7.23463687150838, + "grad_norm": 0.6527316570281982, + "learning_rate": 8.671508379888268e-06, + "loss": 0.2057, + "step": 2590 + }, + { + "epoch": 7.237430167597766, + "grad_norm": 1.2092387676239014, + "learning_rate": 8.674860335195531e-06, + "loss": 0.2048, + "step": 2591 + }, + { + "epoch": 7.240223463687151, + "grad_norm": 1.1991808414459229, + "learning_rate": 8.678212290502794e-06, + "loss": 0.2604, + "step": 2592 + }, + { + "epoch": 7.243016759776537, + "grad_norm": 0.8637111783027649, + "learning_rate": 8.681564245810056e-06, + "loss": 0.2142, + "step": 2593 + }, + { + "epoch": 7.245810055865922, + "grad_norm": 0.873157799243927, + "learning_rate": 8.684916201117319e-06, + "loss": 0.2235, + "step": 2594 + }, + { + "epoch": 7.248603351955307, + "grad_norm": 1.422583818435669, + "learning_rate": 8.688268156424582e-06, + "loss": 0.2792, + "step": 2595 + }, + { + "epoch": 7.251396648044693, + "grad_norm": 1.0097850561141968, + "learning_rate": 8.691620111731845e-06, + "loss": 0.221, + "step": 2596 + }, + { + "epoch": 7.254189944134078, + "grad_norm": 0.8369917273521423, + "learning_rate": 8.694972067039106e-06, + "loss": 0.2358, + "step": 2597 + }, + { + "epoch": 7.256983240223463, + "grad_norm": 1.041198968887329, + "learning_rate": 8.698324022346369e-06, + "loss": 0.2514, + "step": 2598 + }, + { + "epoch": 7.259776536312849, + "grad_norm": 1.4742801189422607, + "learning_rate": 8.701675977653632e-06, + "loss": 0.2149, + "step": 2599 + }, + { + "epoch": 7.262569832402234, + "grad_norm": 1.1736708879470825, + "learning_rate": 8.705027932960893e-06, + "loss": 0.2566, + "step": 2600 + }, + { + "epoch": 7.26536312849162, + "grad_norm": 1.1481573581695557, + "learning_rate": 8.708379888268156e-06, + "loss": 0.2609, + "step": 2601 + }, + { + "epoch": 7.268156424581005, + "grad_norm": 1.3615853786468506, + "learning_rate": 8.71173184357542e-06, + "loss": 0.2644, + "step": 2602 + }, + { + "epoch": 7.270949720670391, + "grad_norm": 7.107853889465332, + "learning_rate": 8.715083798882683e-06, + "loss": 0.2596, + "step": 2603 + }, + { + "epoch": 7.273743016759776, + "grad_norm": 2.640641450881958, + "learning_rate": 8.718435754189944e-06, + "loss": 0.2778, + "step": 2604 + }, + { + "epoch": 7.276536312849162, + "grad_norm": 1.8569828271865845, + "learning_rate": 8.721787709497207e-06, + "loss": 0.245, + "step": 2605 + }, + { + "epoch": 7.279329608938547, + "grad_norm": 2.6743476390838623, + "learning_rate": 8.72513966480447e-06, + "loss": 0.3725, + "step": 2606 + }, + { + "epoch": 7.282122905027933, + "grad_norm": 0.6915119290351868, + "learning_rate": 8.728491620111731e-06, + "loss": 0.2492, + "step": 2607 + }, + { + "epoch": 7.284916201117318, + "grad_norm": 2.5793654918670654, + "learning_rate": 8.731843575418994e-06, + "loss": 0.2963, + "step": 2608 + }, + { + "epoch": 7.287709497206704, + "grad_norm": 1.161983609199524, + "learning_rate": 8.735195530726257e-06, + "loss": 0.225, + "step": 2609 + }, + { + "epoch": 7.290502793296089, + "grad_norm": 0.5729997158050537, + "learning_rate": 8.738547486033519e-06, + "loss": 0.2199, + "step": 2610 + }, + { + "epoch": 7.293296089385475, + "grad_norm": 1.0601327419281006, + "learning_rate": 8.741899441340782e-06, + "loss": 0.2545, + "step": 2611 + }, + { + "epoch": 7.29608938547486, + "grad_norm": 1.1463710069656372, + "learning_rate": 8.745251396648045e-06, + "loss": 0.2249, + "step": 2612 + }, + { + "epoch": 7.298882681564246, + "grad_norm": 1.249761700630188, + "learning_rate": 8.748603351955308e-06, + "loss": 0.2223, + "step": 2613 + }, + { + "epoch": 7.301675977653631, + "grad_norm": 0.7399051189422607, + "learning_rate": 8.75195530726257e-06, + "loss": 0.2342, + "step": 2614 + }, + { + "epoch": 7.304469273743017, + "grad_norm": 1.1701202392578125, + "learning_rate": 8.755307262569832e-06, + "loss": 0.2063, + "step": 2615 + }, + { + "epoch": 7.307262569832402, + "grad_norm": 1.0422338247299194, + "learning_rate": 8.758659217877095e-06, + "loss": 0.293, + "step": 2616 + }, + { + "epoch": 7.310055865921788, + "grad_norm": 0.8673356175422668, + "learning_rate": 8.762011173184357e-06, + "loss": 0.2717, + "step": 2617 + }, + { + "epoch": 7.312849162011173, + "grad_norm": 1.1240768432617188, + "learning_rate": 8.76536312849162e-06, + "loss": 0.2109, + "step": 2618 + }, + { + "epoch": 7.315642458100559, + "grad_norm": 0.8352090120315552, + "learning_rate": 8.768715083798883e-06, + "loss": 0.2467, + "step": 2619 + }, + { + "epoch": 7.318435754189944, + "grad_norm": 0.9892006516456604, + "learning_rate": 8.772067039106144e-06, + "loss": 0.2656, + "step": 2620 + }, + { + "epoch": 7.32122905027933, + "grad_norm": 0.8692042231559753, + "learning_rate": 8.775418994413407e-06, + "loss": 0.2345, + "step": 2621 + }, + { + "epoch": 7.324022346368715, + "grad_norm": 0.9931304454803467, + "learning_rate": 8.77877094972067e-06, + "loss": 0.2381, + "step": 2622 + }, + { + "epoch": 7.326815642458101, + "grad_norm": 1.2663702964782715, + "learning_rate": 8.782122905027933e-06, + "loss": 0.2535, + "step": 2623 + }, + { + "epoch": 7.329608938547486, + "grad_norm": 1.2337603569030762, + "learning_rate": 8.785474860335195e-06, + "loss": 0.2403, + "step": 2624 + }, + { + "epoch": 7.332402234636872, + "grad_norm": 1.0320508480072021, + "learning_rate": 8.788826815642458e-06, + "loss": 0.2857, + "step": 2625 + }, + { + "epoch": 7.335195530726257, + "grad_norm": 0.9710248112678528, + "learning_rate": 8.792178770949722e-06, + "loss": 0.2625, + "step": 2626 + }, + { + "epoch": 7.337988826815643, + "grad_norm": 1.3761886358261108, + "learning_rate": 8.795530726256984e-06, + "loss": 0.2838, + "step": 2627 + }, + { + "epoch": 7.340782122905028, + "grad_norm": 1.4163603782653809, + "learning_rate": 8.798882681564247e-06, + "loss": 0.2428, + "step": 2628 + }, + { + "epoch": 7.343575418994414, + "grad_norm": 1.6485776901245117, + "learning_rate": 8.80223463687151e-06, + "loss": 0.3042, + "step": 2629 + }, + { + "epoch": 7.346368715083799, + "grad_norm": 3.2231788635253906, + "learning_rate": 8.805586592178771e-06, + "loss": 0.3786, + "step": 2630 + }, + { + "epoch": 7.349162011173185, + "grad_norm": 2.6214561462402344, + "learning_rate": 8.808938547486034e-06, + "loss": 0.4007, + "step": 2631 + }, + { + "epoch": 7.351955307262569, + "grad_norm": 0.7820866703987122, + "learning_rate": 8.812290502793297e-06, + "loss": 0.2598, + "step": 2632 + }, + { + "epoch": 7.354748603351955, + "grad_norm": 0.8328871130943298, + "learning_rate": 8.81564245810056e-06, + "loss": 0.2309, + "step": 2633 + }, + { + "epoch": 7.35754189944134, + "grad_norm": 0.8612908124923706, + "learning_rate": 8.818994413407822e-06, + "loss": 0.2214, + "step": 2634 + }, + { + "epoch": 7.360335195530726, + "grad_norm": 0.912079930305481, + "learning_rate": 8.822346368715085e-06, + "loss": 0.2187, + "step": 2635 + }, + { + "epoch": 7.363128491620111, + "grad_norm": 1.0057222843170166, + "learning_rate": 8.825698324022348e-06, + "loss": 0.2626, + "step": 2636 + }, + { + "epoch": 7.365921787709497, + "grad_norm": 0.8844982981681824, + "learning_rate": 8.829050279329609e-06, + "loss": 0.2281, + "step": 2637 + }, + { + "epoch": 7.368715083798882, + "grad_norm": 0.8353021144866943, + "learning_rate": 8.832402234636872e-06, + "loss": 0.2469, + "step": 2638 + }, + { + "epoch": 7.371508379888268, + "grad_norm": 1.1054948568344116, + "learning_rate": 8.835754189944135e-06, + "loss": 0.2402, + "step": 2639 + }, + { + "epoch": 7.374301675977653, + "grad_norm": 0.8973386883735657, + "learning_rate": 8.839106145251396e-06, + "loss": 0.2579, + "step": 2640 + }, + { + "epoch": 7.377094972067039, + "grad_norm": 0.6946473121643066, + "learning_rate": 8.84245810055866e-06, + "loss": 0.2215, + "step": 2641 + }, + { + "epoch": 7.379888268156424, + "grad_norm": 0.6538834571838379, + "learning_rate": 8.845810055865922e-06, + "loss": 0.2018, + "step": 2642 + }, + { + "epoch": 7.38268156424581, + "grad_norm": 0.8582342267036438, + "learning_rate": 8.849162011173185e-06, + "loss": 0.2605, + "step": 2643 + }, + { + "epoch": 7.385474860335195, + "grad_norm": 1.2232666015625, + "learning_rate": 8.852513966480447e-06, + "loss": 0.2329, + "step": 2644 + }, + { + "epoch": 7.388268156424581, + "grad_norm": 2.0072202682495117, + "learning_rate": 8.85586592178771e-06, + "loss": 0.2565, + "step": 2645 + }, + { + "epoch": 7.391061452513966, + "grad_norm": 1.6973122358322144, + "learning_rate": 8.859217877094973e-06, + "loss": 0.2378, + "step": 2646 + }, + { + "epoch": 7.393854748603352, + "grad_norm": 1.215937852859497, + "learning_rate": 8.862569832402234e-06, + "loss": 0.2491, + "step": 2647 + }, + { + "epoch": 7.396648044692737, + "grad_norm": 1.398750901222229, + "learning_rate": 8.865921787709497e-06, + "loss": 0.2683, + "step": 2648 + }, + { + "epoch": 7.399441340782123, + "grad_norm": 1.2886557579040527, + "learning_rate": 8.86927374301676e-06, + "loss": 0.2613, + "step": 2649 + }, + { + "epoch": 7.402234636871508, + "grad_norm": 1.9415011405944824, + "learning_rate": 8.872625698324023e-06, + "loss": 0.2821, + "step": 2650 + }, + { + "epoch": 7.405027932960894, + "grad_norm": 1.682851791381836, + "learning_rate": 8.875977653631285e-06, + "loss": 0.2685, + "step": 2651 + }, + { + "epoch": 7.407821229050279, + "grad_norm": 1.3625119924545288, + "learning_rate": 8.879329608938548e-06, + "loss": 0.3193, + "step": 2652 + }, + { + "epoch": 7.410614525139665, + "grad_norm": 1.823161244392395, + "learning_rate": 8.88268156424581e-06, + "loss": 0.3128, + "step": 2653 + }, + { + "epoch": 7.41340782122905, + "grad_norm": 1.486153483390808, + "learning_rate": 8.886033519553072e-06, + "loss": 0.2804, + "step": 2654 + }, + { + "epoch": 7.416201117318436, + "grad_norm": 2.391831159591675, + "learning_rate": 8.889385474860335e-06, + "loss": 0.3701, + "step": 2655 + }, + { + "epoch": 7.418994413407821, + "grad_norm": 2.7775862216949463, + "learning_rate": 8.892737430167598e-06, + "loss": 0.4341, + "step": 2656 + }, + { + "epoch": 7.421787709497207, + "grad_norm": 0.7702464461326599, + "learning_rate": 8.89608938547486e-06, + "loss": 0.2423, + "step": 2657 + }, + { + "epoch": 7.424581005586592, + "grad_norm": 0.5959042310714722, + "learning_rate": 8.899441340782123e-06, + "loss": 0.2527, + "step": 2658 + }, + { + "epoch": 7.427374301675978, + "grad_norm": 1.0982309579849243, + "learning_rate": 8.902793296089386e-06, + "loss": 0.2283, + "step": 2659 + }, + { + "epoch": 7.430167597765363, + "grad_norm": 0.6605410575866699, + "learning_rate": 8.906145251396649e-06, + "loss": 0.2433, + "step": 2660 + }, + { + "epoch": 7.432960893854749, + "grad_norm": 0.6519520282745361, + "learning_rate": 8.90949720670391e-06, + "loss": 0.2101, + "step": 2661 + }, + { + "epoch": 7.435754189944134, + "grad_norm": 0.9566534757614136, + "learning_rate": 8.912849162011173e-06, + "loss": 0.3153, + "step": 2662 + }, + { + "epoch": 7.43854748603352, + "grad_norm": 0.8309755325317383, + "learning_rate": 8.916201117318436e-06, + "loss": 0.2441, + "step": 2663 + }, + { + "epoch": 7.441340782122905, + "grad_norm": 0.8430695533752441, + "learning_rate": 8.919553072625697e-06, + "loss": 0.2564, + "step": 2664 + }, + { + "epoch": 7.444134078212291, + "grad_norm": 0.8818030953407288, + "learning_rate": 8.92290502793296e-06, + "loss": 0.2151, + "step": 2665 + }, + { + "epoch": 7.446927374301676, + "grad_norm": 0.9063357710838318, + "learning_rate": 8.926256983240224e-06, + "loss": 0.2592, + "step": 2666 + }, + { + "epoch": 7.449720670391061, + "grad_norm": 1.0576109886169434, + "learning_rate": 8.929608938547485e-06, + "loss": 0.2862, + "step": 2667 + }, + { + "epoch": 7.452513966480447, + "grad_norm": 0.7974826693534851, + "learning_rate": 8.932960893854748e-06, + "loss": 0.2291, + "step": 2668 + }, + { + "epoch": 7.455307262569832, + "grad_norm": 0.7883867621421814, + "learning_rate": 8.936312849162011e-06, + "loss": 0.2064, + "step": 2669 + }, + { + "epoch": 7.4581005586592175, + "grad_norm": 2.004612684249878, + "learning_rate": 8.939664804469274e-06, + "loss": 0.2679, + "step": 2670 + }, + { + "epoch": 7.460893854748603, + "grad_norm": 1.4392355680465698, + "learning_rate": 8.943016759776535e-06, + "loss": 0.2586, + "step": 2671 + }, + { + "epoch": 7.4636871508379885, + "grad_norm": 1.307172179222107, + "learning_rate": 8.946368715083798e-06, + "loss": 0.2426, + "step": 2672 + }, + { + "epoch": 7.466480446927374, + "grad_norm": 1.4262943267822266, + "learning_rate": 8.949720670391061e-06, + "loss": 0.2679, + "step": 2673 + }, + { + "epoch": 7.4692737430167595, + "grad_norm": 1.2123335599899292, + "learning_rate": 8.953072625698324e-06, + "loss": 0.2216, + "step": 2674 + }, + { + "epoch": 7.472067039106145, + "grad_norm": 1.1526784896850586, + "learning_rate": 8.956424581005587e-06, + "loss": 0.2279, + "step": 2675 + }, + { + "epoch": 7.4748603351955305, + "grad_norm": 1.5879384279251099, + "learning_rate": 8.95977653631285e-06, + "loss": 0.2948, + "step": 2676 + }, + { + "epoch": 7.477653631284916, + "grad_norm": 1.5036035776138306, + "learning_rate": 8.963128491620112e-06, + "loss": 0.3014, + "step": 2677 + }, + { + "epoch": 7.4804469273743015, + "grad_norm": 1.268452763557434, + "learning_rate": 8.966480446927375e-06, + "loss": 0.2942, + "step": 2678 + }, + { + "epoch": 7.483240223463687, + "grad_norm": 1.897839903831482, + "learning_rate": 8.969832402234638e-06, + "loss": 0.2738, + "step": 2679 + }, + { + "epoch": 7.4860335195530725, + "grad_norm": 1.9028351306915283, + "learning_rate": 8.973184357541901e-06, + "loss": 0.2959, + "step": 2680 + }, + { + "epoch": 7.488826815642458, + "grad_norm": 3.241487979888916, + "learning_rate": 8.976536312849162e-06, + "loss": 0.4302, + "step": 2681 + }, + { + "epoch": 7.4916201117318435, + "grad_norm": 1.0903137922286987, + "learning_rate": 8.979888268156425e-06, + "loss": 0.2772, + "step": 2682 + }, + { + "epoch": 7.494413407821229, + "grad_norm": 0.7278072237968445, + "learning_rate": 8.983240223463688e-06, + "loss": 0.2096, + "step": 2683 + }, + { + "epoch": 7.4972067039106145, + "grad_norm": 1.059012532234192, + "learning_rate": 8.98659217877095e-06, + "loss": 0.2493, + "step": 2684 + }, + { + "epoch": 7.5, + "grad_norm": 0.884428858757019, + "learning_rate": 8.989944134078213e-06, + "loss": 0.2555, + "step": 2685 + }, + { + "epoch": 7.5027932960893855, + "grad_norm": 0.9494547247886658, + "learning_rate": 8.993296089385476e-06, + "loss": 0.2836, + "step": 2686 + }, + { + "epoch": 7.505586592178771, + "grad_norm": 0.9119651913642883, + "learning_rate": 8.996648044692737e-06, + "loss": 0.2592, + "step": 2687 + }, + { + "epoch": 7.5083798882681565, + "grad_norm": 0.7464735507965088, + "learning_rate": 9e-06, + "loss": 0.2625, + "step": 2688 + }, + { + "epoch": 7.511173184357542, + "grad_norm": 0.7162903547286987, + "learning_rate": 9.003351955307263e-06, + "loss": 0.2294, + "step": 2689 + }, + { + "epoch": 7.5139664804469275, + "grad_norm": 1.0555812120437622, + "learning_rate": 9.006703910614526e-06, + "loss": 0.2169, + "step": 2690 + }, + { + "epoch": 7.516759776536313, + "grad_norm": 0.7322402000427246, + "learning_rate": 9.010055865921788e-06, + "loss": 0.2055, + "step": 2691 + }, + { + "epoch": 7.5195530726256985, + "grad_norm": 0.9936563372612, + "learning_rate": 9.01340782122905e-06, + "loss": 0.2473, + "step": 2692 + }, + { + "epoch": 7.522346368715084, + "grad_norm": 1.741931676864624, + "learning_rate": 9.016759776536314e-06, + "loss": 0.2359, + "step": 2693 + }, + { + "epoch": 7.5251396648044695, + "grad_norm": 0.9205590486526489, + "learning_rate": 9.020111731843575e-06, + "loss": 0.2785, + "step": 2694 + }, + { + "epoch": 7.527932960893855, + "grad_norm": 0.8551697731018066, + "learning_rate": 9.023463687150838e-06, + "loss": 0.2522, + "step": 2695 + }, + { + "epoch": 7.5307262569832405, + "grad_norm": 1.0151243209838867, + "learning_rate": 9.026815642458101e-06, + "loss": 0.2509, + "step": 2696 + }, + { + "epoch": 7.533519553072626, + "grad_norm": 2.329097032546997, + "learning_rate": 9.030167597765364e-06, + "loss": 0.2424, + "step": 2697 + }, + { + "epoch": 7.5363128491620115, + "grad_norm": 1.062045931816101, + "learning_rate": 9.033519553072626e-06, + "loss": 0.2291, + "step": 2698 + }, + { + "epoch": 7.539106145251397, + "grad_norm": 1.2063740491867065, + "learning_rate": 9.036871508379889e-06, + "loss": 0.2378, + "step": 2699 + }, + { + "epoch": 7.5418994413407825, + "grad_norm": 1.4687997102737427, + "learning_rate": 9.040223463687152e-06, + "loss": 0.3241, + "step": 2700 + }, + { + "epoch": 7.544692737430168, + "grad_norm": 1.1641380786895752, + "learning_rate": 9.043575418994413e-06, + "loss": 0.2941, + "step": 2701 + }, + { + "epoch": 7.547486033519553, + "grad_norm": 1.0064418315887451, + "learning_rate": 9.046927374301676e-06, + "loss": 0.2596, + "step": 2702 + }, + { + "epoch": 7.550279329608939, + "grad_norm": 1.3695855140686035, + "learning_rate": 9.050279329608939e-06, + "loss": 0.2755, + "step": 2703 + }, + { + "epoch": 7.553072625698324, + "grad_norm": 1.5377060174942017, + "learning_rate": 9.0536312849162e-06, + "loss": 0.3205, + "step": 2704 + }, + { + "epoch": 7.55586592178771, + "grad_norm": 1.739233374595642, + "learning_rate": 9.056983240223463e-06, + "loss": 0.2879, + "step": 2705 + }, + { + "epoch": 7.558659217877095, + "grad_norm": 2.2470383644104004, + "learning_rate": 9.060335195530726e-06, + "loss": 0.4831, + "step": 2706 + }, + { + "epoch": 7.56145251396648, + "grad_norm": 0.6946509480476379, + "learning_rate": 9.06368715083799e-06, + "loss": 0.2412, + "step": 2707 + }, + { + "epoch": 7.564245810055866, + "grad_norm": 0.5582979917526245, + "learning_rate": 9.067039106145251e-06, + "loss": 0.236, + "step": 2708 + }, + { + "epoch": 7.567039106145251, + "grad_norm": 0.6806535124778748, + "learning_rate": 9.070391061452514e-06, + "loss": 0.2036, + "step": 2709 + }, + { + "epoch": 7.569832402234637, + "grad_norm": 1.1556681394577026, + "learning_rate": 9.073743016759777e-06, + "loss": 0.2459, + "step": 2710 + }, + { + "epoch": 7.572625698324022, + "grad_norm": 0.5952070355415344, + "learning_rate": 9.077094972067038e-06, + "loss": 0.2251, + "step": 2711 + }, + { + "epoch": 7.575418994413408, + "grad_norm": 0.7167660593986511, + "learning_rate": 9.080446927374301e-06, + "loss": 0.2137, + "step": 2712 + }, + { + "epoch": 7.578212290502793, + "grad_norm": 0.7014570832252502, + "learning_rate": 9.083798882681564e-06, + "loss": 0.2469, + "step": 2713 + }, + { + "epoch": 7.581005586592179, + "grad_norm": 1.3474777936935425, + "learning_rate": 9.087150837988826e-06, + "loss": 0.2481, + "step": 2714 + }, + { + "epoch": 7.583798882681564, + "grad_norm": 0.7307088375091553, + "learning_rate": 9.090502793296089e-06, + "loss": 0.2419, + "step": 2715 + }, + { + "epoch": 7.58659217877095, + "grad_norm": 0.7105766534805298, + "learning_rate": 9.093854748603352e-06, + "loss": 0.2198, + "step": 2716 + }, + { + "epoch": 7.589385474860335, + "grad_norm": 0.8341657519340515, + "learning_rate": 9.097206703910615e-06, + "loss": 0.2504, + "step": 2717 + }, + { + "epoch": 7.592178770949721, + "grad_norm": 1.0379213094711304, + "learning_rate": 9.100558659217876e-06, + "loss": 0.2382, + "step": 2718 + }, + { + "epoch": 7.594972067039106, + "grad_norm": 1.2497230768203735, + "learning_rate": 9.10391061452514e-06, + "loss": 0.2293, + "step": 2719 + }, + { + "epoch": 7.597765363128492, + "grad_norm": 0.6723450422286987, + "learning_rate": 9.107262569832402e-06, + "loss": 0.2375, + "step": 2720 + }, + { + "epoch": 7.600558659217877, + "grad_norm": 1.001284122467041, + "learning_rate": 9.110614525139664e-06, + "loss": 0.2123, + "step": 2721 + }, + { + "epoch": 7.603351955307263, + "grad_norm": 1.1725212335586548, + "learning_rate": 9.113966480446928e-06, + "loss": 0.2525, + "step": 2722 + }, + { + "epoch": 7.606145251396648, + "grad_norm": 1.058795690536499, + "learning_rate": 9.117318435754191e-06, + "loss": 0.2341, + "step": 2723 + }, + { + "epoch": 7.608938547486034, + "grad_norm": 0.8814569115638733, + "learning_rate": 9.120670391061453e-06, + "loss": 0.2278, + "step": 2724 + }, + { + "epoch": 7.611731843575419, + "grad_norm": 1.3252323865890503, + "learning_rate": 9.124022346368716e-06, + "loss": 0.2166, + "step": 2725 + }, + { + "epoch": 7.614525139664805, + "grad_norm": 1.4900659322738647, + "learning_rate": 9.127374301675979e-06, + "loss": 0.2627, + "step": 2726 + }, + { + "epoch": 7.61731843575419, + "grad_norm": 1.3027222156524658, + "learning_rate": 9.130726256983242e-06, + "loss": 0.2652, + "step": 2727 + }, + { + "epoch": 7.620111731843576, + "grad_norm": 1.2986621856689453, + "learning_rate": 9.134078212290503e-06, + "loss": 0.2218, + "step": 2728 + }, + { + "epoch": 7.622905027932961, + "grad_norm": 1.4154841899871826, + "learning_rate": 9.137430167597766e-06, + "loss": 0.2695, + "step": 2729 + }, + { + "epoch": 7.625698324022347, + "grad_norm": 3.371523141860962, + "learning_rate": 9.14078212290503e-06, + "loss": 0.327, + "step": 2730 + }, + { + "epoch": 7.628491620111732, + "grad_norm": 2.770813226699829, + "learning_rate": 9.14413407821229e-06, + "loss": 0.4104, + "step": 2731 + }, + { + "epoch": 7.631284916201118, + "grad_norm": 0.6162208914756775, + "learning_rate": 9.147486033519554e-06, + "loss": 0.2115, + "step": 2732 + }, + { + "epoch": 7.634078212290503, + "grad_norm": 0.6520435214042664, + "learning_rate": 9.150837988826817e-06, + "loss": 0.2184, + "step": 2733 + }, + { + "epoch": 7.636871508379889, + "grad_norm": 0.6814623475074768, + "learning_rate": 9.154189944134078e-06, + "loss": 0.2456, + "step": 2734 + }, + { + "epoch": 7.639664804469274, + "grad_norm": 0.5959041118621826, + "learning_rate": 9.157541899441341e-06, + "loss": 0.2283, + "step": 2735 + }, + { + "epoch": 7.64245810055866, + "grad_norm": 0.7630435228347778, + "learning_rate": 9.160893854748604e-06, + "loss": 0.2588, + "step": 2736 + }, + { + "epoch": 7.645251396648044, + "grad_norm": 0.8027440309524536, + "learning_rate": 9.164245810055867e-06, + "loss": 0.2596, + "step": 2737 + }, + { + "epoch": 7.648044692737431, + "grad_norm": 1.801526665687561, + "learning_rate": 9.167597765363128e-06, + "loss": 0.2321, + "step": 2738 + }, + { + "epoch": 7.650837988826815, + "grad_norm": 0.6464086771011353, + "learning_rate": 9.170949720670392e-06, + "loss": 0.2436, + "step": 2739 + }, + { + "epoch": 7.653631284916202, + "grad_norm": 0.9868912696838379, + "learning_rate": 9.174301675977655e-06, + "loss": 0.228, + "step": 2740 + }, + { + "epoch": 7.656424581005586, + "grad_norm": 0.9837371110916138, + "learning_rate": 9.177653631284916e-06, + "loss": 0.2454, + "step": 2741 + }, + { + "epoch": 7.659217877094972, + "grad_norm": 0.8641365766525269, + "learning_rate": 9.181005586592179e-06, + "loss": 0.253, + "step": 2742 + }, + { + "epoch": 7.662011173184357, + "grad_norm": 1.3262900114059448, + "learning_rate": 9.184357541899442e-06, + "loss": 0.2532, + "step": 2743 + }, + { + "epoch": 7.664804469273743, + "grad_norm": 0.698014497756958, + "learning_rate": 9.187709497206705e-06, + "loss": 0.2487, + "step": 2744 + }, + { + "epoch": 7.667597765363128, + "grad_norm": 1.575921893119812, + "learning_rate": 9.191061452513966e-06, + "loss": 0.2284, + "step": 2745 + }, + { + "epoch": 7.670391061452514, + "grad_norm": 0.9499934911727905, + "learning_rate": 9.19441340782123e-06, + "loss": 0.2349, + "step": 2746 + }, + { + "epoch": 7.673184357541899, + "grad_norm": 0.7790555357933044, + "learning_rate": 9.197765363128492e-06, + "loss": 0.2407, + "step": 2747 + }, + { + "epoch": 7.675977653631285, + "grad_norm": 1.812994360923767, + "learning_rate": 9.201117318435754e-06, + "loss": 0.2539, + "step": 2748 + }, + { + "epoch": 7.67877094972067, + "grad_norm": 0.9271278381347656, + "learning_rate": 9.204469273743017e-06, + "loss": 0.2659, + "step": 2749 + }, + { + "epoch": 7.681564245810056, + "grad_norm": 1.0529659986495972, + "learning_rate": 9.20782122905028e-06, + "loss": 0.2305, + "step": 2750 + }, + { + "epoch": 7.684357541899441, + "grad_norm": 1.818073034286499, + "learning_rate": 9.211173184357541e-06, + "loss": 0.3121, + "step": 2751 + }, + { + "epoch": 7.687150837988827, + "grad_norm": 1.0420576333999634, + "learning_rate": 9.214525139664804e-06, + "loss": 0.2854, + "step": 2752 + }, + { + "epoch": 7.689944134078212, + "grad_norm": 1.3105320930480957, + "learning_rate": 9.217877094972067e-06, + "loss": 0.3051, + "step": 2753 + }, + { + "epoch": 7.692737430167598, + "grad_norm": 2.4478421211242676, + "learning_rate": 9.22122905027933e-06, + "loss": 0.2868, + "step": 2754 + }, + { + "epoch": 7.695530726256983, + "grad_norm": 1.6930253505706787, + "learning_rate": 9.224581005586592e-06, + "loss": 0.3485, + "step": 2755 + }, + { + "epoch": 7.698324022346369, + "grad_norm": 3.4217753410339355, + "learning_rate": 9.227932960893855e-06, + "loss": 0.386, + "step": 2756 + }, + { + "epoch": 7.701117318435754, + "grad_norm": 0.5814682841300964, + "learning_rate": 9.231284916201118e-06, + "loss": 0.2275, + "step": 2757 + }, + { + "epoch": 7.70391061452514, + "grad_norm": 0.5095791816711426, + "learning_rate": 9.234636871508379e-06, + "loss": 0.2095, + "step": 2758 + }, + { + "epoch": 7.706703910614525, + "grad_norm": 0.617194652557373, + "learning_rate": 9.237988826815642e-06, + "loss": 0.2225, + "step": 2759 + }, + { + "epoch": 7.709497206703911, + "grad_norm": 0.6921057105064392, + "learning_rate": 9.241340782122905e-06, + "loss": 0.2359, + "step": 2760 + }, + { + "epoch": 7.712290502793296, + "grad_norm": 0.5913490056991577, + "learning_rate": 9.244692737430167e-06, + "loss": 0.2135, + "step": 2761 + }, + { + "epoch": 7.715083798882682, + "grad_norm": 0.6599303483963013, + "learning_rate": 9.24804469273743e-06, + "loss": 0.223, + "step": 2762 + }, + { + "epoch": 7.717877094972067, + "grad_norm": 1.0926616191864014, + "learning_rate": 9.251396648044693e-06, + "loss": 0.252, + "step": 2763 + }, + { + "epoch": 7.720670391061453, + "grad_norm": 0.6559785604476929, + "learning_rate": 9.254748603351956e-06, + "loss": 0.2223, + "step": 2764 + }, + { + "epoch": 7.723463687150838, + "grad_norm": 0.9627076387405396, + "learning_rate": 9.258100558659217e-06, + "loss": 0.218, + "step": 2765 + }, + { + "epoch": 7.726256983240224, + "grad_norm": 0.7492142915725708, + "learning_rate": 9.26145251396648e-06, + "loss": 0.2621, + "step": 2766 + }, + { + "epoch": 7.729050279329609, + "grad_norm": 0.9181290864944458, + "learning_rate": 9.264804469273743e-06, + "loss": 0.2237, + "step": 2767 + }, + { + "epoch": 7.731843575418995, + "grad_norm": 0.8641600608825684, + "learning_rate": 9.268156424581004e-06, + "loss": 0.2933, + "step": 2768 + }, + { + "epoch": 7.73463687150838, + "grad_norm": 1.5459678173065186, + "learning_rate": 9.271508379888267e-06, + "loss": 0.2066, + "step": 2769 + }, + { + "epoch": 7.737430167597766, + "grad_norm": 0.8733140230178833, + "learning_rate": 9.274860335195532e-06, + "loss": 0.2314, + "step": 2770 + }, + { + "epoch": 7.740223463687151, + "grad_norm": 1.7758053541183472, + "learning_rate": 9.278212290502794e-06, + "loss": 0.2029, + "step": 2771 + }, + { + "epoch": 7.743016759776537, + "grad_norm": 0.8926159143447876, + "learning_rate": 9.281564245810057e-06, + "loss": 0.2383, + "step": 2772 + }, + { + "epoch": 7.745810055865922, + "grad_norm": 0.9022583961486816, + "learning_rate": 9.28491620111732e-06, + "loss": 0.2922, + "step": 2773 + }, + { + "epoch": 7.748603351955307, + "grad_norm": 1.335143804550171, + "learning_rate": 9.288268156424583e-06, + "loss": 0.2346, + "step": 2774 + }, + { + "epoch": 7.751396648044693, + "grad_norm": 1.276861310005188, + "learning_rate": 9.291620111731844e-06, + "loss": 0.2358, + "step": 2775 + }, + { + "epoch": 7.754189944134078, + "grad_norm": 1.1662808656692505, + "learning_rate": 9.294972067039107e-06, + "loss": 0.2533, + "step": 2776 + }, + { + "epoch": 7.756983240223463, + "grad_norm": 1.5937052965164185, + "learning_rate": 9.29832402234637e-06, + "loss": 0.2645, + "step": 2777 + }, + { + "epoch": 7.759776536312849, + "grad_norm": 1.9294190406799316, + "learning_rate": 9.301675977653631e-06, + "loss": 0.253, + "step": 2778 + }, + { + "epoch": 7.762569832402234, + "grad_norm": 1.3958450555801392, + "learning_rate": 9.305027932960894e-06, + "loss": 0.2637, + "step": 2779 + }, + { + "epoch": 7.76536312849162, + "grad_norm": 1.8608360290527344, + "learning_rate": 9.308379888268157e-06, + "loss": 0.33, + "step": 2780 + }, + { + "epoch": 7.768156424581005, + "grad_norm": 2.8312206268310547, + "learning_rate": 9.311731843575419e-06, + "loss": 0.3671, + "step": 2781 + }, + { + "epoch": 7.770949720670391, + "grad_norm": 0.6473077535629272, + "learning_rate": 9.315083798882682e-06, + "loss": 0.2545, + "step": 2782 + }, + { + "epoch": 7.773743016759776, + "grad_norm": 0.501976728439331, + "learning_rate": 9.318435754189945e-06, + "loss": 0.1994, + "step": 2783 + }, + { + "epoch": 7.776536312849162, + "grad_norm": 1.2209641933441162, + "learning_rate": 9.321787709497208e-06, + "loss": 0.2265, + "step": 2784 + }, + { + "epoch": 7.779329608938547, + "grad_norm": 0.8436747193336487, + "learning_rate": 9.32513966480447e-06, + "loss": 0.2251, + "step": 2785 + }, + { + "epoch": 7.782122905027933, + "grad_norm": 0.6794122457504272, + "learning_rate": 9.328491620111732e-06, + "loss": 0.2412, + "step": 2786 + }, + { + "epoch": 7.784916201117318, + "grad_norm": 1.0301513671875, + "learning_rate": 9.331843575418995e-06, + "loss": 0.2357, + "step": 2787 + }, + { + "epoch": 7.787709497206704, + "grad_norm": 0.6871400475502014, + "learning_rate": 9.335195530726257e-06, + "loss": 0.2342, + "step": 2788 + }, + { + "epoch": 7.790502793296089, + "grad_norm": 0.8454127907752991, + "learning_rate": 9.33854748603352e-06, + "loss": 0.2326, + "step": 2789 + }, + { + "epoch": 7.793296089385475, + "grad_norm": 1.0491009950637817, + "learning_rate": 9.341899441340783e-06, + "loss": 0.2242, + "step": 2790 + }, + { + "epoch": 7.79608938547486, + "grad_norm": 1.311628818511963, + "learning_rate": 9.345251396648046e-06, + "loss": 0.2481, + "step": 2791 + }, + { + "epoch": 7.798882681564246, + "grad_norm": 4.139593124389648, + "learning_rate": 9.348603351955307e-06, + "loss": 0.2825, + "step": 2792 + }, + { + "epoch": 7.801675977653631, + "grad_norm": 1.0153018236160278, + "learning_rate": 9.35195530726257e-06, + "loss": 0.2562, + "step": 2793 + }, + { + "epoch": 7.804469273743017, + "grad_norm": 0.9669975638389587, + "learning_rate": 9.355307262569833e-06, + "loss": 0.2382, + "step": 2794 + }, + { + "epoch": 7.807262569832402, + "grad_norm": 0.7315298914909363, + "learning_rate": 9.358659217877095e-06, + "loss": 0.2406, + "step": 2795 + }, + { + "epoch": 7.810055865921788, + "grad_norm": 1.147596836090088, + "learning_rate": 9.362011173184358e-06, + "loss": 0.1988, + "step": 2796 + }, + { + "epoch": 7.812849162011173, + "grad_norm": 0.9414712190628052, + "learning_rate": 9.36536312849162e-06, + "loss": 0.2414, + "step": 2797 + }, + { + "epoch": 7.815642458100559, + "grad_norm": 1.0673624277114868, + "learning_rate": 9.368715083798882e-06, + "loss": 0.2418, + "step": 2798 + }, + { + "epoch": 7.818435754189944, + "grad_norm": 1.6385141611099243, + "learning_rate": 9.372067039106145e-06, + "loss": 0.2833, + "step": 2799 + }, + { + "epoch": 7.82122905027933, + "grad_norm": 9.928041458129883, + "learning_rate": 9.375418994413408e-06, + "loss": 0.2336, + "step": 2800 + }, + { + "epoch": 7.824022346368715, + "grad_norm": 0.9916796088218689, + "learning_rate": 9.378770949720671e-06, + "loss": 0.2755, + "step": 2801 + }, + { + "epoch": 7.826815642458101, + "grad_norm": 1.3494396209716797, + "learning_rate": 9.382122905027933e-06, + "loss": 0.2416, + "step": 2802 + }, + { + "epoch": 7.829608938547486, + "grad_norm": 1.4427664279937744, + "learning_rate": 9.385474860335196e-06, + "loss": 0.2683, + "step": 2803 + }, + { + "epoch": 7.832402234636872, + "grad_norm": 2.2981674671173096, + "learning_rate": 9.388826815642459e-06, + "loss": 0.275, + "step": 2804 + }, + { + "epoch": 7.835195530726257, + "grad_norm": 2.1910197734832764, + "learning_rate": 9.39217877094972e-06, + "loss": 0.3592, + "step": 2805 + }, + { + "epoch": 7.837988826815643, + "grad_norm": 3.235081434249878, + "learning_rate": 9.395530726256983e-06, + "loss": 0.3538, + "step": 2806 + }, + { + "epoch": 7.840782122905028, + "grad_norm": 0.6609557867050171, + "learning_rate": 9.398882681564246e-06, + "loss": 0.2497, + "step": 2807 + }, + { + "epoch": 7.843575418994414, + "grad_norm": 0.7624232769012451, + "learning_rate": 9.402234636871507e-06, + "loss": 0.2238, + "step": 2808 + }, + { + "epoch": 7.846368715083798, + "grad_norm": 0.6822518110275269, + "learning_rate": 9.40558659217877e-06, + "loss": 0.2694, + "step": 2809 + }, + { + "epoch": 7.849162011173185, + "grad_norm": 0.8797164559364319, + "learning_rate": 9.408938547486033e-06, + "loss": 0.2158, + "step": 2810 + }, + { + "epoch": 7.851955307262569, + "grad_norm": 2.094189167022705, + "learning_rate": 9.412290502793296e-06, + "loss": 0.2625, + "step": 2811 + }, + { + "epoch": 7.854748603351956, + "grad_norm": 0.7812762260437012, + "learning_rate": 9.415642458100558e-06, + "loss": 0.2291, + "step": 2812 + }, + { + "epoch": 7.85754189944134, + "grad_norm": 0.9590010643005371, + "learning_rate": 9.418994413407821e-06, + "loss": 0.2143, + "step": 2813 + }, + { + "epoch": 7.860335195530726, + "grad_norm": 0.7465453147888184, + "learning_rate": 9.422346368715084e-06, + "loss": 0.1975, + "step": 2814 + }, + { + "epoch": 7.863128491620111, + "grad_norm": 1.232529640197754, + "learning_rate": 9.425698324022345e-06, + "loss": 0.2399, + "step": 2815 + }, + { + "epoch": 7.865921787709497, + "grad_norm": 1.5988562107086182, + "learning_rate": 9.429050279329608e-06, + "loss": 0.2775, + "step": 2816 + }, + { + "epoch": 7.868715083798882, + "grad_norm": 0.8076841831207275, + "learning_rate": 9.432402234636871e-06, + "loss": 0.2049, + "step": 2817 + }, + { + "epoch": 7.871508379888268, + "grad_norm": 0.9278537034988403, + "learning_rate": 9.435754189944134e-06, + "loss": 0.3003, + "step": 2818 + }, + { + "epoch": 7.874301675977653, + "grad_norm": 0.9267262816429138, + "learning_rate": 9.439106145251397e-06, + "loss": 0.2052, + "step": 2819 + }, + { + "epoch": 7.877094972067039, + "grad_norm": 1.2062160968780518, + "learning_rate": 9.44245810055866e-06, + "loss": 0.2685, + "step": 2820 + }, + { + "epoch": 7.879888268156424, + "grad_norm": 1.083837866783142, + "learning_rate": 9.445810055865923e-06, + "loss": 0.2719, + "step": 2821 + }, + { + "epoch": 7.88268156424581, + "grad_norm": 1.553768515586853, + "learning_rate": 9.449162011173185e-06, + "loss": 0.2419, + "step": 2822 + }, + { + "epoch": 7.885474860335195, + "grad_norm": 1.8463865518569946, + "learning_rate": 9.452513966480448e-06, + "loss": 0.2688, + "step": 2823 + }, + { + "epoch": 7.888268156424581, + "grad_norm": 1.3606740236282349, + "learning_rate": 9.455865921787711e-06, + "loss": 0.2505, + "step": 2824 + }, + { + "epoch": 7.891061452513966, + "grad_norm": 0.9688084125518799, + "learning_rate": 9.459217877094972e-06, + "loss": 0.2631, + "step": 2825 + }, + { + "epoch": 7.893854748603352, + "grad_norm": 1.0907816886901855, + "learning_rate": 9.462569832402235e-06, + "loss": 0.239, + "step": 2826 + }, + { + "epoch": 7.896648044692737, + "grad_norm": 1.0867913961410522, + "learning_rate": 9.465921787709498e-06, + "loss": 0.2707, + "step": 2827 + }, + { + "epoch": 7.899441340782123, + "grad_norm": 3.199425220489502, + "learning_rate": 9.46927374301676e-06, + "loss": 0.2731, + "step": 2828 + }, + { + "epoch": 7.902234636871508, + "grad_norm": 3.099682569503784, + "learning_rate": 9.472625698324023e-06, + "loss": 0.2687, + "step": 2829 + }, + { + "epoch": 7.905027932960894, + "grad_norm": 1.9276171922683716, + "learning_rate": 9.475977653631286e-06, + "loss": 0.3161, + "step": 2830 + }, + { + "epoch": 7.907821229050279, + "grad_norm": 2.26556658744812, + "learning_rate": 9.479329608938549e-06, + "loss": 0.3816, + "step": 2831 + }, + { + "epoch": 7.910614525139665, + "grad_norm": 0.5601553916931152, + "learning_rate": 9.48268156424581e-06, + "loss": 0.2235, + "step": 2832 + }, + { + "epoch": 7.91340782122905, + "grad_norm": 0.6048760414123535, + "learning_rate": 9.486033519553073e-06, + "loss": 0.2249, + "step": 2833 + }, + { + "epoch": 7.916201117318436, + "grad_norm": 0.9172365069389343, + "learning_rate": 9.489385474860336e-06, + "loss": 0.2307, + "step": 2834 + }, + { + "epoch": 7.918994413407821, + "grad_norm": 0.8502371311187744, + "learning_rate": 9.492737430167598e-06, + "loss": 0.2829, + "step": 2835 + }, + { + "epoch": 7.921787709497207, + "grad_norm": 0.7375717759132385, + "learning_rate": 9.49608938547486e-06, + "loss": 0.2094, + "step": 2836 + }, + { + "epoch": 7.924581005586592, + "grad_norm": 1.9352052211761475, + "learning_rate": 9.499441340782124e-06, + "loss": 0.2157, + "step": 2837 + }, + { + "epoch": 7.927374301675978, + "grad_norm": 0.7618995904922485, + "learning_rate": 9.502793296089387e-06, + "loss": 0.2846, + "step": 2838 + }, + { + "epoch": 7.930167597765363, + "grad_norm": 0.6381585597991943, + "learning_rate": 9.506145251396648e-06, + "loss": 0.2237, + "step": 2839 + }, + { + "epoch": 7.932960893854749, + "grad_norm": 0.9208477139472961, + "learning_rate": 9.509497206703911e-06, + "loss": 0.2215, + "step": 2840 + }, + { + "epoch": 7.935754189944134, + "grad_norm": 0.6449414491653442, + "learning_rate": 9.512849162011174e-06, + "loss": 0.2058, + "step": 2841 + }, + { + "epoch": 7.93854748603352, + "grad_norm": 0.7235997915267944, + "learning_rate": 9.516201117318435e-06, + "loss": 0.2267, + "step": 2842 + }, + { + "epoch": 7.941340782122905, + "grad_norm": 0.9473241567611694, + "learning_rate": 9.519553072625698e-06, + "loss": 0.2088, + "step": 2843 + }, + { + "epoch": 7.94413407821229, + "grad_norm": 0.804825484752655, + "learning_rate": 9.522905027932962e-06, + "loss": 0.2084, + "step": 2844 + }, + { + "epoch": 7.946927374301676, + "grad_norm": 1.26486337184906, + "learning_rate": 9.526256983240223e-06, + "loss": 0.1999, + "step": 2845 + }, + { + "epoch": 7.949720670391061, + "grad_norm": 1.3453733921051025, + "learning_rate": 9.529608938547486e-06, + "loss": 0.2724, + "step": 2846 + }, + { + "epoch": 7.952513966480447, + "grad_norm": 1.0380553007125854, + "learning_rate": 9.532960893854749e-06, + "loss": 0.279, + "step": 2847 + }, + { + "epoch": 7.955307262569832, + "grad_norm": 1.8341478109359741, + "learning_rate": 9.536312849162012e-06, + "loss": 0.2613, + "step": 2848 + }, + { + "epoch": 7.9581005586592175, + "grad_norm": 1.4617564678192139, + "learning_rate": 9.539664804469273e-06, + "loss": 0.2508, + "step": 2849 + }, + { + "epoch": 7.960893854748603, + "grad_norm": 1.4025949239730835, + "learning_rate": 9.543016759776536e-06, + "loss": 0.2572, + "step": 2850 + }, + { + "epoch": 7.9636871508379885, + "grad_norm": 2.3288955688476562, + "learning_rate": 9.5463687150838e-06, + "loss": 0.2643, + "step": 2851 + }, + { + "epoch": 7.966480446927374, + "grad_norm": 1.1137354373931885, + "learning_rate": 9.54972067039106e-06, + "loss": 0.2211, + "step": 2852 + }, + { + "epoch": 7.9692737430167595, + "grad_norm": 1.2303675413131714, + "learning_rate": 9.553072625698324e-06, + "loss": 0.2738, + "step": 2853 + }, + { + "epoch": 7.972067039106145, + "grad_norm": 1.7235690355300903, + "learning_rate": 9.556424581005587e-06, + "loss": 0.2906, + "step": 2854 + }, + { + "epoch": 7.9748603351955305, + "grad_norm": 1.8483260869979858, + "learning_rate": 9.559776536312848e-06, + "loss": 0.3303, + "step": 2855 + }, + { + "epoch": 7.977653631284916, + "grad_norm": 2.07131290435791, + "learning_rate": 9.563128491620111e-06, + "loss": 0.4153, + "step": 2856 + }, + { + "epoch": 7.9804469273743015, + "grad_norm": 0.6586955785751343, + "learning_rate": 9.566480446927374e-06, + "loss": 0.2589, + "step": 2857 + }, + { + "epoch": 7.983240223463687, + "grad_norm": 0.668641984462738, + "learning_rate": 9.569832402234637e-06, + "loss": 0.2311, + "step": 2858 + }, + { + "epoch": 7.9860335195530725, + "grad_norm": 0.6149654388427734, + "learning_rate": 9.573184357541899e-06, + "loss": 0.2153, + "step": 2859 + }, + { + "epoch": 7.988826815642458, + "grad_norm": 0.9377949833869934, + "learning_rate": 9.576536312849162e-06, + "loss": 0.2339, + "step": 2860 + }, + { + "epoch": 7.9916201117318435, + "grad_norm": 0.8185917139053345, + "learning_rate": 9.579888268156425e-06, + "loss": 0.2194, + "step": 2861 + }, + { + "epoch": 7.994413407821229, + "grad_norm": 1.0313695669174194, + "learning_rate": 9.583240223463686e-06, + "loss": 0.2478, + "step": 2862 + }, + { + "epoch": 7.9972067039106145, + "grad_norm": 0.9864821434020996, + "learning_rate": 9.586592178770949e-06, + "loss": 0.2402, + "step": 2863 + }, + { + "epoch": 8.0, + "grad_norm": 1.685511589050293, + "learning_rate": 9.589944134078212e-06, + "loss": 0.3015, + "step": 2864 + }, + { + "epoch": 8.002793296089385, + "grad_norm": 0.6939486861228943, + "learning_rate": 9.593296089385473e-06, + "loss": 0.2296, + "step": 2865 + }, + { + "epoch": 8.005586592178771, + "grad_norm": 0.6062296628952026, + "learning_rate": 9.596648044692738e-06, + "loss": 0.2383, + "step": 2866 + }, + { + "epoch": 8.008379888268156, + "grad_norm": 0.5830140709877014, + "learning_rate": 9.600000000000001e-06, + "loss": 0.2179, + "step": 2867 + }, + { + "epoch": 8.011173184357542, + "grad_norm": 0.7072479128837585, + "learning_rate": 9.603351955307264e-06, + "loss": 0.2225, + "step": 2868 + }, + { + "epoch": 8.013966480446927, + "grad_norm": 0.65046226978302, + "learning_rate": 9.606703910614526e-06, + "loss": 0.1942, + "step": 2869 + }, + { + "epoch": 8.016759776536313, + "grad_norm": 0.8568791151046753, + "learning_rate": 9.610055865921789e-06, + "loss": 0.2365, + "step": 2870 + }, + { + "epoch": 8.019553072625698, + "grad_norm": 0.7171811461448669, + "learning_rate": 9.613407821229052e-06, + "loss": 0.1758, + "step": 2871 + }, + { + "epoch": 8.022346368715084, + "grad_norm": 0.7524972558021545, + "learning_rate": 9.616759776536313e-06, + "loss": 0.2177, + "step": 2872 + }, + { + "epoch": 8.025139664804469, + "grad_norm": 0.7712451219558716, + "learning_rate": 9.620111731843576e-06, + "loss": 0.204, + "step": 2873 + }, + { + "epoch": 8.027932960893855, + "grad_norm": 1.5023229122161865, + "learning_rate": 9.623463687150839e-06, + "loss": 0.256, + "step": 2874 + }, + { + "epoch": 8.03072625698324, + "grad_norm": 1.7019081115722656, + "learning_rate": 9.6268156424581e-06, + "loss": 0.2226, + "step": 2875 + }, + { + "epoch": 8.033519553072626, + "grad_norm": 0.9509430527687073, + "learning_rate": 9.630167597765364e-06, + "loss": 0.2226, + "step": 2876 + }, + { + "epoch": 8.03631284916201, + "grad_norm": 1.1219485998153687, + "learning_rate": 9.633519553072627e-06, + "loss": 0.2044, + "step": 2877 + }, + { + "epoch": 8.039106145251397, + "grad_norm": 0.8191317319869995, + "learning_rate": 9.63687150837989e-06, + "loss": 0.2078, + "step": 2878 + }, + { + "epoch": 8.041899441340782, + "grad_norm": 1.0232034921646118, + "learning_rate": 9.640223463687151e-06, + "loss": 0.2281, + "step": 2879 + }, + { + "epoch": 8.044692737430168, + "grad_norm": 1.112970232963562, + "learning_rate": 9.643575418994414e-06, + "loss": 0.2212, + "step": 2880 + }, + { + "epoch": 8.047486033519553, + "grad_norm": 0.953940749168396, + "learning_rate": 9.646927374301677e-06, + "loss": 0.2095, + "step": 2881 + }, + { + "epoch": 8.050279329608939, + "grad_norm": 1.2026376724243164, + "learning_rate": 9.650279329608938e-06, + "loss": 0.2408, + "step": 2882 + }, + { + "epoch": 8.053072625698324, + "grad_norm": 1.107146978378296, + "learning_rate": 9.653631284916201e-06, + "loss": 0.2469, + "step": 2883 + }, + { + "epoch": 8.05586592178771, + "grad_norm": 0.8951582908630371, + "learning_rate": 9.656983240223464e-06, + "loss": 0.2281, + "step": 2884 + }, + { + "epoch": 8.058659217877095, + "grad_norm": 1.2258003950119019, + "learning_rate": 9.660335195530727e-06, + "loss": 0.2349, + "step": 2885 + }, + { + "epoch": 8.061452513966481, + "grad_norm": 1.3615211248397827, + "learning_rate": 9.663687150837989e-06, + "loss": 0.2713, + "step": 2886 + }, + { + "epoch": 8.064245810055866, + "grad_norm": 1.6446081399917603, + "learning_rate": 9.667039106145252e-06, + "loss": 0.2477, + "step": 2887 + }, + { + "epoch": 8.067039106145252, + "grad_norm": 1.589110255241394, + "learning_rate": 9.670391061452515e-06, + "loss": 0.3126, + "step": 2888 + }, + { + "epoch": 8.069832402234637, + "grad_norm": 2.4356026649475098, + "learning_rate": 9.673743016759776e-06, + "loss": 0.3646, + "step": 2889 + }, + { + "epoch": 8.072625698324023, + "grad_norm": 0.511897087097168, + "learning_rate": 9.67709497206704e-06, + "loss": 0.245, + "step": 2890 + }, + { + "epoch": 8.075418994413408, + "grad_norm": 0.6173885464668274, + "learning_rate": 9.680446927374302e-06, + "loss": 0.2074, + "step": 2891 + }, + { + "epoch": 8.078212290502794, + "grad_norm": 0.5936695337295532, + "learning_rate": 9.683798882681564e-06, + "loss": 0.2414, + "step": 2892 + }, + { + "epoch": 8.081005586592179, + "grad_norm": 0.652831494808197, + "learning_rate": 9.687150837988827e-06, + "loss": 0.2602, + "step": 2893 + }, + { + "epoch": 8.083798882681565, + "grad_norm": 0.905451238155365, + "learning_rate": 9.69050279329609e-06, + "loss": 0.2342, + "step": 2894 + }, + { + "epoch": 8.08659217877095, + "grad_norm": 0.8095020055770874, + "learning_rate": 9.693854748603353e-06, + "loss": 0.2405, + "step": 2895 + }, + { + "epoch": 8.089385474860336, + "grad_norm": 0.9065225124359131, + "learning_rate": 9.697206703910614e-06, + "loss": 0.226, + "step": 2896 + }, + { + "epoch": 8.09217877094972, + "grad_norm": 0.6763017773628235, + "learning_rate": 9.700558659217877e-06, + "loss": 0.2157, + "step": 2897 + }, + { + "epoch": 8.094972067039107, + "grad_norm": 0.9021303057670593, + "learning_rate": 9.70391061452514e-06, + "loss": 0.1961, + "step": 2898 + }, + { + "epoch": 8.097765363128492, + "grad_norm": 1.0509302616119385, + "learning_rate": 9.707262569832402e-06, + "loss": 0.2141, + "step": 2899 + }, + { + "epoch": 8.100558659217878, + "grad_norm": 0.7878400087356567, + "learning_rate": 9.710614525139665e-06, + "loss": 0.1968, + "step": 2900 + }, + { + "epoch": 8.103351955307263, + "grad_norm": 1.04593026638031, + "learning_rate": 9.713966480446928e-06, + "loss": 0.2533, + "step": 2901 + }, + { + "epoch": 8.106145251396647, + "grad_norm": 0.8387154340744019, + "learning_rate": 9.717318435754189e-06, + "loss": 0.2411, + "step": 2902 + }, + { + "epoch": 8.108938547486034, + "grad_norm": 0.7945376634597778, + "learning_rate": 9.720670391061452e-06, + "loss": 0.2597, + "step": 2903 + }, + { + "epoch": 8.111731843575418, + "grad_norm": 1.0276771783828735, + "learning_rate": 9.724022346368715e-06, + "loss": 0.2025, + "step": 2904 + }, + { + "epoch": 8.114525139664805, + "grad_norm": 1.2185399532318115, + "learning_rate": 9.727374301675978e-06, + "loss": 0.2609, + "step": 2905 + }, + { + "epoch": 8.11731843575419, + "grad_norm": 1.7058706283569336, + "learning_rate": 9.73072625698324e-06, + "loss": 0.231, + "step": 2906 + }, + { + "epoch": 8.120111731843576, + "grad_norm": 5.250736236572266, + "learning_rate": 9.734078212290503e-06, + "loss": 0.222, + "step": 2907 + }, + { + "epoch": 8.12290502793296, + "grad_norm": 5.784262180328369, + "learning_rate": 9.737430167597766e-06, + "loss": 0.259, + "step": 2908 + }, + { + "epoch": 8.125698324022347, + "grad_norm": 1.4386264085769653, + "learning_rate": 9.740782122905027e-06, + "loss": 0.2655, + "step": 2909 + }, + { + "epoch": 8.128491620111731, + "grad_norm": 1.1503362655639648, + "learning_rate": 9.74413407821229e-06, + "loss": 0.2284, + "step": 2910 + }, + { + "epoch": 8.131284916201118, + "grad_norm": 1.0780434608459473, + "learning_rate": 9.747486033519553e-06, + "loss": 0.2798, + "step": 2911 + }, + { + "epoch": 8.134078212290502, + "grad_norm": 1.3906681537628174, + "learning_rate": 9.750837988826814e-06, + "loss": 0.2724, + "step": 2912 + }, + { + "epoch": 8.136871508379889, + "grad_norm": 1.6353660821914673, + "learning_rate": 9.754189944134077e-06, + "loss": 0.2464, + "step": 2913 + }, + { + "epoch": 8.139664804469273, + "grad_norm": 2.365215539932251, + "learning_rate": 9.75754189944134e-06, + "loss": 0.4434, + "step": 2914 + }, + { + "epoch": 8.14245810055866, + "grad_norm": 0.7092537879943848, + "learning_rate": 9.760893854748605e-06, + "loss": 0.2603, + "step": 2915 + }, + { + "epoch": 8.145251396648044, + "grad_norm": 0.6176267862319946, + "learning_rate": 9.764245810055866e-06, + "loss": 0.2461, + "step": 2916 + }, + { + "epoch": 8.14804469273743, + "grad_norm": 0.8864729404449463, + "learning_rate": 9.76759776536313e-06, + "loss": 0.217, + "step": 2917 + }, + { + "epoch": 8.150837988826815, + "grad_norm": 0.8639066219329834, + "learning_rate": 9.770949720670393e-06, + "loss": 0.2248, + "step": 2918 + }, + { + "epoch": 8.153631284916202, + "grad_norm": 1.9839472770690918, + "learning_rate": 9.774301675977654e-06, + "loss": 0.2147, + "step": 2919 + }, + { + "epoch": 8.156424581005586, + "grad_norm": 0.8608357310295105, + "learning_rate": 9.777653631284917e-06, + "loss": 0.245, + "step": 2920 + }, + { + "epoch": 8.159217877094973, + "grad_norm": 0.6835077404975891, + "learning_rate": 9.78100558659218e-06, + "loss": 0.2555, + "step": 2921 + }, + { + "epoch": 8.162011173184357, + "grad_norm": 0.6973780393600464, + "learning_rate": 9.784357541899441e-06, + "loss": 0.2037, + "step": 2922 + }, + { + "epoch": 8.164804469273744, + "grad_norm": 0.6855647563934326, + "learning_rate": 9.787709497206704e-06, + "loss": 0.1822, + "step": 2923 + }, + { + "epoch": 8.167597765363128, + "grad_norm": 0.9368285536766052, + "learning_rate": 9.791061452513967e-06, + "loss": 0.2509, + "step": 2924 + }, + { + "epoch": 8.170391061452515, + "grad_norm": 0.8383121490478516, + "learning_rate": 9.79441340782123e-06, + "loss": 0.2078, + "step": 2925 + }, + { + "epoch": 8.1731843575419, + "grad_norm": 1.2396692037582397, + "learning_rate": 9.797765363128492e-06, + "loss": 0.2189, + "step": 2926 + }, + { + "epoch": 8.175977653631286, + "grad_norm": 0.8003161549568176, + "learning_rate": 9.801117318435755e-06, + "loss": 0.2388, + "step": 2927 + }, + { + "epoch": 8.17877094972067, + "grad_norm": 0.9286906719207764, + "learning_rate": 9.804469273743018e-06, + "loss": 0.2573, + "step": 2928 + }, + { + "epoch": 8.181564245810057, + "grad_norm": 0.8946781158447266, + "learning_rate": 9.80782122905028e-06, + "loss": 0.2067, + "step": 2929 + }, + { + "epoch": 8.184357541899441, + "grad_norm": 1.0056356191635132, + "learning_rate": 9.811173184357542e-06, + "loss": 0.2326, + "step": 2930 + }, + { + "epoch": 8.187150837988828, + "grad_norm": 1.3688722848892212, + "learning_rate": 9.814525139664805e-06, + "loss": 0.3057, + "step": 2931 + }, + { + "epoch": 8.189944134078212, + "grad_norm": 1.3620983362197876, + "learning_rate": 9.817877094972068e-06, + "loss": 0.2694, + "step": 2932 + }, + { + "epoch": 8.192737430167599, + "grad_norm": 0.9765917062759399, + "learning_rate": 9.82122905027933e-06, + "loss": 0.2499, + "step": 2933 + }, + { + "epoch": 8.195530726256983, + "grad_norm": 1.7058237791061401, + "learning_rate": 9.824581005586593e-06, + "loss": 0.2816, + "step": 2934 + }, + { + "epoch": 8.19832402234637, + "grad_norm": 1.307533621788025, + "learning_rate": 9.827932960893856e-06, + "loss": 0.3104, + "step": 2935 + }, + { + "epoch": 8.201117318435754, + "grad_norm": 1.3264260292053223, + "learning_rate": 9.831284916201117e-06, + "loss": 0.2615, + "step": 2936 + }, + { + "epoch": 8.203910614525139, + "grad_norm": 2.632157325744629, + "learning_rate": 9.83463687150838e-06, + "loss": 0.285, + "step": 2937 + }, + { + "epoch": 8.206703910614525, + "grad_norm": 1.527962327003479, + "learning_rate": 9.837988826815643e-06, + "loss": 0.3343, + "step": 2938 + }, + { + "epoch": 8.20949720670391, + "grad_norm": 2.1902785301208496, + "learning_rate": 9.841340782122905e-06, + "loss": 0.3654, + "step": 2939 + }, + { + "epoch": 8.212290502793296, + "grad_norm": 0.6361067891120911, + "learning_rate": 9.844692737430168e-06, + "loss": 0.2524, + "step": 2940 + }, + { + "epoch": 8.21508379888268, + "grad_norm": 0.6003903746604919, + "learning_rate": 9.84804469273743e-06, + "loss": 0.24, + "step": 2941 + }, + { + "epoch": 8.217877094972067, + "grad_norm": 0.4744698107242584, + "learning_rate": 9.851396648044694e-06, + "loss": 0.2309, + "step": 2942 + }, + { + "epoch": 8.220670391061452, + "grad_norm": 0.581177294254303, + "learning_rate": 9.854748603351955e-06, + "loss": 0.2173, + "step": 2943 + }, + { + "epoch": 8.223463687150838, + "grad_norm": 0.6973459720611572, + "learning_rate": 9.858100558659218e-06, + "loss": 0.2319, + "step": 2944 + }, + { + "epoch": 8.226256983240223, + "grad_norm": 0.6542634963989258, + "learning_rate": 9.861452513966481e-06, + "loss": 0.2268, + "step": 2945 + }, + { + "epoch": 8.22905027932961, + "grad_norm": 0.6184661388397217, + "learning_rate": 9.864804469273742e-06, + "loss": 0.2291, + "step": 2946 + }, + { + "epoch": 8.231843575418994, + "grad_norm": 0.6873344779014587, + "learning_rate": 9.868156424581005e-06, + "loss": 0.2283, + "step": 2947 + }, + { + "epoch": 8.23463687150838, + "grad_norm": 0.9166136384010315, + "learning_rate": 9.871508379888268e-06, + "loss": 0.221, + "step": 2948 + }, + { + "epoch": 8.237430167597765, + "grad_norm": 0.7768459916114807, + "learning_rate": 9.87486033519553e-06, + "loss": 0.2182, + "step": 2949 + }, + { + "epoch": 8.240223463687151, + "grad_norm": 0.8863435983657837, + "learning_rate": 9.878212290502793e-06, + "loss": 0.2179, + "step": 2950 + }, + { + "epoch": 8.243016759776536, + "grad_norm": 0.8913365006446838, + "learning_rate": 9.881564245810056e-06, + "loss": 0.2376, + "step": 2951 + }, + { + "epoch": 8.245810055865922, + "grad_norm": 1.5076298713684082, + "learning_rate": 9.884916201117319e-06, + "loss": 0.1964, + "step": 2952 + }, + { + "epoch": 8.248603351955307, + "grad_norm": 0.9155826568603516, + "learning_rate": 9.88826815642458e-06, + "loss": 0.2585, + "step": 2953 + }, + { + "epoch": 8.251396648044693, + "grad_norm": 0.8352968692779541, + "learning_rate": 9.891620111731843e-06, + "loss": 0.1875, + "step": 2954 + }, + { + "epoch": 8.254189944134078, + "grad_norm": 0.8252818584442139, + "learning_rate": 9.894972067039106e-06, + "loss": 0.2195, + "step": 2955 + }, + { + "epoch": 8.256983240223464, + "grad_norm": 1.7558505535125732, + "learning_rate": 9.898324022346368e-06, + "loss": 0.2487, + "step": 2956 + }, + { + "epoch": 8.259776536312849, + "grad_norm": 1.018198847770691, + "learning_rate": 9.90167597765363e-06, + "loss": 0.2072, + "step": 2957 + }, + { + "epoch": 8.262569832402235, + "grad_norm": 1.2218213081359863, + "learning_rate": 9.905027932960894e-06, + "loss": 0.2506, + "step": 2958 + }, + { + "epoch": 8.26536312849162, + "grad_norm": 1.2235894203186035, + "learning_rate": 9.908379888268155e-06, + "loss": 0.2034, + "step": 2959 + }, + { + "epoch": 8.268156424581006, + "grad_norm": 1.6719125509262085, + "learning_rate": 9.911731843575418e-06, + "loss": 0.341, + "step": 2960 + }, + { + "epoch": 8.27094972067039, + "grad_norm": 1.0583816766738892, + "learning_rate": 9.915083798882681e-06, + "loss": 0.2602, + "step": 2961 + }, + { + "epoch": 8.273743016759777, + "grad_norm": 1.767026662826538, + "learning_rate": 9.918435754189944e-06, + "loss": 0.2894, + "step": 2962 + }, + { + "epoch": 8.276536312849162, + "grad_norm": 1.4577008485794067, + "learning_rate": 9.921787709497207e-06, + "loss": 0.3445, + "step": 2963 + }, + { + "epoch": 8.279329608938548, + "grad_norm": 2.337756872177124, + "learning_rate": 9.92513966480447e-06, + "loss": 0.3604, + "step": 2964 + }, + { + "epoch": 8.282122905027933, + "grad_norm": 0.6134743094444275, + "learning_rate": 9.928491620111733e-06, + "loss": 0.2251, + "step": 2965 + }, + { + "epoch": 8.28491620111732, + "grad_norm": 0.569677472114563, + "learning_rate": 9.931843575418995e-06, + "loss": 0.2302, + "step": 2966 + }, + { + "epoch": 8.287709497206704, + "grad_norm": 0.583300769329071, + "learning_rate": 9.935195530726258e-06, + "loss": 0.2398, + "step": 2967 + }, + { + "epoch": 8.29050279329609, + "grad_norm": 1.2195100784301758, + "learning_rate": 9.93854748603352e-06, + "loss": 0.2259, + "step": 2968 + }, + { + "epoch": 8.293296089385475, + "grad_norm": 1.247757911682129, + "learning_rate": 9.941899441340782e-06, + "loss": 0.2775, + "step": 2969 + }, + { + "epoch": 8.296089385474861, + "grad_norm": 0.662884533405304, + "learning_rate": 9.945251396648045e-06, + "loss": 0.2199, + "step": 2970 + }, + { + "epoch": 8.298882681564246, + "grad_norm": 0.8083630800247192, + "learning_rate": 9.948603351955308e-06, + "loss": 0.2153, + "step": 2971 + }, + { + "epoch": 8.30167597765363, + "grad_norm": 0.6161537170410156, + "learning_rate": 9.951955307262571e-06, + "loss": 0.2251, + "step": 2972 + }, + { + "epoch": 8.304469273743017, + "grad_norm": 0.5969156622886658, + "learning_rate": 9.955307262569833e-06, + "loss": 0.2042, + "step": 2973 + }, + { + "epoch": 8.307262569832401, + "grad_norm": 0.8740140199661255, + "learning_rate": 9.958659217877096e-06, + "loss": 0.2521, + "step": 2974 + }, + { + "epoch": 8.310055865921788, + "grad_norm": 0.9366194009780884, + "learning_rate": 9.962011173184359e-06, + "loss": 0.2295, + "step": 2975 + }, + { + "epoch": 8.312849162011172, + "grad_norm": 0.9440138339996338, + "learning_rate": 9.96536312849162e-06, + "loss": 0.2218, + "step": 2976 + }, + { + "epoch": 8.315642458100559, + "grad_norm": 0.9322623610496521, + "learning_rate": 9.968715083798883e-06, + "loss": 0.2032, + "step": 2977 + }, + { + "epoch": 8.318435754189943, + "grad_norm": 0.7885984182357788, + "learning_rate": 9.972067039106146e-06, + "loss": 0.226, + "step": 2978 + }, + { + "epoch": 8.32122905027933, + "grad_norm": 0.8539844751358032, + "learning_rate": 9.975418994413409e-06, + "loss": 0.2286, + "step": 2979 + }, + { + "epoch": 8.324022346368714, + "grad_norm": 2.408024787902832, + "learning_rate": 9.97877094972067e-06, + "loss": 0.2335, + "step": 2980 + }, + { + "epoch": 8.3268156424581, + "grad_norm": 1.1452076435089111, + "learning_rate": 9.982122905027934e-06, + "loss": 0.2238, + "step": 2981 + }, + { + "epoch": 8.329608938547485, + "grad_norm": 1.0498756170272827, + "learning_rate": 9.985474860335197e-06, + "loss": 0.231, + "step": 2982 + }, + { + "epoch": 8.332402234636872, + "grad_norm": 1.187779188156128, + "learning_rate": 9.988826815642458e-06, + "loss": 0.2619, + "step": 2983 + }, + { + "epoch": 8.335195530726256, + "grad_norm": 0.9812905192375183, + "learning_rate": 9.992178770949721e-06, + "loss": 0.247, + "step": 2984 + }, + { + "epoch": 8.337988826815643, + "grad_norm": 1.1358792781829834, + "learning_rate": 9.995530726256984e-06, + "loss": 0.2216, + "step": 2985 + }, + { + "epoch": 8.340782122905027, + "grad_norm": 1.6459420919418335, + "learning_rate": 9.998882681564245e-06, + "loss": 0.2541, + "step": 2986 + }, + { + "epoch": 8.343575418994414, + "grad_norm": 1.187185287475586, + "learning_rate": 1.0002234636871508e-05, + "loss": 0.2544, + "step": 2987 + }, + { + "epoch": 8.346368715083798, + "grad_norm": 1.4683363437652588, + "learning_rate": 1.0005586592178771e-05, + "loss": 0.2946, + "step": 2988 + }, + { + "epoch": 8.349162011173185, + "grad_norm": 2.1564249992370605, + "learning_rate": 1.0008938547486034e-05, + "loss": 0.3338, + "step": 2989 + }, + { + "epoch": 8.35195530726257, + "grad_norm": 1.3326091766357422, + "learning_rate": 1.0012290502793296e-05, + "loss": 0.2361, + "step": 2990 + }, + { + "epoch": 8.354748603351956, + "grad_norm": 0.6648940443992615, + "learning_rate": 1.0015642458100559e-05, + "loss": 0.2224, + "step": 2991 + }, + { + "epoch": 8.35754189944134, + "grad_norm": 0.6747652292251587, + "learning_rate": 1.0018994413407822e-05, + "loss": 0.2318, + "step": 2992 + }, + { + "epoch": 8.360335195530727, + "grad_norm": 1.0659197568893433, + "learning_rate": 1.0022346368715083e-05, + "loss": 0.2321, + "step": 2993 + }, + { + "epoch": 8.363128491620111, + "grad_norm": 1.0059983730316162, + "learning_rate": 1.0025698324022346e-05, + "loss": 0.1814, + "step": 2994 + }, + { + "epoch": 8.365921787709498, + "grad_norm": 0.6853020191192627, + "learning_rate": 1.002905027932961e-05, + "loss": 0.2287, + "step": 2995 + }, + { + "epoch": 8.368715083798882, + "grad_norm": 0.7175607085227966, + "learning_rate": 1.003240223463687e-05, + "loss": 0.2126, + "step": 2996 + }, + { + "epoch": 8.371508379888269, + "grad_norm": 0.8915181159973145, + "learning_rate": 1.0035754189944134e-05, + "loss": 0.2331, + "step": 2997 + }, + { + "epoch": 8.374301675977653, + "grad_norm": 0.8935127854347229, + "learning_rate": 1.0039106145251397e-05, + "loss": 0.2497, + "step": 2998 + }, + { + "epoch": 8.37709497206704, + "grad_norm": 0.7590783834457397, + "learning_rate": 1.004245810055866e-05, + "loss": 0.2333, + "step": 2999 + }, + { + "epoch": 8.379888268156424, + "grad_norm": 1.7611160278320312, + "learning_rate": 1.0045810055865921e-05, + "loss": 0.2638, + "step": 3000 + }, + { + "epoch": 8.379888268156424, + "eval_cer": 0.19592921126422486, + "eval_loss": 0.2892208397388458, + "eval_runtime": 23.6623, + "eval_samples_per_second": 191.782, + "eval_steps_per_second": 0.634, + "eval_wer": 0.35557665986983605, + "step": 3000 + }, + { + "epoch": 8.38268156424581, + "grad_norm": 0.78790283203125, + "learning_rate": 1.0049162011173184e-05, + "loss": 0.2062, + "step": 3001 + }, + { + "epoch": 8.385474860335195, + "grad_norm": 0.7174229025840759, + "learning_rate": 1.0052513966480447e-05, + "loss": 0.2531, + "step": 3002 + }, + { + "epoch": 8.388268156424582, + "grad_norm": 2.0383810997009277, + "learning_rate": 1.0055865921787709e-05, + "loss": 0.271, + "step": 3003 + }, + { + "epoch": 8.391061452513966, + "grad_norm": 1.006961703300476, + "learning_rate": 1.0059217877094972e-05, + "loss": 0.233, + "step": 3004 + }, + { + "epoch": 8.393854748603353, + "grad_norm": 1.1598045825958252, + "learning_rate": 1.0062569832402235e-05, + "loss": 0.2384, + "step": 3005 + }, + { + "epoch": 8.396648044692737, + "grad_norm": 1.1870278120040894, + "learning_rate": 1.0065921787709496e-05, + "loss": 0.2316, + "step": 3006 + }, + { + "epoch": 8.399441340782122, + "grad_norm": 1.0041054487228394, + "learning_rate": 1.0069273743016759e-05, + "loss": 0.2163, + "step": 3007 + }, + { + "epoch": 8.402234636871508, + "grad_norm": 1.7301526069641113, + "learning_rate": 1.0072625698324022e-05, + "loss": 0.2349, + "step": 3008 + }, + { + "epoch": 8.405027932960893, + "grad_norm": 1.197363257408142, + "learning_rate": 1.0075977653631285e-05, + "loss": 0.2681, + "step": 3009 + }, + { + "epoch": 8.40782122905028, + "grad_norm": 1.2902823686599731, + "learning_rate": 1.0079329608938546e-05, + "loss": 0.27, + "step": 3010 + }, + { + "epoch": 8.410614525139664, + "grad_norm": 1.5609897375106812, + "learning_rate": 1.0082681564245811e-05, + "loss": 0.2918, + "step": 3011 + }, + { + "epoch": 8.41340782122905, + "grad_norm": 10.606142044067383, + "learning_rate": 1.0086033519553074e-05, + "loss": 0.2739, + "step": 3012 + }, + { + "epoch": 8.416201117318435, + "grad_norm": 2.3816211223602295, + "learning_rate": 1.0089385474860336e-05, + "loss": 0.3426, + "step": 3013 + }, + { + "epoch": 8.418994413407821, + "grad_norm": 1.7594531774520874, + "learning_rate": 1.0092737430167599e-05, + "loss": 0.3631, + "step": 3014 + }, + { + "epoch": 8.421787709497206, + "grad_norm": 0.5547803640365601, + "learning_rate": 1.0096089385474862e-05, + "loss": 0.2353, + "step": 3015 + }, + { + "epoch": 8.424581005586592, + "grad_norm": 0.5972055196762085, + "learning_rate": 1.0099441340782123e-05, + "loss": 0.2389, + "step": 3016 + }, + { + "epoch": 8.427374301675977, + "grad_norm": 0.5709198713302612, + "learning_rate": 1.0102793296089386e-05, + "loss": 0.2211, + "step": 3017 + }, + { + "epoch": 8.430167597765363, + "grad_norm": 0.694983720779419, + "learning_rate": 1.0106145251396649e-05, + "loss": 0.2231, + "step": 3018 + }, + { + "epoch": 8.432960893854748, + "grad_norm": 0.657917320728302, + "learning_rate": 1.0109497206703912e-05, + "loss": 0.219, + "step": 3019 + }, + { + "epoch": 8.435754189944134, + "grad_norm": 1.4571789503097534, + "learning_rate": 1.0112849162011173e-05, + "loss": 0.2175, + "step": 3020 + }, + { + "epoch": 8.438547486033519, + "grad_norm": 0.8168210983276367, + "learning_rate": 1.0116201117318436e-05, + "loss": 0.2296, + "step": 3021 + }, + { + "epoch": 8.441340782122905, + "grad_norm": 0.8654956817626953, + "learning_rate": 1.01195530726257e-05, + "loss": 0.2263, + "step": 3022 + }, + { + "epoch": 8.44413407821229, + "grad_norm": 0.7017212510108948, + "learning_rate": 1.0122905027932961e-05, + "loss": 0.2291, + "step": 3023 + }, + { + "epoch": 8.446927374301676, + "grad_norm": 1.0029232501983643, + "learning_rate": 1.0126256983240224e-05, + "loss": 0.2378, + "step": 3024 + }, + { + "epoch": 8.449720670391061, + "grad_norm": 0.8020324110984802, + "learning_rate": 1.0129608938547487e-05, + "loss": 0.2281, + "step": 3025 + }, + { + "epoch": 8.452513966480447, + "grad_norm": 0.8308049440383911, + "learning_rate": 1.013296089385475e-05, + "loss": 0.2412, + "step": 3026 + }, + { + "epoch": 8.455307262569832, + "grad_norm": 0.9152435064315796, + "learning_rate": 1.0136312849162011e-05, + "loss": 0.2601, + "step": 3027 + }, + { + "epoch": 8.458100558659218, + "grad_norm": 0.8631394505500793, + "learning_rate": 1.0139664804469274e-05, + "loss": 0.2122, + "step": 3028 + }, + { + "epoch": 8.460893854748603, + "grad_norm": 0.8656184673309326, + "learning_rate": 1.0143016759776537e-05, + "loss": 0.2228, + "step": 3029 + }, + { + "epoch": 8.46368715083799, + "grad_norm": 0.9663165211677551, + "learning_rate": 1.0146368715083799e-05, + "loss": 0.2006, + "step": 3030 + }, + { + "epoch": 8.466480446927374, + "grad_norm": 1.6420279741287231, + "learning_rate": 1.0149720670391062e-05, + "loss": 0.2276, + "step": 3031 + }, + { + "epoch": 8.46927374301676, + "grad_norm": 1.2475947141647339, + "learning_rate": 1.0153072625698325e-05, + "loss": 0.2814, + "step": 3032 + }, + { + "epoch": 8.472067039106145, + "grad_norm": 0.8967413902282715, + "learning_rate": 1.0156424581005586e-05, + "loss": 0.2274, + "step": 3033 + }, + { + "epoch": 8.474860335195531, + "grad_norm": 1.0266467332839966, + "learning_rate": 1.015977653631285e-05, + "loss": 0.219, + "step": 3034 + }, + { + "epoch": 8.477653631284916, + "grad_norm": 1.3265153169631958, + "learning_rate": 1.0163128491620112e-05, + "loss": 0.2611, + "step": 3035 + }, + { + "epoch": 8.480446927374302, + "grad_norm": 1.096076250076294, + "learning_rate": 1.0166480446927375e-05, + "loss": 0.2382, + "step": 3036 + }, + { + "epoch": 8.483240223463687, + "grad_norm": 2.129645824432373, + "learning_rate": 1.0169832402234637e-05, + "loss": 0.2779, + "step": 3037 + }, + { + "epoch": 8.486033519553073, + "grad_norm": 1.3483701944351196, + "learning_rate": 1.01731843575419e-05, + "loss": 0.2957, + "step": 3038 + }, + { + "epoch": 8.488826815642458, + "grad_norm": 3.9154720306396484, + "learning_rate": 1.0176536312849163e-05, + "loss": 0.3313, + "step": 3039 + }, + { + "epoch": 8.491620111731844, + "grad_norm": 1.0941253900527954, + "learning_rate": 1.0179888268156424e-05, + "loss": 0.2208, + "step": 3040 + }, + { + "epoch": 8.494413407821229, + "grad_norm": 0.47279593348503113, + "learning_rate": 1.0183240223463687e-05, + "loss": 0.188, + "step": 3041 + }, + { + "epoch": 8.497206703910614, + "grad_norm": 0.6313536763191223, + "learning_rate": 1.018659217877095e-05, + "loss": 0.2278, + "step": 3042 + }, + { + "epoch": 8.5, + "grad_norm": 0.5363937616348267, + "learning_rate": 1.0189944134078211e-05, + "loss": 0.2084, + "step": 3043 + }, + { + "epoch": 8.502793296089386, + "grad_norm": 0.7972652912139893, + "learning_rate": 1.0193296089385475e-05, + "loss": 0.2725, + "step": 3044 + }, + { + "epoch": 8.505586592178771, + "grad_norm": 0.8354148864746094, + "learning_rate": 1.0196648044692738e-05, + "loss": 0.2777, + "step": 3045 + }, + { + "epoch": 8.508379888268156, + "grad_norm": 0.7342433929443359, + "learning_rate": 1.02e-05, + "loss": 0.2048, + "step": 3046 + }, + { + "epoch": 8.511173184357542, + "grad_norm": 1.1770108938217163, + "learning_rate": 1.0203351955307262e-05, + "loss": 0.2197, + "step": 3047 + }, + { + "epoch": 8.513966480446927, + "grad_norm": 0.6633297204971313, + "learning_rate": 1.0206703910614525e-05, + "loss": 0.2066, + "step": 3048 + }, + { + "epoch": 8.516759776536313, + "grad_norm": 0.762598991394043, + "learning_rate": 1.0210055865921788e-05, + "loss": 0.2176, + "step": 3049 + }, + { + "epoch": 8.519553072625698, + "grad_norm": 0.7805562019348145, + "learning_rate": 1.021340782122905e-05, + "loss": 0.2615, + "step": 3050 + }, + { + "epoch": 8.522346368715084, + "grad_norm": 0.811293363571167, + "learning_rate": 1.0216759776536312e-05, + "loss": 0.2208, + "step": 3051 + }, + { + "epoch": 8.525139664804469, + "grad_norm": 0.7932848930358887, + "learning_rate": 1.0220111731843575e-05, + "loss": 0.219, + "step": 3052 + }, + { + "epoch": 8.527932960893855, + "grad_norm": 0.8608736395835876, + "learning_rate": 1.0223463687150837e-05, + "loss": 0.2494, + "step": 3053 + }, + { + "epoch": 8.53072625698324, + "grad_norm": 0.9014523029327393, + "learning_rate": 1.02268156424581e-05, + "loss": 0.2572, + "step": 3054 + }, + { + "epoch": 8.533519553072626, + "grad_norm": 1.239258885383606, + "learning_rate": 1.0230167597765363e-05, + "loss": 0.2489, + "step": 3055 + }, + { + "epoch": 8.53631284916201, + "grad_norm": 1.1792540550231934, + "learning_rate": 1.0233519553072626e-05, + "loss": 0.255, + "step": 3056 + }, + { + "epoch": 8.539106145251397, + "grad_norm": 0.8312408328056335, + "learning_rate": 1.0236871508379887e-05, + "loss": 0.2385, + "step": 3057 + }, + { + "epoch": 8.541899441340782, + "grad_norm": 0.9757577180862427, + "learning_rate": 1.024022346368715e-05, + "loss": 0.2477, + "step": 3058 + }, + { + "epoch": 8.544692737430168, + "grad_norm": 1.2503435611724854, + "learning_rate": 1.0243575418994415e-05, + "loss": 0.2334, + "step": 3059 + }, + { + "epoch": 8.547486033519553, + "grad_norm": 1.0047942399978638, + "learning_rate": 1.0246927374301676e-05, + "loss": 0.2234, + "step": 3060 + }, + { + "epoch": 8.550279329608939, + "grad_norm": 2.173987627029419, + "learning_rate": 1.025027932960894e-05, + "loss": 0.2896, + "step": 3061 + }, + { + "epoch": 8.553072625698324, + "grad_norm": 2.975975513458252, + "learning_rate": 1.0253631284916202e-05, + "loss": 0.2709, + "step": 3062 + }, + { + "epoch": 8.55586592178771, + "grad_norm": 2.1861045360565186, + "learning_rate": 1.0256983240223464e-05, + "loss": 0.2809, + "step": 3063 + }, + { + "epoch": 8.558659217877095, + "grad_norm": 5.5423383712768555, + "learning_rate": 1.0260335195530727e-05, + "loss": 0.4278, + "step": 3064 + }, + { + "epoch": 8.561452513966481, + "grad_norm": 0.9353055357933044, + "learning_rate": 1.026368715083799e-05, + "loss": 0.2263, + "step": 3065 + }, + { + "epoch": 8.564245810055866, + "grad_norm": 0.616034746170044, + "learning_rate": 1.0267039106145253e-05, + "loss": 0.1953, + "step": 3066 + }, + { + "epoch": 8.567039106145252, + "grad_norm": 0.631631076335907, + "learning_rate": 1.0270391061452514e-05, + "loss": 0.2346, + "step": 3067 + }, + { + "epoch": 8.569832402234637, + "grad_norm": 0.5719462037086487, + "learning_rate": 1.0273743016759777e-05, + "loss": 0.1941, + "step": 3068 + }, + { + "epoch": 8.572625698324023, + "grad_norm": 0.6430604457855225, + "learning_rate": 1.027709497206704e-05, + "loss": 0.1996, + "step": 3069 + }, + { + "epoch": 8.575418994413408, + "grad_norm": 0.6971250176429749, + "learning_rate": 1.0280446927374302e-05, + "loss": 0.2323, + "step": 3070 + }, + { + "epoch": 8.578212290502794, + "grad_norm": 0.7100688219070435, + "learning_rate": 1.0283798882681565e-05, + "loss": 0.236, + "step": 3071 + }, + { + "epoch": 8.581005586592179, + "grad_norm": 0.681127667427063, + "learning_rate": 1.0287150837988828e-05, + "loss": 0.2215, + "step": 3072 + }, + { + "epoch": 8.583798882681565, + "grad_norm": 0.7432498335838318, + "learning_rate": 1.029050279329609e-05, + "loss": 0.2163, + "step": 3073 + }, + { + "epoch": 8.58659217877095, + "grad_norm": 0.688161313533783, + "learning_rate": 1.0293854748603352e-05, + "loss": 0.2339, + "step": 3074 + }, + { + "epoch": 8.589385474860336, + "grad_norm": 1.233516812324524, + "learning_rate": 1.0297206703910615e-05, + "loss": 0.2248, + "step": 3075 + }, + { + "epoch": 8.59217877094972, + "grad_norm": 2.4128191471099854, + "learning_rate": 1.0300558659217878e-05, + "loss": 0.2348, + "step": 3076 + }, + { + "epoch": 8.594972067039105, + "grad_norm": 1.2376893758773804, + "learning_rate": 1.030391061452514e-05, + "loss": 0.225, + "step": 3077 + }, + { + "epoch": 8.597765363128492, + "grad_norm": 1.1557990312576294, + "learning_rate": 1.0307262569832403e-05, + "loss": 0.2363, + "step": 3078 + }, + { + "epoch": 8.600558659217878, + "grad_norm": 1.1482207775115967, + "learning_rate": 1.0310614525139666e-05, + "loss": 0.242, + "step": 3079 + }, + { + "epoch": 8.603351955307263, + "grad_norm": 1.3974891901016235, + "learning_rate": 1.0313966480446927e-05, + "loss": 0.2588, + "step": 3080 + }, + { + "epoch": 8.606145251396647, + "grad_norm": 1.2107419967651367, + "learning_rate": 1.031731843575419e-05, + "loss": 0.2236, + "step": 3081 + }, + { + "epoch": 8.608938547486034, + "grad_norm": 1.5000370740890503, + "learning_rate": 1.0320670391061453e-05, + "loss": 0.2701, + "step": 3082 + }, + { + "epoch": 8.611731843575418, + "grad_norm": 1.1805921792984009, + "learning_rate": 1.0324022346368716e-05, + "loss": 0.2417, + "step": 3083 + }, + { + "epoch": 8.614525139664805, + "grad_norm": 7.598458290100098, + "learning_rate": 1.0327374301675977e-05, + "loss": 0.2543, + "step": 3084 + }, + { + "epoch": 8.61731843575419, + "grad_norm": 0.996596097946167, + "learning_rate": 1.033072625698324e-05, + "loss": 0.2684, + "step": 3085 + }, + { + "epoch": 8.620111731843576, + "grad_norm": 1.2928745746612549, + "learning_rate": 1.0334078212290504e-05, + "loss": 0.2843, + "step": 3086 + }, + { + "epoch": 8.62290502793296, + "grad_norm": 1.9579135179519653, + "learning_rate": 1.0337430167597765e-05, + "loss": 0.2901, + "step": 3087 + }, + { + "epoch": 8.625698324022347, + "grad_norm": 1.9157747030258179, + "learning_rate": 1.0340782122905028e-05, + "loss": 0.3298, + "step": 3088 + }, + { + "epoch": 8.628491620111731, + "grad_norm": 2.333484649658203, + "learning_rate": 1.0344134078212291e-05, + "loss": 0.41, + "step": 3089 + }, + { + "epoch": 8.631284916201118, + "grad_norm": 0.8331677317619324, + "learning_rate": 1.0347486033519552e-05, + "loss": 0.2256, + "step": 3090 + }, + { + "epoch": 8.634078212290502, + "grad_norm": 0.5986695885658264, + "learning_rate": 1.0350837988826815e-05, + "loss": 0.2227, + "step": 3091 + }, + { + "epoch": 8.636871508379889, + "grad_norm": 0.6864441633224487, + "learning_rate": 1.0354189944134078e-05, + "loss": 0.2081, + "step": 3092 + }, + { + "epoch": 8.639664804469273, + "grad_norm": 0.7613930702209473, + "learning_rate": 1.0357541899441341e-05, + "loss": 0.2074, + "step": 3093 + }, + { + "epoch": 8.64245810055866, + "grad_norm": 0.8596687316894531, + "learning_rate": 1.0360893854748603e-05, + "loss": 0.2911, + "step": 3094 + }, + { + "epoch": 8.645251396648044, + "grad_norm": 0.5623555779457092, + "learning_rate": 1.0364245810055866e-05, + "loss": 0.2275, + "step": 3095 + }, + { + "epoch": 8.64804469273743, + "grad_norm": 0.6859112977981567, + "learning_rate": 1.0367597765363129e-05, + "loss": 0.2199, + "step": 3096 + }, + { + "epoch": 8.650837988826815, + "grad_norm": 0.7403121590614319, + "learning_rate": 1.037094972067039e-05, + "loss": 0.289, + "step": 3097 + }, + { + "epoch": 8.653631284916202, + "grad_norm": 0.9017990231513977, + "learning_rate": 1.0374301675977653e-05, + "loss": 0.1831, + "step": 3098 + }, + { + "epoch": 8.656424581005586, + "grad_norm": 0.8094789385795593, + "learning_rate": 1.0377653631284916e-05, + "loss": 0.2149, + "step": 3099 + }, + { + "epoch": 8.659217877094973, + "grad_norm": 0.7958694696426392, + "learning_rate": 1.0381005586592178e-05, + "loss": 0.26, + "step": 3100 + }, + { + "epoch": 8.662011173184357, + "grad_norm": 1.0104540586471558, + "learning_rate": 1.038435754189944e-05, + "loss": 0.2222, + "step": 3101 + }, + { + "epoch": 8.664804469273744, + "grad_norm": 1.1946630477905273, + "learning_rate": 1.0387709497206704e-05, + "loss": 0.2071, + "step": 3102 + }, + { + "epoch": 8.667597765363128, + "grad_norm": 2.123600721359253, + "learning_rate": 1.0391061452513967e-05, + "loss": 0.2406, + "step": 3103 + }, + { + "epoch": 8.670391061452515, + "grad_norm": 1.0413761138916016, + "learning_rate": 1.0394413407821228e-05, + "loss": 0.2183, + "step": 3104 + }, + { + "epoch": 8.6731843575419, + "grad_norm": 1.1026817560195923, + "learning_rate": 1.0397765363128491e-05, + "loss": 0.2703, + "step": 3105 + }, + { + "epoch": 8.675977653631286, + "grad_norm": 1.1058155298233032, + "learning_rate": 1.0401117318435754e-05, + "loss": 0.2169, + "step": 3106 + }, + { + "epoch": 8.67877094972067, + "grad_norm": 2.270735502243042, + "learning_rate": 1.0404469273743017e-05, + "loss": 0.1993, + "step": 3107 + }, + { + "epoch": 8.681564245810057, + "grad_norm": 1.6135462522506714, + "learning_rate": 1.040782122905028e-05, + "loss": 0.2687, + "step": 3108 + }, + { + "epoch": 8.684357541899441, + "grad_norm": 1.187232255935669, + "learning_rate": 1.0411173184357543e-05, + "loss": 0.257, + "step": 3109 + }, + { + "epoch": 8.687150837988828, + "grad_norm": 2.211505889892578, + "learning_rate": 1.0414525139664805e-05, + "loss": 0.2531, + "step": 3110 + }, + { + "epoch": 8.689944134078212, + "grad_norm": 2.355846643447876, + "learning_rate": 1.0417877094972068e-05, + "loss": 0.2479, + "step": 3111 + }, + { + "epoch": 8.692737430167599, + "grad_norm": 1.9688565731048584, + "learning_rate": 1.042122905027933e-05, + "loss": 0.2568, + "step": 3112 + }, + { + "epoch": 8.695530726256983, + "grad_norm": 1.548080325126648, + "learning_rate": 1.0424581005586594e-05, + "loss": 0.2787, + "step": 3113 + }, + { + "epoch": 8.69832402234637, + "grad_norm": 2.0901577472686768, + "learning_rate": 1.0427932960893855e-05, + "loss": 0.3899, + "step": 3114 + }, + { + "epoch": 8.701117318435754, + "grad_norm": 0.7123032808303833, + "learning_rate": 1.0431284916201118e-05, + "loss": 0.2248, + "step": 3115 + }, + { + "epoch": 8.703910614525139, + "grad_norm": 0.8311495780944824, + "learning_rate": 1.0434636871508381e-05, + "loss": 0.2632, + "step": 3116 + }, + { + "epoch": 8.706703910614525, + "grad_norm": 0.7514416575431824, + "learning_rate": 1.0437988826815643e-05, + "loss": 0.2229, + "step": 3117 + }, + { + "epoch": 8.70949720670391, + "grad_norm": 1.4017688035964966, + "learning_rate": 1.0441340782122906e-05, + "loss": 0.2258, + "step": 3118 + }, + { + "epoch": 8.712290502793296, + "grad_norm": 0.937626302242279, + "learning_rate": 1.0444692737430169e-05, + "loss": 0.2384, + "step": 3119 + }, + { + "epoch": 8.71508379888268, + "grad_norm": 1.36872398853302, + "learning_rate": 1.0448044692737432e-05, + "loss": 0.2434, + "step": 3120 + }, + { + "epoch": 8.717877094972067, + "grad_norm": 0.660824179649353, + "learning_rate": 1.0451396648044693e-05, + "loss": 0.2358, + "step": 3121 + }, + { + "epoch": 8.720670391061452, + "grad_norm": 0.9180675745010376, + "learning_rate": 1.0454748603351956e-05, + "loss": 0.2418, + "step": 3122 + }, + { + "epoch": 8.723463687150838, + "grad_norm": 0.7859247922897339, + "learning_rate": 1.0458100558659219e-05, + "loss": 0.1991, + "step": 3123 + }, + { + "epoch": 8.726256983240223, + "grad_norm": 0.9615757465362549, + "learning_rate": 1.046145251396648e-05, + "loss": 0.234, + "step": 3124 + }, + { + "epoch": 8.72905027932961, + "grad_norm": 0.8265930414199829, + "learning_rate": 1.0464804469273743e-05, + "loss": 0.2989, + "step": 3125 + }, + { + "epoch": 8.731843575418994, + "grad_norm": 1.1574671268463135, + "learning_rate": 1.0468156424581006e-05, + "loss": 0.2471, + "step": 3126 + }, + { + "epoch": 8.73463687150838, + "grad_norm": 0.8868136405944824, + "learning_rate": 1.0471508379888268e-05, + "loss": 0.1905, + "step": 3127 + }, + { + "epoch": 8.737430167597765, + "grad_norm": 2.0433335304260254, + "learning_rate": 1.0474860335195531e-05, + "loss": 0.2045, + "step": 3128 + }, + { + "epoch": 8.740223463687151, + "grad_norm": 1.0071204900741577, + "learning_rate": 1.0478212290502794e-05, + "loss": 0.2166, + "step": 3129 + }, + { + "epoch": 8.743016759776536, + "grad_norm": 0.8606061339378357, + "learning_rate": 1.0481564245810057e-05, + "loss": 0.2128, + "step": 3130 + }, + { + "epoch": 8.745810055865922, + "grad_norm": 1.6690952777862549, + "learning_rate": 1.0484916201117318e-05, + "loss": 0.263, + "step": 3131 + }, + { + "epoch": 8.748603351955307, + "grad_norm": 1.559032917022705, + "learning_rate": 1.0488268156424581e-05, + "loss": 0.2032, + "step": 3132 + }, + { + "epoch": 8.751396648044693, + "grad_norm": 0.8347238898277283, + "learning_rate": 1.0491620111731844e-05, + "loss": 0.2241, + "step": 3133 + }, + { + "epoch": 8.754189944134078, + "grad_norm": 1.341846227645874, + "learning_rate": 1.0494972067039106e-05, + "loss": 0.2664, + "step": 3134 + }, + { + "epoch": 8.756983240223464, + "grad_norm": 1.6938453912734985, + "learning_rate": 1.0498324022346369e-05, + "loss": 0.262, + "step": 3135 + }, + { + "epoch": 8.759776536312849, + "grad_norm": 1.6497092247009277, + "learning_rate": 1.0501675977653632e-05, + "loss": 0.2709, + "step": 3136 + }, + { + "epoch": 8.762569832402235, + "grad_norm": 1.4694902896881104, + "learning_rate": 1.0505027932960893e-05, + "loss": 0.2463, + "step": 3137 + }, + { + "epoch": 8.76536312849162, + "grad_norm": 2.140068769454956, + "learning_rate": 1.0508379888268156e-05, + "loss": 0.285, + "step": 3138 + }, + { + "epoch": 8.768156424581006, + "grad_norm": 3.6979525089263916, + "learning_rate": 1.051173184357542e-05, + "loss": 0.4812, + "step": 3139 + }, + { + "epoch": 8.77094972067039, + "grad_norm": 0.6257543563842773, + "learning_rate": 1.0515083798882682e-05, + "loss": 0.239, + "step": 3140 + }, + { + "epoch": 8.773743016759777, + "grad_norm": 5.637814044952393, + "learning_rate": 1.0518435754189944e-05, + "loss": 0.2115, + "step": 3141 + }, + { + "epoch": 8.776536312849162, + "grad_norm": 0.5621907711029053, + "learning_rate": 1.0521787709497207e-05, + "loss": 0.2006, + "step": 3142 + }, + { + "epoch": 8.779329608938548, + "grad_norm": 0.7554235458374023, + "learning_rate": 1.052513966480447e-05, + "loss": 0.2232, + "step": 3143 + }, + { + "epoch": 8.782122905027933, + "grad_norm": 0.5796555280685425, + "learning_rate": 1.0528491620111731e-05, + "loss": 0.2319, + "step": 3144 + }, + { + "epoch": 8.78491620111732, + "grad_norm": 0.6232753992080688, + "learning_rate": 1.0531843575418994e-05, + "loss": 0.2081, + "step": 3145 + }, + { + "epoch": 8.787709497206704, + "grad_norm": 0.7719864845275879, + "learning_rate": 1.0535195530726257e-05, + "loss": 0.2796, + "step": 3146 + }, + { + "epoch": 8.79050279329609, + "grad_norm": 0.9676380157470703, + "learning_rate": 1.0538547486033518e-05, + "loss": 0.2239, + "step": 3147 + }, + { + "epoch": 8.793296089385475, + "grad_norm": 0.6195005774497986, + "learning_rate": 1.0541899441340781e-05, + "loss": 0.2235, + "step": 3148 + }, + { + "epoch": 8.796089385474861, + "grad_norm": 0.9336199164390564, + "learning_rate": 1.0545251396648045e-05, + "loss": 0.243, + "step": 3149 + }, + { + "epoch": 8.798882681564246, + "grad_norm": 0.9865847826004028, + "learning_rate": 1.0548603351955308e-05, + "loss": 0.2186, + "step": 3150 + }, + { + "epoch": 8.80167597765363, + "grad_norm": 0.9390310049057007, + "learning_rate": 1.0551955307262569e-05, + "loss": 0.2474, + "step": 3151 + }, + { + "epoch": 8.804469273743017, + "grad_norm": 1.1407688856124878, + "learning_rate": 1.0555307262569832e-05, + "loss": 0.2064, + "step": 3152 + }, + { + "epoch": 8.807262569832401, + "grad_norm": 0.6871207356452942, + "learning_rate": 1.0558659217877095e-05, + "loss": 0.2022, + "step": 3153 + }, + { + "epoch": 8.810055865921788, + "grad_norm": 1.2367340326309204, + "learning_rate": 1.0562011173184356e-05, + "loss": 0.2357, + "step": 3154 + }, + { + "epoch": 8.812849162011172, + "grad_norm": 1.144147276878357, + "learning_rate": 1.0565363128491621e-05, + "loss": 0.2135, + "step": 3155 + }, + { + "epoch": 8.815642458100559, + "grad_norm": 0.9026341438293457, + "learning_rate": 1.0568715083798884e-05, + "loss": 0.2388, + "step": 3156 + }, + { + "epoch": 8.818435754189943, + "grad_norm": 1.0823280811309814, + "learning_rate": 1.0572067039106145e-05, + "loss": 0.2038, + "step": 3157 + }, + { + "epoch": 8.82122905027933, + "grad_norm": 1.0494595766067505, + "learning_rate": 1.0575418994413408e-05, + "loss": 0.2413, + "step": 3158 + }, + { + "epoch": 8.824022346368714, + "grad_norm": 1.1499204635620117, + "learning_rate": 1.0578770949720672e-05, + "loss": 0.2528, + "step": 3159 + }, + { + "epoch": 8.8268156424581, + "grad_norm": 1.297745704650879, + "learning_rate": 1.0582122905027935e-05, + "loss": 0.2094, + "step": 3160 + }, + { + "epoch": 8.829608938547485, + "grad_norm": 1.0844340324401855, + "learning_rate": 1.0585474860335196e-05, + "loss": 0.2306, + "step": 3161 + }, + { + "epoch": 8.832402234636872, + "grad_norm": 2.135915994644165, + "learning_rate": 1.0588826815642459e-05, + "loss": 0.26, + "step": 3162 + }, + { + "epoch": 8.835195530726256, + "grad_norm": 1.2685796022415161, + "learning_rate": 1.0592178770949722e-05, + "loss": 0.2414, + "step": 3163 + }, + { + "epoch": 8.837988826815643, + "grad_norm": 2.0272021293640137, + "learning_rate": 1.0595530726256983e-05, + "loss": 0.394, + "step": 3164 + }, + { + "epoch": 8.840782122905027, + "grad_norm": 0.5999031662940979, + "learning_rate": 1.0598882681564246e-05, + "loss": 0.2147, + "step": 3165 + }, + { + "epoch": 8.843575418994414, + "grad_norm": 0.5733051896095276, + "learning_rate": 1.060223463687151e-05, + "loss": 0.2196, + "step": 3166 + }, + { + "epoch": 8.846368715083798, + "grad_norm": 0.5257775187492371, + "learning_rate": 1.0605586592178772e-05, + "loss": 0.2137, + "step": 3167 + }, + { + "epoch": 8.849162011173185, + "grad_norm": 0.6967881917953491, + "learning_rate": 1.0608938547486034e-05, + "loss": 0.203, + "step": 3168 + }, + { + "epoch": 8.85195530726257, + "grad_norm": 0.7626516819000244, + "learning_rate": 1.0612290502793297e-05, + "loss": 0.2522, + "step": 3169 + }, + { + "epoch": 8.854748603351956, + "grad_norm": 1.240830659866333, + "learning_rate": 1.061564245810056e-05, + "loss": 0.2104, + "step": 3170 + }, + { + "epoch": 8.85754189944134, + "grad_norm": 1.5441384315490723, + "learning_rate": 1.0618994413407821e-05, + "loss": 0.2226, + "step": 3171 + }, + { + "epoch": 8.860335195530727, + "grad_norm": 0.5794720649719238, + "learning_rate": 1.0622346368715084e-05, + "loss": 0.2022, + "step": 3172 + }, + { + "epoch": 8.863128491620111, + "grad_norm": 1.1239813566207886, + "learning_rate": 1.0625698324022347e-05, + "loss": 0.2228, + "step": 3173 + }, + { + "epoch": 8.865921787709498, + "grad_norm": 0.8274424076080322, + "learning_rate": 1.0629050279329609e-05, + "loss": 0.2446, + "step": 3174 + }, + { + "epoch": 8.868715083798882, + "grad_norm": 0.9040669202804565, + "learning_rate": 1.0632402234636872e-05, + "loss": 0.2504, + "step": 3175 + }, + { + "epoch": 8.871508379888269, + "grad_norm": 0.7560740113258362, + "learning_rate": 1.0635754189944135e-05, + "loss": 0.2201, + "step": 3176 + }, + { + "epoch": 8.874301675977653, + "grad_norm": 0.5757977366447449, + "learning_rate": 1.0639106145251398e-05, + "loss": 0.2087, + "step": 3177 + }, + { + "epoch": 8.87709497206704, + "grad_norm": 1.128888487815857, + "learning_rate": 1.0642458100558659e-05, + "loss": 0.2128, + "step": 3178 + }, + { + "epoch": 8.879888268156424, + "grad_norm": 1.0056257247924805, + "learning_rate": 1.0645810055865922e-05, + "loss": 0.2382, + "step": 3179 + }, + { + "epoch": 8.88268156424581, + "grad_norm": 1.0520869493484497, + "learning_rate": 1.0649162011173185e-05, + "loss": 0.2558, + "step": 3180 + }, + { + "epoch": 8.885474860335195, + "grad_norm": 0.780989944934845, + "learning_rate": 1.0652513966480447e-05, + "loss": 0.2089, + "step": 3181 + }, + { + "epoch": 8.888268156424582, + "grad_norm": 1.1164745092391968, + "learning_rate": 1.065586592178771e-05, + "loss": 0.2191, + "step": 3182 + }, + { + "epoch": 8.891061452513966, + "grad_norm": 1.6737505197525024, + "learning_rate": 1.0659217877094973e-05, + "loss": 0.2499, + "step": 3183 + }, + { + "epoch": 8.893854748603353, + "grad_norm": 0.978674590587616, + "learning_rate": 1.0662569832402234e-05, + "loss": 0.2134, + "step": 3184 + }, + { + "epoch": 8.896648044692737, + "grad_norm": 1.1877577304840088, + "learning_rate": 1.0665921787709497e-05, + "loss": 0.2755, + "step": 3185 + }, + { + "epoch": 8.899441340782122, + "grad_norm": 4.8825907707214355, + "learning_rate": 1.066927374301676e-05, + "loss": 0.2235, + "step": 3186 + }, + { + "epoch": 8.902234636871508, + "grad_norm": 2.1930627822875977, + "learning_rate": 1.0672625698324023e-05, + "loss": 0.2529, + "step": 3187 + }, + { + "epoch": 8.905027932960895, + "grad_norm": 2.0230579376220703, + "learning_rate": 1.0675977653631284e-05, + "loss": 0.3929, + "step": 3188 + }, + { + "epoch": 8.90782122905028, + "grad_norm": 1.7383211851119995, + "learning_rate": 1.0679329608938547e-05, + "loss": 0.281, + "step": 3189 + }, + { + "epoch": 8.910614525139664, + "grad_norm": 0.5710355043411255, + "learning_rate": 1.068268156424581e-05, + "loss": 0.2385, + "step": 3190 + }, + { + "epoch": 8.91340782122905, + "grad_norm": 0.6028423309326172, + "learning_rate": 1.0686033519553072e-05, + "loss": 0.2253, + "step": 3191 + }, + { + "epoch": 8.916201117318435, + "grad_norm": 0.5164093971252441, + "learning_rate": 1.0689385474860335e-05, + "loss": 0.2197, + "step": 3192 + }, + { + "epoch": 8.918994413407821, + "grad_norm": 0.8699396848678589, + "learning_rate": 1.0692737430167598e-05, + "loss": 0.2497, + "step": 3193 + }, + { + "epoch": 8.921787709497206, + "grad_norm": 0.6903799772262573, + "learning_rate": 1.069608938547486e-05, + "loss": 0.2715, + "step": 3194 + }, + { + "epoch": 8.924581005586592, + "grad_norm": 0.5981618165969849, + "learning_rate": 1.0699441340782122e-05, + "loss": 0.2077, + "step": 3195 + }, + { + "epoch": 8.927374301675977, + "grad_norm": 1.1702784299850464, + "learning_rate": 1.0702793296089385e-05, + "loss": 0.2348, + "step": 3196 + }, + { + "epoch": 8.930167597765363, + "grad_norm": 0.9630923867225647, + "learning_rate": 1.0706145251396648e-05, + "loss": 0.2478, + "step": 3197 + }, + { + "epoch": 8.932960893854748, + "grad_norm": 0.7544631958007812, + "learning_rate": 1.070949720670391e-05, + "loss": 0.2318, + "step": 3198 + }, + { + "epoch": 8.935754189944134, + "grad_norm": 0.9658316969871521, + "learning_rate": 1.0712849162011173e-05, + "loss": 0.2372, + "step": 3199 + }, + { + "epoch": 8.938547486033519, + "grad_norm": 0.8392074108123779, + "learning_rate": 1.0716201117318436e-05, + "loss": 0.2197, + "step": 3200 + }, + { + "epoch": 8.941340782122905, + "grad_norm": 0.6790133714675903, + "learning_rate": 1.0719553072625697e-05, + "loss": 0.2143, + "step": 3201 + }, + { + "epoch": 8.94413407821229, + "grad_norm": 1.1582915782928467, + "learning_rate": 1.072290502793296e-05, + "loss": 0.2168, + "step": 3202 + }, + { + "epoch": 8.946927374301676, + "grad_norm": 0.9098035097122192, + "learning_rate": 1.0726256983240223e-05, + "loss": 0.2855, + "step": 3203 + }, + { + "epoch": 8.949720670391061, + "grad_norm": 0.6689935326576233, + "learning_rate": 1.0729608938547486e-05, + "loss": 0.2104, + "step": 3204 + }, + { + "epoch": 8.952513966480447, + "grad_norm": 1.0739282369613647, + "learning_rate": 1.073296089385475e-05, + "loss": 0.2373, + "step": 3205 + }, + { + "epoch": 8.955307262569832, + "grad_norm": 1.0035419464111328, + "learning_rate": 1.0736312849162012e-05, + "loss": 0.2369, + "step": 3206 + }, + { + "epoch": 8.958100558659218, + "grad_norm": 1.4360828399658203, + "learning_rate": 1.0739664804469275e-05, + "loss": 0.2369, + "step": 3207 + }, + { + "epoch": 8.960893854748603, + "grad_norm": 3.1000232696533203, + "learning_rate": 1.0743016759776537e-05, + "loss": 0.225, + "step": 3208 + }, + { + "epoch": 8.96368715083799, + "grad_norm": 1.1783204078674316, + "learning_rate": 1.07463687150838e-05, + "loss": 0.2521, + "step": 3209 + }, + { + "epoch": 8.966480446927374, + "grad_norm": 1.269181251525879, + "learning_rate": 1.0749720670391063e-05, + "loss": 0.2597, + "step": 3210 + }, + { + "epoch": 8.96927374301676, + "grad_norm": 5.353310585021973, + "learning_rate": 1.0753072625698324e-05, + "loss": 0.2533, + "step": 3211 + }, + { + "epoch": 8.972067039106145, + "grad_norm": 1.8082395792007446, + "learning_rate": 1.0756424581005587e-05, + "loss": 0.2453, + "step": 3212 + }, + { + "epoch": 8.974860335195531, + "grad_norm": 1.6220979690551758, + "learning_rate": 1.075977653631285e-05, + "loss": 0.245, + "step": 3213 + }, + { + "epoch": 8.977653631284916, + "grad_norm": 6.265468597412109, + "learning_rate": 1.0763128491620113e-05, + "loss": 0.3459, + "step": 3214 + }, + { + "epoch": 8.980446927374302, + "grad_norm": 0.5995340943336487, + "learning_rate": 1.0766480446927375e-05, + "loss": 0.2351, + "step": 3215 + }, + { + "epoch": 8.983240223463687, + "grad_norm": 0.6484401226043701, + "learning_rate": 1.0769832402234638e-05, + "loss": 0.2172, + "step": 3216 + }, + { + "epoch": 8.986033519553073, + "grad_norm": 1.1867040395736694, + "learning_rate": 1.07731843575419e-05, + "loss": 0.235, + "step": 3217 + }, + { + "epoch": 8.988826815642458, + "grad_norm": 1.4771134853363037, + "learning_rate": 1.0776536312849162e-05, + "loss": 0.224, + "step": 3218 + }, + { + "epoch": 8.991620111731844, + "grad_norm": 0.8901110291481018, + "learning_rate": 1.0779888268156425e-05, + "loss": 0.2432, + "step": 3219 + }, + { + "epoch": 8.994413407821229, + "grad_norm": 1.2258399724960327, + "learning_rate": 1.0783240223463688e-05, + "loss": 0.2517, + "step": 3220 + }, + { + "epoch": 8.997206703910614, + "grad_norm": 1.177013874053955, + "learning_rate": 1.078659217877095e-05, + "loss": 0.2455, + "step": 3221 + }, + { + "epoch": 9.0, + "grad_norm": 2.559337615966797, + "learning_rate": 1.0789944134078213e-05, + "loss": 0.3027, + "step": 3222 + }, + { + "epoch": 9.002793296089385, + "grad_norm": 0.837273120880127, + "learning_rate": 1.0793296089385476e-05, + "loss": 0.274, + "step": 3223 + }, + { + "epoch": 9.005586592178771, + "grad_norm": 0.522996187210083, + "learning_rate": 1.0796648044692739e-05, + "loss": 0.2121, + "step": 3224 + }, + { + "epoch": 9.008379888268156, + "grad_norm": 0.9572185277938843, + "learning_rate": 1.08e-05, + "loss": 0.1793, + "step": 3225 + }, + { + "epoch": 9.011173184357542, + "grad_norm": 1.5862764120101929, + "learning_rate": 1.0803351955307263e-05, + "loss": 0.222, + "step": 3226 + }, + { + "epoch": 9.013966480446927, + "grad_norm": 0.7856401801109314, + "learning_rate": 1.0806703910614526e-05, + "loss": 0.224, + "step": 3227 + }, + { + "epoch": 9.016759776536313, + "grad_norm": 0.9093623757362366, + "learning_rate": 1.0810055865921787e-05, + "loss": 0.2043, + "step": 3228 + }, + { + "epoch": 9.019553072625698, + "grad_norm": 0.7950745820999146, + "learning_rate": 1.081340782122905e-05, + "loss": 0.2198, + "step": 3229 + }, + { + "epoch": 9.022346368715084, + "grad_norm": 0.7026410102844238, + "learning_rate": 1.0816759776536313e-05, + "loss": 0.2192, + "step": 3230 + }, + { + "epoch": 9.025139664804469, + "grad_norm": 0.8467795848846436, + "learning_rate": 1.0820111731843575e-05, + "loss": 0.2051, + "step": 3231 + }, + { + "epoch": 9.027932960893855, + "grad_norm": 0.7289355397224426, + "learning_rate": 1.0823463687150838e-05, + "loss": 0.1986, + "step": 3232 + }, + { + "epoch": 9.03072625698324, + "grad_norm": 0.6514780521392822, + "learning_rate": 1.0826815642458101e-05, + "loss": 0.1948, + "step": 3233 + }, + { + "epoch": 9.033519553072626, + "grad_norm": 0.7887862920761108, + "learning_rate": 1.0830167597765364e-05, + "loss": 0.2409, + "step": 3234 + }, + { + "epoch": 9.03631284916201, + "grad_norm": 0.7589343190193176, + "learning_rate": 1.0833519553072625e-05, + "loss": 0.2129, + "step": 3235 + }, + { + "epoch": 9.039106145251397, + "grad_norm": 1.5128223896026611, + "learning_rate": 1.0836871508379888e-05, + "loss": 0.2199, + "step": 3236 + }, + { + "epoch": 9.041899441340782, + "grad_norm": 1.6322835683822632, + "learning_rate": 1.0840223463687151e-05, + "loss": 0.2277, + "step": 3237 + }, + { + "epoch": 9.044692737430168, + "grad_norm": 1.1758242845535278, + "learning_rate": 1.0843575418994413e-05, + "loss": 0.2024, + "step": 3238 + }, + { + "epoch": 9.047486033519553, + "grad_norm": 0.8910885453224182, + "learning_rate": 1.0846927374301676e-05, + "loss": 0.2486, + "step": 3239 + }, + { + "epoch": 9.050279329608939, + "grad_norm": 5.764599323272705, + "learning_rate": 1.0850279329608939e-05, + "loss": 0.1914, + "step": 3240 + }, + { + "epoch": 9.053072625698324, + "grad_norm": 0.8394498229026794, + "learning_rate": 1.08536312849162e-05, + "loss": 0.2199, + "step": 3241 + }, + { + "epoch": 9.05586592178771, + "grad_norm": 1.591081976890564, + "learning_rate": 1.0856983240223463e-05, + "loss": 0.2808, + "step": 3242 + }, + { + "epoch": 9.058659217877095, + "grad_norm": 2.6334924697875977, + "learning_rate": 1.0860335195530726e-05, + "loss": 0.2228, + "step": 3243 + }, + { + "epoch": 9.061452513966481, + "grad_norm": 1.4277338981628418, + "learning_rate": 1.086368715083799e-05, + "loss": 0.2245, + "step": 3244 + }, + { + "epoch": 9.064245810055866, + "grad_norm": 2.022764205932617, + "learning_rate": 1.086703910614525e-05, + "loss": 0.2512, + "step": 3245 + }, + { + "epoch": 9.067039106145252, + "grad_norm": 1.572561502456665, + "learning_rate": 1.0870391061452514e-05, + "loss": 0.2235, + "step": 3246 + }, + { + "epoch": 9.069832402234637, + "grad_norm": 21.594219207763672, + "learning_rate": 1.0873743016759777e-05, + "loss": 0.3314, + "step": 3247 + }, + { + "epoch": 9.072625698324023, + "grad_norm": 0.7046812176704407, + "learning_rate": 1.0877094972067038e-05, + "loss": 0.2022, + "step": 3248 + }, + { + "epoch": 9.075418994413408, + "grad_norm": 0.7015390396118164, + "learning_rate": 1.0880446927374301e-05, + "loss": 0.2158, + "step": 3249 + }, + { + "epoch": 9.078212290502794, + "grad_norm": 0.603828489780426, + "learning_rate": 1.0883798882681564e-05, + "loss": 0.2211, + "step": 3250 + }, + { + "epoch": 9.081005586592179, + "grad_norm": 0.8990091681480408, + "learning_rate": 1.0887150837988827e-05, + "loss": 0.2034, + "step": 3251 + }, + { + "epoch": 9.083798882681565, + "grad_norm": 0.6733452677726746, + "learning_rate": 1.089050279329609e-05, + "loss": 0.2287, + "step": 3252 + }, + { + "epoch": 9.08659217877095, + "grad_norm": 2.62276291847229, + "learning_rate": 1.0893854748603353e-05, + "loss": 0.2227, + "step": 3253 + }, + { + "epoch": 9.089385474860336, + "grad_norm": 0.6666644215583801, + "learning_rate": 1.0897206703910616e-05, + "loss": 0.1896, + "step": 3254 + }, + { + "epoch": 9.09217877094972, + "grad_norm": 0.9197602272033691, + "learning_rate": 1.0900558659217878e-05, + "loss": 0.2069, + "step": 3255 + }, + { + "epoch": 9.094972067039107, + "grad_norm": 0.5614663362503052, + "learning_rate": 1.090391061452514e-05, + "loss": 0.2217, + "step": 3256 + }, + { + "epoch": 9.097765363128492, + "grad_norm": 0.8259888887405396, + "learning_rate": 1.0907262569832404e-05, + "loss": 0.2365, + "step": 3257 + }, + { + "epoch": 9.100558659217878, + "grad_norm": 0.5065736174583435, + "learning_rate": 1.0910614525139665e-05, + "loss": 0.1713, + "step": 3258 + }, + { + "epoch": 9.103351955307263, + "grad_norm": 0.9104751944541931, + "learning_rate": 1.0913966480446928e-05, + "loss": 0.2253, + "step": 3259 + }, + { + "epoch": 9.106145251396647, + "grad_norm": 1.0741955041885376, + "learning_rate": 1.0917318435754191e-05, + "loss": 0.1834, + "step": 3260 + }, + { + "epoch": 9.108938547486034, + "grad_norm": 0.6749626398086548, + "learning_rate": 1.0920670391061454e-05, + "loss": 0.1978, + "step": 3261 + }, + { + "epoch": 9.111731843575418, + "grad_norm": 1.1471261978149414, + "learning_rate": 1.0924022346368715e-05, + "loss": 0.1971, + "step": 3262 + }, + { + "epoch": 9.114525139664805, + "grad_norm": 0.760842502117157, + "learning_rate": 1.0927374301675978e-05, + "loss": 0.2279, + "step": 3263 + }, + { + "epoch": 9.11731843575419, + "grad_norm": 1.1051253080368042, + "learning_rate": 1.0930726256983242e-05, + "loss": 0.2136, + "step": 3264 + }, + { + "epoch": 9.120111731843576, + "grad_norm": 1.3597285747528076, + "learning_rate": 1.0934078212290503e-05, + "loss": 0.2031, + "step": 3265 + }, + { + "epoch": 9.12290502793296, + "grad_norm": 1.0984587669372559, + "learning_rate": 1.0937430167597766e-05, + "loss": 0.2143, + "step": 3266 + }, + { + "epoch": 9.125698324022347, + "grad_norm": 1.0658928155899048, + "learning_rate": 1.0940782122905029e-05, + "loss": 0.2213, + "step": 3267 + }, + { + "epoch": 9.128491620111731, + "grad_norm": 1.3347150087356567, + "learning_rate": 1.094413407821229e-05, + "loss": 0.2539, + "step": 3268 + }, + { + "epoch": 9.131284916201118, + "grad_norm": 1.569738507270813, + "learning_rate": 1.0947486033519553e-05, + "loss": 0.2531, + "step": 3269 + }, + { + "epoch": 9.134078212290502, + "grad_norm": 1.4521617889404297, + "learning_rate": 1.0950837988826816e-05, + "loss": 0.2589, + "step": 3270 + }, + { + "epoch": 9.136871508379889, + "grad_norm": 2.140512228012085, + "learning_rate": 1.095418994413408e-05, + "loss": 0.2413, + "step": 3271 + }, + { + "epoch": 9.139664804469273, + "grad_norm": 2.081078290939331, + "learning_rate": 1.095754189944134e-05, + "loss": 0.3638, + "step": 3272 + }, + { + "epoch": 9.14245810055866, + "grad_norm": 0.5257912874221802, + "learning_rate": 1.0960893854748604e-05, + "loss": 0.249, + "step": 3273 + }, + { + "epoch": 9.145251396648044, + "grad_norm": 0.5327383875846863, + "learning_rate": 1.0964245810055867e-05, + "loss": 0.2313, + "step": 3274 + }, + { + "epoch": 9.14804469273743, + "grad_norm": 0.6967951059341431, + "learning_rate": 1.0967597765363128e-05, + "loss": 0.1981, + "step": 3275 + }, + { + "epoch": 9.150837988826815, + "grad_norm": 0.7772819399833679, + "learning_rate": 1.0970949720670391e-05, + "loss": 0.2269, + "step": 3276 + }, + { + "epoch": 9.153631284916202, + "grad_norm": 0.598335325717926, + "learning_rate": 1.0974301675977654e-05, + "loss": 0.1962, + "step": 3277 + }, + { + "epoch": 9.156424581005586, + "grad_norm": 0.703988790512085, + "learning_rate": 1.0977653631284916e-05, + "loss": 0.2347, + "step": 3278 + }, + { + "epoch": 9.159217877094973, + "grad_norm": 0.7966230511665344, + "learning_rate": 1.0981005586592179e-05, + "loss": 0.2167, + "step": 3279 + }, + { + "epoch": 9.162011173184357, + "grad_norm": 0.7679129838943481, + "learning_rate": 1.0984357541899442e-05, + "loss": 0.1962, + "step": 3280 + }, + { + "epoch": 9.164804469273744, + "grad_norm": 0.7954160571098328, + "learning_rate": 1.0987709497206705e-05, + "loss": 0.1937, + "step": 3281 + }, + { + "epoch": 9.167597765363128, + "grad_norm": 0.6920415759086609, + "learning_rate": 1.0991061452513966e-05, + "loss": 0.2538, + "step": 3282 + }, + { + "epoch": 9.170391061452515, + "grad_norm": 0.7932744026184082, + "learning_rate": 1.0994413407821229e-05, + "loss": 0.2724, + "step": 3283 + }, + { + "epoch": 9.1731843575419, + "grad_norm": 0.6377508044242859, + "learning_rate": 1.0997765363128492e-05, + "loss": 0.2202, + "step": 3284 + }, + { + "epoch": 9.175977653631286, + "grad_norm": 0.7301687598228455, + "learning_rate": 1.1001117318435753e-05, + "loss": 0.2118, + "step": 3285 + }, + { + "epoch": 9.17877094972067, + "grad_norm": 1.442430853843689, + "learning_rate": 1.1004469273743017e-05, + "loss": 0.2086, + "step": 3286 + }, + { + "epoch": 9.181564245810057, + "grad_norm": 0.7890695333480835, + "learning_rate": 1.100782122905028e-05, + "loss": 0.2115, + "step": 3287 + }, + { + "epoch": 9.184357541899441, + "grad_norm": 0.7058039903640747, + "learning_rate": 1.1011173184357541e-05, + "loss": 0.2153, + "step": 3288 + }, + { + "epoch": 9.187150837988828, + "grad_norm": 1.0598541498184204, + "learning_rate": 1.1014525139664804e-05, + "loss": 0.2244, + "step": 3289 + }, + { + "epoch": 9.189944134078212, + "grad_norm": 1.1002511978149414, + "learning_rate": 1.1017877094972067e-05, + "loss": 0.2139, + "step": 3290 + }, + { + "epoch": 9.192737430167599, + "grad_norm": 1.0714150667190552, + "learning_rate": 1.102122905027933e-05, + "loss": 0.2306, + "step": 3291 + }, + { + "epoch": 9.195530726256983, + "grad_norm": 1.701130986213684, + "learning_rate": 1.1024581005586591e-05, + "loss": 0.2396, + "step": 3292 + }, + { + "epoch": 9.19832402234637, + "grad_norm": 1.6348825693130493, + "learning_rate": 1.1027932960893854e-05, + "loss": 0.2319, + "step": 3293 + }, + { + "epoch": 9.201117318435754, + "grad_norm": 1.4952677488327026, + "learning_rate": 1.1031284916201117e-05, + "loss": 0.2531, + "step": 3294 + }, + { + "epoch": 9.203910614525139, + "grad_norm": 1.2213528156280518, + "learning_rate": 1.1034636871508379e-05, + "loss": 0.2716, + "step": 3295 + }, + { + "epoch": 9.206703910614525, + "grad_norm": 1.541985034942627, + "learning_rate": 1.1037988826815642e-05, + "loss": 0.315, + "step": 3296 + }, + { + "epoch": 9.20949720670391, + "grad_norm": 3.327326536178589, + "learning_rate": 1.1041340782122905e-05, + "loss": 0.3437, + "step": 3297 + }, + { + "epoch": 9.212290502793296, + "grad_norm": 0.7568789124488831, + "learning_rate": 1.1044692737430166e-05, + "loss": 0.2315, + "step": 3298 + }, + { + "epoch": 9.21508379888268, + "grad_norm": 0.654945433139801, + "learning_rate": 1.1048044692737431e-05, + "loss": 0.2239, + "step": 3299 + }, + { + "epoch": 9.217877094972067, + "grad_norm": 0.6530112028121948, + "learning_rate": 1.1051396648044694e-05, + "loss": 0.2141, + "step": 3300 + }, + { + "epoch": 9.220670391061452, + "grad_norm": 0.7404937148094177, + "learning_rate": 1.1054748603351957e-05, + "loss": 0.2369, + "step": 3301 + }, + { + "epoch": 9.223463687150838, + "grad_norm": 0.7825579047203064, + "learning_rate": 1.1058100558659218e-05, + "loss": 0.2184, + "step": 3302 + }, + { + "epoch": 9.226256983240223, + "grad_norm": 0.7537280917167664, + "learning_rate": 1.1061452513966481e-05, + "loss": 0.2091, + "step": 3303 + }, + { + "epoch": 9.22905027932961, + "grad_norm": 0.8168483376502991, + "learning_rate": 1.1064804469273744e-05, + "loss": 0.2619, + "step": 3304 + }, + { + "epoch": 9.231843575418994, + "grad_norm": 0.7256521582603455, + "learning_rate": 1.1068156424581006e-05, + "loss": 0.2012, + "step": 3305 + }, + { + "epoch": 9.23463687150838, + "grad_norm": 0.7042166590690613, + "learning_rate": 1.1071508379888269e-05, + "loss": 0.1967, + "step": 3306 + }, + { + "epoch": 9.237430167597765, + "grad_norm": 0.7856917381286621, + "learning_rate": 1.1074860335195532e-05, + "loss": 0.2056, + "step": 3307 + }, + { + "epoch": 9.240223463687151, + "grad_norm": 0.6585147976875305, + "learning_rate": 1.1078212290502795e-05, + "loss": 0.1967, + "step": 3308 + }, + { + "epoch": 9.243016759776536, + "grad_norm": 0.7272319793701172, + "learning_rate": 1.1081564245810056e-05, + "loss": 0.2471, + "step": 3309 + }, + { + "epoch": 9.245810055865922, + "grad_norm": 0.7353860139846802, + "learning_rate": 1.108491620111732e-05, + "loss": 0.202, + "step": 3310 + }, + { + "epoch": 9.248603351955307, + "grad_norm": 0.7268865704536438, + "learning_rate": 1.1088268156424582e-05, + "loss": 0.2073, + "step": 3311 + }, + { + "epoch": 9.251396648044693, + "grad_norm": 0.6943860054016113, + "learning_rate": 1.1091620111731844e-05, + "loss": 0.227, + "step": 3312 + }, + { + "epoch": 9.254189944134078, + "grad_norm": 0.9334520101547241, + "learning_rate": 1.1094972067039107e-05, + "loss": 0.2192, + "step": 3313 + }, + { + "epoch": 9.256983240223464, + "grad_norm": 0.9307050108909607, + "learning_rate": 1.109832402234637e-05, + "loss": 0.2335, + "step": 3314 + }, + { + "epoch": 9.259776536312849, + "grad_norm": 1.0564515590667725, + "learning_rate": 1.1101675977653631e-05, + "loss": 0.2203, + "step": 3315 + }, + { + "epoch": 9.262569832402235, + "grad_norm": 0.8814364671707153, + "learning_rate": 1.1105027932960894e-05, + "loss": 0.2232, + "step": 3316 + }, + { + "epoch": 9.26536312849162, + "grad_norm": 1.1962082386016846, + "learning_rate": 1.1108379888268157e-05, + "loss": 0.2445, + "step": 3317 + }, + { + "epoch": 9.268156424581006, + "grad_norm": 1.236128568649292, + "learning_rate": 1.111173184357542e-05, + "loss": 0.2432, + "step": 3318 + }, + { + "epoch": 9.27094972067039, + "grad_norm": 1.5792659521102905, + "learning_rate": 1.1115083798882682e-05, + "loss": 0.219, + "step": 3319 + }, + { + "epoch": 9.273743016759777, + "grad_norm": 1.4649658203125, + "learning_rate": 1.1118435754189945e-05, + "loss": 0.2215, + "step": 3320 + }, + { + "epoch": 9.276536312849162, + "grad_norm": 1.743159294128418, + "learning_rate": 1.1121787709497208e-05, + "loss": 0.3433, + "step": 3321 + }, + { + "epoch": 9.279329608938548, + "grad_norm": 3.499854803085327, + "learning_rate": 1.1125139664804469e-05, + "loss": 0.3825, + "step": 3322 + }, + { + "epoch": 9.282122905027933, + "grad_norm": 0.6043581366539001, + "learning_rate": 1.1128491620111732e-05, + "loss": 0.2308, + "step": 3323 + }, + { + "epoch": 9.28491620111732, + "grad_norm": 0.6336775422096252, + "learning_rate": 1.1131843575418995e-05, + "loss": 0.2297, + "step": 3324 + }, + { + "epoch": 9.287709497206704, + "grad_norm": 0.6842542290687561, + "learning_rate": 1.1135195530726256e-05, + "loss": 0.2196, + "step": 3325 + }, + { + "epoch": 9.29050279329609, + "grad_norm": 0.5107066631317139, + "learning_rate": 1.113854748603352e-05, + "loss": 0.2338, + "step": 3326 + }, + { + "epoch": 9.293296089385475, + "grad_norm": 0.7188464999198914, + "learning_rate": 1.1141899441340783e-05, + "loss": 0.249, + "step": 3327 + }, + { + "epoch": 9.296089385474861, + "grad_norm": 0.7410153150558472, + "learning_rate": 1.1145251396648046e-05, + "loss": 0.213, + "step": 3328 + }, + { + "epoch": 9.298882681564246, + "grad_norm": 0.6670082807540894, + "learning_rate": 1.1148603351955307e-05, + "loss": 0.237, + "step": 3329 + }, + { + "epoch": 9.30167597765363, + "grad_norm": 0.5466026067733765, + "learning_rate": 1.115195530726257e-05, + "loss": 0.232, + "step": 3330 + }, + { + "epoch": 9.304469273743017, + "grad_norm": 0.7056851983070374, + "learning_rate": 1.1155307262569833e-05, + "loss": 0.2313, + "step": 3331 + }, + { + "epoch": 9.307262569832401, + "grad_norm": 0.7223793864250183, + "learning_rate": 1.1158659217877094e-05, + "loss": 0.1809, + "step": 3332 + }, + { + "epoch": 9.310055865921788, + "grad_norm": 0.8107491731643677, + "learning_rate": 1.1162011173184357e-05, + "loss": 0.2601, + "step": 3333 + }, + { + "epoch": 9.312849162011172, + "grad_norm": 0.7203710079193115, + "learning_rate": 1.116536312849162e-05, + "loss": 0.1921, + "step": 3334 + }, + { + "epoch": 9.315642458100559, + "grad_norm": 0.9094083905220032, + "learning_rate": 1.1168715083798882e-05, + "loss": 0.2201, + "step": 3335 + }, + { + "epoch": 9.318435754189943, + "grad_norm": 0.7239468693733215, + "learning_rate": 1.1172067039106145e-05, + "loss": 0.2385, + "step": 3336 + }, + { + "epoch": 9.32122905027933, + "grad_norm": 0.821891725063324, + "learning_rate": 1.1175418994413408e-05, + "loss": 0.2097, + "step": 3337 + }, + { + "epoch": 9.324022346368714, + "grad_norm": 0.9501901268959045, + "learning_rate": 1.1178770949720671e-05, + "loss": 0.2166, + "step": 3338 + }, + { + "epoch": 9.3268156424581, + "grad_norm": 0.8791401982307434, + "learning_rate": 1.1182122905027932e-05, + "loss": 0.1957, + "step": 3339 + }, + { + "epoch": 9.329608938547485, + "grad_norm": 1.0973224639892578, + "learning_rate": 1.1185474860335195e-05, + "loss": 0.2399, + "step": 3340 + }, + { + "epoch": 9.332402234636872, + "grad_norm": 1.3390226364135742, + "learning_rate": 1.1188826815642458e-05, + "loss": 0.2169, + "step": 3341 + }, + { + "epoch": 9.335195530726256, + "grad_norm": 0.973518431186676, + "learning_rate": 1.119217877094972e-05, + "loss": 0.2402, + "step": 3342 + }, + { + "epoch": 9.337988826815643, + "grad_norm": 2.8300626277923584, + "learning_rate": 1.1195530726256983e-05, + "loss": 0.231, + "step": 3343 + }, + { + "epoch": 9.340782122905027, + "grad_norm": 1.752416968345642, + "learning_rate": 1.1198882681564246e-05, + "loss": 0.2593, + "step": 3344 + }, + { + "epoch": 9.343575418994414, + "grad_norm": 2.009169578552246, + "learning_rate": 1.1202234636871507e-05, + "loss": 0.2776, + "step": 3345 + }, + { + "epoch": 9.346368715083798, + "grad_norm": 2.1851537227630615, + "learning_rate": 1.120558659217877e-05, + "loss": 0.2921, + "step": 3346 + }, + { + "epoch": 9.349162011173185, + "grad_norm": 2.0061075687408447, + "learning_rate": 1.1208938547486033e-05, + "loss": 0.3052, + "step": 3347 + }, + { + "epoch": 9.35195530726257, + "grad_norm": 0.5762128829956055, + "learning_rate": 1.1212290502793298e-05, + "loss": 0.1987, + "step": 3348 + }, + { + "epoch": 9.354748603351956, + "grad_norm": 0.47711822390556335, + "learning_rate": 1.121564245810056e-05, + "loss": 0.2051, + "step": 3349 + }, + { + "epoch": 9.35754189944134, + "grad_norm": 0.6189988255500793, + "learning_rate": 1.1218994413407822e-05, + "loss": 0.2112, + "step": 3350 + }, + { + "epoch": 9.360335195530727, + "grad_norm": 0.7084192633628845, + "learning_rate": 1.1222346368715085e-05, + "loss": 0.256, + "step": 3351 + }, + { + "epoch": 9.363128491620111, + "grad_norm": 0.8251171112060547, + "learning_rate": 1.1225698324022347e-05, + "loss": 0.2028, + "step": 3352 + }, + { + "epoch": 9.365921787709498, + "grad_norm": 0.5340461134910583, + "learning_rate": 1.122905027932961e-05, + "loss": 0.2211, + "step": 3353 + }, + { + "epoch": 9.368715083798882, + "grad_norm": 0.610945999622345, + "learning_rate": 1.1232402234636873e-05, + "loss": 0.203, + "step": 3354 + }, + { + "epoch": 9.371508379888269, + "grad_norm": 0.5916205048561096, + "learning_rate": 1.1235754189944136e-05, + "loss": 0.2221, + "step": 3355 + }, + { + "epoch": 9.374301675977653, + "grad_norm": 0.7304513454437256, + "learning_rate": 1.1239106145251397e-05, + "loss": 0.211, + "step": 3356 + }, + { + "epoch": 9.37709497206704, + "grad_norm": 0.6762126088142395, + "learning_rate": 1.124245810055866e-05, + "loss": 0.208, + "step": 3357 + }, + { + "epoch": 9.379888268156424, + "grad_norm": 0.7928761839866638, + "learning_rate": 1.1245810055865923e-05, + "loss": 0.2091, + "step": 3358 + }, + { + "epoch": 9.38268156424581, + "grad_norm": 1.122127652168274, + "learning_rate": 1.1249162011173185e-05, + "loss": 0.2227, + "step": 3359 + }, + { + "epoch": 9.385474860335195, + "grad_norm": 0.68089759349823, + "learning_rate": 1.1252513966480448e-05, + "loss": 0.2129, + "step": 3360 + }, + { + "epoch": 9.388268156424582, + "grad_norm": 0.858140230178833, + "learning_rate": 1.125586592178771e-05, + "loss": 0.2312, + "step": 3361 + }, + { + "epoch": 9.391061452513966, + "grad_norm": 0.8195881247520447, + "learning_rate": 1.1259217877094972e-05, + "loss": 0.2352, + "step": 3362 + }, + { + "epoch": 9.393854748603353, + "grad_norm": 1.3215466737747192, + "learning_rate": 1.1262569832402235e-05, + "loss": 0.2356, + "step": 3363 + }, + { + "epoch": 9.396648044692737, + "grad_norm": 0.9016291499137878, + "learning_rate": 1.1265921787709498e-05, + "loss": 0.2172, + "step": 3364 + }, + { + "epoch": 9.399441340782122, + "grad_norm": 0.9575898051261902, + "learning_rate": 1.1269273743016761e-05, + "loss": 0.2442, + "step": 3365 + }, + { + "epoch": 9.402234636871508, + "grad_norm": 1.7245588302612305, + "learning_rate": 1.1272625698324022e-05, + "loss": 0.1981, + "step": 3366 + }, + { + "epoch": 9.405027932960893, + "grad_norm": 1.1921453475952148, + "learning_rate": 1.1275977653631285e-05, + "loss": 0.2107, + "step": 3367 + }, + { + "epoch": 9.40782122905028, + "grad_norm": 1.8884599208831787, + "learning_rate": 1.1279329608938548e-05, + "loss": 0.2433, + "step": 3368 + }, + { + "epoch": 9.410614525139664, + "grad_norm": 1.5051143169403076, + "learning_rate": 1.128268156424581e-05, + "loss": 0.2502, + "step": 3369 + }, + { + "epoch": 9.41340782122905, + "grad_norm": 1.9935425519943237, + "learning_rate": 1.1286033519553073e-05, + "loss": 0.2961, + "step": 3370 + }, + { + "epoch": 9.416201117318435, + "grad_norm": 2.1037490367889404, + "learning_rate": 1.1289385474860336e-05, + "loss": 0.3003, + "step": 3371 + }, + { + "epoch": 9.418994413407821, + "grad_norm": 2.888397216796875, + "learning_rate": 1.1292737430167597e-05, + "loss": 0.3869, + "step": 3372 + }, + { + "epoch": 9.421787709497206, + "grad_norm": 0.8936245441436768, + "learning_rate": 1.129608938547486e-05, + "loss": 0.2393, + "step": 3373 + }, + { + "epoch": 9.424581005586592, + "grad_norm": 0.6042218804359436, + "learning_rate": 1.1299441340782123e-05, + "loss": 0.2337, + "step": 3374 + }, + { + "epoch": 9.427374301675977, + "grad_norm": 0.7437140941619873, + "learning_rate": 1.1302793296089386e-05, + "loss": 0.2164, + "step": 3375 + }, + { + "epoch": 9.430167597765363, + "grad_norm": 0.9796949625015259, + "learning_rate": 1.1306145251396648e-05, + "loss": 0.2329, + "step": 3376 + }, + { + "epoch": 9.432960893854748, + "grad_norm": 0.8061531186103821, + "learning_rate": 1.130949720670391e-05, + "loss": 0.2406, + "step": 3377 + }, + { + "epoch": 9.435754189944134, + "grad_norm": 1.8259072303771973, + "learning_rate": 1.1312849162011174e-05, + "loss": 0.2202, + "step": 3378 + }, + { + "epoch": 9.438547486033519, + "grad_norm": 0.7192888855934143, + "learning_rate": 1.1316201117318435e-05, + "loss": 0.2085, + "step": 3379 + }, + { + "epoch": 9.441340782122905, + "grad_norm": 0.6698203086853027, + "learning_rate": 1.1319553072625698e-05, + "loss": 0.224, + "step": 3380 + }, + { + "epoch": 9.44413407821229, + "grad_norm": 0.7796608805656433, + "learning_rate": 1.1322905027932961e-05, + "loss": 0.2146, + "step": 3381 + }, + { + "epoch": 9.446927374301676, + "grad_norm": 2.313292980194092, + "learning_rate": 1.1326256983240223e-05, + "loss": 0.2288, + "step": 3382 + }, + { + "epoch": 9.449720670391061, + "grad_norm": 0.7646060585975647, + "learning_rate": 1.1329608938547486e-05, + "loss": 0.2485, + "step": 3383 + }, + { + "epoch": 9.452513966480447, + "grad_norm": 0.6508960723876953, + "learning_rate": 1.1332960893854749e-05, + "loss": 0.2206, + "step": 3384 + }, + { + "epoch": 9.455307262569832, + "grad_norm": 0.9737193584442139, + "learning_rate": 1.1336312849162012e-05, + "loss": 0.2327, + "step": 3385 + }, + { + "epoch": 9.458100558659218, + "grad_norm": 1.1146893501281738, + "learning_rate": 1.1339664804469273e-05, + "loss": 0.2286, + "step": 3386 + }, + { + "epoch": 9.460893854748603, + "grad_norm": 0.8358224630355835, + "learning_rate": 1.1343016759776536e-05, + "loss": 0.2081, + "step": 3387 + }, + { + "epoch": 9.46368715083799, + "grad_norm": 1.1017320156097412, + "learning_rate": 1.1346368715083799e-05, + "loss": 0.1854, + "step": 3388 + }, + { + "epoch": 9.466480446927374, + "grad_norm": 0.9480065107345581, + "learning_rate": 1.134972067039106e-05, + "loss": 0.1768, + "step": 3389 + }, + { + "epoch": 9.46927374301676, + "grad_norm": 0.9602586030960083, + "learning_rate": 1.1353072625698323e-05, + "loss": 0.1965, + "step": 3390 + }, + { + "epoch": 9.472067039106145, + "grad_norm": 1.1450023651123047, + "learning_rate": 1.1356424581005587e-05, + "loss": 0.2567, + "step": 3391 + }, + { + "epoch": 9.474860335195531, + "grad_norm": 1.892240047454834, + "learning_rate": 1.1359776536312848e-05, + "loss": 0.2678, + "step": 3392 + }, + { + "epoch": 9.477653631284916, + "grad_norm": 1.3181381225585938, + "learning_rate": 1.1363128491620111e-05, + "loss": 0.2554, + "step": 3393 + }, + { + "epoch": 9.480446927374302, + "grad_norm": 1.3370344638824463, + "learning_rate": 1.1366480446927374e-05, + "loss": 0.2538, + "step": 3394 + }, + { + "epoch": 9.483240223463687, + "grad_norm": 3.4092836380004883, + "learning_rate": 1.1369832402234637e-05, + "loss": 0.2538, + "step": 3395 + }, + { + "epoch": 9.486033519553073, + "grad_norm": 1.6904330253601074, + "learning_rate": 1.13731843575419e-05, + "loss": 0.2622, + "step": 3396 + }, + { + "epoch": 9.488826815642458, + "grad_norm": 7.825469493865967, + "learning_rate": 1.1376536312849163e-05, + "loss": 0.4168, + "step": 3397 + }, + { + "epoch": 9.491620111731844, + "grad_norm": 0.4891795516014099, + "learning_rate": 1.1379888268156426e-05, + "loss": 0.2063, + "step": 3398 + }, + { + "epoch": 9.494413407821229, + "grad_norm": 0.4954397678375244, + "learning_rate": 1.1383240223463687e-05, + "loss": 0.193, + "step": 3399 + }, + { + "epoch": 9.497206703910614, + "grad_norm": 0.5899249315261841, + "learning_rate": 1.138659217877095e-05, + "loss": 0.2159, + "step": 3400 + }, + { + "epoch": 9.5, + "grad_norm": 0.5530983805656433, + "learning_rate": 1.1389944134078214e-05, + "loss": 0.1977, + "step": 3401 + }, + { + "epoch": 9.502793296089386, + "grad_norm": 0.7307820916175842, + "learning_rate": 1.1393296089385477e-05, + "loss": 0.2165, + "step": 3402 + }, + { + "epoch": 9.505586592178771, + "grad_norm": 0.6158713698387146, + "learning_rate": 1.1396648044692738e-05, + "loss": 0.2086, + "step": 3403 + }, + { + "epoch": 9.508379888268156, + "grad_norm": 0.6566550731658936, + "learning_rate": 1.1400000000000001e-05, + "loss": 0.265, + "step": 3404 + }, + { + "epoch": 9.511173184357542, + "grad_norm": 0.6347560286521912, + "learning_rate": 1.1403351955307264e-05, + "loss": 0.2309, + "step": 3405 + }, + { + "epoch": 9.513966480446927, + "grad_norm": 0.8636351823806763, + "learning_rate": 1.1406703910614525e-05, + "loss": 0.1895, + "step": 3406 + }, + { + "epoch": 9.516759776536313, + "grad_norm": 2.7124898433685303, + "learning_rate": 1.1410055865921788e-05, + "loss": 0.2505, + "step": 3407 + }, + { + "epoch": 9.519553072625698, + "grad_norm": 0.6774752140045166, + "learning_rate": 1.1413407821229051e-05, + "loss": 0.2062, + "step": 3408 + }, + { + "epoch": 9.522346368715084, + "grad_norm": 0.9621689319610596, + "learning_rate": 1.1416759776536313e-05, + "loss": 0.2549, + "step": 3409 + }, + { + "epoch": 9.525139664804469, + "grad_norm": 0.9795498847961426, + "learning_rate": 1.1420111731843576e-05, + "loss": 0.1926, + "step": 3410 + }, + { + "epoch": 9.527932960893855, + "grad_norm": 0.8298664093017578, + "learning_rate": 1.1423463687150839e-05, + "loss": 0.2524, + "step": 3411 + }, + { + "epoch": 9.53072625698324, + "grad_norm": 1.3194166421890259, + "learning_rate": 1.1426815642458102e-05, + "loss": 0.1734, + "step": 3412 + }, + { + "epoch": 9.533519553072626, + "grad_norm": 0.8806458115577698, + "learning_rate": 1.1430167597765363e-05, + "loss": 0.1896, + "step": 3413 + }, + { + "epoch": 9.53631284916201, + "grad_norm": 1.2185496091842651, + "learning_rate": 1.1433519553072626e-05, + "loss": 0.2132, + "step": 3414 + }, + { + "epoch": 9.539106145251397, + "grad_norm": 3.0598347187042236, + "learning_rate": 1.143687150837989e-05, + "loss": 0.227, + "step": 3415 + }, + { + "epoch": 9.541899441340782, + "grad_norm": 1.6002599000930786, + "learning_rate": 1.144022346368715e-05, + "loss": 0.2251, + "step": 3416 + }, + { + "epoch": 9.544692737430168, + "grad_norm": 1.6951258182525635, + "learning_rate": 1.1443575418994414e-05, + "loss": 0.256, + "step": 3417 + }, + { + "epoch": 9.547486033519553, + "grad_norm": 4.708590507507324, + "learning_rate": 1.1446927374301677e-05, + "loss": 0.292, + "step": 3418 + }, + { + "epoch": 9.550279329608939, + "grad_norm": 1.4946887493133545, + "learning_rate": 1.1450279329608938e-05, + "loss": 0.2125, + "step": 3419 + }, + { + "epoch": 9.553072625698324, + "grad_norm": 1.4416627883911133, + "learning_rate": 1.1453631284916201e-05, + "loss": 0.2028, + "step": 3420 + }, + { + "epoch": 9.55586592178771, + "grad_norm": 3.9530301094055176, + "learning_rate": 1.1456983240223464e-05, + "loss": 0.2557, + "step": 3421 + }, + { + "epoch": 9.558659217877095, + "grad_norm": 2.3426480293273926, + "learning_rate": 1.1460335195530727e-05, + "loss": 0.326, + "step": 3422 + }, + { + "epoch": 9.561452513966481, + "grad_norm": 0.507257342338562, + "learning_rate": 1.1463687150837989e-05, + "loss": 0.2035, + "step": 3423 + }, + { + "epoch": 9.564245810055866, + "grad_norm": 0.6511245369911194, + "learning_rate": 1.1467039106145252e-05, + "loss": 0.2341, + "step": 3424 + }, + { + "epoch": 9.567039106145252, + "grad_norm": 0.7226115465164185, + "learning_rate": 1.1470391061452515e-05, + "loss": 0.2374, + "step": 3425 + }, + { + "epoch": 9.569832402234637, + "grad_norm": 0.756916880607605, + "learning_rate": 1.1473743016759776e-05, + "loss": 0.2024, + "step": 3426 + }, + { + "epoch": 9.572625698324023, + "grad_norm": 0.7182049751281738, + "learning_rate": 1.1477094972067039e-05, + "loss": 0.2347, + "step": 3427 + }, + { + "epoch": 9.575418994413408, + "grad_norm": 0.6548609137535095, + "learning_rate": 1.1480446927374302e-05, + "loss": 0.2532, + "step": 3428 + }, + { + "epoch": 9.578212290502794, + "grad_norm": 0.7530311346054077, + "learning_rate": 1.1483798882681563e-05, + "loss": 0.1929, + "step": 3429 + }, + { + "epoch": 9.581005586592179, + "grad_norm": 1.0861663818359375, + "learning_rate": 1.1487150837988826e-05, + "loss": 0.1963, + "step": 3430 + }, + { + "epoch": 9.583798882681565, + "grad_norm": 0.636437714099884, + "learning_rate": 1.149050279329609e-05, + "loss": 0.2167, + "step": 3431 + }, + { + "epoch": 9.58659217877095, + "grad_norm": 0.6519576907157898, + "learning_rate": 1.1493854748603353e-05, + "loss": 0.235, + "step": 3432 + }, + { + "epoch": 9.589385474860336, + "grad_norm": 0.9471127390861511, + "learning_rate": 1.1497206703910614e-05, + "loss": 0.203, + "step": 3433 + }, + { + "epoch": 9.59217877094972, + "grad_norm": 0.6647396683692932, + "learning_rate": 1.1500558659217877e-05, + "loss": 0.2086, + "step": 3434 + }, + { + "epoch": 9.594972067039105, + "grad_norm": 0.7392998933792114, + "learning_rate": 1.150391061452514e-05, + "loss": 0.2205, + "step": 3435 + }, + { + "epoch": 9.597765363128492, + "grad_norm": 1.0956239700317383, + "learning_rate": 1.1507262569832401e-05, + "loss": 0.2105, + "step": 3436 + }, + { + "epoch": 9.600558659217878, + "grad_norm": 0.8524746894836426, + "learning_rate": 1.1510614525139664e-05, + "loss": 0.2323, + "step": 3437 + }, + { + "epoch": 9.603351955307263, + "grad_norm": 1.5840239524841309, + "learning_rate": 1.1513966480446927e-05, + "loss": 0.238, + "step": 3438 + }, + { + "epoch": 9.606145251396647, + "grad_norm": 1.043882966041565, + "learning_rate": 1.1517318435754189e-05, + "loss": 0.2393, + "step": 3439 + }, + { + "epoch": 9.608938547486034, + "grad_norm": 0.9229077100753784, + "learning_rate": 1.1520670391061452e-05, + "loss": 0.2168, + "step": 3440 + }, + { + "epoch": 9.611731843575418, + "grad_norm": 1.7630749940872192, + "learning_rate": 1.1524022346368715e-05, + "loss": 0.2376, + "step": 3441 + }, + { + "epoch": 9.614525139664805, + "grad_norm": 1.0293365716934204, + "learning_rate": 1.1527374301675978e-05, + "loss": 0.218, + "step": 3442 + }, + { + "epoch": 9.61731843575419, + "grad_norm": 1.429356575012207, + "learning_rate": 1.153072625698324e-05, + "loss": 0.2189, + "step": 3443 + }, + { + "epoch": 9.620111731843576, + "grad_norm": 1.3917161226272583, + "learning_rate": 1.1534078212290504e-05, + "loss": 0.2619, + "step": 3444 + }, + { + "epoch": 9.62290502793296, + "grad_norm": 1.2930704355239868, + "learning_rate": 1.1537430167597767e-05, + "loss": 0.2912, + "step": 3445 + }, + { + "epoch": 9.625698324022347, + "grad_norm": 1.3136250972747803, + "learning_rate": 1.1540782122905028e-05, + "loss": 0.2501, + "step": 3446 + }, + { + "epoch": 9.628491620111731, + "grad_norm": 2.5741775035858154, + "learning_rate": 1.1544134078212291e-05, + "loss": 0.3182, + "step": 3447 + }, + { + "epoch": 9.631284916201118, + "grad_norm": 1.2309497594833374, + "learning_rate": 1.1547486033519554e-05, + "loss": 0.1961, + "step": 3448 + }, + { + "epoch": 9.634078212290502, + "grad_norm": 0.5115754008293152, + "learning_rate": 1.1550837988826817e-05, + "loss": 0.1896, + "step": 3449 + }, + { + "epoch": 9.636871508379889, + "grad_norm": 1.1045223474502563, + "learning_rate": 1.1554189944134079e-05, + "loss": 0.2599, + "step": 3450 + }, + { + "epoch": 9.639664804469273, + "grad_norm": 0.49422502517700195, + "learning_rate": 1.1557541899441342e-05, + "loss": 0.1885, + "step": 3451 + }, + { + "epoch": 9.64245810055866, + "grad_norm": 0.6925631165504456, + "learning_rate": 1.1560893854748605e-05, + "loss": 0.2076, + "step": 3452 + }, + { + "epoch": 9.645251396648044, + "grad_norm": 0.6683768630027771, + "learning_rate": 1.1564245810055866e-05, + "loss": 0.1933, + "step": 3453 + }, + { + "epoch": 9.64804469273743, + "grad_norm": 0.8326707482337952, + "learning_rate": 1.156759776536313e-05, + "loss": 0.2265, + "step": 3454 + }, + { + "epoch": 9.650837988826815, + "grad_norm": 1.0652562379837036, + "learning_rate": 1.1570949720670392e-05, + "loss": 0.2141, + "step": 3455 + }, + { + "epoch": 9.653631284916202, + "grad_norm": 0.7336655855178833, + "learning_rate": 1.1574301675977654e-05, + "loss": 0.1987, + "step": 3456 + }, + { + "epoch": 9.656424581005586, + "grad_norm": 0.9110353589057922, + "learning_rate": 1.1577653631284917e-05, + "loss": 0.1794, + "step": 3457 + }, + { + "epoch": 9.659217877094973, + "grad_norm": 1.829257845878601, + "learning_rate": 1.158100558659218e-05, + "loss": 0.2087, + "step": 3458 + }, + { + "epoch": 9.662011173184357, + "grad_norm": 0.7416763305664062, + "learning_rate": 1.1584357541899443e-05, + "loss": 0.1876, + "step": 3459 + }, + { + "epoch": 9.664804469273744, + "grad_norm": 0.7036628723144531, + "learning_rate": 1.1587709497206704e-05, + "loss": 0.1905, + "step": 3460 + }, + { + "epoch": 9.667597765363128, + "grad_norm": 0.8629494309425354, + "learning_rate": 1.1591061452513967e-05, + "loss": 0.2529, + "step": 3461 + }, + { + "epoch": 9.670391061452515, + "grad_norm": 0.8874157667160034, + "learning_rate": 1.159441340782123e-05, + "loss": 0.2313, + "step": 3462 + }, + { + "epoch": 9.6731843575419, + "grad_norm": 0.7909137606620789, + "learning_rate": 1.1597765363128491e-05, + "loss": 0.2163, + "step": 3463 + }, + { + "epoch": 9.675977653631286, + "grad_norm": 0.7723409533500671, + "learning_rate": 1.1601117318435755e-05, + "loss": 0.2197, + "step": 3464 + }, + { + "epoch": 9.67877094972067, + "grad_norm": 2.0867130756378174, + "learning_rate": 1.1604469273743018e-05, + "loss": 0.2646, + "step": 3465 + }, + { + "epoch": 9.681564245810057, + "grad_norm": 1.3229670524597168, + "learning_rate": 1.1607821229050279e-05, + "loss": 0.2662, + "step": 3466 + }, + { + "epoch": 9.684357541899441, + "grad_norm": 6.613818168640137, + "learning_rate": 1.1611173184357542e-05, + "loss": 0.2781, + "step": 3467 + }, + { + "epoch": 9.687150837988828, + "grad_norm": 1.2637778520584106, + "learning_rate": 1.1614525139664805e-05, + "loss": 0.251, + "step": 3468 + }, + { + "epoch": 9.689944134078212, + "grad_norm": 0.9955556988716125, + "learning_rate": 1.1617877094972068e-05, + "loss": 0.2508, + "step": 3469 + }, + { + "epoch": 9.692737430167599, + "grad_norm": 1.2294881343841553, + "learning_rate": 1.162122905027933e-05, + "loss": 0.2588, + "step": 3470 + }, + { + "epoch": 9.695530726256983, + "grad_norm": 1.6554834842681885, + "learning_rate": 1.1624581005586592e-05, + "loss": 0.329, + "step": 3471 + }, + { + "epoch": 9.69832402234637, + "grad_norm": 1.7825944423675537, + "learning_rate": 1.1627932960893855e-05, + "loss": 0.3149, + "step": 3472 + }, + { + "epoch": 9.701117318435754, + "grad_norm": 0.5775532722473145, + "learning_rate": 1.1631284916201117e-05, + "loss": 0.2407, + "step": 3473 + }, + { + "epoch": 9.703910614525139, + "grad_norm": 0.622852087020874, + "learning_rate": 1.163463687150838e-05, + "loss": 0.2305, + "step": 3474 + }, + { + "epoch": 9.706703910614525, + "grad_norm": 0.7875474095344543, + "learning_rate": 1.1637988826815643e-05, + "loss": 0.2267, + "step": 3475 + }, + { + "epoch": 9.70949720670391, + "grad_norm": 1.1945244073867798, + "learning_rate": 1.1641340782122904e-05, + "loss": 0.1881, + "step": 3476 + }, + { + "epoch": 9.712290502793296, + "grad_norm": 0.7535152435302734, + "learning_rate": 1.1644692737430167e-05, + "loss": 0.2705, + "step": 3477 + }, + { + "epoch": 9.71508379888268, + "grad_norm": 1.4315757751464844, + "learning_rate": 1.164804469273743e-05, + "loss": 0.2909, + "step": 3478 + }, + { + "epoch": 9.717877094972067, + "grad_norm": 0.9350800514221191, + "learning_rate": 1.1651396648044693e-05, + "loss": 0.2285, + "step": 3479 + }, + { + "epoch": 9.720670391061452, + "grad_norm": 0.6880145072937012, + "learning_rate": 1.1654748603351955e-05, + "loss": 0.2061, + "step": 3480 + }, + { + "epoch": 9.723463687150838, + "grad_norm": 0.7626796960830688, + "learning_rate": 1.1658100558659218e-05, + "loss": 0.1928, + "step": 3481 + }, + { + "epoch": 9.726256983240223, + "grad_norm": 0.884381890296936, + "learning_rate": 1.166145251396648e-05, + "loss": 0.1991, + "step": 3482 + }, + { + "epoch": 9.72905027932961, + "grad_norm": 0.8550395965576172, + "learning_rate": 1.1664804469273742e-05, + "loss": 0.2951, + "step": 3483 + }, + { + "epoch": 9.731843575418994, + "grad_norm": 0.8769298195838928, + "learning_rate": 1.1668156424581005e-05, + "loss": 0.2109, + "step": 3484 + }, + { + "epoch": 9.73463687150838, + "grad_norm": 1.0281908512115479, + "learning_rate": 1.1671508379888268e-05, + "loss": 0.2178, + "step": 3485 + }, + { + "epoch": 9.737430167597765, + "grad_norm": 0.9397208094596863, + "learning_rate": 1.167486033519553e-05, + "loss": 0.2235, + "step": 3486 + }, + { + "epoch": 9.740223463687151, + "grad_norm": 0.8262137770652771, + "learning_rate": 1.1678212290502793e-05, + "loss": 0.1973, + "step": 3487 + }, + { + "epoch": 9.743016759776536, + "grad_norm": 0.9919820427894592, + "learning_rate": 1.1681564245810056e-05, + "loss": 0.1874, + "step": 3488 + }, + { + "epoch": 9.745810055865922, + "grad_norm": 1.7297152280807495, + "learning_rate": 1.1684916201117319e-05, + "loss": 0.2481, + "step": 3489 + }, + { + "epoch": 9.748603351955307, + "grad_norm": 0.7874985933303833, + "learning_rate": 1.168826815642458e-05, + "loss": 0.2149, + "step": 3490 + }, + { + "epoch": 9.751396648044693, + "grad_norm": 1.1287271976470947, + "learning_rate": 1.1691620111731843e-05, + "loss": 0.2316, + "step": 3491 + }, + { + "epoch": 9.754189944134078, + "grad_norm": 1.2977136373519897, + "learning_rate": 1.1694972067039108e-05, + "loss": 0.2772, + "step": 3492 + }, + { + "epoch": 9.756983240223464, + "grad_norm": 1.7115302085876465, + "learning_rate": 1.1698324022346369e-05, + "loss": 0.2349, + "step": 3493 + }, + { + "epoch": 9.759776536312849, + "grad_norm": 1.7310980558395386, + "learning_rate": 1.1701675977653632e-05, + "loss": 0.2434, + "step": 3494 + }, + { + "epoch": 9.762569832402235, + "grad_norm": 1.382678508758545, + "learning_rate": 1.1705027932960895e-05, + "loss": 0.2566, + "step": 3495 + }, + { + "epoch": 9.76536312849162, + "grad_norm": 1.2077713012695312, + "learning_rate": 1.1708379888268158e-05, + "loss": 0.2948, + "step": 3496 + }, + { + "epoch": 9.768156424581006, + "grad_norm": 3.1994681358337402, + "learning_rate": 1.171173184357542e-05, + "loss": 0.4435, + "step": 3497 + }, + { + "epoch": 9.77094972067039, + "grad_norm": 1.4668664932250977, + "learning_rate": 1.1715083798882683e-05, + "loss": 0.2653, + "step": 3498 + }, + { + "epoch": 9.773743016759777, + "grad_norm": 1.1390244960784912, + "learning_rate": 1.1718435754189946e-05, + "loss": 0.2169, + "step": 3499 + }, + { + "epoch": 9.776536312849162, + "grad_norm": 0.7042890787124634, + "learning_rate": 1.1721787709497207e-05, + "loss": 0.1998, + "step": 3500 + }, + { + "epoch": 9.779329608938548, + "grad_norm": 0.6650234460830688, + "learning_rate": 1.172513966480447e-05, + "loss": 0.2009, + "step": 3501 + }, + { + "epoch": 9.782122905027933, + "grad_norm": 1.8162131309509277, + "learning_rate": 1.1728491620111733e-05, + "loss": 0.2018, + "step": 3502 + }, + { + "epoch": 9.78491620111732, + "grad_norm": 0.7987037897109985, + "learning_rate": 1.1731843575418994e-05, + "loss": 0.213, + "step": 3503 + }, + { + "epoch": 9.787709497206704, + "grad_norm": 0.9289015531539917, + "learning_rate": 1.1735195530726257e-05, + "loss": 0.2135, + "step": 3504 + }, + { + "epoch": 9.79050279329609, + "grad_norm": 3.002763032913208, + "learning_rate": 1.173854748603352e-05, + "loss": 0.2962, + "step": 3505 + }, + { + "epoch": 9.793296089385475, + "grad_norm": 0.6672846674919128, + "learning_rate": 1.1741899441340784e-05, + "loss": 0.2153, + "step": 3506 + }, + { + "epoch": 9.796089385474861, + "grad_norm": 1.0101759433746338, + "learning_rate": 1.1745251396648045e-05, + "loss": 0.2432, + "step": 3507 + }, + { + "epoch": 9.798882681564246, + "grad_norm": 0.7373185753822327, + "learning_rate": 1.1748603351955308e-05, + "loss": 0.2336, + "step": 3508 + }, + { + "epoch": 9.80167597765363, + "grad_norm": 1.3644604682922363, + "learning_rate": 1.1751955307262571e-05, + "loss": 0.227, + "step": 3509 + }, + { + "epoch": 9.804469273743017, + "grad_norm": 0.690997838973999, + "learning_rate": 1.1755307262569832e-05, + "loss": 0.2083, + "step": 3510 + }, + { + "epoch": 9.807262569832401, + "grad_norm": 0.7926396727561951, + "learning_rate": 1.1758659217877095e-05, + "loss": 0.2363, + "step": 3511 + }, + { + "epoch": 9.810055865921788, + "grad_norm": 0.7886630892753601, + "learning_rate": 1.1762011173184358e-05, + "loss": 0.2421, + "step": 3512 + }, + { + "epoch": 9.812849162011172, + "grad_norm": 0.8462393283843994, + "learning_rate": 1.176536312849162e-05, + "loss": 0.2182, + "step": 3513 + }, + { + "epoch": 9.815642458100559, + "grad_norm": 1.0349289178848267, + "learning_rate": 1.1768715083798883e-05, + "loss": 0.221, + "step": 3514 + }, + { + "epoch": 9.818435754189943, + "grad_norm": 1.338098406791687, + "learning_rate": 1.1772067039106146e-05, + "loss": 0.2724, + "step": 3515 + }, + { + "epoch": 9.82122905027933, + "grad_norm": 1.4430867433547974, + "learning_rate": 1.1775418994413409e-05, + "loss": 0.267, + "step": 3516 + }, + { + "epoch": 9.824022346368714, + "grad_norm": 0.9728423357009888, + "learning_rate": 1.177877094972067e-05, + "loss": 0.1968, + "step": 3517 + }, + { + "epoch": 9.8268156424581, + "grad_norm": 1.456271767616272, + "learning_rate": 1.1782122905027933e-05, + "loss": 0.2426, + "step": 3518 + }, + { + "epoch": 9.829608938547485, + "grad_norm": 1.403412103652954, + "learning_rate": 1.1785474860335196e-05, + "loss": 0.2608, + "step": 3519 + }, + { + "epoch": 9.832402234636872, + "grad_norm": 1.889710783958435, + "learning_rate": 1.1788826815642458e-05, + "loss": 0.2556, + "step": 3520 + }, + { + "epoch": 9.835195530726256, + "grad_norm": 1.6136256456375122, + "learning_rate": 1.179217877094972e-05, + "loss": 0.288, + "step": 3521 + }, + { + "epoch": 9.837988826815643, + "grad_norm": 5.010725021362305, + "learning_rate": 1.1795530726256984e-05, + "loss": 0.4133, + "step": 3522 + }, + { + "epoch": 9.840782122905027, + "grad_norm": 0.6816205382347107, + "learning_rate": 1.1798882681564245e-05, + "loss": 0.217, + "step": 3523 + }, + { + "epoch": 9.843575418994414, + "grad_norm": 0.7129043340682983, + "learning_rate": 1.1802234636871508e-05, + "loss": 0.2112, + "step": 3524 + }, + { + "epoch": 9.846368715083798, + "grad_norm": 0.4914708733558655, + "learning_rate": 1.1805586592178771e-05, + "loss": 0.2313, + "step": 3525 + }, + { + "epoch": 9.849162011173185, + "grad_norm": 0.6799536943435669, + "learning_rate": 1.1808938547486034e-05, + "loss": 0.2337, + "step": 3526 + }, + { + "epoch": 9.85195530726257, + "grad_norm": 2.226503372192383, + "learning_rate": 1.1812290502793296e-05, + "loss": 0.2624, + "step": 3527 + }, + { + "epoch": 9.854748603351956, + "grad_norm": 0.6379780769348145, + "learning_rate": 1.1815642458100559e-05, + "loss": 0.2297, + "step": 3528 + }, + { + "epoch": 9.85754189944134, + "grad_norm": 0.6831942796707153, + "learning_rate": 1.1818994413407822e-05, + "loss": 0.2156, + "step": 3529 + }, + { + "epoch": 9.860335195530727, + "grad_norm": 0.7575421929359436, + "learning_rate": 1.1822346368715083e-05, + "loss": 0.186, + "step": 3530 + }, + { + "epoch": 9.863128491620111, + "grad_norm": 0.7877178192138672, + "learning_rate": 1.1825698324022346e-05, + "loss": 0.2154, + "step": 3531 + }, + { + "epoch": 9.865921787709498, + "grad_norm": 0.6919330358505249, + "learning_rate": 1.1829050279329609e-05, + "loss": 0.2269, + "step": 3532 + }, + { + "epoch": 9.868715083798882, + "grad_norm": 0.8559656739234924, + "learning_rate": 1.183240223463687e-05, + "loss": 0.2672, + "step": 3533 + }, + { + "epoch": 9.871508379888269, + "grad_norm": 0.7664428353309631, + "learning_rate": 1.1835754189944133e-05, + "loss": 0.2372, + "step": 3534 + }, + { + "epoch": 9.874301675977653, + "grad_norm": 0.6797202229499817, + "learning_rate": 1.1839106145251396e-05, + "loss": 0.2324, + "step": 3535 + }, + { + "epoch": 9.87709497206704, + "grad_norm": 0.9140998125076294, + "learning_rate": 1.184245810055866e-05, + "loss": 0.1922, + "step": 3536 + }, + { + "epoch": 9.879888268156424, + "grad_norm": 1.355432391166687, + "learning_rate": 1.184581005586592e-05, + "loss": 0.192, + "step": 3537 + }, + { + "epoch": 9.88268156424581, + "grad_norm": 1.0151152610778809, + "learning_rate": 1.1849162011173184e-05, + "loss": 0.2579, + "step": 3538 + }, + { + "epoch": 9.885474860335195, + "grad_norm": 1.689656138420105, + "learning_rate": 1.1852513966480447e-05, + "loss": 0.2329, + "step": 3539 + }, + { + "epoch": 9.888268156424582, + "grad_norm": 0.7989908456802368, + "learning_rate": 1.185586592178771e-05, + "loss": 0.2105, + "step": 3540 + }, + { + "epoch": 9.891061452513966, + "grad_norm": 1.4534533023834229, + "learning_rate": 1.1859217877094973e-05, + "loss": 0.2659, + "step": 3541 + }, + { + "epoch": 9.893854748603353, + "grad_norm": 1.0916996002197266, + "learning_rate": 1.1862569832402236e-05, + "loss": 0.2525, + "step": 3542 + }, + { + "epoch": 9.896648044692737, + "grad_norm": 0.9850709438323975, + "learning_rate": 1.1865921787709499e-05, + "loss": 0.255, + "step": 3543 + }, + { + "epoch": 9.899441340782122, + "grad_norm": 1.3158437013626099, + "learning_rate": 1.186927374301676e-05, + "loss": 0.2526, + "step": 3544 + }, + { + "epoch": 9.902234636871508, + "grad_norm": 1.8463704586029053, + "learning_rate": 1.1872625698324023e-05, + "loss": 0.2637, + "step": 3545 + }, + { + "epoch": 9.905027932960895, + "grad_norm": 2.9256174564361572, + "learning_rate": 1.1875977653631286e-05, + "loss": 0.2984, + "step": 3546 + }, + { + "epoch": 9.90782122905028, + "grad_norm": 2.1065666675567627, + "learning_rate": 1.1879329608938548e-05, + "loss": 0.3117, + "step": 3547 + }, + { + "epoch": 9.910614525139664, + "grad_norm": 1.1027003526687622, + "learning_rate": 1.1882681564245811e-05, + "loss": 0.2626, + "step": 3548 + }, + { + "epoch": 9.91340782122905, + "grad_norm": 0.48616039752960205, + "learning_rate": 1.1886033519553074e-05, + "loss": 0.2304, + "step": 3549 + }, + { + "epoch": 9.916201117318435, + "grad_norm": 0.7399222254753113, + "learning_rate": 1.1889385474860335e-05, + "loss": 0.2374, + "step": 3550 + }, + { + "epoch": 9.918994413407821, + "grad_norm": 0.8886821866035461, + "learning_rate": 1.1892737430167598e-05, + "loss": 0.2181, + "step": 3551 + }, + { + "epoch": 9.921787709497206, + "grad_norm": 0.5305331349372864, + "learning_rate": 1.1896089385474861e-05, + "loss": 0.2199, + "step": 3552 + }, + { + "epoch": 9.924581005586592, + "grad_norm": 0.5852490663528442, + "learning_rate": 1.1899441340782124e-05, + "loss": 0.2119, + "step": 3553 + }, + { + "epoch": 9.927374301675977, + "grad_norm": 1.9170349836349487, + "learning_rate": 1.1902793296089386e-05, + "loss": 0.236, + "step": 3554 + }, + { + "epoch": 9.930167597765363, + "grad_norm": 0.8642455339431763, + "learning_rate": 1.1906145251396649e-05, + "loss": 0.211, + "step": 3555 + }, + { + "epoch": 9.932960893854748, + "grad_norm": 0.8649130463600159, + "learning_rate": 1.1909497206703912e-05, + "loss": 0.1994, + "step": 3556 + }, + { + "epoch": 9.935754189944134, + "grad_norm": 1.0535130500793457, + "learning_rate": 1.1912849162011173e-05, + "loss": 0.2747, + "step": 3557 + }, + { + "epoch": 9.938547486033519, + "grad_norm": 1.8255743980407715, + "learning_rate": 1.1916201117318436e-05, + "loss": 0.1924, + "step": 3558 + }, + { + "epoch": 9.941340782122905, + "grad_norm": 0.8682335615158081, + "learning_rate": 1.19195530726257e-05, + "loss": 0.1943, + "step": 3559 + }, + { + "epoch": 9.94413407821229, + "grad_norm": 0.763672947883606, + "learning_rate": 1.192290502793296e-05, + "loss": 0.2193, + "step": 3560 + }, + { + "epoch": 9.946927374301676, + "grad_norm": 0.7429617047309875, + "learning_rate": 1.1926256983240224e-05, + "loss": 0.2088, + "step": 3561 + }, + { + "epoch": 9.949720670391061, + "grad_norm": 0.8111502528190613, + "learning_rate": 1.1929608938547487e-05, + "loss": 0.21, + "step": 3562 + }, + { + "epoch": 9.952513966480447, + "grad_norm": 0.8577755093574524, + "learning_rate": 1.193296089385475e-05, + "loss": 0.2215, + "step": 3563 + }, + { + "epoch": 9.955307262569832, + "grad_norm": 1.0656200647354126, + "learning_rate": 1.1936312849162011e-05, + "loss": 0.2311, + "step": 3564 + }, + { + "epoch": 9.958100558659218, + "grad_norm": 1.2147340774536133, + "learning_rate": 1.1939664804469274e-05, + "loss": 0.2539, + "step": 3565 + }, + { + "epoch": 9.960893854748603, + "grad_norm": 0.7622268795967102, + "learning_rate": 1.1943016759776537e-05, + "loss": 0.2036, + "step": 3566 + }, + { + "epoch": 9.96368715083799, + "grad_norm": 1.1948163509368896, + "learning_rate": 1.1946368715083798e-05, + "loss": 0.2175, + "step": 3567 + }, + { + "epoch": 9.966480446927374, + "grad_norm": 0.8999443650245667, + "learning_rate": 1.1949720670391061e-05, + "loss": 0.1932, + "step": 3568 + }, + { + "epoch": 9.96927374301676, + "grad_norm": 1.1571476459503174, + "learning_rate": 1.1953072625698325e-05, + "loss": 0.197, + "step": 3569 + }, + { + "epoch": 9.972067039106145, + "grad_norm": 1.4269123077392578, + "learning_rate": 1.1956424581005586e-05, + "loss": 0.2471, + "step": 3570 + }, + { + "epoch": 9.974860335195531, + "grad_norm": 1.9069164991378784, + "learning_rate": 1.1959776536312849e-05, + "loss": 0.2444, + "step": 3571 + }, + { + "epoch": 9.977653631284916, + "grad_norm": 2.3827695846557617, + "learning_rate": 1.1963128491620112e-05, + "loss": 0.3491, + "step": 3572 + }, + { + "epoch": 9.980446927374302, + "grad_norm": 0.6851401329040527, + "learning_rate": 1.1966480446927375e-05, + "loss": 0.1986, + "step": 3573 + }, + { + "epoch": 9.983240223463687, + "grad_norm": 0.9285342693328857, + "learning_rate": 1.1969832402234636e-05, + "loss": 0.2363, + "step": 3574 + }, + { + "epoch": 9.986033519553073, + "grad_norm": 0.9993571639060974, + "learning_rate": 1.19731843575419e-05, + "loss": 0.2142, + "step": 3575 + }, + { + "epoch": 9.988826815642458, + "grad_norm": 0.8438435196876526, + "learning_rate": 1.1976536312849162e-05, + "loss": 0.219, + "step": 3576 + }, + { + "epoch": 9.991620111731844, + "grad_norm": 1.5742393732070923, + "learning_rate": 1.1979888268156424e-05, + "loss": 0.2754, + "step": 3577 + }, + { + "epoch": 9.994413407821229, + "grad_norm": 1.0437345504760742, + "learning_rate": 1.1983240223463687e-05, + "loss": 0.2485, + "step": 3578 + }, + { + "epoch": 9.997206703910614, + "grad_norm": 0.9481368660926819, + "learning_rate": 1.198659217877095e-05, + "loss": 0.2325, + "step": 3579 + }, + { + "epoch": 10.0, + "grad_norm": 1.3606388568878174, + "learning_rate": 1.1989944134078211e-05, + "loss": 0.2904, + "step": 3580 + }, + { + "epoch": 10.002793296089385, + "grad_norm": 0.7777466773986816, + "learning_rate": 1.1993296089385474e-05, + "loss": 0.235, + "step": 3581 + }, + { + "epoch": 10.005586592178771, + "grad_norm": 0.48575329780578613, + "learning_rate": 1.1996648044692737e-05, + "loss": 0.2144, + "step": 3582 + }, + { + "epoch": 10.008379888268156, + "grad_norm": 0.49576935172080994, + "learning_rate": 1.2e-05, + "loss": 0.2273, + "step": 3583 + }, + { + "epoch": 10.011173184357542, + "grad_norm": 0.5119617581367493, + "learning_rate": 1.2003351955307262e-05, + "loss": 0.2397, + "step": 3584 + }, + { + "epoch": 10.013966480446927, + "grad_norm": 0.5575078725814819, + "learning_rate": 1.2006703910614525e-05, + "loss": 0.1779, + "step": 3585 + }, + { + "epoch": 10.016759776536313, + "grad_norm": 0.6276967525482178, + "learning_rate": 1.2010055865921788e-05, + "loss": 0.2182, + "step": 3586 + }, + { + "epoch": 10.019553072625698, + "grad_norm": 0.7123677730560303, + "learning_rate": 1.2013407821229049e-05, + "loss": 0.2044, + "step": 3587 + }, + { + "epoch": 10.022346368715084, + "grad_norm": 1.0012688636779785, + "learning_rate": 1.2016759776536314e-05, + "loss": 0.2407, + "step": 3588 + }, + { + "epoch": 10.025139664804469, + "grad_norm": 0.6957440972328186, + "learning_rate": 1.2020111731843577e-05, + "loss": 0.1997, + "step": 3589 + }, + { + "epoch": 10.027932960893855, + "grad_norm": 1.0610817670822144, + "learning_rate": 1.202346368715084e-05, + "loss": 0.2034, + "step": 3590 + }, + { + "epoch": 10.03072625698324, + "grad_norm": 0.800797164440155, + "learning_rate": 1.2026815642458101e-05, + "loss": 0.2034, + "step": 3591 + }, + { + "epoch": 10.033519553072626, + "grad_norm": 0.589097261428833, + "learning_rate": 1.2030167597765364e-05, + "loss": 0.1957, + "step": 3592 + }, + { + "epoch": 10.03631284916201, + "grad_norm": 0.7589951157569885, + "learning_rate": 1.2033519553072627e-05, + "loss": 0.2191, + "step": 3593 + }, + { + "epoch": 10.039106145251397, + "grad_norm": 1.0691592693328857, + "learning_rate": 1.2036871508379889e-05, + "loss": 0.217, + "step": 3594 + }, + { + "epoch": 10.041899441340782, + "grad_norm": 1.9572279453277588, + "learning_rate": 1.2040223463687152e-05, + "loss": 0.1862, + "step": 3595 + }, + { + "epoch": 10.044692737430168, + "grad_norm": 0.7929636836051941, + "learning_rate": 1.2043575418994415e-05, + "loss": 0.1826, + "step": 3596 + }, + { + "epoch": 10.047486033519553, + "grad_norm": 1.1244556903839111, + "learning_rate": 1.2046927374301676e-05, + "loss": 0.2184, + "step": 3597 + }, + { + "epoch": 10.050279329608939, + "grad_norm": 0.9861982464790344, + "learning_rate": 1.2050279329608939e-05, + "loss": 0.2114, + "step": 3598 + }, + { + "epoch": 10.053072625698324, + "grad_norm": 1.6393800973892212, + "learning_rate": 1.2053631284916202e-05, + "loss": 0.2101, + "step": 3599 + }, + { + "epoch": 10.05586592178771, + "grad_norm": 1.0223537683486938, + "learning_rate": 1.2056983240223465e-05, + "loss": 0.2157, + "step": 3600 + }, + { + "epoch": 10.058659217877095, + "grad_norm": 1.7443739175796509, + "learning_rate": 1.2060335195530727e-05, + "loss": 0.2663, + "step": 3601 + }, + { + "epoch": 10.061452513966481, + "grad_norm": 1.3903300762176514, + "learning_rate": 1.206368715083799e-05, + "loss": 0.2479, + "step": 3602 + }, + { + "epoch": 10.064245810055866, + "grad_norm": 1.0691038370132446, + "learning_rate": 1.2067039106145253e-05, + "loss": 0.2009, + "step": 3603 + }, + { + "epoch": 10.067039106145252, + "grad_norm": 3.0505518913269043, + "learning_rate": 1.2070391061452514e-05, + "loss": 0.2816, + "step": 3604 + }, + { + "epoch": 10.069832402234637, + "grad_norm": 2.3637771606445312, + "learning_rate": 1.2073743016759777e-05, + "loss": 0.2979, + "step": 3605 + }, + { + "epoch": 10.072625698324023, + "grad_norm": 0.5571272969245911, + "learning_rate": 1.207709497206704e-05, + "loss": 0.2046, + "step": 3606 + }, + { + "epoch": 10.075418994413408, + "grad_norm": 0.6100501418113708, + "learning_rate": 1.2080446927374301e-05, + "loss": 0.1957, + "step": 3607 + }, + { + "epoch": 10.078212290502794, + "grad_norm": 0.5556634664535522, + "learning_rate": 1.2083798882681564e-05, + "loss": 0.1807, + "step": 3608 + }, + { + "epoch": 10.081005586592179, + "grad_norm": 0.7523097991943359, + "learning_rate": 1.2087150837988827e-05, + "loss": 0.248, + "step": 3609 + }, + { + "epoch": 10.083798882681565, + "grad_norm": 0.7464683055877686, + "learning_rate": 1.209050279329609e-05, + "loss": 0.1891, + "step": 3610 + }, + { + "epoch": 10.08659217877095, + "grad_norm": 0.5889096260070801, + "learning_rate": 1.2093854748603352e-05, + "loss": 0.2025, + "step": 3611 + }, + { + "epoch": 10.089385474860336, + "grad_norm": 0.8088960647583008, + "learning_rate": 1.2097206703910615e-05, + "loss": 0.2259, + "step": 3612 + }, + { + "epoch": 10.09217877094972, + "grad_norm": 0.6622597575187683, + "learning_rate": 1.2100558659217878e-05, + "loss": 0.1921, + "step": 3613 + }, + { + "epoch": 10.094972067039107, + "grad_norm": 0.8501612544059753, + "learning_rate": 1.210391061452514e-05, + "loss": 0.2041, + "step": 3614 + }, + { + "epoch": 10.097765363128492, + "grad_norm": 0.9076066613197327, + "learning_rate": 1.2107262569832402e-05, + "loss": 0.2529, + "step": 3615 + }, + { + "epoch": 10.100558659217878, + "grad_norm": 0.6645806431770325, + "learning_rate": 1.2110614525139665e-05, + "loss": 0.2188, + "step": 3616 + }, + { + "epoch": 10.103351955307263, + "grad_norm": 0.755318820476532, + "learning_rate": 1.2113966480446927e-05, + "loss": 0.2096, + "step": 3617 + }, + { + "epoch": 10.106145251396647, + "grad_norm": 0.9326803088188171, + "learning_rate": 1.211731843575419e-05, + "loss": 0.1983, + "step": 3618 + }, + { + "epoch": 10.108938547486034, + "grad_norm": 0.8297111392021179, + "learning_rate": 1.2120670391061453e-05, + "loss": 0.225, + "step": 3619 + }, + { + "epoch": 10.111731843575418, + "grad_norm": 0.735913872718811, + "learning_rate": 1.2124022346368716e-05, + "loss": 0.2017, + "step": 3620 + }, + { + "epoch": 10.114525139664805, + "grad_norm": 1.2787997722625732, + "learning_rate": 1.2127374301675977e-05, + "loss": 0.238, + "step": 3621 + }, + { + "epoch": 10.11731843575419, + "grad_norm": 1.0701167583465576, + "learning_rate": 1.213072625698324e-05, + "loss": 0.201, + "step": 3622 + }, + { + "epoch": 10.120111731843576, + "grad_norm": 0.8525843620300293, + "learning_rate": 1.2134078212290503e-05, + "loss": 0.207, + "step": 3623 + }, + { + "epoch": 10.12290502793296, + "grad_norm": 0.9682866930961609, + "learning_rate": 1.2137430167597765e-05, + "loss": 0.2311, + "step": 3624 + }, + { + "epoch": 10.125698324022347, + "grad_norm": 1.2857869863510132, + "learning_rate": 1.2140782122905028e-05, + "loss": 0.2119, + "step": 3625 + }, + { + "epoch": 10.128491620111731, + "grad_norm": 1.0836164951324463, + "learning_rate": 1.214413407821229e-05, + "loss": 0.2097, + "step": 3626 + }, + { + "epoch": 10.131284916201118, + "grad_norm": 1.0256963968276978, + "learning_rate": 1.2147486033519552e-05, + "loss": 0.2198, + "step": 3627 + }, + { + "epoch": 10.134078212290502, + "grad_norm": 2.0696473121643066, + "learning_rate": 1.2150837988826815e-05, + "loss": 0.2366, + "step": 3628 + }, + { + "epoch": 10.136871508379889, + "grad_norm": 1.701114296913147, + "learning_rate": 1.2154189944134078e-05, + "loss": 0.2405, + "step": 3629 + }, + { + "epoch": 10.139664804469273, + "grad_norm": 2.8006725311279297, + "learning_rate": 1.2157541899441341e-05, + "loss": 0.3697, + "step": 3630 + }, + { + "epoch": 10.14245810055866, + "grad_norm": 0.44499894976615906, + "learning_rate": 1.2160893854748602e-05, + "loss": 0.238, + "step": 3631 + }, + { + "epoch": 10.145251396648044, + "grad_norm": 0.4458119869232178, + "learning_rate": 1.2164245810055866e-05, + "loss": 0.1854, + "step": 3632 + }, + { + "epoch": 10.14804469273743, + "grad_norm": 0.5522883534431458, + "learning_rate": 1.2167597765363129e-05, + "loss": 0.2356, + "step": 3633 + }, + { + "epoch": 10.150837988826815, + "grad_norm": 0.7505648136138916, + "learning_rate": 1.217094972067039e-05, + "loss": 0.2116, + "step": 3634 + }, + { + "epoch": 10.153631284916202, + "grad_norm": 0.9219403266906738, + "learning_rate": 1.2174301675977653e-05, + "loss": 0.2459, + "step": 3635 + }, + { + "epoch": 10.156424581005586, + "grad_norm": 0.6064212322235107, + "learning_rate": 1.2177653631284916e-05, + "loss": 0.2202, + "step": 3636 + }, + { + "epoch": 10.159217877094973, + "grad_norm": 0.6955708265304565, + "learning_rate": 1.218100558659218e-05, + "loss": 0.1874, + "step": 3637 + }, + { + "epoch": 10.162011173184357, + "grad_norm": 0.5624136328697205, + "learning_rate": 1.2184357541899442e-05, + "loss": 0.1915, + "step": 3638 + }, + { + "epoch": 10.164804469273744, + "grad_norm": 0.6631973385810852, + "learning_rate": 1.2187709497206705e-05, + "loss": 0.1937, + "step": 3639 + }, + { + "epoch": 10.167597765363128, + "grad_norm": 0.5934213995933533, + "learning_rate": 1.2191061452513968e-05, + "loss": 0.2031, + "step": 3640 + }, + { + "epoch": 10.170391061452515, + "grad_norm": 0.6585385203361511, + "learning_rate": 1.219441340782123e-05, + "loss": 0.242, + "step": 3641 + }, + { + "epoch": 10.1731843575419, + "grad_norm": 1.2266507148742676, + "learning_rate": 1.2197765363128493e-05, + "loss": 0.2112, + "step": 3642 + }, + { + "epoch": 10.175977653631286, + "grad_norm": 0.8374020457267761, + "learning_rate": 1.2201117318435756e-05, + "loss": 0.1915, + "step": 3643 + }, + { + "epoch": 10.17877094972067, + "grad_norm": 1.0336438417434692, + "learning_rate": 1.2204469273743017e-05, + "loss": 0.1847, + "step": 3644 + }, + { + "epoch": 10.181564245810057, + "grad_norm": 0.8092039227485657, + "learning_rate": 1.220782122905028e-05, + "loss": 0.1739, + "step": 3645 + }, + { + "epoch": 10.184357541899441, + "grad_norm": 0.7280057072639465, + "learning_rate": 1.2211173184357543e-05, + "loss": 0.1872, + "step": 3646 + }, + { + "epoch": 10.187150837988828, + "grad_norm": 0.9272576570510864, + "learning_rate": 1.2214525139664806e-05, + "loss": 0.2006, + "step": 3647 + }, + { + "epoch": 10.189944134078212, + "grad_norm": 0.954886794090271, + "learning_rate": 1.2217877094972067e-05, + "loss": 0.2389, + "step": 3648 + }, + { + "epoch": 10.192737430167599, + "grad_norm": 2.4168903827667236, + "learning_rate": 1.222122905027933e-05, + "loss": 0.214, + "step": 3649 + }, + { + "epoch": 10.195530726256983, + "grad_norm": 1.407798409461975, + "learning_rate": 1.2224581005586593e-05, + "loss": 0.2335, + "step": 3650 + }, + { + "epoch": 10.19832402234637, + "grad_norm": 1.1671501398086548, + "learning_rate": 1.2227932960893855e-05, + "loss": 0.2483, + "step": 3651 + }, + { + "epoch": 10.201117318435754, + "grad_norm": 2.585448741912842, + "learning_rate": 1.2231284916201118e-05, + "loss": 0.2193, + "step": 3652 + }, + { + "epoch": 10.203910614525139, + "grad_norm": 1.1948705911636353, + "learning_rate": 1.2234636871508381e-05, + "loss": 0.2005, + "step": 3653 + }, + { + "epoch": 10.206703910614525, + "grad_norm": 2.1015000343322754, + "learning_rate": 1.2237988826815642e-05, + "loss": 0.2534, + "step": 3654 + }, + { + "epoch": 10.20949720670391, + "grad_norm": 2.96916127204895, + "learning_rate": 1.2241340782122905e-05, + "loss": 0.3502, + "step": 3655 + }, + { + "epoch": 10.212290502793296, + "grad_norm": 0.5878201127052307, + "learning_rate": 1.2244692737430168e-05, + "loss": 0.237, + "step": 3656 + }, + { + "epoch": 10.21508379888268, + "grad_norm": 0.5462366342544556, + "learning_rate": 1.2248044692737431e-05, + "loss": 0.2388, + "step": 3657 + }, + { + "epoch": 10.217877094972067, + "grad_norm": 0.5150467157363892, + "learning_rate": 1.2251396648044693e-05, + "loss": 0.2377, + "step": 3658 + }, + { + "epoch": 10.220670391061452, + "grad_norm": 0.6019988059997559, + "learning_rate": 1.2254748603351956e-05, + "loss": 0.2225, + "step": 3659 + }, + { + "epoch": 10.223463687150838, + "grad_norm": 0.5793697237968445, + "learning_rate": 1.2258100558659219e-05, + "loss": 0.2229, + "step": 3660 + }, + { + "epoch": 10.226256983240223, + "grad_norm": 0.5953251719474792, + "learning_rate": 1.226145251396648e-05, + "loss": 0.2092, + "step": 3661 + }, + { + "epoch": 10.22905027932961, + "grad_norm": 0.693217396736145, + "learning_rate": 1.2264804469273743e-05, + "loss": 0.2477, + "step": 3662 + }, + { + "epoch": 10.231843575418994, + "grad_norm": 0.6313478946685791, + "learning_rate": 1.2268156424581006e-05, + "loss": 0.2074, + "step": 3663 + }, + { + "epoch": 10.23463687150838, + "grad_norm": 0.6355196833610535, + "learning_rate": 1.2271508379888268e-05, + "loss": 0.1802, + "step": 3664 + }, + { + "epoch": 10.237430167597765, + "grad_norm": 0.6921277642250061, + "learning_rate": 1.227486033519553e-05, + "loss": 0.2002, + "step": 3665 + }, + { + "epoch": 10.240223463687151, + "grad_norm": 0.9600462317466736, + "learning_rate": 1.2278212290502794e-05, + "loss": 0.1949, + "step": 3666 + }, + { + "epoch": 10.243016759776536, + "grad_norm": 0.8364211320877075, + "learning_rate": 1.2281564245810057e-05, + "loss": 0.2003, + "step": 3667 + }, + { + "epoch": 10.245810055865922, + "grad_norm": 0.7124790549278259, + "learning_rate": 1.2284916201117318e-05, + "loss": 0.1736, + "step": 3668 + }, + { + "epoch": 10.248603351955307, + "grad_norm": 0.811384379863739, + "learning_rate": 1.2288268156424581e-05, + "loss": 0.2332, + "step": 3669 + }, + { + "epoch": 10.251396648044693, + "grad_norm": 0.645215630531311, + "learning_rate": 1.2291620111731844e-05, + "loss": 0.2071, + "step": 3670 + }, + { + "epoch": 10.254189944134078, + "grad_norm": 1.8298753499984741, + "learning_rate": 1.2294972067039105e-05, + "loss": 0.2062, + "step": 3671 + }, + { + "epoch": 10.256983240223464, + "grad_norm": 1.0828630924224854, + "learning_rate": 1.2298324022346368e-05, + "loss": 0.2253, + "step": 3672 + }, + { + "epoch": 10.259776536312849, + "grad_norm": 1.1966016292572021, + "learning_rate": 1.2301675977653631e-05, + "loss": 0.2185, + "step": 3673 + }, + { + "epoch": 10.262569832402235, + "grad_norm": 1.0650032758712769, + "learning_rate": 1.2305027932960893e-05, + "loss": 0.2174, + "step": 3674 + }, + { + "epoch": 10.26536312849162, + "grad_norm": 0.7727355360984802, + "learning_rate": 1.2308379888268156e-05, + "loss": 0.1981, + "step": 3675 + }, + { + "epoch": 10.268156424581006, + "grad_norm": 1.0585269927978516, + "learning_rate": 1.2311731843575419e-05, + "loss": 0.2285, + "step": 3676 + }, + { + "epoch": 10.27094972067039, + "grad_norm": 1.1806613206863403, + "learning_rate": 1.2315083798882682e-05, + "loss": 0.2892, + "step": 3677 + }, + { + "epoch": 10.273743016759777, + "grad_norm": 1.6302803754806519, + "learning_rate": 1.2318435754189943e-05, + "loss": 0.2494, + "step": 3678 + }, + { + "epoch": 10.276536312849162, + "grad_norm": 1.7607359886169434, + "learning_rate": 1.2321787709497206e-05, + "loss": 0.2672, + "step": 3679 + }, + { + "epoch": 10.279329608938548, + "grad_norm": 2.016079902648926, + "learning_rate": 1.232513966480447e-05, + "loss": 0.3084, + "step": 3680 + }, + { + "epoch": 10.282122905027933, + "grad_norm": 0.8894931077957153, + "learning_rate": 1.232849162011173e-05, + "loss": 0.2247, + "step": 3681 + }, + { + "epoch": 10.28491620111732, + "grad_norm": 0.8840950131416321, + "learning_rate": 1.2331843575418994e-05, + "loss": 0.212, + "step": 3682 + }, + { + "epoch": 10.287709497206704, + "grad_norm": 0.554338276386261, + "learning_rate": 1.2335195530726257e-05, + "loss": 0.2219, + "step": 3683 + }, + { + "epoch": 10.29050279329609, + "grad_norm": 0.7130484580993652, + "learning_rate": 1.233854748603352e-05, + "loss": 0.1958, + "step": 3684 + }, + { + "epoch": 10.293296089385475, + "grad_norm": 0.8836376667022705, + "learning_rate": 1.2341899441340783e-05, + "loss": 0.2181, + "step": 3685 + }, + { + "epoch": 10.296089385474861, + "grad_norm": 0.8683385848999023, + "learning_rate": 1.2345251396648046e-05, + "loss": 0.2347, + "step": 3686 + }, + { + "epoch": 10.298882681564246, + "grad_norm": 0.9902687668800354, + "learning_rate": 1.2348603351955309e-05, + "loss": 0.2522, + "step": 3687 + }, + { + "epoch": 10.30167597765363, + "grad_norm": 0.7130879163742065, + "learning_rate": 1.235195530726257e-05, + "loss": 0.2048, + "step": 3688 + }, + { + "epoch": 10.304469273743017, + "grad_norm": 1.1900534629821777, + "learning_rate": 1.2355307262569833e-05, + "loss": 0.1987, + "step": 3689 + }, + { + "epoch": 10.307262569832401, + "grad_norm": 0.5822724103927612, + "learning_rate": 1.2358659217877096e-05, + "loss": 0.2288, + "step": 3690 + }, + { + "epoch": 10.310055865921788, + "grad_norm": 0.9549341797828674, + "learning_rate": 1.2362011173184358e-05, + "loss": 0.2178, + "step": 3691 + }, + { + "epoch": 10.312849162011172, + "grad_norm": 0.8303830623626709, + "learning_rate": 1.236536312849162e-05, + "loss": 0.2342, + "step": 3692 + }, + { + "epoch": 10.315642458100559, + "grad_norm": 0.6857410669326782, + "learning_rate": 1.2368715083798884e-05, + "loss": 0.1942, + "step": 3693 + }, + { + "epoch": 10.318435754189943, + "grad_norm": 1.758040189743042, + "learning_rate": 1.2372067039106147e-05, + "loss": 0.2783, + "step": 3694 + }, + { + "epoch": 10.32122905027933, + "grad_norm": 1.0044769048690796, + "learning_rate": 1.2375418994413408e-05, + "loss": 0.2146, + "step": 3695 + }, + { + "epoch": 10.324022346368714, + "grad_norm": 0.8128578662872314, + "learning_rate": 1.2378770949720671e-05, + "loss": 0.2073, + "step": 3696 + }, + { + "epoch": 10.3268156424581, + "grad_norm": 1.1807305812835693, + "learning_rate": 1.2382122905027934e-05, + "loss": 0.2064, + "step": 3697 + }, + { + "epoch": 10.329608938547485, + "grad_norm": 1.7778369188308716, + "learning_rate": 1.2385474860335196e-05, + "loss": 0.1998, + "step": 3698 + }, + { + "epoch": 10.332402234636872, + "grad_norm": 1.7764147520065308, + "learning_rate": 1.2388826815642459e-05, + "loss": 0.2336, + "step": 3699 + }, + { + "epoch": 10.335195530726256, + "grad_norm": 0.9215977787971497, + "learning_rate": 1.2392178770949722e-05, + "loss": 0.2433, + "step": 3700 + }, + { + "epoch": 10.337988826815643, + "grad_norm": 1.0054335594177246, + "learning_rate": 1.2395530726256983e-05, + "loss": 0.2216, + "step": 3701 + }, + { + "epoch": 10.340782122905027, + "grad_norm": 1.5006905794143677, + "learning_rate": 1.2398882681564246e-05, + "loss": 0.2583, + "step": 3702 + }, + { + "epoch": 10.343575418994414, + "grad_norm": 1.7023459672927856, + "learning_rate": 1.2402234636871509e-05, + "loss": 0.2652, + "step": 3703 + }, + { + "epoch": 10.346368715083798, + "grad_norm": 1.6747102737426758, + "learning_rate": 1.2405586592178772e-05, + "loss": 0.34, + "step": 3704 + }, + { + "epoch": 10.349162011173185, + "grad_norm": 2.4625370502471924, + "learning_rate": 1.2408938547486034e-05, + "loss": 0.3317, + "step": 3705 + }, + { + "epoch": 10.35195530726257, + "grad_norm": 0.8724327683448792, + "learning_rate": 1.2412290502793297e-05, + "loss": 0.2235, + "step": 3706 + }, + { + "epoch": 10.354748603351956, + "grad_norm": 0.7962023019790649, + "learning_rate": 1.241564245810056e-05, + "loss": 0.2542, + "step": 3707 + }, + { + "epoch": 10.35754189944134, + "grad_norm": 0.6447317004203796, + "learning_rate": 1.2418994413407821e-05, + "loss": 0.2141, + "step": 3708 + }, + { + "epoch": 10.360335195530727, + "grad_norm": 0.6427066326141357, + "learning_rate": 1.2422346368715084e-05, + "loss": 0.1991, + "step": 3709 + }, + { + "epoch": 10.363128491620111, + "grad_norm": 0.7533539533615112, + "learning_rate": 1.2425698324022347e-05, + "loss": 0.2117, + "step": 3710 + }, + { + "epoch": 10.365921787709498, + "grad_norm": 0.7507545948028564, + "learning_rate": 1.2429050279329608e-05, + "loss": 0.2002, + "step": 3711 + }, + { + "epoch": 10.368715083798882, + "grad_norm": 0.7759698629379272, + "learning_rate": 1.2432402234636871e-05, + "loss": 0.1841, + "step": 3712 + }, + { + "epoch": 10.371508379888269, + "grad_norm": 0.9699993133544922, + "learning_rate": 1.2435754189944134e-05, + "loss": 0.227, + "step": 3713 + }, + { + "epoch": 10.374301675977653, + "grad_norm": 1.044074535369873, + "learning_rate": 1.2439106145251397e-05, + "loss": 0.2316, + "step": 3714 + }, + { + "epoch": 10.37709497206704, + "grad_norm": 1.1940120458602905, + "learning_rate": 1.2442458100558659e-05, + "loss": 0.182, + "step": 3715 + }, + { + "epoch": 10.379888268156424, + "grad_norm": 0.6531934142112732, + "learning_rate": 1.2445810055865922e-05, + "loss": 0.2349, + "step": 3716 + }, + { + "epoch": 10.38268156424581, + "grad_norm": 0.6834475994110107, + "learning_rate": 1.2449162011173185e-05, + "loss": 0.2327, + "step": 3717 + }, + { + "epoch": 10.385474860335195, + "grad_norm": 0.7824273109436035, + "learning_rate": 1.2452513966480446e-05, + "loss": 0.2272, + "step": 3718 + }, + { + "epoch": 10.388268156424582, + "grad_norm": 1.2224855422973633, + "learning_rate": 1.245586592178771e-05, + "loss": 0.2283, + "step": 3719 + }, + { + "epoch": 10.391061452513966, + "grad_norm": 0.8626258373260498, + "learning_rate": 1.2459217877094972e-05, + "loss": 0.2173, + "step": 3720 + }, + { + "epoch": 10.393854748603353, + "grad_norm": 1.1236425638198853, + "learning_rate": 1.2462569832402234e-05, + "loss": 0.218, + "step": 3721 + }, + { + "epoch": 10.396648044692737, + "grad_norm": 1.7763522863388062, + "learning_rate": 1.2465921787709497e-05, + "loss": 0.2488, + "step": 3722 + }, + { + "epoch": 10.399441340782122, + "grad_norm": 0.9227075576782227, + "learning_rate": 1.246927374301676e-05, + "loss": 0.1795, + "step": 3723 + }, + { + "epoch": 10.402234636871508, + "grad_norm": 0.9385399222373962, + "learning_rate": 1.2472625698324023e-05, + "loss": 0.2004, + "step": 3724 + }, + { + "epoch": 10.405027932960893, + "grad_norm": 1.5372470617294312, + "learning_rate": 1.2475977653631284e-05, + "loss": 0.2369, + "step": 3725 + }, + { + "epoch": 10.40782122905028, + "grad_norm": 1.6991891860961914, + "learning_rate": 1.2479329608938547e-05, + "loss": 0.2215, + "step": 3726 + }, + { + "epoch": 10.410614525139664, + "grad_norm": 1.1443939208984375, + "learning_rate": 1.248268156424581e-05, + "loss": 0.2285, + "step": 3727 + }, + { + "epoch": 10.41340782122905, + "grad_norm": 1.8636822700500488, + "learning_rate": 1.2486033519553072e-05, + "loss": 0.2391, + "step": 3728 + }, + { + "epoch": 10.416201117318435, + "grad_norm": 3.813899278640747, + "learning_rate": 1.2489385474860335e-05, + "loss": 0.2462, + "step": 3729 + }, + { + "epoch": 10.418994413407821, + "grad_norm": 2.8191516399383545, + "learning_rate": 1.2492737430167598e-05, + "loss": 0.3934, + "step": 3730 + }, + { + "epoch": 10.421787709497206, + "grad_norm": 0.6679432988166809, + "learning_rate": 1.249608938547486e-05, + "loss": 0.2249, + "step": 3731 + }, + { + "epoch": 10.424581005586592, + "grad_norm": 0.7160091996192932, + "learning_rate": 1.2499441340782124e-05, + "loss": 0.1816, + "step": 3732 + }, + { + "epoch": 10.427374301675977, + "grad_norm": 0.4947071075439453, + "learning_rate": 1.2502793296089387e-05, + "loss": 0.1869, + "step": 3733 + }, + { + "epoch": 10.430167597765363, + "grad_norm": 0.6507582068443298, + "learning_rate": 1.250614525139665e-05, + "loss": 0.2252, + "step": 3734 + }, + { + "epoch": 10.432960893854748, + "grad_norm": 1.470965027809143, + "learning_rate": 1.2509497206703911e-05, + "loss": 0.2525, + "step": 3735 + }, + { + "epoch": 10.435754189944134, + "grad_norm": 0.6355251669883728, + "learning_rate": 1.2512849162011174e-05, + "loss": 0.1995, + "step": 3736 + }, + { + "epoch": 10.438547486033519, + "grad_norm": 0.6568571925163269, + "learning_rate": 1.2516201117318437e-05, + "loss": 0.1927, + "step": 3737 + }, + { + "epoch": 10.441340782122905, + "grad_norm": 0.6059680581092834, + "learning_rate": 1.2519553072625699e-05, + "loss": 0.2082, + "step": 3738 + }, + { + "epoch": 10.44413407821229, + "grad_norm": 0.8141254186630249, + "learning_rate": 1.2522905027932962e-05, + "loss": 0.1853, + "step": 3739 + }, + { + "epoch": 10.446927374301676, + "grad_norm": 0.6982534527778625, + "learning_rate": 1.2526256983240225e-05, + "loss": 0.2186, + "step": 3740 + }, + { + "epoch": 10.449720670391061, + "grad_norm": 0.7170388102531433, + "learning_rate": 1.2529608938547488e-05, + "loss": 0.2081, + "step": 3741 + }, + { + "epoch": 10.452513966480447, + "grad_norm": 0.6758825778961182, + "learning_rate": 1.2532960893854749e-05, + "loss": 0.1781, + "step": 3742 + }, + { + "epoch": 10.455307262569832, + "grad_norm": 0.6830357313156128, + "learning_rate": 1.2536312849162012e-05, + "loss": 0.2302, + "step": 3743 + }, + { + "epoch": 10.458100558659218, + "grad_norm": 0.8822519779205322, + "learning_rate": 1.2539664804469275e-05, + "loss": 0.2044, + "step": 3744 + }, + { + "epoch": 10.460893854748603, + "grad_norm": 0.7208335995674133, + "learning_rate": 1.2543016759776536e-05, + "loss": 0.2195, + "step": 3745 + }, + { + "epoch": 10.46368715083799, + "grad_norm": 1.9983065128326416, + "learning_rate": 1.25463687150838e-05, + "loss": 0.2225, + "step": 3746 + }, + { + "epoch": 10.466480446927374, + "grad_norm": 0.9981288909912109, + "learning_rate": 1.2549720670391063e-05, + "loss": 0.2411, + "step": 3747 + }, + { + "epoch": 10.46927374301676, + "grad_norm": 1.0403460264205933, + "learning_rate": 1.2553072625698324e-05, + "loss": 0.2074, + "step": 3748 + }, + { + "epoch": 10.472067039106145, + "grad_norm": 1.060196876525879, + "learning_rate": 1.2556424581005587e-05, + "loss": 0.1992, + "step": 3749 + }, + { + "epoch": 10.474860335195531, + "grad_norm": 1.7426940202713013, + "learning_rate": 1.255977653631285e-05, + "loss": 0.2218, + "step": 3750 + }, + { + "epoch": 10.477653631284916, + "grad_norm": 1.064080834388733, + "learning_rate": 1.2563128491620113e-05, + "loss": 0.2486, + "step": 3751 + }, + { + "epoch": 10.480446927374302, + "grad_norm": 1.298956036567688, + "learning_rate": 1.2566480446927374e-05, + "loss": 0.2458, + "step": 3752 + }, + { + "epoch": 10.483240223463687, + "grad_norm": 1.2325199842453003, + "learning_rate": 1.2569832402234637e-05, + "loss": 0.2493, + "step": 3753 + }, + { + "epoch": 10.486033519553073, + "grad_norm": 1.509262204170227, + "learning_rate": 1.25731843575419e-05, + "loss": 0.3035, + "step": 3754 + }, + { + "epoch": 10.488826815642458, + "grad_norm": 2.5113518238067627, + "learning_rate": 1.2576536312849162e-05, + "loss": 0.3347, + "step": 3755 + }, + { + "epoch": 10.491620111731844, + "grad_norm": 0.7280263304710388, + "learning_rate": 1.2579888268156425e-05, + "loss": 0.2755, + "step": 3756 + }, + { + "epoch": 10.494413407821229, + "grad_norm": 0.5418054461479187, + "learning_rate": 1.2583240223463688e-05, + "loss": 0.2096, + "step": 3757 + }, + { + "epoch": 10.497206703910614, + "grad_norm": 0.5273208618164062, + "learning_rate": 1.258659217877095e-05, + "loss": 0.2123, + "step": 3758 + }, + { + "epoch": 10.5, + "grad_norm": 0.5765011310577393, + "learning_rate": 1.2589944134078212e-05, + "loss": 0.2175, + "step": 3759 + }, + { + "epoch": 10.502793296089386, + "grad_norm": 0.706550121307373, + "learning_rate": 1.2593296089385475e-05, + "loss": 0.1769, + "step": 3760 + }, + { + "epoch": 10.505586592178771, + "grad_norm": 0.7754123210906982, + "learning_rate": 1.2596648044692738e-05, + "loss": 0.2333, + "step": 3761 + }, + { + "epoch": 10.508379888268156, + "grad_norm": 1.0571595430374146, + "learning_rate": 1.26e-05, + "loss": 0.2243, + "step": 3762 + }, + { + "epoch": 10.511173184357542, + "grad_norm": 0.5929199457168579, + "learning_rate": 1.2603351955307263e-05, + "loss": 0.1954, + "step": 3763 + }, + { + "epoch": 10.513966480446927, + "grad_norm": 0.6998215317726135, + "learning_rate": 1.2606703910614526e-05, + "loss": 0.2046, + "step": 3764 + }, + { + "epoch": 10.516759776536313, + "grad_norm": 0.9153921008110046, + "learning_rate": 1.2610055865921787e-05, + "loss": 0.2142, + "step": 3765 + }, + { + "epoch": 10.519553072625698, + "grad_norm": 1.2960528135299683, + "learning_rate": 1.261340782122905e-05, + "loss": 0.2196, + "step": 3766 + }, + { + "epoch": 10.522346368715084, + "grad_norm": 0.8117894530296326, + "learning_rate": 1.2616759776536313e-05, + "loss": 0.2328, + "step": 3767 + }, + { + "epoch": 10.525139664804469, + "grad_norm": 0.6739299297332764, + "learning_rate": 1.2620111731843574e-05, + "loss": 0.1967, + "step": 3768 + }, + { + "epoch": 10.527932960893855, + "grad_norm": 0.9621912837028503, + "learning_rate": 1.2623463687150838e-05, + "loss": 0.2111, + "step": 3769 + }, + { + "epoch": 10.53072625698324, + "grad_norm": 0.7745662927627563, + "learning_rate": 1.26268156424581e-05, + "loss": 0.1873, + "step": 3770 + }, + { + "epoch": 10.533519553072626, + "grad_norm": 0.8976259827613831, + "learning_rate": 1.2630167597765364e-05, + "loss": 0.1887, + "step": 3771 + }, + { + "epoch": 10.53631284916201, + "grad_norm": 0.9401381015777588, + "learning_rate": 1.2633519553072625e-05, + "loss": 0.2388, + "step": 3772 + }, + { + "epoch": 10.539106145251397, + "grad_norm": 1.1641933917999268, + "learning_rate": 1.2636871508379888e-05, + "loss": 0.2062, + "step": 3773 + }, + { + "epoch": 10.541899441340782, + "grad_norm": 1.082189917564392, + "learning_rate": 1.2640223463687151e-05, + "loss": 0.2592, + "step": 3774 + }, + { + "epoch": 10.544692737430168, + "grad_norm": 1.1882898807525635, + "learning_rate": 1.2643575418994412e-05, + "loss": 0.1982, + "step": 3775 + }, + { + "epoch": 10.547486033519553, + "grad_norm": 1.7273212671279907, + "learning_rate": 1.2646927374301675e-05, + "loss": 0.2363, + "step": 3776 + }, + { + "epoch": 10.550279329608939, + "grad_norm": 1.2006393671035767, + "learning_rate": 1.2650279329608938e-05, + "loss": 0.2212, + "step": 3777 + }, + { + "epoch": 10.553072625698324, + "grad_norm": 2.60697865486145, + "learning_rate": 1.2653631284916201e-05, + "loss": 0.2486, + "step": 3778 + }, + { + "epoch": 10.55586592178771, + "grad_norm": 3.1729323863983154, + "learning_rate": 1.2656983240223463e-05, + "loss": 0.2896, + "step": 3779 + }, + { + "epoch": 10.558659217877095, + "grad_norm": 2.9192662239074707, + "learning_rate": 1.2660335195530726e-05, + "loss": 0.3792, + "step": 3780 + }, + { + "epoch": 10.561452513966481, + "grad_norm": 0.9551946520805359, + "learning_rate": 1.266368715083799e-05, + "loss": 0.2271, + "step": 3781 + }, + { + "epoch": 10.564245810055866, + "grad_norm": 0.5861912369728088, + "learning_rate": 1.2667039106145252e-05, + "loss": 0.2326, + "step": 3782 + }, + { + "epoch": 10.567039106145252, + "grad_norm": 0.5192221403121948, + "learning_rate": 1.2670391061452515e-05, + "loss": 0.18, + "step": 3783 + }, + { + "epoch": 10.569832402234637, + "grad_norm": 0.5568036437034607, + "learning_rate": 1.2673743016759778e-05, + "loss": 0.1942, + "step": 3784 + }, + { + "epoch": 10.572625698324023, + "grad_norm": 0.45381930470466614, + "learning_rate": 1.267709497206704e-05, + "loss": 0.1832, + "step": 3785 + }, + { + "epoch": 10.575418994413408, + "grad_norm": 0.6978787779808044, + "learning_rate": 1.2680446927374302e-05, + "loss": 0.2314, + "step": 3786 + }, + { + "epoch": 10.578212290502794, + "grad_norm": 0.6960095167160034, + "learning_rate": 1.2683798882681565e-05, + "loss": 0.1931, + "step": 3787 + }, + { + "epoch": 10.581005586592179, + "grad_norm": 0.5697109699249268, + "learning_rate": 1.2687150837988828e-05, + "loss": 0.1927, + "step": 3788 + }, + { + "epoch": 10.583798882681565, + "grad_norm": 0.6475376486778259, + "learning_rate": 1.269050279329609e-05, + "loss": 0.199, + "step": 3789 + }, + { + "epoch": 10.58659217877095, + "grad_norm": 0.6828493475914001, + "learning_rate": 1.2693854748603353e-05, + "loss": 0.2166, + "step": 3790 + }, + { + "epoch": 10.589385474860336, + "grad_norm": 0.8032903075218201, + "learning_rate": 1.2697206703910616e-05, + "loss": 0.2278, + "step": 3791 + }, + { + "epoch": 10.59217877094972, + "grad_norm": 0.9505577683448792, + "learning_rate": 1.2700558659217877e-05, + "loss": 0.2239, + "step": 3792 + }, + { + "epoch": 10.594972067039105, + "grad_norm": 0.6349625587463379, + "learning_rate": 1.270391061452514e-05, + "loss": 0.1823, + "step": 3793 + }, + { + "epoch": 10.597765363128492, + "grad_norm": 0.6360328197479248, + "learning_rate": 1.2707262569832403e-05, + "loss": 0.2352, + "step": 3794 + }, + { + "epoch": 10.600558659217878, + "grad_norm": 1.0583397150039673, + "learning_rate": 1.2710614525139665e-05, + "loss": 0.1767, + "step": 3795 + }, + { + "epoch": 10.603351955307263, + "grad_norm": 0.7373006343841553, + "learning_rate": 1.2713966480446928e-05, + "loss": 0.22, + "step": 3796 + }, + { + "epoch": 10.606145251396647, + "grad_norm": 0.7576307058334351, + "learning_rate": 1.271731843575419e-05, + "loss": 0.2516, + "step": 3797 + }, + { + "epoch": 10.608938547486034, + "grad_norm": 4.590017318725586, + "learning_rate": 1.2720670391061454e-05, + "loss": 0.1925, + "step": 3798 + }, + { + "epoch": 10.611731843575418, + "grad_norm": 0.9053855538368225, + "learning_rate": 1.2724022346368715e-05, + "loss": 0.2135, + "step": 3799 + }, + { + "epoch": 10.614525139664805, + "grad_norm": 0.8154563903808594, + "learning_rate": 1.2727374301675978e-05, + "loss": 0.2281, + "step": 3800 + }, + { + "epoch": 10.61731843575419, + "grad_norm": 0.8832427263259888, + "learning_rate": 1.2730726256983241e-05, + "loss": 0.1973, + "step": 3801 + }, + { + "epoch": 10.620111731843576, + "grad_norm": 1.0317167043685913, + "learning_rate": 1.2734078212290503e-05, + "loss": 0.2452, + "step": 3802 + }, + { + "epoch": 10.62290502793296, + "grad_norm": 3.71354603767395, + "learning_rate": 1.2737430167597766e-05, + "loss": 0.2695, + "step": 3803 + }, + { + "epoch": 10.625698324022347, + "grad_norm": 1.8564016819000244, + "learning_rate": 1.2740782122905029e-05, + "loss": 0.2521, + "step": 3804 + }, + { + "epoch": 10.628491620111731, + "grad_norm": 2.274324417114258, + "learning_rate": 1.274413407821229e-05, + "loss": 0.3213, + "step": 3805 + }, + { + "epoch": 10.631284916201118, + "grad_norm": 0.8748537302017212, + "learning_rate": 1.2747486033519553e-05, + "loss": 0.2421, + "step": 3806 + }, + { + "epoch": 10.634078212290502, + "grad_norm": 0.6516913175582886, + "learning_rate": 1.2750837988826816e-05, + "loss": 0.1767, + "step": 3807 + }, + { + "epoch": 10.636871508379889, + "grad_norm": 0.539890468120575, + "learning_rate": 1.2754189944134079e-05, + "loss": 0.2003, + "step": 3808 + }, + { + "epoch": 10.639664804469273, + "grad_norm": 0.6699892282485962, + "learning_rate": 1.275754189944134e-05, + "loss": 0.2137, + "step": 3809 + }, + { + "epoch": 10.64245810055866, + "grad_norm": 0.6511008739471436, + "learning_rate": 1.2760893854748604e-05, + "loss": 0.2016, + "step": 3810 + }, + { + "epoch": 10.645251396648044, + "grad_norm": 1.3467512130737305, + "learning_rate": 1.2764245810055867e-05, + "loss": 0.1888, + "step": 3811 + }, + { + "epoch": 10.64804469273743, + "grad_norm": 0.6521399021148682, + "learning_rate": 1.2767597765363128e-05, + "loss": 0.213, + "step": 3812 + }, + { + "epoch": 10.650837988826815, + "grad_norm": 0.6177999377250671, + "learning_rate": 1.2770949720670391e-05, + "loss": 0.2088, + "step": 3813 + }, + { + "epoch": 10.653631284916202, + "grad_norm": 0.8969693779945374, + "learning_rate": 1.2774301675977654e-05, + "loss": 0.1991, + "step": 3814 + }, + { + "epoch": 10.656424581005586, + "grad_norm": 0.6688258051872253, + "learning_rate": 1.2777653631284915e-05, + "loss": 0.2359, + "step": 3815 + }, + { + "epoch": 10.659217877094973, + "grad_norm": 0.6541182994842529, + "learning_rate": 1.2781005586592178e-05, + "loss": 0.1847, + "step": 3816 + }, + { + "epoch": 10.662011173184357, + "grad_norm": 0.6433907151222229, + "learning_rate": 1.2784357541899441e-05, + "loss": 0.2358, + "step": 3817 + }, + { + "epoch": 10.664804469273744, + "grad_norm": 1.100403904914856, + "learning_rate": 1.2787709497206704e-05, + "loss": 0.2001, + "step": 3818 + }, + { + "epoch": 10.667597765363128, + "grad_norm": 2.9561421871185303, + "learning_rate": 1.2791061452513966e-05, + "loss": 0.1878, + "step": 3819 + }, + { + "epoch": 10.670391061452515, + "grad_norm": 0.7579318284988403, + "learning_rate": 1.2794413407821229e-05, + "loss": 0.2268, + "step": 3820 + }, + { + "epoch": 10.6731843575419, + "grad_norm": 0.8342117667198181, + "learning_rate": 1.2797765363128492e-05, + "loss": 0.2381, + "step": 3821 + }, + { + "epoch": 10.675977653631286, + "grad_norm": 1.016027569770813, + "learning_rate": 1.2801117318435753e-05, + "loss": 0.193, + "step": 3822 + }, + { + "epoch": 10.67877094972067, + "grad_norm": 1.90312659740448, + "learning_rate": 1.2804469273743016e-05, + "loss": 0.2157, + "step": 3823 + }, + { + "epoch": 10.681564245810057, + "grad_norm": 1.1738243103027344, + "learning_rate": 1.280782122905028e-05, + "loss": 0.2392, + "step": 3824 + }, + { + "epoch": 10.684357541899441, + "grad_norm": 1.1743561029434204, + "learning_rate": 1.2811173184357542e-05, + "loss": 0.2024, + "step": 3825 + }, + { + "epoch": 10.687150837988828, + "grad_norm": 2.7827699184417725, + "learning_rate": 1.2814525139664804e-05, + "loss": 0.2317, + "step": 3826 + }, + { + "epoch": 10.689944134078212, + "grad_norm": 1.679596185684204, + "learning_rate": 1.2817877094972067e-05, + "loss": 0.2155, + "step": 3827 + }, + { + "epoch": 10.692737430167599, + "grad_norm": 1.0824395418167114, + "learning_rate": 1.282122905027933e-05, + "loss": 0.2403, + "step": 3828 + }, + { + "epoch": 10.695530726256983, + "grad_norm": 7.307173252105713, + "learning_rate": 1.2824581005586593e-05, + "loss": 0.2838, + "step": 3829 + }, + { + "epoch": 10.69832402234637, + "grad_norm": 2.8894381523132324, + "learning_rate": 1.2827932960893856e-05, + "loss": 0.3895, + "step": 3830 + }, + { + "epoch": 10.701117318435754, + "grad_norm": 0.5374540090560913, + "learning_rate": 1.2831284916201119e-05, + "loss": 0.199, + "step": 3831 + }, + { + "epoch": 10.703910614525139, + "grad_norm": 0.4714552164077759, + "learning_rate": 1.283463687150838e-05, + "loss": 0.201, + "step": 3832 + }, + { + "epoch": 10.706703910614525, + "grad_norm": 0.656848669052124, + "learning_rate": 1.2837988826815643e-05, + "loss": 0.2093, + "step": 3833 + }, + { + "epoch": 10.70949720670391, + "grad_norm": 0.8321217894554138, + "learning_rate": 1.2841340782122906e-05, + "loss": 0.1973, + "step": 3834 + }, + { + "epoch": 10.712290502793296, + "grad_norm": 0.867364764213562, + "learning_rate": 1.284469273743017e-05, + "loss": 0.2225, + "step": 3835 + }, + { + "epoch": 10.71508379888268, + "grad_norm": 0.6925740242004395, + "learning_rate": 1.284804469273743e-05, + "loss": 0.2036, + "step": 3836 + }, + { + "epoch": 10.717877094972067, + "grad_norm": 0.7183954119682312, + "learning_rate": 1.2851396648044694e-05, + "loss": 0.2503, + "step": 3837 + }, + { + "epoch": 10.720670391061452, + "grad_norm": 0.7314544320106506, + "learning_rate": 1.2854748603351957e-05, + "loss": 0.2096, + "step": 3838 + }, + { + "epoch": 10.723463687150838, + "grad_norm": 0.6448348164558411, + "learning_rate": 1.2858100558659218e-05, + "loss": 0.2082, + "step": 3839 + }, + { + "epoch": 10.726256983240223, + "grad_norm": 0.6047266721725464, + "learning_rate": 1.2861452513966481e-05, + "loss": 0.1919, + "step": 3840 + }, + { + "epoch": 10.72905027932961, + "grad_norm": 0.7337368726730347, + "learning_rate": 1.2864804469273744e-05, + "loss": 0.2132, + "step": 3841 + }, + { + "epoch": 10.731843575418994, + "grad_norm": 0.7058912515640259, + "learning_rate": 1.2868156424581006e-05, + "loss": 0.2159, + "step": 3842 + }, + { + "epoch": 10.73463687150838, + "grad_norm": 0.7640464305877686, + "learning_rate": 1.2871508379888269e-05, + "loss": 0.214, + "step": 3843 + }, + { + "epoch": 10.737430167597765, + "grad_norm": 0.8692854642868042, + "learning_rate": 1.2874860335195532e-05, + "loss": 0.2104, + "step": 3844 + }, + { + "epoch": 10.740223463687151, + "grad_norm": 0.821365237236023, + "learning_rate": 1.2878212290502795e-05, + "loss": 0.1895, + "step": 3845 + }, + { + "epoch": 10.743016759776536, + "grad_norm": 1.252829670906067, + "learning_rate": 1.2881564245810056e-05, + "loss": 0.1783, + "step": 3846 + }, + { + "epoch": 10.745810055865922, + "grad_norm": 0.847938597202301, + "learning_rate": 1.2884916201117319e-05, + "loss": 0.2149, + "step": 3847 + }, + { + "epoch": 10.748603351955307, + "grad_norm": 1.1866005659103394, + "learning_rate": 1.2888268156424582e-05, + "loss": 0.2026, + "step": 3848 + }, + { + "epoch": 10.751396648044693, + "grad_norm": 1.1497517824172974, + "learning_rate": 1.2891620111731843e-05, + "loss": 0.2213, + "step": 3849 + }, + { + "epoch": 10.754189944134078, + "grad_norm": 1.2760660648345947, + "learning_rate": 1.2894972067039106e-05, + "loss": 0.2697, + "step": 3850 + }, + { + "epoch": 10.756983240223464, + "grad_norm": 1.3026838302612305, + "learning_rate": 1.289832402234637e-05, + "loss": 0.2079, + "step": 3851 + }, + { + "epoch": 10.759776536312849, + "grad_norm": 0.9143239259719849, + "learning_rate": 1.290167597765363e-05, + "loss": 0.2306, + "step": 3852 + }, + { + "epoch": 10.762569832402235, + "grad_norm": 1.8348908424377441, + "learning_rate": 1.2905027932960894e-05, + "loss": 0.2662, + "step": 3853 + }, + { + "epoch": 10.76536312849162, + "grad_norm": 2.1458613872528076, + "learning_rate": 1.2908379888268157e-05, + "loss": 0.2591, + "step": 3854 + }, + { + "epoch": 10.768156424581006, + "grad_norm": 2.346968412399292, + "learning_rate": 1.291173184357542e-05, + "loss": 0.3545, + "step": 3855 + }, + { + "epoch": 10.77094972067039, + "grad_norm": Infinity, + "learning_rate": 1.291173184357542e-05, + "loss": 0.1996, + "step": 3856 + }, + { + "epoch": 10.773743016759777, + "grad_norm": 0.5375748872756958, + "learning_rate": 1.2915083798882681e-05, + "loss": 0.2132, + "step": 3857 + }, + { + "epoch": 10.776536312849162, + "grad_norm": 0.6494588255882263, + "learning_rate": 1.2918435754189944e-05, + "loss": 0.2166, + "step": 3858 + }, + { + "epoch": 10.779329608938548, + "grad_norm": 0.6067144274711609, + "learning_rate": 1.2921787709497207e-05, + "loss": 0.1931, + "step": 3859 + }, + { + "epoch": 10.782122905027933, + "grad_norm": 0.6001162528991699, + "learning_rate": 1.2925139664804469e-05, + "loss": 0.2351, + "step": 3860 + }, + { + "epoch": 10.78491620111732, + "grad_norm": 1.0989937782287598, + "learning_rate": 1.2928491620111732e-05, + "loss": 0.2147, + "step": 3861 + }, + { + "epoch": 10.787709497206704, + "grad_norm": 0.8464879989624023, + "learning_rate": 1.2931843575418995e-05, + "loss": 0.2296, + "step": 3862 + }, + { + "epoch": 10.79050279329609, + "grad_norm": 0.8227540254592896, + "learning_rate": 1.2935195530726256e-05, + "loss": 0.1841, + "step": 3863 + }, + { + "epoch": 10.793296089385475, + "grad_norm": 1.489974021911621, + "learning_rate": 1.293854748603352e-05, + "loss": 0.1898, + "step": 3864 + }, + { + "epoch": 10.796089385474861, + "grad_norm": 0.7693848013877869, + "learning_rate": 1.2941899441340782e-05, + "loss": 0.2303, + "step": 3865 + }, + { + "epoch": 10.798882681564246, + "grad_norm": 0.7948211431503296, + "learning_rate": 1.2945251396648045e-05, + "loss": 0.216, + "step": 3866 + }, + { + "epoch": 10.80167597765363, + "grad_norm": 0.7577455639839172, + "learning_rate": 1.2948603351955307e-05, + "loss": 0.2231, + "step": 3867 + }, + { + "epoch": 10.804469273743017, + "grad_norm": 0.9064778089523315, + "learning_rate": 1.295195530726257e-05, + "loss": 0.2082, + "step": 3868 + }, + { + "epoch": 10.807262569832401, + "grad_norm": 0.8344342708587646, + "learning_rate": 1.2955307262569833e-05, + "loss": 0.1921, + "step": 3869 + }, + { + "epoch": 10.810055865921788, + "grad_norm": 1.2911068201065063, + "learning_rate": 1.2958659217877094e-05, + "loss": 0.2381, + "step": 3870 + }, + { + "epoch": 10.812849162011172, + "grad_norm": 0.8481412529945374, + "learning_rate": 1.2962011173184357e-05, + "loss": 0.1808, + "step": 3871 + }, + { + "epoch": 10.815642458100559, + "grad_norm": 0.8114192485809326, + "learning_rate": 1.296536312849162e-05, + "loss": 0.1951, + "step": 3872 + }, + { + "epoch": 10.818435754189943, + "grad_norm": 1.0794395208358765, + "learning_rate": 1.2968715083798883e-05, + "loss": 0.2563, + "step": 3873 + }, + { + "epoch": 10.82122905027933, + "grad_norm": 0.976500391960144, + "learning_rate": 1.2972067039106144e-05, + "loss": 0.2277, + "step": 3874 + }, + { + "epoch": 10.824022346368714, + "grad_norm": 1.1792607307434082, + "learning_rate": 1.2975418994413408e-05, + "loss": 0.2083, + "step": 3875 + }, + { + "epoch": 10.8268156424581, + "grad_norm": 1.153577208518982, + "learning_rate": 1.297877094972067e-05, + "loss": 0.2182, + "step": 3876 + }, + { + "epoch": 10.829608938547485, + "grad_norm": 1.8446016311645508, + "learning_rate": 1.2982122905027932e-05, + "loss": 0.2452, + "step": 3877 + }, + { + "epoch": 10.832402234636872, + "grad_norm": 1.3943886756896973, + "learning_rate": 1.2985474860335197e-05, + "loss": 0.2388, + "step": 3878 + }, + { + "epoch": 10.835195530726256, + "grad_norm": 1.6693472862243652, + "learning_rate": 1.298882681564246e-05, + "loss": 0.3159, + "step": 3879 + }, + { + "epoch": 10.837988826815643, + "grad_norm": 3.3095147609710693, + "learning_rate": 1.2992178770949721e-05, + "loss": 0.3113, + "step": 3880 + }, + { + "epoch": 10.840782122905027, + "grad_norm": 0.4930731952190399, + "learning_rate": 1.2995530726256984e-05, + "loss": 0.2008, + "step": 3881 + }, + { + "epoch": 10.843575418994414, + "grad_norm": 0.570561408996582, + "learning_rate": 1.2998882681564247e-05, + "loss": 0.2286, + "step": 3882 + }, + { + "epoch": 10.846368715083798, + "grad_norm": 0.7626252770423889, + "learning_rate": 1.300223463687151e-05, + "loss": 0.2275, + "step": 3883 + }, + { + "epoch": 10.849162011173185, + "grad_norm": 0.60189288854599, + "learning_rate": 1.3005586592178771e-05, + "loss": 0.1978, + "step": 3884 + }, + { + "epoch": 10.85195530726257, + "grad_norm": 0.8341513872146606, + "learning_rate": 1.3008938547486035e-05, + "loss": 0.2571, + "step": 3885 + }, + { + "epoch": 10.854748603351956, + "grad_norm": 0.62429279088974, + "learning_rate": 1.3012290502793298e-05, + "loss": 0.1959, + "step": 3886 + }, + { + "epoch": 10.85754189944134, + "grad_norm": 2.9129724502563477, + "learning_rate": 1.3015642458100559e-05, + "loss": 0.2188, + "step": 3887 + }, + { + "epoch": 10.860335195530727, + "grad_norm": 0.6366764903068542, + "learning_rate": 1.3018994413407822e-05, + "loss": 0.232, + "step": 3888 + }, + { + "epoch": 10.863128491620111, + "grad_norm": 0.660622239112854, + "learning_rate": 1.3022346368715085e-05, + "loss": 0.214, + "step": 3889 + }, + { + "epoch": 10.865921787709498, + "grad_norm": 0.7940386533737183, + "learning_rate": 1.3025698324022346e-05, + "loss": 0.2223, + "step": 3890 + }, + { + "epoch": 10.868715083798882, + "grad_norm": 1.3321053981781006, + "learning_rate": 1.302905027932961e-05, + "loss": 0.2028, + "step": 3891 + }, + { + "epoch": 10.871508379888269, + "grad_norm": 0.9569311141967773, + "learning_rate": 1.3032402234636872e-05, + "loss": 0.1979, + "step": 3892 + }, + { + "epoch": 10.874301675977653, + "grad_norm": 0.7009804248809814, + "learning_rate": 1.3035754189944135e-05, + "loss": 0.1927, + "step": 3893 + }, + { + "epoch": 10.87709497206704, + "grad_norm": 0.873306930065155, + "learning_rate": 1.3039106145251397e-05, + "loss": 0.198, + "step": 3894 + }, + { + "epoch": 10.879888268156424, + "grad_norm": 1.5338014364242554, + "learning_rate": 1.304245810055866e-05, + "loss": 0.1951, + "step": 3895 + }, + { + "epoch": 10.88268156424581, + "grad_norm": 1.5627508163452148, + "learning_rate": 1.3045810055865923e-05, + "loss": 0.2138, + "step": 3896 + }, + { + "epoch": 10.885474860335195, + "grad_norm": 1.2165473699569702, + "learning_rate": 1.3049162011173184e-05, + "loss": 0.2046, + "step": 3897 + }, + { + "epoch": 10.888268156424582, + "grad_norm": 1.158333659172058, + "learning_rate": 1.3052513966480447e-05, + "loss": 0.2331, + "step": 3898 + }, + { + "epoch": 10.891061452513966, + "grad_norm": 0.8501939177513123, + "learning_rate": 1.305586592178771e-05, + "loss": 0.1974, + "step": 3899 + }, + { + "epoch": 10.893854748603353, + "grad_norm": 1.510158658027649, + "learning_rate": 1.3059217877094972e-05, + "loss": 0.2181, + "step": 3900 + }, + { + "epoch": 10.896648044692737, + "grad_norm": 0.9931496977806091, + "learning_rate": 1.3062569832402235e-05, + "loss": 0.2394, + "step": 3901 + }, + { + "epoch": 10.899441340782122, + "grad_norm": 1.472537875175476, + "learning_rate": 1.3065921787709498e-05, + "loss": 0.2045, + "step": 3902 + }, + { + "epoch": 10.902234636871508, + "grad_norm": 1.3375368118286133, + "learning_rate": 1.306927374301676e-05, + "loss": 0.2253, + "step": 3903 + }, + { + "epoch": 10.905027932960895, + "grad_norm": 1.2440787553787231, + "learning_rate": 1.3072625698324022e-05, + "loss": 0.2539, + "step": 3904 + }, + { + "epoch": 10.90782122905028, + "grad_norm": 2.623786211013794, + "learning_rate": 1.3075977653631285e-05, + "loss": 0.326, + "step": 3905 + }, + { + "epoch": 10.910614525139664, + "grad_norm": 0.6709384322166443, + "learning_rate": 1.3079329608938548e-05, + "loss": 0.201, + "step": 3906 + }, + { + "epoch": 10.91340782122905, + "grad_norm": 0.6203086972236633, + "learning_rate": 1.308268156424581e-05, + "loss": 0.1995, + "step": 3907 + }, + { + "epoch": 10.916201117318435, + "grad_norm": 0.5362895727157593, + "learning_rate": 1.3086033519553073e-05, + "loss": 0.1988, + "step": 3908 + }, + { + "epoch": 10.918994413407821, + "grad_norm": 0.7019741535186768, + "learning_rate": 1.3089385474860336e-05, + "loss": 0.2167, + "step": 3909 + }, + { + "epoch": 10.921787709497206, + "grad_norm": 0.603715717792511, + "learning_rate": 1.3092737430167597e-05, + "loss": 0.2157, + "step": 3910 + }, + { + "epoch": 10.924581005586592, + "grad_norm": 0.7056169509887695, + "learning_rate": 1.309608938547486e-05, + "loss": 0.2339, + "step": 3911 + }, + { + "epoch": 10.927374301675977, + "grad_norm": 0.7039938569068909, + "learning_rate": 1.3099441340782123e-05, + "loss": 0.212, + "step": 3912 + }, + { + "epoch": 10.930167597765363, + "grad_norm": 0.9781815409660339, + "learning_rate": 1.3102793296089386e-05, + "loss": 0.195, + "step": 3913 + }, + { + "epoch": 10.932960893854748, + "grad_norm": 0.6511630415916443, + "learning_rate": 1.3106145251396647e-05, + "loss": 0.1947, + "step": 3914 + }, + { + "epoch": 10.935754189944134, + "grad_norm": 0.7758023142814636, + "learning_rate": 1.310949720670391e-05, + "loss": 0.233, + "step": 3915 + }, + { + "epoch": 10.938547486033519, + "grad_norm": 1.0468790531158447, + "learning_rate": 1.3112849162011174e-05, + "loss": 0.2021, + "step": 3916 + }, + { + "epoch": 10.941340782122905, + "grad_norm": 0.709183931350708, + "learning_rate": 1.3116201117318435e-05, + "loss": 0.2112, + "step": 3917 + }, + { + "epoch": 10.94413407821229, + "grad_norm": 0.7826236486434937, + "learning_rate": 1.3119553072625698e-05, + "loss": 0.2034, + "step": 3918 + }, + { + "epoch": 10.946927374301676, + "grad_norm": 3.2121975421905518, + "learning_rate": 1.3122905027932961e-05, + "loss": 0.2356, + "step": 3919 + }, + { + "epoch": 10.949720670391061, + "grad_norm": 0.9546288847923279, + "learning_rate": 1.3126256983240224e-05, + "loss": 0.2019, + "step": 3920 + }, + { + "epoch": 10.952513966480447, + "grad_norm": 1.1267602443695068, + "learning_rate": 1.3129608938547485e-05, + "loss": 0.2092, + "step": 3921 + }, + { + "epoch": 10.955307262569832, + "grad_norm": 1.6067185401916504, + "learning_rate": 1.3132960893854748e-05, + "loss": 0.2101, + "step": 3922 + }, + { + "epoch": 10.958100558659218, + "grad_norm": 1.5760096311569214, + "learning_rate": 1.3136312849162011e-05, + "loss": 0.2128, + "step": 3923 + }, + { + "epoch": 10.960893854748603, + "grad_norm": 0.8861494064331055, + "learning_rate": 1.3139664804469273e-05, + "loss": 0.2279, + "step": 3924 + }, + { + "epoch": 10.96368715083799, + "grad_norm": 0.9535877108573914, + "learning_rate": 1.3143016759776536e-05, + "loss": 0.2334, + "step": 3925 + }, + { + "epoch": 10.966480446927374, + "grad_norm": 1.1394846439361572, + "learning_rate": 1.31463687150838e-05, + "loss": 0.1944, + "step": 3926 + }, + { + "epoch": 10.96927374301676, + "grad_norm": 1.5828227996826172, + "learning_rate": 1.3149720670391062e-05, + "loss": 0.2158, + "step": 3927 + }, + { + "epoch": 10.972067039106145, + "grad_norm": 0.9845840930938721, + "learning_rate": 1.3153072625698325e-05, + "loss": 0.1991, + "step": 3928 + }, + { + "epoch": 10.974860335195531, + "grad_norm": 1.354945421218872, + "learning_rate": 1.3156424581005588e-05, + "loss": 0.2302, + "step": 3929 + }, + { + "epoch": 10.977653631284916, + "grad_norm": 2.161477565765381, + "learning_rate": 1.3159776536312851e-05, + "loss": 0.2672, + "step": 3930 + }, + { + "epoch": 10.980446927374302, + "grad_norm": 0.6602393388748169, + "learning_rate": 1.3163128491620112e-05, + "loss": 0.2096, + "step": 3931 + }, + { + "epoch": 10.983240223463687, + "grad_norm": 0.6836141347885132, + "learning_rate": 1.3166480446927375e-05, + "loss": 0.1795, + "step": 3932 + }, + { + "epoch": 10.986033519553073, + "grad_norm": 0.6825992465019226, + "learning_rate": 1.3169832402234638e-05, + "loss": 0.2219, + "step": 3933 + }, + { + "epoch": 10.988826815642458, + "grad_norm": 0.9697242379188538, + "learning_rate": 1.31731843575419e-05, + "loss": 0.2622, + "step": 3934 + }, + { + "epoch": 10.991620111731844, + "grad_norm": 0.718993067741394, + "learning_rate": 1.3176536312849163e-05, + "loss": 0.2099, + "step": 3935 + }, + { + "epoch": 10.994413407821229, + "grad_norm": 1.0676536560058594, + "learning_rate": 1.3179888268156426e-05, + "loss": 0.2346, + "step": 3936 + }, + { + "epoch": 10.997206703910614, + "grad_norm": 1.1550952196121216, + "learning_rate": 1.3183240223463687e-05, + "loss": 0.2847, + "step": 3937 + }, + { + "epoch": 11.0, + "grad_norm": 1.887609839439392, + "learning_rate": 1.318659217877095e-05, + "loss": 0.2701, + "step": 3938 + }, + { + "epoch": 11.002793296089385, + "grad_norm": 0.5862767100334167, + "learning_rate": 1.3189944134078213e-05, + "loss": 0.226, + "step": 3939 + }, + { + "epoch": 11.005586592178771, + "grad_norm": 0.467701256275177, + "learning_rate": 1.3193296089385476e-05, + "loss": 0.1973, + "step": 3940 + }, + { + "epoch": 11.008379888268156, + "grad_norm": 0.5132733583450317, + "learning_rate": 1.3196648044692738e-05, + "loss": 0.2279, + "step": 3941 + }, + { + "epoch": 11.011173184357542, + "grad_norm": 0.7073836326599121, + "learning_rate": 1.32e-05, + "loss": 0.2724, + "step": 3942 + }, + { + "epoch": 11.013966480446927, + "grad_norm": 0.7695304751396179, + "learning_rate": 1.3203351955307264e-05, + "loss": 0.2386, + "step": 3943 + }, + { + "epoch": 11.016759776536313, + "grad_norm": 0.4284460246562958, + "learning_rate": 1.3206703910614525e-05, + "loss": 0.1744, + "step": 3944 + }, + { + "epoch": 11.019553072625698, + "grad_norm": 7.879252910614014, + "learning_rate": 1.3210055865921788e-05, + "loss": 0.211, + "step": 3945 + }, + { + "epoch": 11.022346368715084, + "grad_norm": 0.5236763954162598, + "learning_rate": 1.3213407821229051e-05, + "loss": 0.1754, + "step": 3946 + }, + { + "epoch": 11.025139664804469, + "grad_norm": 0.7212948203086853, + "learning_rate": 1.3216759776536312e-05, + "loss": 0.2029, + "step": 3947 + }, + { + "epoch": 11.027932960893855, + "grad_norm": 0.8462468981742859, + "learning_rate": 1.3220111731843576e-05, + "loss": 0.2079, + "step": 3948 + }, + { + "epoch": 11.03072625698324, + "grad_norm": 0.5458770394325256, + "learning_rate": 1.3223463687150839e-05, + "loss": 0.1953, + "step": 3949 + }, + { + "epoch": 11.033519553072626, + "grad_norm": 0.6898829936981201, + "learning_rate": 1.3226815642458102e-05, + "loss": 0.2045, + "step": 3950 + }, + { + "epoch": 11.03631284916201, + "grad_norm": 0.8280276656150818, + "learning_rate": 1.3230167597765363e-05, + "loss": 0.2202, + "step": 3951 + }, + { + "epoch": 11.039106145251397, + "grad_norm": 0.7394163608551025, + "learning_rate": 1.3233519553072626e-05, + "loss": 0.1827, + "step": 3952 + }, + { + "epoch": 11.041899441340782, + "grad_norm": 0.71742844581604, + "learning_rate": 1.3236871508379889e-05, + "loss": 0.1995, + "step": 3953 + }, + { + "epoch": 11.044692737430168, + "grad_norm": 0.9348833560943604, + "learning_rate": 1.324022346368715e-05, + "loss": 0.2133, + "step": 3954 + }, + { + "epoch": 11.047486033519553, + "grad_norm": 0.920644223690033, + "learning_rate": 1.3243575418994413e-05, + "loss": 0.1973, + "step": 3955 + }, + { + "epoch": 11.050279329608939, + "grad_norm": 1.0965607166290283, + "learning_rate": 1.3246927374301676e-05, + "loss": 0.2287, + "step": 3956 + }, + { + "epoch": 11.053072625698324, + "grad_norm": 0.722436249256134, + "learning_rate": 1.3250279329608938e-05, + "loss": 0.2479, + "step": 3957 + }, + { + "epoch": 11.05586592178771, + "grad_norm": 0.9079165458679199, + "learning_rate": 1.32536312849162e-05, + "loss": 0.1905, + "step": 3958 + }, + { + "epoch": 11.058659217877095, + "grad_norm": 1.3687103986740112, + "learning_rate": 1.3256983240223464e-05, + "loss": 0.2039, + "step": 3959 + }, + { + "epoch": 11.061452513966481, + "grad_norm": 1.3063981533050537, + "learning_rate": 1.3260335195530727e-05, + "loss": 0.2621, + "step": 3960 + }, + { + "epoch": 11.064245810055866, + "grad_norm": 1.0190733671188354, + "learning_rate": 1.3263687150837988e-05, + "loss": 0.2178, + "step": 3961 + }, + { + "epoch": 11.067039106145252, + "grad_norm": 1.4244400262832642, + "learning_rate": 1.3267039106145251e-05, + "loss": 0.2907, + "step": 3962 + }, + { + "epoch": 11.069832402234637, + "grad_norm": 2.1310997009277344, + "learning_rate": 1.3270391061452514e-05, + "loss": 0.3616, + "step": 3963 + }, + { + "epoch": 11.072625698324023, + "grad_norm": 0.6609694957733154, + "learning_rate": 1.3273743016759776e-05, + "loss": 0.1983, + "step": 3964 + }, + { + "epoch": 11.075418994413408, + "grad_norm": 0.45003747940063477, + "learning_rate": 1.3277094972067039e-05, + "loss": 0.1839, + "step": 3965 + }, + { + "epoch": 11.078212290502794, + "grad_norm": 0.5343344211578369, + "learning_rate": 1.3280446927374302e-05, + "loss": 0.2198, + "step": 3966 + }, + { + "epoch": 11.081005586592179, + "grad_norm": 0.6870099902153015, + "learning_rate": 1.3283798882681565e-05, + "loss": 0.2098, + "step": 3967 + }, + { + "epoch": 11.083798882681565, + "grad_norm": 0.586830735206604, + "learning_rate": 1.3287150837988826e-05, + "loss": 0.2253, + "step": 3968 + }, + { + "epoch": 11.08659217877095, + "grad_norm": 0.7955715656280518, + "learning_rate": 1.329050279329609e-05, + "loss": 0.2312, + "step": 3969 + }, + { + "epoch": 11.089385474860336, + "grad_norm": 0.5843416452407837, + "learning_rate": 1.3293854748603352e-05, + "loss": 0.208, + "step": 3970 + }, + { + "epoch": 11.09217877094972, + "grad_norm": 0.795759379863739, + "learning_rate": 1.3297206703910614e-05, + "loss": 0.2044, + "step": 3971 + }, + { + "epoch": 11.094972067039107, + "grad_norm": 0.9647024869918823, + "learning_rate": 1.3300558659217877e-05, + "loss": 0.1977, + "step": 3972 + }, + { + "epoch": 11.097765363128492, + "grad_norm": 0.9124093651771545, + "learning_rate": 1.330391061452514e-05, + "loss": 0.2232, + "step": 3973 + }, + { + "epoch": 11.100558659217878, + "grad_norm": 0.6216570138931274, + "learning_rate": 1.3307262569832403e-05, + "loss": 0.1785, + "step": 3974 + }, + { + "epoch": 11.103351955307263, + "grad_norm": 0.898984432220459, + "learning_rate": 1.3310614525139666e-05, + "loss": 0.2403, + "step": 3975 + }, + { + "epoch": 11.106145251396647, + "grad_norm": 1.2527614831924438, + "learning_rate": 1.3313966480446929e-05, + "loss": 0.19, + "step": 3976 + }, + { + "epoch": 11.108938547486034, + "grad_norm": 0.9334877133369446, + "learning_rate": 1.3317318435754192e-05, + "loss": 0.2271, + "step": 3977 + }, + { + "epoch": 11.111731843575418, + "grad_norm": 0.9226744174957275, + "learning_rate": 1.3320670391061453e-05, + "loss": 0.241, + "step": 3978 + }, + { + "epoch": 11.114525139664805, + "grad_norm": 1.081394910812378, + "learning_rate": 1.3324022346368716e-05, + "loss": 0.1949, + "step": 3979 + }, + { + "epoch": 11.11731843575419, + "grad_norm": 0.7130879163742065, + "learning_rate": 1.332737430167598e-05, + "loss": 0.2311, + "step": 3980 + }, + { + "epoch": 11.120111731843576, + "grad_norm": 1.2244608402252197, + "learning_rate": 1.333072625698324e-05, + "loss": 0.2177, + "step": 3981 + }, + { + "epoch": 11.12290502793296, + "grad_norm": 0.9151128530502319, + "learning_rate": 1.3334078212290504e-05, + "loss": 0.216, + "step": 3982 + }, + { + "epoch": 11.125698324022347, + "grad_norm": 1.278940200805664, + "learning_rate": 1.3337430167597767e-05, + "loss": 0.2085, + "step": 3983 + }, + { + "epoch": 11.128491620111731, + "grad_norm": 1.374871850013733, + "learning_rate": 1.3340782122905028e-05, + "loss": 0.2334, + "step": 3984 + }, + { + "epoch": 11.131284916201118, + "grad_norm": 1.507913589477539, + "learning_rate": 1.3344134078212291e-05, + "loss": 0.2265, + "step": 3985 + }, + { + "epoch": 11.134078212290502, + "grad_norm": 1.4145610332489014, + "learning_rate": 1.3347486033519554e-05, + "loss": 0.2548, + "step": 3986 + }, + { + "epoch": 11.136871508379889, + "grad_norm": 1.4787328243255615, + "learning_rate": 1.3350837988826817e-05, + "loss": 0.2536, + "step": 3987 + }, + { + "epoch": 11.139664804469273, + "grad_norm": 2.317295789718628, + "learning_rate": 1.3354189944134078e-05, + "loss": 0.2973, + "step": 3988 + }, + { + "epoch": 11.14245810055866, + "grad_norm": 0.8048063516616821, + "learning_rate": 1.3357541899441341e-05, + "loss": 0.242, + "step": 3989 + }, + { + "epoch": 11.145251396648044, + "grad_norm": 0.7521153092384338, + "learning_rate": 1.3360893854748605e-05, + "loss": 0.2303, + "step": 3990 + }, + { + "epoch": 11.14804469273743, + "grad_norm": 0.7551659345626831, + "learning_rate": 1.3364245810055866e-05, + "loss": 0.2316, + "step": 3991 + }, + { + "epoch": 11.150837988826815, + "grad_norm": 0.5148569941520691, + "learning_rate": 1.3367597765363129e-05, + "loss": 0.1848, + "step": 3992 + }, + { + "epoch": 11.153631284916202, + "grad_norm": 0.850338339805603, + "learning_rate": 1.3370949720670392e-05, + "loss": 0.1825, + "step": 3993 + }, + { + "epoch": 11.156424581005586, + "grad_norm": 0.6699392795562744, + "learning_rate": 1.3374301675977653e-05, + "loss": 0.1839, + "step": 3994 + }, + { + "epoch": 11.159217877094973, + "grad_norm": 0.8000361919403076, + "learning_rate": 1.3377653631284916e-05, + "loss": 0.2485, + "step": 3995 + }, + { + "epoch": 11.162011173184357, + "grad_norm": 0.5427797436714172, + "learning_rate": 1.338100558659218e-05, + "loss": 0.1878, + "step": 3996 + }, + { + "epoch": 11.164804469273744, + "grad_norm": 0.6579750776290894, + "learning_rate": 1.3384357541899442e-05, + "loss": 0.2035, + "step": 3997 + }, + { + "epoch": 11.167597765363128, + "grad_norm": 0.7462218999862671, + "learning_rate": 1.3387709497206704e-05, + "loss": 0.1765, + "step": 3998 + }, + { + "epoch": 11.170391061452515, + "grad_norm": 1.538865566253662, + "learning_rate": 1.3391061452513967e-05, + "loss": 0.1981, + "step": 3999 + }, + { + "epoch": 11.1731843575419, + "grad_norm": 0.7742692828178406, + "learning_rate": 1.339441340782123e-05, + "loss": 0.2308, + "step": 4000 + }, + { + "epoch": 11.1731843575419, + "eval_cer": 0.1938179874090319, + "eval_loss": 0.2728058993816376, + "eval_runtime": 23.8627, + "eval_samples_per_second": 190.171, + "eval_steps_per_second": 0.629, + "eval_wer": 0.347141141308902, + "step": 4000 + }, + { + "epoch": 11.175977653631286, + "grad_norm": 1.0513888597488403, + "learning_rate": 1.3397765363128491e-05, + "loss": 0.1862, + "step": 4001 + }, + { + "epoch": 11.17877094972067, + "grad_norm": 1.7332825660705566, + "learning_rate": 1.3401117318435754e-05, + "loss": 0.2141, + "step": 4002 + }, + { + "epoch": 11.181564245810057, + "grad_norm": 1.0404878854751587, + "learning_rate": 1.3404469273743017e-05, + "loss": 0.1813, + "step": 4003 + }, + { + "epoch": 11.184357541899441, + "grad_norm": 0.9090386629104614, + "learning_rate": 1.3407821229050279e-05, + "loss": 0.1912, + "step": 4004 + }, + { + "epoch": 11.187150837988828, + "grad_norm": 5.820863723754883, + "learning_rate": 1.3411173184357542e-05, + "loss": 0.1819, + "step": 4005 + }, + { + "epoch": 11.189944134078212, + "grad_norm": 1.0044251680374146, + "learning_rate": 1.3414525139664805e-05, + "loss": 0.2061, + "step": 4006 + }, + { + "epoch": 11.192737430167599, + "grad_norm": 1.708371877670288, + "learning_rate": 1.3417877094972068e-05, + "loss": 0.1938, + "step": 4007 + }, + { + "epoch": 11.195530726256983, + "grad_norm": 1.2813067436218262, + "learning_rate": 1.3421229050279329e-05, + "loss": 0.1812, + "step": 4008 + }, + { + "epoch": 11.19832402234637, + "grad_norm": 1.114315152168274, + "learning_rate": 1.3424581005586592e-05, + "loss": 0.2652, + "step": 4009 + }, + { + "epoch": 11.201117318435754, + "grad_norm": 1.39530611038208, + "learning_rate": 1.3427932960893855e-05, + "loss": 0.231, + "step": 4010 + }, + { + "epoch": 11.203910614525139, + "grad_norm": 1.09405517578125, + "learning_rate": 1.3431284916201117e-05, + "loss": 0.2268, + "step": 4011 + }, + { + "epoch": 11.206703910614525, + "grad_norm": 1.7125049829483032, + "learning_rate": 1.343463687150838e-05, + "loss": 0.2566, + "step": 4012 + }, + { + "epoch": 11.20949720670391, + "grad_norm": 2.523878812789917, + "learning_rate": 1.3437988826815643e-05, + "loss": 0.3747, + "step": 4013 + }, + { + "epoch": 11.212290502793296, + "grad_norm": 0.6914112567901611, + "learning_rate": 1.3441340782122906e-05, + "loss": 0.2064, + "step": 4014 + }, + { + "epoch": 11.21508379888268, + "grad_norm": 1.170747995376587, + "learning_rate": 1.3444692737430167e-05, + "loss": 0.2404, + "step": 4015 + }, + { + "epoch": 11.217877094972067, + "grad_norm": 0.767882764339447, + "learning_rate": 1.344804469273743e-05, + "loss": 0.2728, + "step": 4016 + }, + { + "epoch": 11.220670391061452, + "grad_norm": 1.158822774887085, + "learning_rate": 1.3451396648044693e-05, + "loss": 0.1686, + "step": 4017 + }, + { + "epoch": 11.223463687150838, + "grad_norm": 0.8378045558929443, + "learning_rate": 1.3454748603351954e-05, + "loss": 0.2052, + "step": 4018 + }, + { + "epoch": 11.226256983240223, + "grad_norm": 0.6397417187690735, + "learning_rate": 1.3458100558659217e-05, + "loss": 0.1944, + "step": 4019 + }, + { + "epoch": 11.22905027932961, + "grad_norm": 0.6556358933448792, + "learning_rate": 1.346145251396648e-05, + "loss": 0.1827, + "step": 4020 + }, + { + "epoch": 11.231843575418994, + "grad_norm": 0.7163035273551941, + "learning_rate": 1.3464804469273742e-05, + "loss": 0.2103, + "step": 4021 + }, + { + "epoch": 11.23463687150838, + "grad_norm": 0.7969279289245605, + "learning_rate": 1.3468156424581007e-05, + "loss": 0.1944, + "step": 4022 + }, + { + "epoch": 11.237430167597765, + "grad_norm": 0.7951824069023132, + "learning_rate": 1.347150837988827e-05, + "loss": 0.2102, + "step": 4023 + }, + { + "epoch": 11.240223463687151, + "grad_norm": 0.7800863981246948, + "learning_rate": 1.3474860335195533e-05, + "loss": 0.2272, + "step": 4024 + }, + { + "epoch": 11.243016759776536, + "grad_norm": 0.8467881083488464, + "learning_rate": 1.3478212290502794e-05, + "loss": 0.1968, + "step": 4025 + }, + { + "epoch": 11.245810055865922, + "grad_norm": 0.6946139931678772, + "learning_rate": 1.3481564245810057e-05, + "loss": 0.2004, + "step": 4026 + }, + { + "epoch": 11.248603351955307, + "grad_norm": 0.8806976079940796, + "learning_rate": 1.348491620111732e-05, + "loss": 0.2106, + "step": 4027 + }, + { + "epoch": 11.251396648044693, + "grad_norm": 0.9260680675506592, + "learning_rate": 1.3488268156424581e-05, + "loss": 0.1938, + "step": 4028 + }, + { + "epoch": 11.254189944134078, + "grad_norm": 1.3296794891357422, + "learning_rate": 1.3491620111731844e-05, + "loss": 0.1669, + "step": 4029 + }, + { + "epoch": 11.256983240223464, + "grad_norm": 0.9570960402488708, + "learning_rate": 1.3494972067039107e-05, + "loss": 0.203, + "step": 4030 + }, + { + "epoch": 11.259776536312849, + "grad_norm": 1.0783292055130005, + "learning_rate": 1.3498324022346369e-05, + "loss": 0.2, + "step": 4031 + }, + { + "epoch": 11.262569832402235, + "grad_norm": 0.9381405711174011, + "learning_rate": 1.3501675977653632e-05, + "loss": 0.2106, + "step": 4032 + }, + { + "epoch": 11.26536312849162, + "grad_norm": 1.1600027084350586, + "learning_rate": 1.3505027932960895e-05, + "loss": 0.2053, + "step": 4033 + }, + { + "epoch": 11.268156424581006, + "grad_norm": 1.05730140209198, + "learning_rate": 1.3508379888268158e-05, + "loss": 0.194, + "step": 4034 + }, + { + "epoch": 11.27094972067039, + "grad_norm": 1.6548954248428345, + "learning_rate": 1.351173184357542e-05, + "loss": 0.2444, + "step": 4035 + }, + { + "epoch": 11.273743016759777, + "grad_norm": 0.9302733540534973, + "learning_rate": 1.3515083798882682e-05, + "loss": 0.2254, + "step": 4036 + }, + { + "epoch": 11.276536312849162, + "grad_norm": 1.6692345142364502, + "learning_rate": 1.3518435754189945e-05, + "loss": 0.2834, + "step": 4037 + }, + { + "epoch": 11.279329608938548, + "grad_norm": 3.446348190307617, + "learning_rate": 1.3521787709497207e-05, + "loss": 0.3035, + "step": 4038 + }, + { + "epoch": 11.282122905027933, + "grad_norm": 0.6124748587608337, + "learning_rate": 1.352513966480447e-05, + "loss": 0.2143, + "step": 4039 + }, + { + "epoch": 11.28491620111732, + "grad_norm": 0.5334251523017883, + "learning_rate": 1.3528491620111733e-05, + "loss": 0.2026, + "step": 4040 + }, + { + "epoch": 11.287709497206704, + "grad_norm": 0.5408712029457092, + "learning_rate": 1.3531843575418994e-05, + "loss": 0.2018, + "step": 4041 + }, + { + "epoch": 11.29050279329609, + "grad_norm": 0.5273786783218384, + "learning_rate": 1.3535195530726257e-05, + "loss": 0.1927, + "step": 4042 + }, + { + "epoch": 11.293296089385475, + "grad_norm": 0.6135551929473877, + "learning_rate": 1.353854748603352e-05, + "loss": 0.2174, + "step": 4043 + }, + { + "epoch": 11.296089385474861, + "grad_norm": 0.895301878452301, + "learning_rate": 1.3541899441340783e-05, + "loss": 0.2358, + "step": 4044 + }, + { + "epoch": 11.298882681564246, + "grad_norm": 0.6624784469604492, + "learning_rate": 1.3545251396648045e-05, + "loss": 0.2217, + "step": 4045 + }, + { + "epoch": 11.30167597765363, + "grad_norm": 0.5921584963798523, + "learning_rate": 1.3548603351955308e-05, + "loss": 0.1992, + "step": 4046 + }, + { + "epoch": 11.304469273743017, + "grad_norm": 0.7723993062973022, + "learning_rate": 1.355195530726257e-05, + "loss": 0.2131, + "step": 4047 + }, + { + "epoch": 11.307262569832401, + "grad_norm": 0.7874276638031006, + "learning_rate": 1.3555307262569832e-05, + "loss": 0.2165, + "step": 4048 + }, + { + "epoch": 11.310055865921788, + "grad_norm": 0.7755721807479858, + "learning_rate": 1.3558659217877095e-05, + "loss": 0.2, + "step": 4049 + }, + { + "epoch": 11.312849162011172, + "grad_norm": 1.0183528661727905, + "learning_rate": 1.3562011173184358e-05, + "loss": 0.2065, + "step": 4050 + }, + { + "epoch": 11.315642458100559, + "grad_norm": 0.7325595021247864, + "learning_rate": 1.356536312849162e-05, + "loss": 0.2046, + "step": 4051 + }, + { + "epoch": 11.318435754189943, + "grad_norm": 0.8873467445373535, + "learning_rate": 1.3568715083798882e-05, + "loss": 0.1944, + "step": 4052 + }, + { + "epoch": 11.32122905027933, + "grad_norm": 0.8404275178909302, + "learning_rate": 1.3572067039106146e-05, + "loss": 0.1774, + "step": 4053 + }, + { + "epoch": 11.324022346368714, + "grad_norm": 0.8309682011604309, + "learning_rate": 1.3575418994413409e-05, + "loss": 0.1797, + "step": 4054 + }, + { + "epoch": 11.3268156424581, + "grad_norm": 1.0275739431381226, + "learning_rate": 1.357877094972067e-05, + "loss": 0.2181, + "step": 4055 + }, + { + "epoch": 11.329608938547485, + "grad_norm": 0.7425996661186218, + "learning_rate": 1.3582122905027933e-05, + "loss": 0.1978, + "step": 4056 + }, + { + "epoch": 11.332402234636872, + "grad_norm": 0.9131567478179932, + "learning_rate": 1.3585474860335196e-05, + "loss": 0.2375, + "step": 4057 + }, + { + "epoch": 11.335195530726256, + "grad_norm": 1.495879888534546, + "learning_rate": 1.3588826815642457e-05, + "loss": 0.221, + "step": 4058 + }, + { + "epoch": 11.337988826815643, + "grad_norm": 0.8316770195960999, + "learning_rate": 1.359217877094972e-05, + "loss": 0.2006, + "step": 4059 + }, + { + "epoch": 11.340782122905027, + "grad_norm": 1.4018558263778687, + "learning_rate": 1.3595530726256983e-05, + "loss": 0.209, + "step": 4060 + }, + { + "epoch": 11.343575418994414, + "grad_norm": 1.6122995615005493, + "learning_rate": 1.3598882681564246e-05, + "loss": 0.2698, + "step": 4061 + }, + { + "epoch": 11.346368715083798, + "grad_norm": 1.8858078718185425, + "learning_rate": 1.3602234636871508e-05, + "loss": 0.2446, + "step": 4062 + }, + { + "epoch": 11.349162011173185, + "grad_norm": 9.590666770935059, + "learning_rate": 1.360558659217877e-05, + "loss": 0.3227, + "step": 4063 + }, + { + "epoch": 11.35195530726257, + "grad_norm": 0.5400997996330261, + "learning_rate": 1.3608938547486034e-05, + "loss": 0.2213, + "step": 4064 + }, + { + "epoch": 11.354748603351956, + "grad_norm": 0.8341555595397949, + "learning_rate": 1.3612290502793295e-05, + "loss": 0.2003, + "step": 4065 + }, + { + "epoch": 11.35754189944134, + "grad_norm": 0.7821341753005981, + "learning_rate": 1.3615642458100558e-05, + "loss": 0.178, + "step": 4066 + }, + { + "epoch": 11.360335195530727, + "grad_norm": 0.6204046010971069, + "learning_rate": 1.3618994413407821e-05, + "loss": 0.1875, + "step": 4067 + }, + { + "epoch": 11.363128491620111, + "grad_norm": 0.5709865093231201, + "learning_rate": 1.3622346368715083e-05, + "loss": 0.2154, + "step": 4068 + }, + { + "epoch": 11.365921787709498, + "grad_norm": 0.686971127986908, + "learning_rate": 1.3625698324022346e-05, + "loss": 0.2189, + "step": 4069 + }, + { + "epoch": 11.368715083798882, + "grad_norm": 0.5646462440490723, + "learning_rate": 1.3629050279329609e-05, + "loss": 0.2115, + "step": 4070 + }, + { + "epoch": 11.371508379888269, + "grad_norm": 1.1103320121765137, + "learning_rate": 1.3632402234636873e-05, + "loss": 0.2234, + "step": 4071 + }, + { + "epoch": 11.374301675977653, + "grad_norm": 1.6968952417373657, + "learning_rate": 1.3635754189944135e-05, + "loss": 0.1841, + "step": 4072 + }, + { + "epoch": 11.37709497206704, + "grad_norm": 1.6497113704681396, + "learning_rate": 1.3639106145251398e-05, + "loss": 0.2105, + "step": 4073 + }, + { + "epoch": 11.379888268156424, + "grad_norm": 0.7465371489524841, + "learning_rate": 1.3642458100558661e-05, + "loss": 0.2042, + "step": 4074 + }, + { + "epoch": 11.38268156424581, + "grad_norm": 0.6451117992401123, + "learning_rate": 1.3645810055865922e-05, + "loss": 0.2009, + "step": 4075 + }, + { + "epoch": 11.385474860335195, + "grad_norm": 0.859650731086731, + "learning_rate": 1.3649162011173185e-05, + "loss": 0.2069, + "step": 4076 + }, + { + "epoch": 11.388268156424582, + "grad_norm": 0.7531914114952087, + "learning_rate": 1.3652513966480448e-05, + "loss": 0.2146, + "step": 4077 + }, + { + "epoch": 11.391061452513966, + "grad_norm": 0.6082303524017334, + "learning_rate": 1.365586592178771e-05, + "loss": 0.1429, + "step": 4078 + }, + { + "epoch": 11.393854748603353, + "grad_norm": 1.1245766878128052, + "learning_rate": 1.3659217877094973e-05, + "loss": 0.1927, + "step": 4079 + }, + { + "epoch": 11.396648044692737, + "grad_norm": 1.6319048404693604, + "learning_rate": 1.3662569832402236e-05, + "loss": 0.1785, + "step": 4080 + }, + { + "epoch": 11.399441340782122, + "grad_norm": 1.044173240661621, + "learning_rate": 1.3665921787709499e-05, + "loss": 0.2336, + "step": 4081 + }, + { + "epoch": 11.402234636871508, + "grad_norm": 1.1060720682144165, + "learning_rate": 1.366927374301676e-05, + "loss": 0.2111, + "step": 4082 + }, + { + "epoch": 11.405027932960893, + "grad_norm": 1.5362995862960815, + "learning_rate": 1.3672625698324023e-05, + "loss": 0.2197, + "step": 4083 + }, + { + "epoch": 11.40782122905028, + "grad_norm": 1.6944321393966675, + "learning_rate": 1.3675977653631286e-05, + "loss": 0.2436, + "step": 4084 + }, + { + "epoch": 11.410614525139664, + "grad_norm": 1.2073578834533691, + "learning_rate": 1.3679329608938548e-05, + "loss": 0.2158, + "step": 4085 + }, + { + "epoch": 11.41340782122905, + "grad_norm": 2.788217306137085, + "learning_rate": 1.368268156424581e-05, + "loss": 0.2359, + "step": 4086 + }, + { + "epoch": 11.416201117318435, + "grad_norm": 1.4965609312057495, + "learning_rate": 1.3686033519553074e-05, + "loss": 0.2449, + "step": 4087 + }, + { + "epoch": 11.418994413407821, + "grad_norm": 2.0139005184173584, + "learning_rate": 1.3689385474860335e-05, + "loss": 0.3017, + "step": 4088 + }, + { + "epoch": 11.421787709497206, + "grad_norm": 0.4834789037704468, + "learning_rate": 1.3692737430167598e-05, + "loss": 0.1956, + "step": 4089 + }, + { + "epoch": 11.424581005586592, + "grad_norm": 0.75612872838974, + "learning_rate": 1.3696089385474861e-05, + "loss": 0.2346, + "step": 4090 + }, + { + "epoch": 11.427374301675977, + "grad_norm": 0.49596208333969116, + "learning_rate": 1.3699441340782124e-05, + "loss": 0.1709, + "step": 4091 + }, + { + "epoch": 11.430167597765363, + "grad_norm": 0.9730912446975708, + "learning_rate": 1.3702793296089385e-05, + "loss": 0.1987, + "step": 4092 + }, + { + "epoch": 11.432960893854748, + "grad_norm": 0.8223089575767517, + "learning_rate": 1.3706145251396648e-05, + "loss": 0.2598, + "step": 4093 + }, + { + "epoch": 11.435754189944134, + "grad_norm": 0.6973732113838196, + "learning_rate": 1.3709497206703912e-05, + "loss": 0.1677, + "step": 4094 + }, + { + "epoch": 11.438547486033519, + "grad_norm": 0.7725614905357361, + "learning_rate": 1.3712849162011173e-05, + "loss": 0.1855, + "step": 4095 + }, + { + "epoch": 11.441340782122905, + "grad_norm": 0.7731137871742249, + "learning_rate": 1.3716201117318436e-05, + "loss": 0.2319, + "step": 4096 + }, + { + "epoch": 11.44413407821229, + "grad_norm": 0.6625702977180481, + "learning_rate": 1.3719553072625699e-05, + "loss": 0.1924, + "step": 4097 + }, + { + "epoch": 11.446927374301676, + "grad_norm": 0.7633686065673828, + "learning_rate": 1.372290502793296e-05, + "loss": 0.2144, + "step": 4098 + }, + { + "epoch": 11.449720670391061, + "grad_norm": 0.8605445623397827, + "learning_rate": 1.3726256983240223e-05, + "loss": 0.2496, + "step": 4099 + }, + { + "epoch": 11.452513966480447, + "grad_norm": 0.7797690033912659, + "learning_rate": 1.3729608938547486e-05, + "loss": 0.1795, + "step": 4100 + }, + { + "epoch": 11.455307262569832, + "grad_norm": 0.6461515426635742, + "learning_rate": 1.373296089385475e-05, + "loss": 0.1787, + "step": 4101 + }, + { + "epoch": 11.458100558659218, + "grad_norm": 0.9607236385345459, + "learning_rate": 1.373631284916201e-05, + "loss": 0.2203, + "step": 4102 + }, + { + "epoch": 11.460893854748603, + "grad_norm": 2.292288303375244, + "learning_rate": 1.3739664804469274e-05, + "loss": 0.2085, + "step": 4103 + }, + { + "epoch": 11.46368715083799, + "grad_norm": 0.7937337160110474, + "learning_rate": 1.3743016759776537e-05, + "loss": 0.2009, + "step": 4104 + }, + { + "epoch": 11.466480446927374, + "grad_norm": 1.0182522535324097, + "learning_rate": 1.3746368715083798e-05, + "loss": 0.2439, + "step": 4105 + }, + { + "epoch": 11.46927374301676, + "grad_norm": 0.8636438846588135, + "learning_rate": 1.3749720670391061e-05, + "loss": 0.1722, + "step": 4106 + }, + { + "epoch": 11.472067039106145, + "grad_norm": 1.231769323348999, + "learning_rate": 1.3753072625698324e-05, + "loss": 0.2206, + "step": 4107 + }, + { + "epoch": 11.474860335195531, + "grad_norm": 0.9978561401367188, + "learning_rate": 1.3756424581005587e-05, + "loss": 0.1955, + "step": 4108 + }, + { + "epoch": 11.477653631284916, + "grad_norm": 0.8739961981773376, + "learning_rate": 1.3759776536312849e-05, + "loss": 0.2352, + "step": 4109 + }, + { + "epoch": 11.480446927374302, + "grad_norm": 1.0992028713226318, + "learning_rate": 1.3763128491620112e-05, + "loss": 0.2079, + "step": 4110 + }, + { + "epoch": 11.483240223463687, + "grad_norm": 1.2312241792678833, + "learning_rate": 1.3766480446927375e-05, + "loss": 0.1955, + "step": 4111 + }, + { + "epoch": 11.486033519553073, + "grad_norm": 2.173661470413208, + "learning_rate": 1.3769832402234636e-05, + "loss": 0.2455, + "step": 4112 + }, + { + "epoch": 11.488826815642458, + "grad_norm": 2.149138927459717, + "learning_rate": 1.3773184357541899e-05, + "loss": 0.352, + "step": 4113 + }, + { + "epoch": 11.491620111731844, + "grad_norm": 0.5806239247322083, + "learning_rate": 1.3776536312849162e-05, + "loss": 0.2048, + "step": 4114 + }, + { + "epoch": 11.494413407821229, + "grad_norm": 0.5933106541633606, + "learning_rate": 1.3779888268156423e-05, + "loss": 0.1966, + "step": 4115 + }, + { + "epoch": 11.497206703910614, + "grad_norm": 0.5334256291389465, + "learning_rate": 1.3783240223463687e-05, + "loss": 0.1896, + "step": 4116 + }, + { + "epoch": 11.5, + "grad_norm": 0.6325591206550598, + "learning_rate": 1.378659217877095e-05, + "loss": 0.2172, + "step": 4117 + }, + { + "epoch": 11.502793296089386, + "grad_norm": 0.6572709679603577, + "learning_rate": 1.3789944134078213e-05, + "loss": 0.2211, + "step": 4118 + }, + { + "epoch": 11.505586592178771, + "grad_norm": 1.018464207649231, + "learning_rate": 1.3793296089385476e-05, + "loss": 0.2335, + "step": 4119 + }, + { + "epoch": 11.508379888268156, + "grad_norm": 0.910568118095398, + "learning_rate": 1.3796648044692739e-05, + "loss": 0.2374, + "step": 4120 + }, + { + "epoch": 11.511173184357542, + "grad_norm": 0.6250022649765015, + "learning_rate": 1.3800000000000002e-05, + "loss": 0.2366, + "step": 4121 + }, + { + "epoch": 11.513966480446927, + "grad_norm": 0.5495905876159668, + "learning_rate": 1.3803351955307263e-05, + "loss": 0.1694, + "step": 4122 + }, + { + "epoch": 11.516759776536313, + "grad_norm": 0.7749677896499634, + "learning_rate": 1.3806703910614526e-05, + "loss": 0.1846, + "step": 4123 + }, + { + "epoch": 11.519553072625698, + "grad_norm": 0.7177734375, + "learning_rate": 1.3810055865921789e-05, + "loss": 0.2141, + "step": 4124 + }, + { + "epoch": 11.522346368715084, + "grad_norm": 0.6040139198303223, + "learning_rate": 1.381340782122905e-05, + "loss": 0.171, + "step": 4125 + }, + { + "epoch": 11.525139664804469, + "grad_norm": 0.7182684540748596, + "learning_rate": 1.3816759776536314e-05, + "loss": 0.1992, + "step": 4126 + }, + { + "epoch": 11.527932960893855, + "grad_norm": 1.1585952043533325, + "learning_rate": 1.3820111731843577e-05, + "loss": 0.2002, + "step": 4127 + }, + { + "epoch": 11.53072625698324, + "grad_norm": 0.8957107067108154, + "learning_rate": 1.382346368715084e-05, + "loss": 0.2168, + "step": 4128 + }, + { + "epoch": 11.533519553072626, + "grad_norm": 2.2711610794067383, + "learning_rate": 1.3826815642458101e-05, + "loss": 0.1924, + "step": 4129 + }, + { + "epoch": 11.53631284916201, + "grad_norm": 0.7632351517677307, + "learning_rate": 1.3830167597765364e-05, + "loss": 0.1822, + "step": 4130 + }, + { + "epoch": 11.539106145251397, + "grad_norm": 1.0028700828552246, + "learning_rate": 1.3833519553072627e-05, + "loss": 0.2124, + "step": 4131 + }, + { + "epoch": 11.541899441340782, + "grad_norm": 1.0352027416229248, + "learning_rate": 1.3836871508379888e-05, + "loss": 0.1973, + "step": 4132 + }, + { + "epoch": 11.544692737430168, + "grad_norm": 1.4457128047943115, + "learning_rate": 1.3840223463687151e-05, + "loss": 0.2163, + "step": 4133 + }, + { + "epoch": 11.547486033519553, + "grad_norm": 1.5017282962799072, + "learning_rate": 1.3843575418994414e-05, + "loss": 0.2113, + "step": 4134 + }, + { + "epoch": 11.550279329608939, + "grad_norm": 1.3494590520858765, + "learning_rate": 1.3846927374301676e-05, + "loss": 0.2129, + "step": 4135 + }, + { + "epoch": 11.553072625698324, + "grad_norm": 1.3025578260421753, + "learning_rate": 1.3850279329608939e-05, + "loss": 0.2119, + "step": 4136 + }, + { + "epoch": 11.55586592178771, + "grad_norm": 1.7305552959442139, + "learning_rate": 1.3853631284916202e-05, + "loss": 0.2383, + "step": 4137 + }, + { + "epoch": 11.558659217877095, + "grad_norm": 1.8851938247680664, + "learning_rate": 1.3856983240223465e-05, + "loss": 0.3381, + "step": 4138 + }, + { + "epoch": 11.561452513966481, + "grad_norm": 0.5047016739845276, + "learning_rate": 1.3860335195530726e-05, + "loss": 0.2023, + "step": 4139 + }, + { + "epoch": 11.564245810055866, + "grad_norm": 0.5914696455001831, + "learning_rate": 1.386368715083799e-05, + "loss": 0.2366, + "step": 4140 + }, + { + "epoch": 11.567039106145252, + "grad_norm": 0.6461318731307983, + "learning_rate": 1.3867039106145252e-05, + "loss": 0.1816, + "step": 4141 + }, + { + "epoch": 11.569832402234637, + "grad_norm": 0.707044243812561, + "learning_rate": 1.3870391061452514e-05, + "loss": 0.2215, + "step": 4142 + }, + { + "epoch": 11.572625698324023, + "grad_norm": 0.6381350159645081, + "learning_rate": 1.3873743016759777e-05, + "loss": 0.2158, + "step": 4143 + }, + { + "epoch": 11.575418994413408, + "grad_norm": 0.7071874737739563, + "learning_rate": 1.387709497206704e-05, + "loss": 0.2039, + "step": 4144 + }, + { + "epoch": 11.578212290502794, + "grad_norm": 0.7587505578994751, + "learning_rate": 1.3880446927374301e-05, + "loss": 0.2006, + "step": 4145 + }, + { + "epoch": 11.581005586592179, + "grad_norm": 0.723935067653656, + "learning_rate": 1.3883798882681564e-05, + "loss": 0.1984, + "step": 4146 + }, + { + "epoch": 11.583798882681565, + "grad_norm": 0.669569730758667, + "learning_rate": 1.3887150837988827e-05, + "loss": 0.1762, + "step": 4147 + }, + { + "epoch": 11.58659217877095, + "grad_norm": 0.809384286403656, + "learning_rate": 1.389050279329609e-05, + "loss": 0.2309, + "step": 4148 + }, + { + "epoch": 11.589385474860336, + "grad_norm": 0.9735549092292786, + "learning_rate": 1.3893854748603352e-05, + "loss": 0.2344, + "step": 4149 + }, + { + "epoch": 11.59217877094972, + "grad_norm": 0.6281841993331909, + "learning_rate": 1.3897206703910615e-05, + "loss": 0.1745, + "step": 4150 + }, + { + "epoch": 11.594972067039105, + "grad_norm": 1.3987245559692383, + "learning_rate": 1.3900558659217878e-05, + "loss": 0.2104, + "step": 4151 + }, + { + "epoch": 11.597765363128492, + "grad_norm": 0.8882980346679688, + "learning_rate": 1.3903910614525139e-05, + "loss": 0.1952, + "step": 4152 + }, + { + "epoch": 11.600558659217878, + "grad_norm": 0.8162394762039185, + "learning_rate": 1.3907262569832402e-05, + "loss": 0.1936, + "step": 4153 + }, + { + "epoch": 11.603351955307263, + "grad_norm": 0.7361008524894714, + "learning_rate": 1.3910614525139665e-05, + "loss": 0.174, + "step": 4154 + }, + { + "epoch": 11.606145251396647, + "grad_norm": 1.0889947414398193, + "learning_rate": 1.3913966480446928e-05, + "loss": 0.1979, + "step": 4155 + }, + { + "epoch": 11.608938547486034, + "grad_norm": 2.3793375492095947, + "learning_rate": 1.391731843575419e-05, + "loss": 0.2314, + "step": 4156 + }, + { + "epoch": 11.611731843575418, + "grad_norm": 0.8957722187042236, + "learning_rate": 1.3920670391061452e-05, + "loss": 0.2057, + "step": 4157 + }, + { + "epoch": 11.614525139664805, + "grad_norm": 1.4126979112625122, + "learning_rate": 1.3924022346368716e-05, + "loss": 0.2345, + "step": 4158 + }, + { + "epoch": 11.61731843575419, + "grad_norm": 1.3551032543182373, + "learning_rate": 1.3927374301675977e-05, + "loss": 0.2544, + "step": 4159 + }, + { + "epoch": 11.620111731843576, + "grad_norm": 1.5093094110488892, + "learning_rate": 1.393072625698324e-05, + "loss": 0.2275, + "step": 4160 + }, + { + "epoch": 11.62290502793296, + "grad_norm": 1.6668133735656738, + "learning_rate": 1.3934078212290503e-05, + "loss": 0.226, + "step": 4161 + }, + { + "epoch": 11.625698324022347, + "grad_norm": 1.6618421077728271, + "learning_rate": 1.3937430167597764e-05, + "loss": 0.287, + "step": 4162 + }, + { + "epoch": 11.628491620111731, + "grad_norm": 1.8601270914077759, + "learning_rate": 1.3940782122905027e-05, + "loss": 0.2879, + "step": 4163 + }, + { + "epoch": 11.631284916201118, + "grad_norm": 0.508267879486084, + "learning_rate": 1.394413407821229e-05, + "loss": 0.2249, + "step": 4164 + }, + { + "epoch": 11.634078212290502, + "grad_norm": 0.45978039503097534, + "learning_rate": 1.3947486033519553e-05, + "loss": 0.1879, + "step": 4165 + }, + { + "epoch": 11.636871508379889, + "grad_norm": 0.7328731417655945, + "learning_rate": 1.3950837988826815e-05, + "loss": 0.2124, + "step": 4166 + }, + { + "epoch": 11.639664804469273, + "grad_norm": 0.6718019247055054, + "learning_rate": 1.395418994413408e-05, + "loss": 0.1922, + "step": 4167 + }, + { + "epoch": 11.64245810055866, + "grad_norm": 0.6805366277694702, + "learning_rate": 1.3957541899441343e-05, + "loss": 0.2191, + "step": 4168 + }, + { + "epoch": 11.645251396648044, + "grad_norm": 0.6413515210151672, + "learning_rate": 1.3960893854748604e-05, + "loss": 0.2133, + "step": 4169 + }, + { + "epoch": 11.64804469273743, + "grad_norm": 0.6488085389137268, + "learning_rate": 1.3964245810055867e-05, + "loss": 0.233, + "step": 4170 + }, + { + "epoch": 11.650837988826815, + "grad_norm": 0.5461202263832092, + "learning_rate": 1.396759776536313e-05, + "loss": 0.2024, + "step": 4171 + }, + { + "epoch": 11.653631284916202, + "grad_norm": 0.6749232411384583, + "learning_rate": 1.3970949720670391e-05, + "loss": 0.1808, + "step": 4172 + }, + { + "epoch": 11.656424581005586, + "grad_norm": 0.8528192043304443, + "learning_rate": 1.3974301675977654e-05, + "loss": 0.2114, + "step": 4173 + }, + { + "epoch": 11.659217877094973, + "grad_norm": 0.7200974225997925, + "learning_rate": 1.3977653631284917e-05, + "loss": 0.21, + "step": 4174 + }, + { + "epoch": 11.662011173184357, + "grad_norm": 1.1491097211837769, + "learning_rate": 1.398100558659218e-05, + "loss": 0.2005, + "step": 4175 + }, + { + "epoch": 11.664804469273744, + "grad_norm": 0.7287048697471619, + "learning_rate": 1.3984357541899442e-05, + "loss": 0.2163, + "step": 4176 + }, + { + "epoch": 11.667597765363128, + "grad_norm": 0.7344403862953186, + "learning_rate": 1.3987709497206705e-05, + "loss": 0.2293, + "step": 4177 + }, + { + "epoch": 11.670391061452515, + "grad_norm": 0.756414532661438, + "learning_rate": 1.3991061452513968e-05, + "loss": 0.2586, + "step": 4178 + }, + { + "epoch": 11.6731843575419, + "grad_norm": 1.0490955114364624, + "learning_rate": 1.399441340782123e-05, + "loss": 0.2045, + "step": 4179 + }, + { + "epoch": 11.675977653631286, + "grad_norm": 0.8149511218070984, + "learning_rate": 1.3997765363128492e-05, + "loss": 0.2238, + "step": 4180 + }, + { + "epoch": 11.67877094972067, + "grad_norm": 0.877972424030304, + "learning_rate": 1.4001117318435755e-05, + "loss": 0.2147, + "step": 4181 + }, + { + "epoch": 11.681564245810057, + "grad_norm": 1.3454962968826294, + "learning_rate": 1.4004469273743017e-05, + "loss": 0.1942, + "step": 4182 + }, + { + "epoch": 11.684357541899441, + "grad_norm": 0.8054750561714172, + "learning_rate": 1.400782122905028e-05, + "loss": 0.2187, + "step": 4183 + }, + { + "epoch": 11.687150837988828, + "grad_norm": 0.9039391875267029, + "learning_rate": 1.4011173184357543e-05, + "loss": 0.2211, + "step": 4184 + }, + { + "epoch": 11.689944134078212, + "grad_norm": 1.1334691047668457, + "learning_rate": 1.4014525139664806e-05, + "loss": 0.204, + "step": 4185 + }, + { + "epoch": 11.692737430167599, + "grad_norm": 1.4050031900405884, + "learning_rate": 1.4017877094972067e-05, + "loss": 0.2864, + "step": 4186 + }, + { + "epoch": 11.695530726256983, + "grad_norm": 1.7209722995758057, + "learning_rate": 1.402122905027933e-05, + "loss": 0.2601, + "step": 4187 + }, + { + "epoch": 11.69832402234637, + "grad_norm": 2.1883223056793213, + "learning_rate": 1.4024581005586593e-05, + "loss": 0.3187, + "step": 4188 + }, + { + "epoch": 11.701117318435754, + "grad_norm": 0.38740476965904236, + "learning_rate": 1.4027932960893854e-05, + "loss": 0.1657, + "step": 4189 + }, + { + "epoch": 11.703910614525139, + "grad_norm": 0.4551706612110138, + "learning_rate": 1.4031284916201118e-05, + "loss": 0.2075, + "step": 4190 + }, + { + "epoch": 11.706703910614525, + "grad_norm": 0.5100874304771423, + "learning_rate": 1.403463687150838e-05, + "loss": 0.1964, + "step": 4191 + }, + { + "epoch": 11.70949720670391, + "grad_norm": 0.5064406991004944, + "learning_rate": 1.4037988826815642e-05, + "loss": 0.2015, + "step": 4192 + }, + { + "epoch": 11.712290502793296, + "grad_norm": 0.815907895565033, + "learning_rate": 1.4041340782122905e-05, + "loss": 0.2005, + "step": 4193 + }, + { + "epoch": 11.71508379888268, + "grad_norm": 0.6854702234268188, + "learning_rate": 1.4044692737430168e-05, + "loss": 0.2391, + "step": 4194 + }, + { + "epoch": 11.717877094972067, + "grad_norm": 0.6241378784179688, + "learning_rate": 1.4048044692737431e-05, + "loss": 0.1993, + "step": 4195 + }, + { + "epoch": 11.720670391061452, + "grad_norm": 0.6707154512405396, + "learning_rate": 1.4051396648044692e-05, + "loss": 0.1825, + "step": 4196 + }, + { + "epoch": 11.723463687150838, + "grad_norm": 2.080415964126587, + "learning_rate": 1.4054748603351955e-05, + "loss": 0.1854, + "step": 4197 + }, + { + "epoch": 11.726256983240223, + "grad_norm": 1.1379252672195435, + "learning_rate": 1.4058100558659218e-05, + "loss": 0.214, + "step": 4198 + }, + { + "epoch": 11.72905027932961, + "grad_norm": 0.8462045788764954, + "learning_rate": 1.406145251396648e-05, + "loss": 0.2211, + "step": 4199 + }, + { + "epoch": 11.731843575418994, + "grad_norm": 0.9694442749023438, + "learning_rate": 1.4064804469273743e-05, + "loss": 0.1939, + "step": 4200 + }, + { + "epoch": 11.73463687150838, + "grad_norm": 0.6312135457992554, + "learning_rate": 1.4068156424581006e-05, + "loss": 0.2177, + "step": 4201 + }, + { + "epoch": 11.737430167597765, + "grad_norm": 1.4037820100784302, + "learning_rate": 1.4071508379888269e-05, + "loss": 0.1966, + "step": 4202 + }, + { + "epoch": 11.740223463687151, + "grad_norm": 1.747483253479004, + "learning_rate": 1.407486033519553e-05, + "loss": 0.2397, + "step": 4203 + }, + { + "epoch": 11.743016759776536, + "grad_norm": 1.050775170326233, + "learning_rate": 1.4078212290502793e-05, + "loss": 0.2233, + "step": 4204 + }, + { + "epoch": 11.745810055865922, + "grad_norm": 0.7118331789970398, + "learning_rate": 1.4081564245810056e-05, + "loss": 0.2025, + "step": 4205 + }, + { + "epoch": 11.748603351955307, + "grad_norm": 1.115011215209961, + "learning_rate": 1.4084916201117318e-05, + "loss": 0.2004, + "step": 4206 + }, + { + "epoch": 11.751396648044693, + "grad_norm": 1.2401782274246216, + "learning_rate": 1.408826815642458e-05, + "loss": 0.2287, + "step": 4207 + }, + { + "epoch": 11.754189944134078, + "grad_norm": 0.9934053421020508, + "learning_rate": 1.4091620111731844e-05, + "loss": 0.2332, + "step": 4208 + }, + { + "epoch": 11.756983240223464, + "grad_norm": 1.2804861068725586, + "learning_rate": 1.4094972067039105e-05, + "loss": 0.21, + "step": 4209 + }, + { + "epoch": 11.759776536312849, + "grad_norm": 1.0978848934173584, + "learning_rate": 1.4098324022346368e-05, + "loss": 0.213, + "step": 4210 + }, + { + "epoch": 11.762569832402235, + "grad_norm": 1.0513588190078735, + "learning_rate": 1.4101675977653631e-05, + "loss": 0.2048, + "step": 4211 + }, + { + "epoch": 11.76536312849162, + "grad_norm": 1.1346147060394287, + "learning_rate": 1.4105027932960894e-05, + "loss": 0.2035, + "step": 4212 + }, + { + "epoch": 11.768156424581006, + "grad_norm": 1.9029372930526733, + "learning_rate": 1.4108379888268156e-05, + "loss": 0.2815, + "step": 4213 + }, + { + "epoch": 11.77094972067039, + "grad_norm": 0.5316303372383118, + "learning_rate": 1.4111731843575419e-05, + "loss": 0.2506, + "step": 4214 + }, + { + "epoch": 11.773743016759777, + "grad_norm": 0.47003406286239624, + "learning_rate": 1.4115083798882683e-05, + "loss": 0.1809, + "step": 4215 + }, + { + "epoch": 11.776536312849162, + "grad_norm": 0.5904784202575684, + "learning_rate": 1.4118435754189945e-05, + "loss": 0.1998, + "step": 4216 + }, + { + "epoch": 11.779329608938548, + "grad_norm": 0.8915541172027588, + "learning_rate": 1.4121787709497208e-05, + "loss": 0.2066, + "step": 4217 + }, + { + "epoch": 11.782122905027933, + "grad_norm": 0.6360481381416321, + "learning_rate": 1.412513966480447e-05, + "loss": 0.2003, + "step": 4218 + }, + { + "epoch": 11.78491620111732, + "grad_norm": 0.6991106271743774, + "learning_rate": 1.4128491620111732e-05, + "loss": 0.1709, + "step": 4219 + }, + { + "epoch": 11.787709497206704, + "grad_norm": 0.6568682789802551, + "learning_rate": 1.4131843575418995e-05, + "loss": 0.2106, + "step": 4220 + }, + { + "epoch": 11.79050279329609, + "grad_norm": 0.7400347590446472, + "learning_rate": 1.4135195530726258e-05, + "loss": 0.2443, + "step": 4221 + }, + { + "epoch": 11.793296089385475, + "grad_norm": 0.6935071349143982, + "learning_rate": 1.4138547486033521e-05, + "loss": 0.1775, + "step": 4222 + }, + { + "epoch": 11.796089385474861, + "grad_norm": 1.1344391107559204, + "learning_rate": 1.4141899441340783e-05, + "loss": 0.2094, + "step": 4223 + }, + { + "epoch": 11.798882681564246, + "grad_norm": 0.9015337824821472, + "learning_rate": 1.4145251396648046e-05, + "loss": 0.2235, + "step": 4224 + }, + { + "epoch": 11.80167597765363, + "grad_norm": 1.1837191581726074, + "learning_rate": 1.4148603351955309e-05, + "loss": 0.2044, + "step": 4225 + }, + { + "epoch": 11.804469273743017, + "grad_norm": 1.2922115325927734, + "learning_rate": 1.415195530726257e-05, + "loss": 0.2073, + "step": 4226 + }, + { + "epoch": 11.807262569832401, + "grad_norm": 0.6188051104545593, + "learning_rate": 1.4155307262569833e-05, + "loss": 0.1901, + "step": 4227 + }, + { + "epoch": 11.810055865921788, + "grad_norm": 1.0663951635360718, + "learning_rate": 1.4158659217877096e-05, + "loss": 0.2193, + "step": 4228 + }, + { + "epoch": 11.812849162011172, + "grad_norm": 1.6659513711929321, + "learning_rate": 1.4162011173184357e-05, + "loss": 0.2457, + "step": 4229 + }, + { + "epoch": 11.815642458100559, + "grad_norm": 1.0195322036743164, + "learning_rate": 1.416536312849162e-05, + "loss": 0.2049, + "step": 4230 + }, + { + "epoch": 11.818435754189943, + "grad_norm": 0.7565490007400513, + "learning_rate": 1.4168715083798884e-05, + "loss": 0.2138, + "step": 4231 + }, + { + "epoch": 11.82122905027933, + "grad_norm": 0.8204647898674011, + "learning_rate": 1.4172067039106147e-05, + "loss": 0.2312, + "step": 4232 + }, + { + "epoch": 11.824022346368714, + "grad_norm": 1.0051945447921753, + "learning_rate": 1.4175418994413408e-05, + "loss": 0.2022, + "step": 4233 + }, + { + "epoch": 11.8268156424581, + "grad_norm": 1.1961734294891357, + "learning_rate": 1.4178770949720671e-05, + "loss": 0.2467, + "step": 4234 + }, + { + "epoch": 11.829608938547485, + "grad_norm": 1.3616344928741455, + "learning_rate": 1.4182122905027934e-05, + "loss": 0.264, + "step": 4235 + }, + { + "epoch": 11.832402234636872, + "grad_norm": 1.9246386289596558, + "learning_rate": 1.4185474860335195e-05, + "loss": 0.2427, + "step": 4236 + }, + { + "epoch": 11.835195530726256, + "grad_norm": 1.9286946058273315, + "learning_rate": 1.4188826815642458e-05, + "loss": 0.2526, + "step": 4237 + }, + { + "epoch": 11.837988826815643, + "grad_norm": 2.7604496479034424, + "learning_rate": 1.4192178770949721e-05, + "loss": 0.3699, + "step": 4238 + }, + { + "epoch": 11.840782122905027, + "grad_norm": 0.4239022731781006, + "learning_rate": 1.4195530726256983e-05, + "loss": 0.195, + "step": 4239 + }, + { + "epoch": 11.843575418994414, + "grad_norm": 0.5617244839668274, + "learning_rate": 1.4198882681564246e-05, + "loss": 0.2226, + "step": 4240 + }, + { + "epoch": 11.846368715083798, + "grad_norm": 0.5908072590827942, + "learning_rate": 1.4202234636871509e-05, + "loss": 0.1947, + "step": 4241 + }, + { + "epoch": 11.849162011173185, + "grad_norm": 0.6268956661224365, + "learning_rate": 1.4205586592178772e-05, + "loss": 0.2335, + "step": 4242 + }, + { + "epoch": 11.85195530726257, + "grad_norm": 0.503300666809082, + "learning_rate": 1.4208938547486033e-05, + "loss": 0.1855, + "step": 4243 + }, + { + "epoch": 11.854748603351956, + "grad_norm": 1.0256236791610718, + "learning_rate": 1.4212290502793296e-05, + "loss": 0.2131, + "step": 4244 + }, + { + "epoch": 11.85754189944134, + "grad_norm": 0.5096432566642761, + "learning_rate": 1.421564245810056e-05, + "loss": 0.2052, + "step": 4245 + }, + { + "epoch": 11.860335195530727, + "grad_norm": 0.7600059509277344, + "learning_rate": 1.421899441340782e-05, + "loss": 0.1866, + "step": 4246 + }, + { + "epoch": 11.863128491620111, + "grad_norm": 0.8974477648735046, + "learning_rate": 1.4222346368715084e-05, + "loss": 0.2005, + "step": 4247 + }, + { + "epoch": 11.865921787709498, + "grad_norm": 0.7566686272621155, + "learning_rate": 1.4225698324022347e-05, + "loss": 0.2277, + "step": 4248 + }, + { + "epoch": 11.868715083798882, + "grad_norm": 0.6358734965324402, + "learning_rate": 1.422905027932961e-05, + "loss": 0.1852, + "step": 4249 + }, + { + "epoch": 11.871508379888269, + "grad_norm": 0.6614810228347778, + "learning_rate": 1.4232402234636871e-05, + "loss": 0.2048, + "step": 4250 + }, + { + "epoch": 11.874301675977653, + "grad_norm": 0.9977600574493408, + "learning_rate": 1.4235754189944134e-05, + "loss": 0.2004, + "step": 4251 + }, + { + "epoch": 11.87709497206704, + "grad_norm": 0.8741793036460876, + "learning_rate": 1.4239106145251397e-05, + "loss": 0.1866, + "step": 4252 + }, + { + "epoch": 11.879888268156424, + "grad_norm": 0.867394745349884, + "learning_rate": 1.4242458100558659e-05, + "loss": 0.1867, + "step": 4253 + }, + { + "epoch": 11.88268156424581, + "grad_norm": 1.0987889766693115, + "learning_rate": 1.4245810055865922e-05, + "loss": 0.1948, + "step": 4254 + }, + { + "epoch": 11.885474860335195, + "grad_norm": 0.7223462462425232, + "learning_rate": 1.4249162011173185e-05, + "loss": 0.2017, + "step": 4255 + }, + { + "epoch": 11.888268156424582, + "grad_norm": 0.9895256757736206, + "learning_rate": 1.4252513966480446e-05, + "loss": 0.1964, + "step": 4256 + }, + { + "epoch": 11.891061452513966, + "grad_norm": 1.1350228786468506, + "learning_rate": 1.4255865921787709e-05, + "loss": 0.2073, + "step": 4257 + }, + { + "epoch": 11.893854748603353, + "grad_norm": 0.8526801466941833, + "learning_rate": 1.4259217877094972e-05, + "loss": 0.2276, + "step": 4258 + }, + { + "epoch": 11.896648044692737, + "grad_norm": 0.9498152732849121, + "learning_rate": 1.4262569832402235e-05, + "loss": 0.2111, + "step": 4259 + }, + { + "epoch": 11.899441340782122, + "grad_norm": 1.1576930284500122, + "learning_rate": 1.4265921787709496e-05, + "loss": 0.2134, + "step": 4260 + }, + { + "epoch": 11.902234636871508, + "grad_norm": 1.2781752347946167, + "learning_rate": 1.426927374301676e-05, + "loss": 0.2392, + "step": 4261 + }, + { + "epoch": 11.905027932960895, + "grad_norm": 2.6240437030792236, + "learning_rate": 1.4272625698324022e-05, + "loss": 0.2294, + "step": 4262 + }, + { + "epoch": 11.90782122905028, + "grad_norm": 1.7114840745925903, + "learning_rate": 1.4275977653631286e-05, + "loss": 0.3211, + "step": 4263 + }, + { + "epoch": 11.910614525139664, + "grad_norm": 0.4784245491027832, + "learning_rate": 1.4279329608938549e-05, + "loss": 0.2594, + "step": 4264 + }, + { + "epoch": 11.91340782122905, + "grad_norm": 0.5431984663009644, + "learning_rate": 1.4282681564245812e-05, + "loss": 0.1998, + "step": 4265 + }, + { + "epoch": 11.916201117318435, + "grad_norm": 0.539050817489624, + "learning_rate": 1.4286033519553073e-05, + "loss": 0.183, + "step": 4266 + }, + { + "epoch": 11.918994413407821, + "grad_norm": 0.7266884446144104, + "learning_rate": 1.4289385474860336e-05, + "loss": 0.1853, + "step": 4267 + }, + { + "epoch": 11.921787709497206, + "grad_norm": 0.5976480841636658, + "learning_rate": 1.4292737430167599e-05, + "loss": 0.238, + "step": 4268 + }, + { + "epoch": 11.924581005586592, + "grad_norm": 0.5746777653694153, + "learning_rate": 1.4296089385474862e-05, + "loss": 0.2145, + "step": 4269 + }, + { + "epoch": 11.927374301675977, + "grad_norm": 0.8020143508911133, + "learning_rate": 1.4299441340782123e-05, + "loss": 0.1932, + "step": 4270 + }, + { + "epoch": 11.930167597765363, + "grad_norm": 0.5837132930755615, + "learning_rate": 1.4302793296089386e-05, + "loss": 0.2087, + "step": 4271 + }, + { + "epoch": 11.932960893854748, + "grad_norm": 0.6461705565452576, + "learning_rate": 1.430614525139665e-05, + "loss": 0.186, + "step": 4272 + }, + { + "epoch": 11.935754189944134, + "grad_norm": 0.6053308844566345, + "learning_rate": 1.430949720670391e-05, + "loss": 0.1837, + "step": 4273 + }, + { + "epoch": 11.938547486033519, + "grad_norm": 0.6520276069641113, + "learning_rate": 1.4312849162011174e-05, + "loss": 0.227, + "step": 4274 + }, + { + "epoch": 11.941340782122905, + "grad_norm": 0.6076303124427795, + "learning_rate": 1.4316201117318437e-05, + "loss": 0.2252, + "step": 4275 + }, + { + "epoch": 11.94413407821229, + "grad_norm": 0.7426614165306091, + "learning_rate": 1.4319553072625698e-05, + "loss": 0.1978, + "step": 4276 + }, + { + "epoch": 11.946927374301676, + "grad_norm": 0.7093616127967834, + "learning_rate": 1.4322905027932961e-05, + "loss": 0.2013, + "step": 4277 + }, + { + "epoch": 11.949720670391061, + "grad_norm": 0.7004756927490234, + "learning_rate": 1.4326256983240224e-05, + "loss": 0.1852, + "step": 4278 + }, + { + "epoch": 11.952513966480447, + "grad_norm": 1.1242246627807617, + "learning_rate": 1.4329608938547487e-05, + "loss": 0.2077, + "step": 4279 + }, + { + "epoch": 11.955307262569832, + "grad_norm": 0.6972532868385315, + "learning_rate": 1.4332960893854749e-05, + "loss": 0.2176, + "step": 4280 + }, + { + "epoch": 11.958100558659218, + "grad_norm": 0.7003270387649536, + "learning_rate": 1.4336312849162012e-05, + "loss": 0.1777, + "step": 4281 + }, + { + "epoch": 11.960893854748603, + "grad_norm": 1.0098588466644287, + "learning_rate": 1.4339664804469275e-05, + "loss": 0.2224, + "step": 4282 + }, + { + "epoch": 11.96368715083799, + "grad_norm": 1.3091596364974976, + "learning_rate": 1.4343016759776536e-05, + "loss": 0.2335, + "step": 4283 + }, + { + "epoch": 11.966480446927374, + "grad_norm": 1.2631763219833374, + "learning_rate": 1.43463687150838e-05, + "loss": 0.2387, + "step": 4284 + }, + { + "epoch": 11.96927374301676, + "grad_norm": 1.6533713340759277, + "learning_rate": 1.4349720670391062e-05, + "loss": 0.1976, + "step": 4285 + }, + { + "epoch": 11.972067039106145, + "grad_norm": 1.4780359268188477, + "learning_rate": 1.4353072625698324e-05, + "loss": 0.2111, + "step": 4286 + }, + { + "epoch": 11.974860335195531, + "grad_norm": 1.5817300081253052, + "learning_rate": 1.4356424581005587e-05, + "loss": 0.2785, + "step": 4287 + }, + { + "epoch": 11.977653631284916, + "grad_norm": 2.3832106590270996, + "learning_rate": 1.435977653631285e-05, + "loss": 0.3163, + "step": 4288 + }, + { + "epoch": 11.980446927374302, + "grad_norm": 0.5442178249359131, + "learning_rate": 1.4363128491620113e-05, + "loss": 0.2104, + "step": 4289 + }, + { + "epoch": 11.983240223463687, + "grad_norm": 0.6855880618095398, + "learning_rate": 1.4366480446927374e-05, + "loss": 0.187, + "step": 4290 + }, + { + "epoch": 11.986033519553073, + "grad_norm": 1.227260708808899, + "learning_rate": 1.4369832402234637e-05, + "loss": 0.1852, + "step": 4291 + }, + { + "epoch": 11.988826815642458, + "grad_norm": 1.145648717880249, + "learning_rate": 1.43731843575419e-05, + "loss": 0.2321, + "step": 4292 + }, + { + "epoch": 11.991620111731844, + "grad_norm": 0.7923678159713745, + "learning_rate": 1.4376536312849161e-05, + "loss": 0.1831, + "step": 4293 + }, + { + "epoch": 11.994413407821229, + "grad_norm": 0.8593906164169312, + "learning_rate": 1.4379888268156425e-05, + "loss": 0.1909, + "step": 4294 + }, + { + "epoch": 11.997206703910614, + "grad_norm": 0.9124041199684143, + "learning_rate": 1.4383240223463688e-05, + "loss": 0.2008, + "step": 4295 + }, + { + "epoch": 12.0, + "grad_norm": 2.575317621231079, + "learning_rate": 1.438659217877095e-05, + "loss": 0.2231, + "step": 4296 + }, + { + "epoch": 12.002793296089385, + "grad_norm": 0.5701280236244202, + "learning_rate": 1.4389944134078212e-05, + "loss": 0.1926, + "step": 4297 + }, + { + "epoch": 12.005586592178771, + "grad_norm": 0.5287161469459534, + "learning_rate": 1.4393296089385475e-05, + "loss": 0.1784, + "step": 4298 + }, + { + "epoch": 12.008379888268156, + "grad_norm": 0.5811320543289185, + "learning_rate": 1.4396648044692738e-05, + "loss": 0.2007, + "step": 4299 + }, + { + "epoch": 12.011173184357542, + "grad_norm": 0.5460929274559021, + "learning_rate": 1.44e-05, + "loss": 0.207, + "step": 4300 + }, + { + "epoch": 12.013966480446927, + "grad_norm": 0.5884343981742859, + "learning_rate": 1.4403351955307262e-05, + "loss": 0.1679, + "step": 4301 + }, + { + "epoch": 12.016759776536313, + "grad_norm": 0.6230310201644897, + "learning_rate": 1.4406703910614525e-05, + "loss": 0.196, + "step": 4302 + }, + { + "epoch": 12.019553072625698, + "grad_norm": 0.818419873714447, + "learning_rate": 1.4410055865921787e-05, + "loss": 0.1936, + "step": 4303 + }, + { + "epoch": 12.022346368715084, + "grad_norm": 0.6220111846923828, + "learning_rate": 1.441340782122905e-05, + "loss": 0.1939, + "step": 4304 + }, + { + "epoch": 12.025139664804469, + "grad_norm": 0.5393975973129272, + "learning_rate": 1.4416759776536313e-05, + "loss": 0.1807, + "step": 4305 + }, + { + "epoch": 12.027932960893855, + "grad_norm": 0.8132918477058411, + "learning_rate": 1.4420111731843576e-05, + "loss": 0.2136, + "step": 4306 + }, + { + "epoch": 12.03072625698324, + "grad_norm": 0.7404342889785767, + "learning_rate": 1.4423463687150837e-05, + "loss": 0.2078, + "step": 4307 + }, + { + "epoch": 12.033519553072626, + "grad_norm": 0.6111890077590942, + "learning_rate": 1.44268156424581e-05, + "loss": 0.2208, + "step": 4308 + }, + { + "epoch": 12.03631284916201, + "grad_norm": 0.620938241481781, + "learning_rate": 1.4430167597765363e-05, + "loss": 0.194, + "step": 4309 + }, + { + "epoch": 12.039106145251397, + "grad_norm": 0.7607131004333496, + "learning_rate": 1.4433519553072625e-05, + "loss": 0.2376, + "step": 4310 + }, + { + "epoch": 12.041899441340782, + "grad_norm": 0.9670799374580383, + "learning_rate": 1.443687150837989e-05, + "loss": 0.1859, + "step": 4311 + }, + { + "epoch": 12.044692737430168, + "grad_norm": 0.712932288646698, + "learning_rate": 1.4440223463687152e-05, + "loss": 0.1874, + "step": 4312 + }, + { + "epoch": 12.047486033519553, + "grad_norm": 0.8977457880973816, + "learning_rate": 1.4443575418994414e-05, + "loss": 0.1848, + "step": 4313 + }, + { + "epoch": 12.050279329608939, + "grad_norm": 0.9252563714981079, + "learning_rate": 1.4446927374301677e-05, + "loss": 0.1739, + "step": 4314 + }, + { + "epoch": 12.053072625698324, + "grad_norm": 0.9187907576560974, + "learning_rate": 1.445027932960894e-05, + "loss": 0.1898, + "step": 4315 + }, + { + "epoch": 12.05586592178771, + "grad_norm": 1.406650185585022, + "learning_rate": 1.4453631284916203e-05, + "loss": 0.2103, + "step": 4316 + }, + { + "epoch": 12.058659217877095, + "grad_norm": 1.4319268465042114, + "learning_rate": 1.4456983240223464e-05, + "loss": 0.2154, + "step": 4317 + }, + { + "epoch": 12.061452513966481, + "grad_norm": 1.2332234382629395, + "learning_rate": 1.4460335195530727e-05, + "loss": 0.2067, + "step": 4318 + }, + { + "epoch": 12.064245810055866, + "grad_norm": 1.0024884939193726, + "learning_rate": 1.446368715083799e-05, + "loss": 0.2146, + "step": 4319 + }, + { + "epoch": 12.067039106145252, + "grad_norm": 1.474947452545166, + "learning_rate": 1.4467039106145252e-05, + "loss": 0.267, + "step": 4320 + }, + { + "epoch": 12.069832402234637, + "grad_norm": 3.249751091003418, + "learning_rate": 1.4470391061452515e-05, + "loss": 0.3032, + "step": 4321 + }, + { + "epoch": 12.072625698324023, + "grad_norm": 0.56801837682724, + "learning_rate": 1.4473743016759778e-05, + "loss": 0.2056, + "step": 4322 + }, + { + "epoch": 12.075418994413408, + "grad_norm": 0.7008383274078369, + "learning_rate": 1.4477094972067039e-05, + "loss": 0.2344, + "step": 4323 + }, + { + "epoch": 12.078212290502794, + "grad_norm": 0.506429135799408, + "learning_rate": 1.4480446927374302e-05, + "loss": 0.207, + "step": 4324 + }, + { + "epoch": 12.081005586592179, + "grad_norm": 0.595622181892395, + "learning_rate": 1.4483798882681565e-05, + "loss": 0.1875, + "step": 4325 + }, + { + "epoch": 12.083798882681565, + "grad_norm": 0.8263293504714966, + "learning_rate": 1.4487150837988828e-05, + "loss": 0.2088, + "step": 4326 + }, + { + "epoch": 12.08659217877095, + "grad_norm": 0.597183346748352, + "learning_rate": 1.449050279329609e-05, + "loss": 0.2099, + "step": 4327 + }, + { + "epoch": 12.089385474860336, + "grad_norm": 0.9108871817588806, + "learning_rate": 1.4493854748603353e-05, + "loss": 0.2103, + "step": 4328 + }, + { + "epoch": 12.09217877094972, + "grad_norm": 0.5984592437744141, + "learning_rate": 1.4497206703910616e-05, + "loss": 0.1902, + "step": 4329 + }, + { + "epoch": 12.094972067039107, + "grad_norm": 0.7433319687843323, + "learning_rate": 1.4500558659217877e-05, + "loss": 0.1918, + "step": 4330 + }, + { + "epoch": 12.097765363128492, + "grad_norm": 0.7343323230743408, + "learning_rate": 1.450391061452514e-05, + "loss": 0.1679, + "step": 4331 + }, + { + "epoch": 12.100558659217878, + "grad_norm": 0.5343948006629944, + "learning_rate": 1.4507262569832403e-05, + "loss": 0.1696, + "step": 4332 + }, + { + "epoch": 12.103351955307263, + "grad_norm": 0.6162701845169067, + "learning_rate": 1.4510614525139664e-05, + "loss": 0.1857, + "step": 4333 + }, + { + "epoch": 12.106145251396647, + "grad_norm": 0.8178154230117798, + "learning_rate": 1.4513966480446927e-05, + "loss": 0.1786, + "step": 4334 + }, + { + "epoch": 12.108938547486034, + "grad_norm": 0.7007404565811157, + "learning_rate": 1.451731843575419e-05, + "loss": 0.2113, + "step": 4335 + }, + { + "epoch": 12.111731843575418, + "grad_norm": 0.6311195492744446, + "learning_rate": 1.4520670391061454e-05, + "loss": 0.1931, + "step": 4336 + }, + { + "epoch": 12.114525139664805, + "grad_norm": 1.0090184211730957, + "learning_rate": 1.4524022346368715e-05, + "loss": 0.1736, + "step": 4337 + }, + { + "epoch": 12.11731843575419, + "grad_norm": 0.7271620035171509, + "learning_rate": 1.4527374301675978e-05, + "loss": 0.2244, + "step": 4338 + }, + { + "epoch": 12.120111731843576, + "grad_norm": 1.0456836223602295, + "learning_rate": 1.4530726256983241e-05, + "loss": 0.1765, + "step": 4339 + }, + { + "epoch": 12.12290502793296, + "grad_norm": 0.967180609703064, + "learning_rate": 1.4534078212290502e-05, + "loss": 0.1768, + "step": 4340 + }, + { + "epoch": 12.125698324022347, + "grad_norm": 0.8857141733169556, + "learning_rate": 1.4537430167597765e-05, + "loss": 0.1984, + "step": 4341 + }, + { + "epoch": 12.128491620111731, + "grad_norm": 1.155062198638916, + "learning_rate": 1.4540782122905028e-05, + "loss": 0.2006, + "step": 4342 + }, + { + "epoch": 12.131284916201118, + "grad_norm": 1.4345053434371948, + "learning_rate": 1.4544134078212291e-05, + "loss": 0.2347, + "step": 4343 + }, + { + "epoch": 12.134078212290502, + "grad_norm": 1.3636527061462402, + "learning_rate": 1.4547486033519553e-05, + "loss": 0.2103, + "step": 4344 + }, + { + "epoch": 12.136871508379889, + "grad_norm": 1.5316269397735596, + "learning_rate": 1.4550837988826816e-05, + "loss": 0.2272, + "step": 4345 + }, + { + "epoch": 12.139664804469273, + "grad_norm": 1.3898754119873047, + "learning_rate": 1.4554189944134079e-05, + "loss": 0.2722, + "step": 4346 + }, + { + "epoch": 12.14245810055866, + "grad_norm": 1.1883351802825928, + "learning_rate": 1.455754189944134e-05, + "loss": 0.2457, + "step": 4347 + }, + { + "epoch": 12.145251396648044, + "grad_norm": 0.48546460270881653, + "learning_rate": 1.4560893854748603e-05, + "loss": 0.1934, + "step": 4348 + }, + { + "epoch": 12.14804469273743, + "grad_norm": 1.601118564605713, + "learning_rate": 1.4564245810055866e-05, + "loss": 0.2397, + "step": 4349 + }, + { + "epoch": 12.150837988826815, + "grad_norm": 0.6562268733978271, + "learning_rate": 1.4567597765363128e-05, + "loss": 0.2235, + "step": 4350 + }, + { + "epoch": 12.153631284916202, + "grad_norm": 0.5131312608718872, + "learning_rate": 1.457094972067039e-05, + "loss": 0.2166, + "step": 4351 + }, + { + "epoch": 12.156424581005586, + "grad_norm": 0.7792536020278931, + "learning_rate": 1.4574301675977654e-05, + "loss": 0.2293, + "step": 4352 + }, + { + "epoch": 12.159217877094973, + "grad_norm": 0.7430758476257324, + "learning_rate": 1.4577653631284917e-05, + "loss": 0.1832, + "step": 4353 + }, + { + "epoch": 12.162011173184357, + "grad_norm": 0.6720899939537048, + "learning_rate": 1.4581005586592178e-05, + "loss": 0.1882, + "step": 4354 + }, + { + "epoch": 12.164804469273744, + "grad_norm": 0.8047145009040833, + "learning_rate": 1.4584357541899441e-05, + "loss": 0.2017, + "step": 4355 + }, + { + "epoch": 12.167597765363128, + "grad_norm": 0.8365569114685059, + "learning_rate": 1.4587709497206704e-05, + "loss": 0.2023, + "step": 4356 + }, + { + "epoch": 12.170391061452515, + "grad_norm": 0.9139548540115356, + "learning_rate": 1.4591061452513965e-05, + "loss": 0.2162, + "step": 4357 + }, + { + "epoch": 12.1731843575419, + "grad_norm": 0.7912387251853943, + "learning_rate": 1.4594413407821229e-05, + "loss": 0.2048, + "step": 4358 + }, + { + "epoch": 12.175977653631286, + "grad_norm": 1.1685038805007935, + "learning_rate": 1.4597765363128493e-05, + "loss": 0.2346, + "step": 4359 + }, + { + "epoch": 12.17877094972067, + "grad_norm": 0.7811694145202637, + "learning_rate": 1.4601117318435755e-05, + "loss": 0.19, + "step": 4360 + }, + { + "epoch": 12.181564245810057, + "grad_norm": 0.7129538059234619, + "learning_rate": 1.4604469273743018e-05, + "loss": 0.2065, + "step": 4361 + }, + { + "epoch": 12.184357541899441, + "grad_norm": 1.1316689252853394, + "learning_rate": 1.460782122905028e-05, + "loss": 0.2065, + "step": 4362 + }, + { + "epoch": 12.187150837988828, + "grad_norm": 2.3530654907226562, + "learning_rate": 1.4611173184357544e-05, + "loss": 0.1866, + "step": 4363 + }, + { + "epoch": 12.189944134078212, + "grad_norm": 0.9063121676445007, + "learning_rate": 1.4614525139664805e-05, + "loss": 0.2115, + "step": 4364 + }, + { + "epoch": 12.192737430167599, + "grad_norm": 1.6003060340881348, + "learning_rate": 1.4617877094972068e-05, + "loss": 0.2378, + "step": 4365 + }, + { + "epoch": 12.195530726256983, + "grad_norm": 1.8679453134536743, + "learning_rate": 1.4621229050279331e-05, + "loss": 0.2167, + "step": 4366 + }, + { + "epoch": 12.19832402234637, + "grad_norm": 1.8398301601409912, + "learning_rate": 1.4624581005586592e-05, + "loss": 0.191, + "step": 4367 + }, + { + "epoch": 12.201117318435754, + "grad_norm": 2.258784294128418, + "learning_rate": 1.4627932960893856e-05, + "loss": 0.195, + "step": 4368 + }, + { + "epoch": 12.203910614525139, + "grad_norm": 1.0871474742889404, + "learning_rate": 1.4631284916201119e-05, + "loss": 0.2174, + "step": 4369 + }, + { + "epoch": 12.206703910614525, + "grad_norm": 1.557913064956665, + "learning_rate": 1.463463687150838e-05, + "loss": 0.2173, + "step": 4370 + }, + { + "epoch": 12.20949720670391, + "grad_norm": 1.9724067449569702, + "learning_rate": 1.4637988826815643e-05, + "loss": 0.4285, + "step": 4371 + }, + { + "epoch": 12.212290502793296, + "grad_norm": 0.483511745929718, + "learning_rate": 1.4641340782122906e-05, + "loss": 0.1832, + "step": 4372 + }, + { + "epoch": 12.21508379888268, + "grad_norm": 0.48292338848114014, + "learning_rate": 1.4644692737430169e-05, + "loss": 0.1876, + "step": 4373 + }, + { + "epoch": 12.217877094972067, + "grad_norm": 0.5067977905273438, + "learning_rate": 1.464804469273743e-05, + "loss": 0.2057, + "step": 4374 + }, + { + "epoch": 12.220670391061452, + "grad_norm": 0.45905107259750366, + "learning_rate": 1.4651396648044693e-05, + "loss": 0.1903, + "step": 4375 + }, + { + "epoch": 12.223463687150838, + "grad_norm": 0.6526184678077698, + "learning_rate": 1.4654748603351956e-05, + "loss": 0.1945, + "step": 4376 + }, + { + "epoch": 12.226256983240223, + "grad_norm": 0.7530298829078674, + "learning_rate": 1.4658100558659218e-05, + "loss": 0.2001, + "step": 4377 + }, + { + "epoch": 12.22905027932961, + "grad_norm": 0.6265946626663208, + "learning_rate": 1.466145251396648e-05, + "loss": 0.1961, + "step": 4378 + }, + { + "epoch": 12.231843575418994, + "grad_norm": 0.6002511978149414, + "learning_rate": 1.4664804469273744e-05, + "loss": 0.1917, + "step": 4379 + }, + { + "epoch": 12.23463687150838, + "grad_norm": 0.6065762042999268, + "learning_rate": 1.4668156424581005e-05, + "loss": 0.1957, + "step": 4380 + }, + { + "epoch": 12.237430167597765, + "grad_norm": 0.6126528382301331, + "learning_rate": 1.4671508379888268e-05, + "loss": 0.2245, + "step": 4381 + }, + { + "epoch": 12.240223463687151, + "grad_norm": 1.1398215293884277, + "learning_rate": 1.4674860335195531e-05, + "loss": 0.2052, + "step": 4382 + }, + { + "epoch": 12.243016759776536, + "grad_norm": 0.5686419606208801, + "learning_rate": 1.4678212290502794e-05, + "loss": 0.1813, + "step": 4383 + }, + { + "epoch": 12.245810055865922, + "grad_norm": 0.6750921607017517, + "learning_rate": 1.4681564245810056e-05, + "loss": 0.2043, + "step": 4384 + }, + { + "epoch": 12.248603351955307, + "grad_norm": 0.7806252837181091, + "learning_rate": 1.4684916201117319e-05, + "loss": 0.2019, + "step": 4385 + }, + { + "epoch": 12.251396648044693, + "grad_norm": 0.8575930595397949, + "learning_rate": 1.4688268156424582e-05, + "loss": 0.1871, + "step": 4386 + }, + { + "epoch": 12.254189944134078, + "grad_norm": 0.7783242464065552, + "learning_rate": 1.4691620111731843e-05, + "loss": 0.2079, + "step": 4387 + }, + { + "epoch": 12.256983240223464, + "grad_norm": 0.9273446202278137, + "learning_rate": 1.4694972067039106e-05, + "loss": 0.2125, + "step": 4388 + }, + { + "epoch": 12.259776536312849, + "grad_norm": 1.109700322151184, + "learning_rate": 1.469832402234637e-05, + "loss": 0.2071, + "step": 4389 + }, + { + "epoch": 12.262569832402235, + "grad_norm": 0.8430426716804504, + "learning_rate": 1.4701675977653632e-05, + "loss": 0.2035, + "step": 4390 + }, + { + "epoch": 12.26536312849162, + "grad_norm": 1.0399705171585083, + "learning_rate": 1.4705027932960894e-05, + "loss": 0.191, + "step": 4391 + }, + { + "epoch": 12.268156424581006, + "grad_norm": 1.1758097410202026, + "learning_rate": 1.4708379888268157e-05, + "loss": 0.2179, + "step": 4392 + }, + { + "epoch": 12.27094972067039, + "grad_norm": 1.083457112312317, + "learning_rate": 1.471173184357542e-05, + "loss": 0.2027, + "step": 4393 + }, + { + "epoch": 12.273743016759777, + "grad_norm": 1.0454399585723877, + "learning_rate": 1.4715083798882681e-05, + "loss": 0.2155, + "step": 4394 + }, + { + "epoch": 12.276536312849162, + "grad_norm": 1.4200448989868164, + "learning_rate": 1.4718435754189944e-05, + "loss": 0.264, + "step": 4395 + }, + { + "epoch": 12.279329608938548, + "grad_norm": 2.757337808609009, + "learning_rate": 1.4721787709497207e-05, + "loss": 0.2811, + "step": 4396 + }, + { + "epoch": 12.282122905027933, + "grad_norm": 0.7513442039489746, + "learning_rate": 1.4725139664804468e-05, + "loss": 0.2013, + "step": 4397 + }, + { + "epoch": 12.28491620111732, + "grad_norm": 0.5155564546585083, + "learning_rate": 1.4728491620111731e-05, + "loss": 0.1771, + "step": 4398 + }, + { + "epoch": 12.287709497206704, + "grad_norm": 0.6417407393455505, + "learning_rate": 1.4731843575418995e-05, + "loss": 0.19, + "step": 4399 + }, + { + "epoch": 12.29050279329609, + "grad_norm": 0.6517201662063599, + "learning_rate": 1.4735195530726258e-05, + "loss": 0.2064, + "step": 4400 + }, + { + "epoch": 12.293296089385475, + "grad_norm": 0.7164453864097595, + "learning_rate": 1.4738547486033519e-05, + "loss": 0.1768, + "step": 4401 + }, + { + "epoch": 12.296089385474861, + "grad_norm": 0.7101495265960693, + "learning_rate": 1.4741899441340782e-05, + "loss": 0.2149, + "step": 4402 + }, + { + "epoch": 12.298882681564246, + "grad_norm": 0.6574653387069702, + "learning_rate": 1.4745251396648045e-05, + "loss": 0.2487, + "step": 4403 + }, + { + "epoch": 12.30167597765363, + "grad_norm": 0.6531410217285156, + "learning_rate": 1.4748603351955306e-05, + "loss": 0.194, + "step": 4404 + }, + { + "epoch": 12.304469273743017, + "grad_norm": 0.8442827463150024, + "learning_rate": 1.475195530726257e-05, + "loss": 0.2007, + "step": 4405 + }, + { + "epoch": 12.307262569832401, + "grad_norm": 0.7326877117156982, + "learning_rate": 1.4755307262569832e-05, + "loss": 0.2011, + "step": 4406 + }, + { + "epoch": 12.310055865921788, + "grad_norm": 0.709720253944397, + "learning_rate": 1.4758659217877095e-05, + "loss": 0.2006, + "step": 4407 + }, + { + "epoch": 12.312849162011172, + "grad_norm": 0.5626229047775269, + "learning_rate": 1.4762011173184358e-05, + "loss": 0.1627, + "step": 4408 + }, + { + "epoch": 12.315642458100559, + "grad_norm": 1.83566415309906, + "learning_rate": 1.4765363128491622e-05, + "loss": 0.1952, + "step": 4409 + }, + { + "epoch": 12.318435754189943, + "grad_norm": 0.9992433190345764, + "learning_rate": 1.4768715083798885e-05, + "loss": 0.1836, + "step": 4410 + }, + { + "epoch": 12.32122905027933, + "grad_norm": 0.8999152779579163, + "learning_rate": 1.4772067039106146e-05, + "loss": 0.1811, + "step": 4411 + }, + { + "epoch": 12.324022346368714, + "grad_norm": 0.8426945805549622, + "learning_rate": 1.4775418994413409e-05, + "loss": 0.1624, + "step": 4412 + }, + { + "epoch": 12.3268156424581, + "grad_norm": 0.698954164981842, + "learning_rate": 1.4778770949720672e-05, + "loss": 0.1797, + "step": 4413 + }, + { + "epoch": 12.329608938547485, + "grad_norm": 3.8583788871765137, + "learning_rate": 1.4782122905027933e-05, + "loss": 0.1935, + "step": 4414 + }, + { + "epoch": 12.332402234636872, + "grad_norm": 1.092725396156311, + "learning_rate": 1.4785474860335196e-05, + "loss": 0.1915, + "step": 4415 + }, + { + "epoch": 12.335195530726256, + "grad_norm": 1.3409242630004883, + "learning_rate": 1.478882681564246e-05, + "loss": 0.1868, + "step": 4416 + }, + { + "epoch": 12.337988826815643, + "grad_norm": 1.525146245956421, + "learning_rate": 1.479217877094972e-05, + "loss": 0.1881, + "step": 4417 + }, + { + "epoch": 12.340782122905027, + "grad_norm": 1.3342758417129517, + "learning_rate": 1.4795530726256984e-05, + "loss": 0.1877, + "step": 4418 + }, + { + "epoch": 12.343575418994414, + "grad_norm": 8.099832534790039, + "learning_rate": 1.4798882681564247e-05, + "loss": 0.2075, + "step": 4419 + }, + { + "epoch": 12.346368715083798, + "grad_norm": 1.6658705472946167, + "learning_rate": 1.480223463687151e-05, + "loss": 0.2851, + "step": 4420 + }, + { + "epoch": 12.349162011173185, + "grad_norm": 2.654348850250244, + "learning_rate": 1.4805586592178771e-05, + "loss": 0.2934, + "step": 4421 + }, + { + "epoch": 12.35195530726257, + "grad_norm": 0.6927871704101562, + "learning_rate": 1.4808938547486034e-05, + "loss": 0.1838, + "step": 4422 + }, + { + "epoch": 12.354748603351956, + "grad_norm": 0.7489193677902222, + "learning_rate": 1.4812290502793297e-05, + "loss": 0.2201, + "step": 4423 + }, + { + "epoch": 12.35754189944134, + "grad_norm": 0.7141326665878296, + "learning_rate": 1.4815642458100559e-05, + "loss": 0.2309, + "step": 4424 + }, + { + "epoch": 12.360335195530727, + "grad_norm": 0.544442355632782, + "learning_rate": 1.4818994413407822e-05, + "loss": 0.1896, + "step": 4425 + }, + { + "epoch": 12.363128491620111, + "grad_norm": 0.8487884402275085, + "learning_rate": 1.4822346368715085e-05, + "loss": 0.2759, + "step": 4426 + }, + { + "epoch": 12.365921787709498, + "grad_norm": 0.5734151601791382, + "learning_rate": 1.4825698324022346e-05, + "loss": 0.204, + "step": 4427 + }, + { + "epoch": 12.368715083798882, + "grad_norm": 0.7643246054649353, + "learning_rate": 1.4829050279329609e-05, + "loss": 0.2219, + "step": 4428 + }, + { + "epoch": 12.371508379888269, + "grad_norm": 0.7335776090621948, + "learning_rate": 1.4832402234636872e-05, + "loss": 0.2085, + "step": 4429 + }, + { + "epoch": 12.374301675977653, + "grad_norm": 0.7637225985527039, + "learning_rate": 1.4835754189944135e-05, + "loss": 0.1843, + "step": 4430 + }, + { + "epoch": 12.37709497206704, + "grad_norm": 0.6102579236030579, + "learning_rate": 1.4839106145251397e-05, + "loss": 0.1865, + "step": 4431 + }, + { + "epoch": 12.379888268156424, + "grad_norm": 1.0501325130462646, + "learning_rate": 1.484245810055866e-05, + "loss": 0.2128, + "step": 4432 + }, + { + "epoch": 12.38268156424581, + "grad_norm": 0.6567489504814148, + "learning_rate": 1.4845810055865923e-05, + "loss": 0.1832, + "step": 4433 + }, + { + "epoch": 12.385474860335195, + "grad_norm": 1.1676807403564453, + "learning_rate": 1.4849162011173184e-05, + "loss": 0.1951, + "step": 4434 + }, + { + "epoch": 12.388268156424582, + "grad_norm": 1.0728678703308105, + "learning_rate": 1.4852513966480447e-05, + "loss": 0.1957, + "step": 4435 + }, + { + "epoch": 12.391061452513966, + "grad_norm": 0.5818567872047424, + "learning_rate": 1.485586592178771e-05, + "loss": 0.156, + "step": 4436 + }, + { + "epoch": 12.393854748603353, + "grad_norm": 0.979996919631958, + "learning_rate": 1.4859217877094973e-05, + "loss": 0.204, + "step": 4437 + }, + { + "epoch": 12.396648044692737, + "grad_norm": 0.8547490239143372, + "learning_rate": 1.4862569832402234e-05, + "loss": 0.2098, + "step": 4438 + }, + { + "epoch": 12.399441340782122, + "grad_norm": 0.9409061074256897, + "learning_rate": 1.4865921787709497e-05, + "loss": 0.1759, + "step": 4439 + }, + { + "epoch": 12.402234636871508, + "grad_norm": 2.4066429138183594, + "learning_rate": 1.486927374301676e-05, + "loss": 0.1929, + "step": 4440 + }, + { + "epoch": 12.405027932960893, + "grad_norm": 1.0368059873580933, + "learning_rate": 1.4872625698324022e-05, + "loss": 0.2285, + "step": 4441 + }, + { + "epoch": 12.40782122905028, + "grad_norm": 1.068766713142395, + "learning_rate": 1.4875977653631285e-05, + "loss": 0.265, + "step": 4442 + }, + { + "epoch": 12.410614525139664, + "grad_norm": 1.1351075172424316, + "learning_rate": 1.4879329608938548e-05, + "loss": 0.2477, + "step": 4443 + }, + { + "epoch": 12.41340782122905, + "grad_norm": 1.2287211418151855, + "learning_rate": 1.488268156424581e-05, + "loss": 0.2182, + "step": 4444 + }, + { + "epoch": 12.416201117318435, + "grad_norm": 1.7355027198791504, + "learning_rate": 1.4886033519553072e-05, + "loss": 0.2402, + "step": 4445 + }, + { + "epoch": 12.418994413407821, + "grad_norm": 5.700326919555664, + "learning_rate": 1.4889385474860335e-05, + "loss": 0.2769, + "step": 4446 + }, + { + "epoch": 12.421787709497206, + "grad_norm": 0.44549793004989624, + "learning_rate": 1.4892737430167598e-05, + "loss": 0.1938, + "step": 4447 + }, + { + "epoch": 12.424581005586592, + "grad_norm": 0.7880837917327881, + "learning_rate": 1.489608938547486e-05, + "loss": 0.1885, + "step": 4448 + }, + { + "epoch": 12.427374301675977, + "grad_norm": 0.5455135107040405, + "learning_rate": 1.4899441340782123e-05, + "loss": 0.1783, + "step": 4449 + }, + { + "epoch": 12.430167597765363, + "grad_norm": 0.6900818347930908, + "learning_rate": 1.4902793296089386e-05, + "loss": 0.2065, + "step": 4450 + }, + { + "epoch": 12.432960893854748, + "grad_norm": 0.5398233532905579, + "learning_rate": 1.4906145251396647e-05, + "loss": 0.2108, + "step": 4451 + }, + { + "epoch": 12.435754189944134, + "grad_norm": 0.4176795482635498, + "learning_rate": 1.490949720670391e-05, + "loss": 0.1669, + "step": 4452 + }, + { + "epoch": 12.438547486033519, + "grad_norm": 0.575719952583313, + "learning_rate": 1.4912849162011173e-05, + "loss": 0.1483, + "step": 4453 + }, + { + "epoch": 12.441340782122905, + "grad_norm": 1.4010776281356812, + "learning_rate": 1.4916201117318435e-05, + "loss": 0.2177, + "step": 4454 + }, + { + "epoch": 12.44413407821229, + "grad_norm": 0.6896913051605225, + "learning_rate": 1.49195530726257e-05, + "loss": 0.1742, + "step": 4455 + }, + { + "epoch": 12.446927374301676, + "grad_norm": 0.6756688952445984, + "learning_rate": 1.4922905027932962e-05, + "loss": 0.2168, + "step": 4456 + }, + { + "epoch": 12.449720670391061, + "grad_norm": 1.0433979034423828, + "learning_rate": 1.4926256983240225e-05, + "loss": 0.1761, + "step": 4457 + }, + { + "epoch": 12.452513966480447, + "grad_norm": 0.7196043133735657, + "learning_rate": 1.4929608938547487e-05, + "loss": 0.217, + "step": 4458 + }, + { + "epoch": 12.455307262569832, + "grad_norm": 0.6519199013710022, + "learning_rate": 1.493296089385475e-05, + "loss": 0.1648, + "step": 4459 + }, + { + "epoch": 12.458100558659218, + "grad_norm": 1.2389644384384155, + "learning_rate": 1.4936312849162013e-05, + "loss": 0.2077, + "step": 4460 + }, + { + "epoch": 12.460893854748603, + "grad_norm": 0.6747907996177673, + "learning_rate": 1.4939664804469274e-05, + "loss": 0.188, + "step": 4461 + }, + { + "epoch": 12.46368715083799, + "grad_norm": 0.8768600225448608, + "learning_rate": 1.4943016759776537e-05, + "loss": 0.1942, + "step": 4462 + }, + { + "epoch": 12.466480446927374, + "grad_norm": 1.1082429885864258, + "learning_rate": 1.49463687150838e-05, + "loss": 0.2261, + "step": 4463 + }, + { + "epoch": 12.46927374301676, + "grad_norm": 0.6705049276351929, + "learning_rate": 1.4949720670391062e-05, + "loss": 0.1712, + "step": 4464 + }, + { + "epoch": 12.472067039106145, + "grad_norm": 1.544150471687317, + "learning_rate": 1.4953072625698325e-05, + "loss": 0.1902, + "step": 4465 + }, + { + "epoch": 12.474860335195531, + "grad_norm": 0.8267551064491272, + "learning_rate": 1.4956424581005588e-05, + "loss": 0.2193, + "step": 4466 + }, + { + "epoch": 12.477653631284916, + "grad_norm": 1.0797518491744995, + "learning_rate": 1.495977653631285e-05, + "loss": 0.2484, + "step": 4467 + }, + { + "epoch": 12.480446927374302, + "grad_norm": 1.292366862297058, + "learning_rate": 1.4963128491620112e-05, + "loss": 0.226, + "step": 4468 + }, + { + "epoch": 12.483240223463687, + "grad_norm": 1.2308772802352905, + "learning_rate": 1.4966480446927375e-05, + "loss": 0.2345, + "step": 4469 + }, + { + "epoch": 12.486033519553073, + "grad_norm": 10.870575904846191, + "learning_rate": 1.4969832402234638e-05, + "loss": 0.262, + "step": 4470 + }, + { + "epoch": 12.488826815642458, + "grad_norm": 2.830641269683838, + "learning_rate": 1.49731843575419e-05, + "loss": 0.2668, + "step": 4471 + }, + { + "epoch": 12.491620111731844, + "grad_norm": 0.7189430594444275, + "learning_rate": 1.4976536312849162e-05, + "loss": 0.234, + "step": 4472 + }, + { + "epoch": 12.494413407821229, + "grad_norm": 1.1254626512527466, + "learning_rate": 1.4979888268156426e-05, + "loss": 0.2007, + "step": 4473 + }, + { + "epoch": 12.497206703910614, + "grad_norm": 0.7398231029510498, + "learning_rate": 1.4983240223463687e-05, + "loss": 0.2014, + "step": 4474 + }, + { + "epoch": 12.5, + "grad_norm": 0.5524194836616516, + "learning_rate": 1.498659217877095e-05, + "loss": 0.1744, + "step": 4475 + }, + { + "epoch": 12.502793296089386, + "grad_norm": 0.7265095710754395, + "learning_rate": 1.4989944134078213e-05, + "loss": 0.2035, + "step": 4476 + }, + { + "epoch": 12.505586592178771, + "grad_norm": 0.7433755993843079, + "learning_rate": 1.4993296089385476e-05, + "loss": 0.227, + "step": 4477 + }, + { + "epoch": 12.508379888268156, + "grad_norm": 0.8668108582496643, + "learning_rate": 1.4996648044692737e-05, + "loss": 0.2143, + "step": 4478 + }, + { + "epoch": 12.511173184357542, + "grad_norm": 1.0490649938583374, + "learning_rate": 1.5e-05, + "loss": 0.2043, + "step": 4479 + }, + { + "epoch": 12.513966480446927, + "grad_norm": 0.6639259457588196, + "learning_rate": 1.5003351955307262e-05, + "loss": 0.2115, + "step": 4480 + }, + { + "epoch": 12.516759776536313, + "grad_norm": 0.8968842029571533, + "learning_rate": 1.5006703910614526e-05, + "loss": 0.2238, + "step": 4481 + }, + { + "epoch": 12.519553072625698, + "grad_norm": 1.1075211763381958, + "learning_rate": 1.5010055865921788e-05, + "loss": 0.1972, + "step": 4482 + }, + { + "epoch": 12.522346368715084, + "grad_norm": 0.8779720664024353, + "learning_rate": 1.5013407821229049e-05, + "loss": 0.2393, + "step": 4483 + }, + { + "epoch": 12.525139664804469, + "grad_norm": 0.7383827567100525, + "learning_rate": 1.5016759776536314e-05, + "loss": 0.1803, + "step": 4484 + }, + { + "epoch": 12.527932960893855, + "grad_norm": 0.8795973062515259, + "learning_rate": 1.5020111731843575e-05, + "loss": 0.1896, + "step": 4485 + }, + { + "epoch": 12.53072625698324, + "grad_norm": 1.1176806688308716, + "learning_rate": 1.5023463687150837e-05, + "loss": 0.2053, + "step": 4486 + }, + { + "epoch": 12.533519553072626, + "grad_norm": 0.7209283113479614, + "learning_rate": 1.5026815642458101e-05, + "loss": 0.1904, + "step": 4487 + }, + { + "epoch": 12.53631284916201, + "grad_norm": 1.0142663717269897, + "learning_rate": 1.5030167597765363e-05, + "loss": 0.1982, + "step": 4488 + }, + { + "epoch": 12.539106145251397, + "grad_norm": 1.2570526599884033, + "learning_rate": 1.5033519553072627e-05, + "loss": 0.2154, + "step": 4489 + }, + { + "epoch": 12.541899441340782, + "grad_norm": 1.2465057373046875, + "learning_rate": 1.5036871508379889e-05, + "loss": 0.1801, + "step": 4490 + }, + { + "epoch": 12.544692737430168, + "grad_norm": 1.144689679145813, + "learning_rate": 1.504022346368715e-05, + "loss": 0.2181, + "step": 4491 + }, + { + "epoch": 12.547486033519553, + "grad_norm": 1.1636781692504883, + "learning_rate": 1.5043575418994415e-05, + "loss": 0.2415, + "step": 4492 + }, + { + "epoch": 12.550279329608939, + "grad_norm": 1.6834187507629395, + "learning_rate": 1.5046927374301676e-05, + "loss": 0.1864, + "step": 4493 + }, + { + "epoch": 12.553072625698324, + "grad_norm": 1.0676994323730469, + "learning_rate": 1.5050279329608937e-05, + "loss": 0.2085, + "step": 4494 + }, + { + "epoch": 12.55586592178771, + "grad_norm": 2.219329357147217, + "learning_rate": 1.5053631284916202e-05, + "loss": 0.2777, + "step": 4495 + }, + { + "epoch": 12.558659217877095, + "grad_norm": 2.75081467628479, + "learning_rate": 1.5056983240223464e-05, + "loss": 0.3067, + "step": 4496 + }, + { + "epoch": 12.561452513966481, + "grad_norm": 0.7044459581375122, + "learning_rate": 1.5060335195530725e-05, + "loss": 0.257, + "step": 4497 + }, + { + "epoch": 12.564245810055866, + "grad_norm": 0.6350319981575012, + "learning_rate": 1.506368715083799e-05, + "loss": 0.2159, + "step": 4498 + }, + { + "epoch": 12.567039106145252, + "grad_norm": 0.5797572135925293, + "learning_rate": 1.5067039106145251e-05, + "loss": 0.1594, + "step": 4499 + }, + { + "epoch": 12.569832402234637, + "grad_norm": 0.7599446177482605, + "learning_rate": 1.5070391061452512e-05, + "loss": 0.2298, + "step": 4500 + }, + { + "epoch": 12.572625698324023, + "grad_norm": 1.2710996866226196, + "learning_rate": 1.5073743016759777e-05, + "loss": 0.1876, + "step": 4501 + }, + { + "epoch": 12.575418994413408, + "grad_norm": 0.9597491025924683, + "learning_rate": 1.5077094972067038e-05, + "loss": 0.2046, + "step": 4502 + }, + { + "epoch": 12.578212290502794, + "grad_norm": 0.8032236695289612, + "learning_rate": 1.5080446927374301e-05, + "loss": 0.227, + "step": 4503 + }, + { + "epoch": 12.581005586592179, + "grad_norm": 0.64569491147995, + "learning_rate": 1.5083798882681566e-05, + "loss": 0.1789, + "step": 4504 + }, + { + "epoch": 12.583798882681565, + "grad_norm": 2.136589527130127, + "learning_rate": 1.5087150837988828e-05, + "loss": 0.1829, + "step": 4505 + }, + { + "epoch": 12.58659217877095, + "grad_norm": 0.7580874562263489, + "learning_rate": 1.5090502793296089e-05, + "loss": 0.2015, + "step": 4506 + }, + { + "epoch": 12.589385474860336, + "grad_norm": 0.7149032950401306, + "learning_rate": 1.5093854748603354e-05, + "loss": 0.2047, + "step": 4507 + }, + { + "epoch": 12.59217877094972, + "grad_norm": 0.807081401348114, + "learning_rate": 1.5097206703910615e-05, + "loss": 0.187, + "step": 4508 + }, + { + "epoch": 12.594972067039105, + "grad_norm": 0.9434452652931213, + "learning_rate": 1.510055865921788e-05, + "loss": 0.1814, + "step": 4509 + }, + { + "epoch": 12.597765363128492, + "grad_norm": 0.5943213701248169, + "learning_rate": 1.5103910614525141e-05, + "loss": 0.1671, + "step": 4510 + }, + { + "epoch": 12.600558659217878, + "grad_norm": 1.0138347148895264, + "learning_rate": 1.5107262569832402e-05, + "loss": 0.1801, + "step": 4511 + }, + { + "epoch": 12.603351955307263, + "grad_norm": 1.2414380311965942, + "learning_rate": 1.5110614525139667e-05, + "loss": 0.174, + "step": 4512 + }, + { + "epoch": 12.606145251396647, + "grad_norm": 2.1342954635620117, + "learning_rate": 1.5113966480446928e-05, + "loss": 0.2474, + "step": 4513 + }, + { + "epoch": 12.608938547486034, + "grad_norm": 1.0039218664169312, + "learning_rate": 1.511731843575419e-05, + "loss": 0.2429, + "step": 4514 + }, + { + "epoch": 12.611731843575418, + "grad_norm": 0.8074836134910583, + "learning_rate": 1.5120670391061455e-05, + "loss": 0.1693, + "step": 4515 + }, + { + "epoch": 12.614525139664805, + "grad_norm": 1.086037516593933, + "learning_rate": 1.5124022346368716e-05, + "loss": 0.1913, + "step": 4516 + }, + { + "epoch": 12.61731843575419, + "grad_norm": 0.9583508968353271, + "learning_rate": 1.5127374301675977e-05, + "loss": 0.2218, + "step": 4517 + }, + { + "epoch": 12.620111731843576, + "grad_norm": 1.1834479570388794, + "learning_rate": 1.5130726256983242e-05, + "loss": 0.2084, + "step": 4518 + }, + { + "epoch": 12.62290502793296, + "grad_norm": 1.2376885414123535, + "learning_rate": 1.5134078212290503e-05, + "loss": 0.237, + "step": 4519 + }, + { + "epoch": 12.625698324022347, + "grad_norm": 1.556406855583191, + "learning_rate": 1.5137430167597765e-05, + "loss": 0.2363, + "step": 4520 + }, + { + "epoch": 12.628491620111731, + "grad_norm": 2.5961899757385254, + "learning_rate": 1.514078212290503e-05, + "loss": 0.29, + "step": 4521 + }, + { + "epoch": 12.631284916201118, + "grad_norm": 0.4620635211467743, + "learning_rate": 1.514413407821229e-05, + "loss": 0.1886, + "step": 4522 + }, + { + "epoch": 12.634078212290502, + "grad_norm": 0.5127986669540405, + "learning_rate": 1.5147486033519552e-05, + "loss": 0.1847, + "step": 4523 + }, + { + "epoch": 12.636871508379889, + "grad_norm": 0.5624856352806091, + "learning_rate": 1.5150837988826817e-05, + "loss": 0.1822, + "step": 4524 + }, + { + "epoch": 12.639664804469273, + "grad_norm": 1.5517487525939941, + "learning_rate": 1.5154189944134078e-05, + "loss": 0.2214, + "step": 4525 + }, + { + "epoch": 12.64245810055866, + "grad_norm": 0.5903233289718628, + "learning_rate": 1.515754189944134e-05, + "loss": 0.2158, + "step": 4526 + }, + { + "epoch": 12.645251396648044, + "grad_norm": 0.6842474937438965, + "learning_rate": 1.5160893854748604e-05, + "loss": 0.2073, + "step": 4527 + }, + { + "epoch": 12.64804469273743, + "grad_norm": 0.641769289970398, + "learning_rate": 1.5164245810055866e-05, + "loss": 0.1948, + "step": 4528 + }, + { + "epoch": 12.650837988826815, + "grad_norm": 0.7802746891975403, + "learning_rate": 1.516759776536313e-05, + "loss": 0.1808, + "step": 4529 + }, + { + "epoch": 12.653631284916202, + "grad_norm": 0.5750057697296143, + "learning_rate": 1.5170949720670392e-05, + "loss": 0.1901, + "step": 4530 + }, + { + "epoch": 12.656424581005586, + "grad_norm": 5.958463668823242, + "learning_rate": 1.5174301675977653e-05, + "loss": 0.2097, + "step": 4531 + }, + { + "epoch": 12.659217877094973, + "grad_norm": 0.928707480430603, + "learning_rate": 1.5177653631284918e-05, + "loss": 0.1865, + "step": 4532 + }, + { + "epoch": 12.662011173184357, + "grad_norm": 0.8213618397712708, + "learning_rate": 1.5181005586592179e-05, + "loss": 0.184, + "step": 4533 + }, + { + "epoch": 12.664804469273744, + "grad_norm": 0.9405850172042847, + "learning_rate": 1.518435754189944e-05, + "loss": 0.1828, + "step": 4534 + }, + { + "epoch": 12.667597765363128, + "grad_norm": 1.059122920036316, + "learning_rate": 1.5187709497206705e-05, + "loss": 0.1833, + "step": 4535 + }, + { + "epoch": 12.670391061452515, + "grad_norm": 2.0529918670654297, + "learning_rate": 1.5191061452513967e-05, + "loss": 0.1934, + "step": 4536 + }, + { + "epoch": 12.6731843575419, + "grad_norm": 0.9893752932548523, + "learning_rate": 1.5194413407821228e-05, + "loss": 0.2027, + "step": 4537 + }, + { + "epoch": 12.675977653631286, + "grad_norm": 1.3155326843261719, + "learning_rate": 1.5197765363128493e-05, + "loss": 0.1913, + "step": 4538 + }, + { + "epoch": 12.67877094972067, + "grad_norm": 1.1143429279327393, + "learning_rate": 1.5201117318435754e-05, + "loss": 0.2115, + "step": 4539 + }, + { + "epoch": 12.681564245810057, + "grad_norm": 0.9846829771995544, + "learning_rate": 1.5204469273743015e-05, + "loss": 0.2079, + "step": 4540 + }, + { + "epoch": 12.684357541899441, + "grad_norm": 1.5577446222305298, + "learning_rate": 1.520782122905028e-05, + "loss": 0.2507, + "step": 4541 + }, + { + "epoch": 12.687150837988828, + "grad_norm": 5.413323402404785, + "learning_rate": 1.5211173184357541e-05, + "loss": 0.2179, + "step": 4542 + }, + { + "epoch": 12.689944134078212, + "grad_norm": 1.2232946157455444, + "learning_rate": 1.5214525139664803e-05, + "loss": 0.2363, + "step": 4543 + }, + { + "epoch": 12.692737430167599, + "grad_norm": 1.3033910989761353, + "learning_rate": 1.5217877094972067e-05, + "loss": 0.2313, + "step": 4544 + }, + { + "epoch": 12.695530726256983, + "grad_norm": 1.3909498453140259, + "learning_rate": 1.5221229050279329e-05, + "loss": 0.2477, + "step": 4545 + }, + { + "epoch": 12.69832402234637, + "grad_norm": 3.7873241901397705, + "learning_rate": 1.5224581005586594e-05, + "loss": 0.3003, + "step": 4546 + }, + { + "epoch": 12.701117318435754, + "grad_norm": 0.630515992641449, + "learning_rate": 1.5227932960893855e-05, + "loss": 0.2002, + "step": 4547 + }, + { + "epoch": 12.703910614525139, + "grad_norm": 0.5738760232925415, + "learning_rate": 1.5231284916201116e-05, + "loss": 0.2125, + "step": 4548 + }, + { + "epoch": 12.706703910614525, + "grad_norm": 0.7756384015083313, + "learning_rate": 1.5234636871508381e-05, + "loss": 0.2187, + "step": 4549 + }, + { + "epoch": 12.70949720670391, + "grad_norm": 0.5125120282173157, + "learning_rate": 1.5237988826815642e-05, + "loss": 0.2061, + "step": 4550 + }, + { + "epoch": 12.712290502793296, + "grad_norm": 0.6436158418655396, + "learning_rate": 1.5241340782122905e-05, + "loss": 0.1926, + "step": 4551 + }, + { + "epoch": 12.71508379888268, + "grad_norm": 0.6612012386322021, + "learning_rate": 1.524469273743017e-05, + "loss": 0.1724, + "step": 4552 + }, + { + "epoch": 12.717877094972067, + "grad_norm": 0.7715325951576233, + "learning_rate": 1.5248044692737431e-05, + "loss": 0.1986, + "step": 4553 + }, + { + "epoch": 12.720670391061452, + "grad_norm": 0.6656082272529602, + "learning_rate": 1.5251396648044693e-05, + "loss": 0.1953, + "step": 4554 + }, + { + "epoch": 12.723463687150838, + "grad_norm": 0.7770776748657227, + "learning_rate": 1.5254748603351957e-05, + "loss": 0.1964, + "step": 4555 + }, + { + "epoch": 12.726256983240223, + "grad_norm": 0.6732997894287109, + "learning_rate": 1.5258100558659219e-05, + "loss": 0.2002, + "step": 4556 + }, + { + "epoch": 12.72905027932961, + "grad_norm": 0.6999877095222473, + "learning_rate": 1.526145251396648e-05, + "loss": 0.1994, + "step": 4557 + }, + { + "epoch": 12.731843575418994, + "grad_norm": 0.914544939994812, + "learning_rate": 1.5264804469273743e-05, + "loss": 0.2095, + "step": 4558 + }, + { + "epoch": 12.73463687150838, + "grad_norm": 0.8854045271873474, + "learning_rate": 1.5268156424581005e-05, + "loss": 0.1872, + "step": 4559 + }, + { + "epoch": 12.737430167597765, + "grad_norm": 1.0671356916427612, + "learning_rate": 1.5271508379888266e-05, + "loss": 0.202, + "step": 4560 + }, + { + "epoch": 12.740223463687151, + "grad_norm": 1.8594655990600586, + "learning_rate": 1.527486033519553e-05, + "loss": 0.2093, + "step": 4561 + }, + { + "epoch": 12.743016759776536, + "grad_norm": 0.71195387840271, + "learning_rate": 1.5278212290502792e-05, + "loss": 0.1555, + "step": 4562 + }, + { + "epoch": 12.745810055865922, + "grad_norm": 0.7431557178497314, + "learning_rate": 1.5281564245810053e-05, + "loss": 0.2104, + "step": 4563 + }, + { + "epoch": 12.748603351955307, + "grad_norm": 0.9272972345352173, + "learning_rate": 1.5284916201117318e-05, + "loss": 0.2015, + "step": 4564 + }, + { + "epoch": 12.751396648044693, + "grad_norm": 1.7579233646392822, + "learning_rate": 1.528826815642458e-05, + "loss": 0.2286, + "step": 4565 + }, + { + "epoch": 12.754189944134078, + "grad_norm": 0.9037145376205444, + "learning_rate": 1.5291620111731844e-05, + "loss": 0.1886, + "step": 4566 + }, + { + "epoch": 12.756983240223464, + "grad_norm": 0.9424108862876892, + "learning_rate": 1.5294972067039105e-05, + "loss": 0.2086, + "step": 4567 + }, + { + "epoch": 12.759776536312849, + "grad_norm": 1.8641583919525146, + "learning_rate": 1.5298324022346367e-05, + "loss": 0.2491, + "step": 4568 + }, + { + "epoch": 12.762569832402235, + "grad_norm": 2.726172685623169, + "learning_rate": 1.530167597765363e-05, + "loss": 0.1848, + "step": 4569 + }, + { + "epoch": 12.76536312849162, + "grad_norm": 1.9069452285766602, + "learning_rate": 1.5305027932960893e-05, + "loss": 0.1864, + "step": 4570 + }, + { + "epoch": 12.768156424581006, + "grad_norm": 4.395595073699951, + "learning_rate": 1.5308379888268154e-05, + "loss": 0.3653, + "step": 4571 + }, + { + "epoch": 12.77094972067039, + "grad_norm": 0.6757256984710693, + "learning_rate": 1.531173184357542e-05, + "loss": 0.2168, + "step": 4572 + }, + { + "epoch": 12.773743016759777, + "grad_norm": 0.5520175695419312, + "learning_rate": 1.531508379888268e-05, + "loss": 0.203, + "step": 4573 + }, + { + "epoch": 12.776536312849162, + "grad_norm": 1.4157021045684814, + "learning_rate": 1.531843575418994e-05, + "loss": 0.215, + "step": 4574 + }, + { + "epoch": 12.779329608938548, + "grad_norm": 0.5925761461257935, + "learning_rate": 1.5321787709497206e-05, + "loss": 0.1801, + "step": 4575 + }, + { + "epoch": 12.782122905027933, + "grad_norm": 0.5875861644744873, + "learning_rate": 1.532513966480447e-05, + "loss": 0.2, + "step": 4576 + }, + { + "epoch": 12.78491620111732, + "grad_norm": 0.6178386211395264, + "learning_rate": 1.5328491620111732e-05, + "loss": 0.1852, + "step": 4577 + }, + { + "epoch": 12.787709497206704, + "grad_norm": 0.6339854598045349, + "learning_rate": 1.5331843575418997e-05, + "loss": 0.1856, + "step": 4578 + }, + { + "epoch": 12.79050279329609, + "grad_norm": 0.5489737391471863, + "learning_rate": 1.533519553072626e-05, + "loss": 0.1907, + "step": 4579 + }, + { + "epoch": 12.793296089385475, + "grad_norm": 0.6308954954147339, + "learning_rate": 1.533854748603352e-05, + "loss": 0.1718, + "step": 4580 + }, + { + "epoch": 12.796089385474861, + "grad_norm": 0.950888991355896, + "learning_rate": 1.5341899441340785e-05, + "loss": 0.2509, + "step": 4581 + }, + { + "epoch": 12.798882681564246, + "grad_norm": 2.213301181793213, + "learning_rate": 1.5345251396648046e-05, + "loss": 0.1893, + "step": 4582 + }, + { + "epoch": 12.80167597765363, + "grad_norm": 1.1146209239959717, + "learning_rate": 1.534860335195531e-05, + "loss": 0.2286, + "step": 4583 + }, + { + "epoch": 12.804469273743017, + "grad_norm": 0.6673130989074707, + "learning_rate": 1.5351955307262572e-05, + "loss": 0.2048, + "step": 4584 + }, + { + "epoch": 12.807262569832401, + "grad_norm": 0.7842928171157837, + "learning_rate": 1.5355307262569833e-05, + "loss": 0.217, + "step": 4585 + }, + { + "epoch": 12.810055865921788, + "grad_norm": 0.8054066300392151, + "learning_rate": 1.5358659217877098e-05, + "loss": 0.2294, + "step": 4586 + }, + { + "epoch": 12.812849162011172, + "grad_norm": 1.4514687061309814, + "learning_rate": 1.536201117318436e-05, + "loss": 0.2305, + "step": 4587 + }, + { + "epoch": 12.815642458100559, + "grad_norm": 0.7892318964004517, + "learning_rate": 1.536536312849162e-05, + "loss": 0.1811, + "step": 4588 + }, + { + "epoch": 12.818435754189943, + "grad_norm": 1.9931440353393555, + "learning_rate": 1.5368715083798886e-05, + "loss": 0.2391, + "step": 4589 + }, + { + "epoch": 12.82122905027933, + "grad_norm": 1.0224398374557495, + "learning_rate": 1.5372067039106147e-05, + "loss": 0.1797, + "step": 4590 + }, + { + "epoch": 12.824022346368714, + "grad_norm": 0.987949013710022, + "learning_rate": 1.5375418994413408e-05, + "loss": 0.1929, + "step": 4591 + }, + { + "epoch": 12.8268156424581, + "grad_norm": 1.2291994094848633, + "learning_rate": 1.5378770949720673e-05, + "loss": 0.1855, + "step": 4592 + }, + { + "epoch": 12.829608938547485, + "grad_norm": 1.2118107080459595, + "learning_rate": 1.5382122905027934e-05, + "loss": 0.2011, + "step": 4593 + }, + { + "epoch": 12.832402234636872, + "grad_norm": 1.2764337062835693, + "learning_rate": 1.5385474860335196e-05, + "loss": 0.2221, + "step": 4594 + }, + { + "epoch": 12.835195530726256, + "grad_norm": 7.10743522644043, + "learning_rate": 1.538882681564246e-05, + "loss": 0.2152, + "step": 4595 + }, + { + "epoch": 12.837988826815643, + "grad_norm": 2.034189462661743, + "learning_rate": 1.5392178770949722e-05, + "loss": 0.2963, + "step": 4596 + }, + { + "epoch": 12.840782122905027, + "grad_norm": 2.0242857933044434, + "learning_rate": 1.5395530726256983e-05, + "loss": 0.2192, + "step": 4597 + }, + { + "epoch": 12.843575418994414, + "grad_norm": 0.7604122757911682, + "learning_rate": 1.5398882681564248e-05, + "loss": 0.2249, + "step": 4598 + }, + { + "epoch": 12.846368715083798, + "grad_norm": 0.6256192922592163, + "learning_rate": 1.540223463687151e-05, + "loss": 0.2046, + "step": 4599 + }, + { + "epoch": 12.849162011173185, + "grad_norm": 0.717348039150238, + "learning_rate": 1.540558659217877e-05, + "loss": 0.1932, + "step": 4600 + }, + { + "epoch": 12.85195530726257, + "grad_norm": 0.6753129959106445, + "learning_rate": 1.5408938547486035e-05, + "loss": 0.229, + "step": 4601 + }, + { + "epoch": 12.854748603351956, + "grad_norm": 0.7360239028930664, + "learning_rate": 1.5412290502793297e-05, + "loss": 0.2087, + "step": 4602 + }, + { + "epoch": 12.85754189944134, + "grad_norm": 0.7384194731712341, + "learning_rate": 1.541564245810056e-05, + "loss": 0.2097, + "step": 4603 + }, + { + "epoch": 12.860335195530727, + "grad_norm": 0.5891803503036499, + "learning_rate": 1.5418994413407823e-05, + "loss": 0.2364, + "step": 4604 + }, + { + "epoch": 12.863128491620111, + "grad_norm": 0.7006487250328064, + "learning_rate": 1.5422346368715084e-05, + "loss": 0.2052, + "step": 4605 + }, + { + "epoch": 12.865921787709498, + "grad_norm": 0.6371285319328308, + "learning_rate": 1.542569832402235e-05, + "loss": 0.1841, + "step": 4606 + }, + { + "epoch": 12.868715083798882, + "grad_norm": 0.8616012334823608, + "learning_rate": 1.542905027932961e-05, + "loss": 0.2177, + "step": 4607 + }, + { + "epoch": 12.871508379888269, + "grad_norm": 1.073510766029358, + "learning_rate": 1.543240223463687e-05, + "loss": 0.1796, + "step": 4608 + }, + { + "epoch": 12.874301675977653, + "grad_norm": 0.7134238481521606, + "learning_rate": 1.5435754189944136e-05, + "loss": 0.1808, + "step": 4609 + }, + { + "epoch": 12.87709497206704, + "grad_norm": 1.1997520923614502, + "learning_rate": 1.5439106145251398e-05, + "loss": 0.2128, + "step": 4610 + }, + { + "epoch": 12.879888268156424, + "grad_norm": 1.186794400215149, + "learning_rate": 1.544245810055866e-05, + "loss": 0.2648, + "step": 4611 + }, + { + "epoch": 12.88268156424581, + "grad_norm": 1.3412911891937256, + "learning_rate": 1.5445810055865924e-05, + "loss": 0.2013, + "step": 4612 + }, + { + "epoch": 12.885474860335195, + "grad_norm": 0.909196138381958, + "learning_rate": 1.5449162011173185e-05, + "loss": 0.1926, + "step": 4613 + }, + { + "epoch": 12.888268156424582, + "grad_norm": 1.2015358209609985, + "learning_rate": 1.5452513966480446e-05, + "loss": 0.1716, + "step": 4614 + }, + { + "epoch": 12.891061452513966, + "grad_norm": 0.8964020609855652, + "learning_rate": 1.545586592178771e-05, + "loss": 0.2008, + "step": 4615 + }, + { + "epoch": 12.893854748603353, + "grad_norm": 0.8544033169746399, + "learning_rate": 1.5459217877094972e-05, + "loss": 0.2419, + "step": 4616 + }, + { + "epoch": 12.896648044692737, + "grad_norm": 1.184876799583435, + "learning_rate": 1.5462569832402234e-05, + "loss": 0.2465, + "step": 4617 + }, + { + "epoch": 12.899441340782122, + "grad_norm": 1.1161446571350098, + "learning_rate": 1.54659217877095e-05, + "loss": 0.2614, + "step": 4618 + }, + { + "epoch": 12.902234636871508, + "grad_norm": 1.449416160583496, + "learning_rate": 1.546927374301676e-05, + "loss": 0.2012, + "step": 4619 + }, + { + "epoch": 12.905027932960895, + "grad_norm": 1.6446787118911743, + "learning_rate": 1.547262569832402e-05, + "loss": 0.267, + "step": 4620 + }, + { + "epoch": 12.90782122905028, + "grad_norm": 7.1978888511657715, + "learning_rate": 1.5475977653631286e-05, + "loss": 0.3224, + "step": 4621 + }, + { + "epoch": 12.910614525139664, + "grad_norm": 0.5302404761314392, + "learning_rate": 1.5479329608938547e-05, + "loss": 0.2049, + "step": 4622 + }, + { + "epoch": 12.91340782122905, + "grad_norm": 0.5768572092056274, + "learning_rate": 1.5482681564245812e-05, + "loss": 0.2029, + "step": 4623 + }, + { + "epoch": 12.916201117318435, + "grad_norm": 0.6215446591377258, + "learning_rate": 1.5486033519553073e-05, + "loss": 0.1673, + "step": 4624 + }, + { + "epoch": 12.918994413407821, + "grad_norm": 0.7465837597846985, + "learning_rate": 1.5489385474860335e-05, + "loss": 0.2264, + "step": 4625 + }, + { + "epoch": 12.921787709497206, + "grad_norm": 1.0359233617782593, + "learning_rate": 1.54927374301676e-05, + "loss": 0.2123, + "step": 4626 + }, + { + "epoch": 12.924581005586592, + "grad_norm": 1.130145788192749, + "learning_rate": 1.549608938547486e-05, + "loss": 0.1815, + "step": 4627 + }, + { + "epoch": 12.927374301675977, + "grad_norm": 0.7232991456985474, + "learning_rate": 1.5499441340782122e-05, + "loss": 0.2193, + "step": 4628 + }, + { + "epoch": 12.930167597765363, + "grad_norm": 0.5345622301101685, + "learning_rate": 1.5502793296089387e-05, + "loss": 0.2052, + "step": 4629 + }, + { + "epoch": 12.932960893854748, + "grad_norm": 1.083169937133789, + "learning_rate": 1.5506145251396648e-05, + "loss": 0.1817, + "step": 4630 + }, + { + "epoch": 12.935754189944134, + "grad_norm": 1.599916934967041, + "learning_rate": 1.550949720670391e-05, + "loss": 0.1875, + "step": 4631 + }, + { + "epoch": 12.938547486033519, + "grad_norm": 0.8208882212638855, + "learning_rate": 1.5512849162011174e-05, + "loss": 0.2152, + "step": 4632 + }, + { + "epoch": 12.941340782122905, + "grad_norm": 0.8549173474311829, + "learning_rate": 1.5516201117318436e-05, + "loss": 0.2045, + "step": 4633 + }, + { + "epoch": 12.94413407821229, + "grad_norm": 1.0756707191467285, + "learning_rate": 1.5519553072625697e-05, + "loss": 0.2187, + "step": 4634 + }, + { + "epoch": 12.946927374301676, + "grad_norm": 0.9438807964324951, + "learning_rate": 1.552290502793296e-05, + "loss": 0.1891, + "step": 4635 + }, + { + "epoch": 12.949720670391061, + "grad_norm": 1.3996555805206299, + "learning_rate": 1.5526256983240223e-05, + "loss": 0.1955, + "step": 4636 + }, + { + "epoch": 12.952513966480447, + "grad_norm": 0.7733123898506165, + "learning_rate": 1.5529608938547484e-05, + "loss": 0.1917, + "step": 4637 + }, + { + "epoch": 12.955307262569832, + "grad_norm": 1.3009006977081299, + "learning_rate": 1.553296089385475e-05, + "loss": 0.2102, + "step": 4638 + }, + { + "epoch": 12.958100558659218, + "grad_norm": 0.7826189398765564, + "learning_rate": 1.553631284916201e-05, + "loss": 0.2036, + "step": 4639 + }, + { + "epoch": 12.960893854748603, + "grad_norm": 0.8804622292518616, + "learning_rate": 1.5539664804469275e-05, + "loss": 0.1882, + "step": 4640 + }, + { + "epoch": 12.96368715083799, + "grad_norm": 1.682826280593872, + "learning_rate": 1.5543016759776537e-05, + "loss": 0.2459, + "step": 4641 + }, + { + "epoch": 12.966480446927374, + "grad_norm": 1.2753267288208008, + "learning_rate": 1.5546368715083798e-05, + "loss": 0.1737, + "step": 4642 + }, + { + "epoch": 12.96927374301676, + "grad_norm": 2.853429079055786, + "learning_rate": 1.5549720670391063e-05, + "loss": 0.2075, + "step": 4643 + }, + { + "epoch": 12.972067039106145, + "grad_norm": 1.4344110488891602, + "learning_rate": 1.5553072625698324e-05, + "loss": 0.2415, + "step": 4644 + }, + { + "epoch": 12.974860335195531, + "grad_norm": 1.6163359880447388, + "learning_rate": 1.5556424581005585e-05, + "loss": 0.2205, + "step": 4645 + }, + { + "epoch": 12.977653631284916, + "grad_norm": 2.843275547027588, + "learning_rate": 1.555977653631285e-05, + "loss": 0.2989, + "step": 4646 + }, + { + "epoch": 12.980446927374302, + "grad_norm": 0.6586806178092957, + "learning_rate": 1.556312849162011e-05, + "loss": 0.2059, + "step": 4647 + }, + { + "epoch": 12.983240223463687, + "grad_norm": 1.1225495338439941, + "learning_rate": 1.5566480446927373e-05, + "loss": 0.2055, + "step": 4648 + }, + { + "epoch": 12.986033519553073, + "grad_norm": 0.6346359848976135, + "learning_rate": 1.5569832402234637e-05, + "loss": 0.1968, + "step": 4649 + }, + { + "epoch": 12.988826815642458, + "grad_norm": 0.875960111618042, + "learning_rate": 1.55731843575419e-05, + "loss": 0.2594, + "step": 4650 + }, + { + "epoch": 12.991620111731844, + "grad_norm": 0.9056214094161987, + "learning_rate": 1.557653631284916e-05, + "loss": 0.1884, + "step": 4651 + }, + { + "epoch": 12.994413407821229, + "grad_norm": 0.8303931951522827, + "learning_rate": 1.5579888268156425e-05, + "loss": 0.1797, + "step": 4652 + }, + { + "epoch": 12.997206703910614, + "grad_norm": 0.9106590151786804, + "learning_rate": 1.5583240223463686e-05, + "loss": 0.1727, + "step": 4653 + }, + { + "epoch": 13.0, + "grad_norm": 1.801419973373413, + "learning_rate": 1.5586592178770948e-05, + "loss": 0.2114, + "step": 4654 + }, + { + "epoch": 13.002793296089385, + "grad_norm": 0.6166894435882568, + "learning_rate": 1.5589944134078212e-05, + "loss": 0.2136, + "step": 4655 + }, + { + "epoch": 13.005586592178771, + "grad_norm": 0.6202844381332397, + "learning_rate": 1.5593296089385474e-05, + "loss": 0.2131, + "step": 4656 + }, + { + "epoch": 13.008379888268156, + "grad_norm": 0.5812720656394958, + "learning_rate": 1.5596648044692735e-05, + "loss": 0.1932, + "step": 4657 + }, + { + "epoch": 13.011173184357542, + "grad_norm": 0.5939750075340271, + "learning_rate": 1.56e-05, + "loss": 0.2243, + "step": 4658 + }, + { + "epoch": 13.013966480446927, + "grad_norm": 0.9058940410614014, + "learning_rate": 1.560335195530726e-05, + "loss": 0.1972, + "step": 4659 + }, + { + "epoch": 13.016759776536313, + "grad_norm": 2.26969838142395, + "learning_rate": 1.5606703910614526e-05, + "loss": 0.1829, + "step": 4660 + }, + { + "epoch": 13.019553072625698, + "grad_norm": 0.6690644025802612, + "learning_rate": 1.5610055865921787e-05, + "loss": 0.2007, + "step": 4661 + }, + { + "epoch": 13.022346368715084, + "grad_norm": 0.7294594049453735, + "learning_rate": 1.561340782122905e-05, + "loss": 0.1987, + "step": 4662 + }, + { + "epoch": 13.025139664804469, + "grad_norm": 0.7402754426002502, + "learning_rate": 1.5616759776536313e-05, + "loss": 0.2017, + "step": 4663 + }, + { + "epoch": 13.027932960893855, + "grad_norm": 1.1012238264083862, + "learning_rate": 1.5620111731843575e-05, + "loss": 0.2057, + "step": 4664 + }, + { + "epoch": 13.03072625698324, + "grad_norm": 0.666743814945221, + "learning_rate": 1.5623463687150836e-05, + "loss": 0.1782, + "step": 4665 + }, + { + "epoch": 13.033519553072626, + "grad_norm": 0.6916714310646057, + "learning_rate": 1.56268156424581e-05, + "loss": 0.1928, + "step": 4666 + }, + { + "epoch": 13.03631284916201, + "grad_norm": 0.657856822013855, + "learning_rate": 1.5630167597765362e-05, + "loss": 0.1808, + "step": 4667 + }, + { + "epoch": 13.039106145251397, + "grad_norm": 1.0827819108963013, + "learning_rate": 1.5633519553072623e-05, + "loss": 0.2053, + "step": 4668 + }, + { + "epoch": 13.041899441340782, + "grad_norm": 0.8894603252410889, + "learning_rate": 1.5636871508379888e-05, + "loss": 0.1841, + "step": 4669 + }, + { + "epoch": 13.044692737430168, + "grad_norm": 0.9048619866371155, + "learning_rate": 1.564022346368715e-05, + "loss": 0.1801, + "step": 4670 + }, + { + "epoch": 13.047486033519553, + "grad_norm": 0.9195515513420105, + "learning_rate": 1.5643575418994414e-05, + "loss": 0.2037, + "step": 4671 + }, + { + "epoch": 13.050279329608939, + "grad_norm": 1.9849917888641357, + "learning_rate": 1.564692737430168e-05, + "loss": 0.2229, + "step": 4672 + }, + { + "epoch": 13.053072625698324, + "grad_norm": 2.378312110900879, + "learning_rate": 1.565027932960894e-05, + "loss": 0.2039, + "step": 4673 + }, + { + "epoch": 13.05586592178771, + "grad_norm": 1.1395339965820312, + "learning_rate": 1.56536312849162e-05, + "loss": 0.2251, + "step": 4674 + }, + { + "epoch": 13.058659217877095, + "grad_norm": 1.308855414390564, + "learning_rate": 1.5656983240223466e-05, + "loss": 0.2623, + "step": 4675 + }, + { + "epoch": 13.061452513966481, + "grad_norm": 1.1780433654785156, + "learning_rate": 1.5660335195530728e-05, + "loss": 0.2109, + "step": 4676 + }, + { + "epoch": 13.064245810055866, + "grad_norm": 1.6419007778167725, + "learning_rate": 1.5663687150837992e-05, + "loss": 0.2043, + "step": 4677 + }, + { + "epoch": 13.067039106145252, + "grad_norm": 1.1678357124328613, + "learning_rate": 1.5667039106145254e-05, + "loss": 0.2309, + "step": 4678 + }, + { + "epoch": 13.069832402234637, + "grad_norm": 3.456221103668213, + "learning_rate": 1.5670391061452515e-05, + "loss": 0.2492, + "step": 4679 + }, + { + "epoch": 13.072625698324023, + "grad_norm": 0.6803950667381287, + "learning_rate": 1.567374301675978e-05, + "loss": 0.2126, + "step": 4680 + }, + { + "epoch": 13.075418994413408, + "grad_norm": 0.597181499004364, + "learning_rate": 1.567709497206704e-05, + "loss": 0.1918, + "step": 4681 + }, + { + "epoch": 13.078212290502794, + "grad_norm": 0.6651896834373474, + "learning_rate": 1.5680446927374302e-05, + "loss": 0.1874, + "step": 4682 + }, + { + "epoch": 13.081005586592179, + "grad_norm": 1.1741054058074951, + "learning_rate": 1.5683798882681567e-05, + "loss": 0.2123, + "step": 4683 + }, + { + "epoch": 13.083798882681565, + "grad_norm": 1.5331664085388184, + "learning_rate": 1.568715083798883e-05, + "loss": 0.2007, + "step": 4684 + }, + { + "epoch": 13.08659217877095, + "grad_norm": 1.8650803565979004, + "learning_rate": 1.569050279329609e-05, + "loss": 0.1845, + "step": 4685 + }, + { + "epoch": 13.089385474860336, + "grad_norm": 0.7429161667823792, + "learning_rate": 1.5693854748603355e-05, + "loss": 0.1901, + "step": 4686 + }, + { + "epoch": 13.09217877094972, + "grad_norm": 1.1010832786560059, + "learning_rate": 1.5697206703910616e-05, + "loss": 0.1886, + "step": 4687 + }, + { + "epoch": 13.094972067039107, + "grad_norm": 1.542515754699707, + "learning_rate": 1.5700558659217877e-05, + "loss": 0.1933, + "step": 4688 + }, + { + "epoch": 13.097765363128492, + "grad_norm": 0.7891400456428528, + "learning_rate": 1.5703910614525142e-05, + "loss": 0.2014, + "step": 4689 + }, + { + "epoch": 13.100558659217878, + "grad_norm": 0.9151551723480225, + "learning_rate": 1.5707262569832403e-05, + "loss": 0.1774, + "step": 4690 + }, + { + "epoch": 13.103351955307263, + "grad_norm": 0.7000437378883362, + "learning_rate": 1.5710614525139665e-05, + "loss": 0.2032, + "step": 4691 + }, + { + "epoch": 13.106145251396647, + "grad_norm": 1.088547706604004, + "learning_rate": 1.571396648044693e-05, + "loss": 0.1724, + "step": 4692 + }, + { + "epoch": 13.108938547486034, + "grad_norm": 1.3438962697982788, + "learning_rate": 1.571731843575419e-05, + "loss": 0.1851, + "step": 4693 + }, + { + "epoch": 13.111731843575418, + "grad_norm": 0.9679244160652161, + "learning_rate": 1.5720670391061452e-05, + "loss": 0.2158, + "step": 4694 + }, + { + "epoch": 13.114525139664805, + "grad_norm": 0.7700936794281006, + "learning_rate": 1.5724022346368717e-05, + "loss": 0.174, + "step": 4695 + }, + { + "epoch": 13.11731843575419, + "grad_norm": 1.1892937421798706, + "learning_rate": 1.5727374301675978e-05, + "loss": 0.2038, + "step": 4696 + }, + { + "epoch": 13.120111731843576, + "grad_norm": 1.0773591995239258, + "learning_rate": 1.5730726256983243e-05, + "loss": 0.2036, + "step": 4697 + }, + { + "epoch": 13.12290502793296, + "grad_norm": 3.8385283946990967, + "learning_rate": 1.5734078212290504e-05, + "loss": 0.1947, + "step": 4698 + }, + { + "epoch": 13.125698324022347, + "grad_norm": 1.055977463722229, + "learning_rate": 1.5737430167597766e-05, + "loss": 0.1597, + "step": 4699 + }, + { + "epoch": 13.128491620111731, + "grad_norm": 0.7955960035324097, + "learning_rate": 1.574078212290503e-05, + "loss": 0.1841, + "step": 4700 + }, + { + "epoch": 13.131284916201118, + "grad_norm": 1.3788976669311523, + "learning_rate": 1.5744134078212292e-05, + "loss": 0.2339, + "step": 4701 + }, + { + "epoch": 13.134078212290502, + "grad_norm": 0.9190235733985901, + "learning_rate": 1.5747486033519553e-05, + "loss": 0.2121, + "step": 4702 + }, + { + "epoch": 13.136871508379889, + "grad_norm": 1.5323315858840942, + "learning_rate": 1.5750837988826818e-05, + "loss": 0.2512, + "step": 4703 + }, + { + "epoch": 13.139664804469273, + "grad_norm": 3.4467999935150146, + "learning_rate": 1.575418994413408e-05, + "loss": 0.3173, + "step": 4704 + }, + { + "epoch": 13.14245810055866, + "grad_norm": 0.7116386294364929, + "learning_rate": 1.575754189944134e-05, + "loss": 0.2069, + "step": 4705 + }, + { + "epoch": 13.145251396648044, + "grad_norm": 0.5209029316902161, + "learning_rate": 1.5760893854748605e-05, + "loss": 0.1598, + "step": 4706 + }, + { + "epoch": 13.14804469273743, + "grad_norm": 1.3354238271713257, + "learning_rate": 1.5764245810055867e-05, + "loss": 0.2094, + "step": 4707 + }, + { + "epoch": 13.150837988826815, + "grad_norm": 1.1206400394439697, + "learning_rate": 1.5767597765363128e-05, + "loss": 0.1879, + "step": 4708 + }, + { + "epoch": 13.153631284916202, + "grad_norm": 0.9114437103271484, + "learning_rate": 1.5770949720670393e-05, + "loss": 0.2053, + "step": 4709 + }, + { + "epoch": 13.156424581005586, + "grad_norm": 0.8197435140609741, + "learning_rate": 1.5774301675977654e-05, + "loss": 0.22, + "step": 4710 + }, + { + "epoch": 13.159217877094973, + "grad_norm": 0.7229783535003662, + "learning_rate": 1.5777653631284915e-05, + "loss": 0.1916, + "step": 4711 + }, + { + "epoch": 13.162011173184357, + "grad_norm": 0.7535585165023804, + "learning_rate": 1.578100558659218e-05, + "loss": 0.2063, + "step": 4712 + }, + { + "epoch": 13.164804469273744, + "grad_norm": 0.974663496017456, + "learning_rate": 1.578435754189944e-05, + "loss": 0.1873, + "step": 4713 + }, + { + "epoch": 13.167597765363128, + "grad_norm": 1.5345878601074219, + "learning_rate": 1.5787709497206703e-05, + "loss": 0.2385, + "step": 4714 + }, + { + "epoch": 13.170391061452515, + "grad_norm": 1.313805103302002, + "learning_rate": 1.5791061452513968e-05, + "loss": 0.2068, + "step": 4715 + }, + { + "epoch": 13.1731843575419, + "grad_norm": 0.7959429025650024, + "learning_rate": 1.579441340782123e-05, + "loss": 0.1844, + "step": 4716 + }, + { + "epoch": 13.175977653631286, + "grad_norm": 0.8026633262634277, + "learning_rate": 1.5797765363128494e-05, + "loss": 0.187, + "step": 4717 + }, + { + "epoch": 13.17877094972067, + "grad_norm": 1.2349560260772705, + "learning_rate": 1.5801117318435755e-05, + "loss": 0.2112, + "step": 4718 + }, + { + "epoch": 13.181564245810057, + "grad_norm": 0.9775865077972412, + "learning_rate": 1.5804469273743016e-05, + "loss": 0.1901, + "step": 4719 + }, + { + "epoch": 13.184357541899441, + "grad_norm": 1.6138007640838623, + "learning_rate": 1.580782122905028e-05, + "loss": 0.1937, + "step": 4720 + }, + { + "epoch": 13.187150837988828, + "grad_norm": 1.1012957096099854, + "learning_rate": 1.5811173184357542e-05, + "loss": 0.2022, + "step": 4721 + }, + { + "epoch": 13.189944134078212, + "grad_norm": 0.9249135255813599, + "learning_rate": 1.5814525139664804e-05, + "loss": 0.1896, + "step": 4722 + }, + { + "epoch": 13.192737430167599, + "grad_norm": 0.9934033751487732, + "learning_rate": 1.581787709497207e-05, + "loss": 0.1931, + "step": 4723 + }, + { + "epoch": 13.195530726256983, + "grad_norm": 3.1052839756011963, + "learning_rate": 1.582122905027933e-05, + "loss": 0.1951, + "step": 4724 + }, + { + "epoch": 13.19832402234637, + "grad_norm": 0.9606661200523376, + "learning_rate": 1.582458100558659e-05, + "loss": 0.2469, + "step": 4725 + }, + { + "epoch": 13.201117318435754, + "grad_norm": 1.7827167510986328, + "learning_rate": 1.5827932960893856e-05, + "loss": 0.2206, + "step": 4726 + }, + { + "epoch": 13.203910614525139, + "grad_norm": 1.1446841955184937, + "learning_rate": 1.5831284916201117e-05, + "loss": 0.2017, + "step": 4727 + }, + { + "epoch": 13.206703910614525, + "grad_norm": 1.1832813024520874, + "learning_rate": 1.583463687150838e-05, + "loss": 0.2658, + "step": 4728 + }, + { + "epoch": 13.20949720670391, + "grad_norm": 3.051379680633545, + "learning_rate": 1.5837988826815643e-05, + "loss": 0.3172, + "step": 4729 + }, + { + "epoch": 13.212290502793296, + "grad_norm": 0.42256617546081543, + "learning_rate": 1.5841340782122905e-05, + "loss": 0.1899, + "step": 4730 + }, + { + "epoch": 13.21508379888268, + "grad_norm": 0.5582777857780457, + "learning_rate": 1.5844692737430166e-05, + "loss": 0.2091, + "step": 4731 + }, + { + "epoch": 13.217877094972067, + "grad_norm": 0.5399238467216492, + "learning_rate": 1.584804469273743e-05, + "loss": 0.1859, + "step": 4732 + }, + { + "epoch": 13.220670391061452, + "grad_norm": 0.5145843029022217, + "learning_rate": 1.5851396648044692e-05, + "loss": 0.1965, + "step": 4733 + }, + { + "epoch": 13.223463687150838, + "grad_norm": 0.7261435389518738, + "learning_rate": 1.5854748603351957e-05, + "loss": 0.2067, + "step": 4734 + }, + { + "epoch": 13.226256983240223, + "grad_norm": 0.5264418125152588, + "learning_rate": 1.5858100558659218e-05, + "loss": 0.1724, + "step": 4735 + }, + { + "epoch": 13.22905027932961, + "grad_norm": 1.7375001907348633, + "learning_rate": 1.586145251396648e-05, + "loss": 0.2162, + "step": 4736 + }, + { + "epoch": 13.231843575418994, + "grad_norm": 0.8217145204544067, + "learning_rate": 1.5864804469273744e-05, + "loss": 0.2264, + "step": 4737 + }, + { + "epoch": 13.23463687150838, + "grad_norm": 0.6734139323234558, + "learning_rate": 1.5868156424581006e-05, + "loss": 0.1989, + "step": 4738 + }, + { + "epoch": 13.237430167597765, + "grad_norm": 0.7252740263938904, + "learning_rate": 1.5871508379888267e-05, + "loss": 0.1653, + "step": 4739 + }, + { + "epoch": 13.240223463687151, + "grad_norm": 0.6780953407287598, + "learning_rate": 1.587486033519553e-05, + "loss": 0.1726, + "step": 4740 + }, + { + "epoch": 13.243016759776536, + "grad_norm": 0.9457089304924011, + "learning_rate": 1.5878212290502793e-05, + "loss": 0.1737, + "step": 4741 + }, + { + "epoch": 13.245810055865922, + "grad_norm": 0.8742204904556274, + "learning_rate": 1.5881564245810054e-05, + "loss": 0.1862, + "step": 4742 + }, + { + "epoch": 13.248603351955307, + "grad_norm": 1.1436315774917603, + "learning_rate": 1.588491620111732e-05, + "loss": 0.1823, + "step": 4743 + }, + { + "epoch": 13.251396648044693, + "grad_norm": 1.0593018531799316, + "learning_rate": 1.588826815642458e-05, + "loss": 0.1982, + "step": 4744 + }, + { + "epoch": 13.254189944134078, + "grad_norm": 1.0438194274902344, + "learning_rate": 1.5891620111731842e-05, + "loss": 0.1804, + "step": 4745 + }, + { + "epoch": 13.256983240223464, + "grad_norm": 0.7083249092102051, + "learning_rate": 1.5894972067039107e-05, + "loss": 0.187, + "step": 4746 + }, + { + "epoch": 13.259776536312849, + "grad_norm": 1.2586318254470825, + "learning_rate": 1.5898324022346368e-05, + "loss": 0.2117, + "step": 4747 + }, + { + "epoch": 13.262569832402235, + "grad_norm": 1.2478784322738647, + "learning_rate": 1.590167597765363e-05, + "loss": 0.2111, + "step": 4748 + }, + { + "epoch": 13.26536312849162, + "grad_norm": 0.8364942669868469, + "learning_rate": 1.5905027932960894e-05, + "loss": 0.2018, + "step": 4749 + }, + { + "epoch": 13.268156424581006, + "grad_norm": 1.3968884944915771, + "learning_rate": 1.5908379888268155e-05, + "loss": 0.2238, + "step": 4750 + }, + { + "epoch": 13.27094972067039, + "grad_norm": 0.9345178008079529, + "learning_rate": 1.5911731843575417e-05, + "loss": 0.1867, + "step": 4751 + }, + { + "epoch": 13.273743016759777, + "grad_norm": 1.5155377388000488, + "learning_rate": 1.591508379888268e-05, + "loss": 0.224, + "step": 4752 + }, + { + "epoch": 13.276536312849162, + "grad_norm": 1.6415551900863647, + "learning_rate": 1.5918435754189943e-05, + "loss": 0.24, + "step": 4753 + }, + { + "epoch": 13.279329608938548, + "grad_norm": 2.402129650115967, + "learning_rate": 1.5921787709497207e-05, + "loss": 0.346, + "step": 4754 + }, + { + "epoch": 13.282122905027933, + "grad_norm": 0.516152024269104, + "learning_rate": 1.592513966480447e-05, + "loss": 0.207, + "step": 4755 + }, + { + "epoch": 13.28491620111732, + "grad_norm": 0.5076278448104858, + "learning_rate": 1.592849162011173e-05, + "loss": 0.2, + "step": 4756 + }, + { + "epoch": 13.287709497206704, + "grad_norm": 0.4614071249961853, + "learning_rate": 1.5931843575418995e-05, + "loss": 0.185, + "step": 4757 + }, + { + "epoch": 13.29050279329609, + "grad_norm": 0.761559009552002, + "learning_rate": 1.5935195530726256e-05, + "loss": 0.2283, + "step": 4758 + }, + { + "epoch": 13.293296089385475, + "grad_norm": 0.6620398759841919, + "learning_rate": 1.5938547486033518e-05, + "loss": 0.1823, + "step": 4759 + }, + { + "epoch": 13.296089385474861, + "grad_norm": 0.9769040942192078, + "learning_rate": 1.5941899441340782e-05, + "loss": 0.1908, + "step": 4760 + }, + { + "epoch": 13.298882681564246, + "grad_norm": 0.7494993805885315, + "learning_rate": 1.5945251396648044e-05, + "loss": 0.2015, + "step": 4761 + }, + { + "epoch": 13.30167597765363, + "grad_norm": 0.5593118071556091, + "learning_rate": 1.5948603351955305e-05, + "loss": 0.1802, + "step": 4762 + }, + { + "epoch": 13.304469273743017, + "grad_norm": 1.3243787288665771, + "learning_rate": 1.595195530726257e-05, + "loss": 0.185, + "step": 4763 + }, + { + "epoch": 13.307262569832401, + "grad_norm": 0.6742425560951233, + "learning_rate": 1.595530726256983e-05, + "loss": 0.2091, + "step": 4764 + }, + { + "epoch": 13.310055865921788, + "grad_norm": 0.6281177997589111, + "learning_rate": 1.5958659217877092e-05, + "loss": 0.1838, + "step": 4765 + }, + { + "epoch": 13.312849162011172, + "grad_norm": 0.7077625393867493, + "learning_rate": 1.5962011173184357e-05, + "loss": 0.2005, + "step": 4766 + }, + { + "epoch": 13.315642458100559, + "grad_norm": 0.7133495211601257, + "learning_rate": 1.596536312849162e-05, + "loss": 0.171, + "step": 4767 + }, + { + "epoch": 13.318435754189943, + "grad_norm": 0.6566075682640076, + "learning_rate": 1.5968715083798883e-05, + "loss": 0.1994, + "step": 4768 + }, + { + "epoch": 13.32122905027933, + "grad_norm": 1.0809876918792725, + "learning_rate": 1.5972067039106148e-05, + "loss": 0.1918, + "step": 4769 + }, + { + "epoch": 13.324022346368714, + "grad_norm": 0.8216356039047241, + "learning_rate": 1.597541899441341e-05, + "loss": 0.1828, + "step": 4770 + }, + { + "epoch": 13.3268156424581, + "grad_norm": 0.7890827059745789, + "learning_rate": 1.5978770949720674e-05, + "loss": 0.1852, + "step": 4771 + }, + { + "epoch": 13.329608938547485, + "grad_norm": 0.8628719449043274, + "learning_rate": 1.5982122905027935e-05, + "loss": 0.1916, + "step": 4772 + }, + { + "epoch": 13.332402234636872, + "grad_norm": 0.6726382374763489, + "learning_rate": 1.5985474860335197e-05, + "loss": 0.1711, + "step": 4773 + }, + { + "epoch": 13.335195530726256, + "grad_norm": 0.9519588947296143, + "learning_rate": 1.598882681564246e-05, + "loss": 0.1914, + "step": 4774 + }, + { + "epoch": 13.337988826815643, + "grad_norm": 0.9266947507858276, + "learning_rate": 1.5992178770949723e-05, + "loss": 0.1939, + "step": 4775 + }, + { + "epoch": 13.340782122905027, + "grad_norm": 0.9684043526649475, + "learning_rate": 1.5995530726256984e-05, + "loss": 0.2078, + "step": 4776 + }, + { + "epoch": 13.343575418994414, + "grad_norm": 1.2138890027999878, + "learning_rate": 1.599888268156425e-05, + "loss": 0.1894, + "step": 4777 + }, + { + "epoch": 13.346368715083798, + "grad_norm": 2.311849355697632, + "learning_rate": 1.600223463687151e-05, + "loss": 0.2268, + "step": 4778 + }, + { + "epoch": 13.349162011173185, + "grad_norm": 2.6976332664489746, + "learning_rate": 1.600558659217877e-05, + "loss": 0.2791, + "step": 4779 + }, + { + "epoch": 13.35195530726257, + "grad_norm": 0.5508947968482971, + "learning_rate": 1.6008938547486036e-05, + "loss": 0.1976, + "step": 4780 + }, + { + "epoch": 13.354748603351956, + "grad_norm": 0.6511574387550354, + "learning_rate": 1.6012290502793298e-05, + "loss": 0.1944, + "step": 4781 + }, + { + "epoch": 13.35754189944134, + "grad_norm": 0.798784077167511, + "learning_rate": 1.601564245810056e-05, + "loss": 0.2177, + "step": 4782 + }, + { + "epoch": 13.360335195530727, + "grad_norm": 0.5871323943138123, + "learning_rate": 1.6018994413407824e-05, + "loss": 0.1784, + "step": 4783 + }, + { + "epoch": 13.363128491620111, + "grad_norm": 0.8562290668487549, + "learning_rate": 1.6022346368715085e-05, + "loss": 0.2255, + "step": 4784 + }, + { + "epoch": 13.365921787709498, + "grad_norm": 0.6618053317070007, + "learning_rate": 1.6025698324022346e-05, + "loss": 0.1621, + "step": 4785 + }, + { + "epoch": 13.368715083798882, + "grad_norm": 0.8682106733322144, + "learning_rate": 1.602905027932961e-05, + "loss": 0.1847, + "step": 4786 + }, + { + "epoch": 13.371508379888269, + "grad_norm": 0.9096164703369141, + "learning_rate": 1.6032402234636872e-05, + "loss": 0.2047, + "step": 4787 + }, + { + "epoch": 13.374301675977653, + "grad_norm": 0.7922865748405457, + "learning_rate": 1.6035754189944134e-05, + "loss": 0.1685, + "step": 4788 + }, + { + "epoch": 13.37709497206704, + "grad_norm": 0.9352543950080872, + "learning_rate": 1.60391061452514e-05, + "loss": 0.1857, + "step": 4789 + }, + { + "epoch": 13.379888268156424, + "grad_norm": 0.8572918772697449, + "learning_rate": 1.604245810055866e-05, + "loss": 0.191, + "step": 4790 + }, + { + "epoch": 13.38268156424581, + "grad_norm": 1.0468302965164185, + "learning_rate": 1.6045810055865925e-05, + "loss": 0.2219, + "step": 4791 + }, + { + "epoch": 13.385474860335195, + "grad_norm": 0.6840357184410095, + "learning_rate": 1.6049162011173186e-05, + "loss": 0.1857, + "step": 4792 + }, + { + "epoch": 13.388268156424582, + "grad_norm": 0.878677248954773, + "learning_rate": 1.6052513966480447e-05, + "loss": 0.1786, + "step": 4793 + }, + { + "epoch": 13.391061452513966, + "grad_norm": 0.6764219403266907, + "learning_rate": 1.6055865921787712e-05, + "loss": 0.1806, + "step": 4794 + }, + { + "epoch": 13.393854748603353, + "grad_norm": 1.881650686264038, + "learning_rate": 1.6059217877094973e-05, + "loss": 0.172, + "step": 4795 + }, + { + "epoch": 13.396648044692737, + "grad_norm": 1.0919965505599976, + "learning_rate": 1.6062569832402235e-05, + "loss": 0.2305, + "step": 4796 + }, + { + "epoch": 13.399441340782122, + "grad_norm": 0.942333459854126, + "learning_rate": 1.60659217877095e-05, + "loss": 0.1842, + "step": 4797 + }, + { + "epoch": 13.402234636871508, + "grad_norm": 0.8606573939323425, + "learning_rate": 1.606927374301676e-05, + "loss": 0.1667, + "step": 4798 + }, + { + "epoch": 13.405027932960893, + "grad_norm": 1.068237066268921, + "learning_rate": 1.6072625698324022e-05, + "loss": 0.2372, + "step": 4799 + }, + { + "epoch": 13.40782122905028, + "grad_norm": 2.513293981552124, + "learning_rate": 1.6075977653631287e-05, + "loss": 0.2719, + "step": 4800 + }, + { + "epoch": 13.410614525139664, + "grad_norm": 1.3651134967803955, + "learning_rate": 1.6079329608938548e-05, + "loss": 0.2022, + "step": 4801 + }, + { + "epoch": 13.41340782122905, + "grad_norm": 1.8024693727493286, + "learning_rate": 1.608268156424581e-05, + "loss": 0.2101, + "step": 4802 + }, + { + "epoch": 13.416201117318435, + "grad_norm": 1.5809106826782227, + "learning_rate": 1.6086033519553074e-05, + "loss": 0.264, + "step": 4803 + }, + { + "epoch": 13.418994413407821, + "grad_norm": 2.2597386837005615, + "learning_rate": 1.6089385474860336e-05, + "loss": 0.249, + "step": 4804 + }, + { + "epoch": 13.421787709497206, + "grad_norm": 0.49252399802207947, + "learning_rate": 1.6092737430167597e-05, + "loss": 0.1772, + "step": 4805 + }, + { + "epoch": 13.424581005586592, + "grad_norm": 1.074953317642212, + "learning_rate": 1.6096089385474862e-05, + "loss": 0.1986, + "step": 4806 + }, + { + "epoch": 13.427374301675977, + "grad_norm": 0.7280862927436829, + "learning_rate": 1.6099441340782123e-05, + "loss": 0.1818, + "step": 4807 + }, + { + "epoch": 13.430167597765363, + "grad_norm": 0.6528027057647705, + "learning_rate": 1.6102793296089384e-05, + "loss": 0.2325, + "step": 4808 + }, + { + "epoch": 13.432960893854748, + "grad_norm": 0.5856274366378784, + "learning_rate": 1.610614525139665e-05, + "loss": 0.1836, + "step": 4809 + }, + { + "epoch": 13.435754189944134, + "grad_norm": 0.44622865319252014, + "learning_rate": 1.610949720670391e-05, + "loss": 0.1994, + "step": 4810 + }, + { + "epoch": 13.438547486033519, + "grad_norm": 0.7826400399208069, + "learning_rate": 1.6112849162011175e-05, + "loss": 0.2248, + "step": 4811 + }, + { + "epoch": 13.441340782122905, + "grad_norm": 0.6461165547370911, + "learning_rate": 1.6116201117318437e-05, + "loss": 0.188, + "step": 4812 + }, + { + "epoch": 13.44413407821229, + "grad_norm": 0.6105127334594727, + "learning_rate": 1.6119553072625698e-05, + "loss": 0.1776, + "step": 4813 + }, + { + "epoch": 13.446927374301676, + "grad_norm": 0.7952030301094055, + "learning_rate": 1.6122905027932963e-05, + "loss": 0.2199, + "step": 4814 + }, + { + "epoch": 13.449720670391061, + "grad_norm": 0.7677476406097412, + "learning_rate": 1.6126256983240224e-05, + "loss": 0.228, + "step": 4815 + }, + { + "epoch": 13.452513966480447, + "grad_norm": 0.7719614505767822, + "learning_rate": 1.6129608938547485e-05, + "loss": 0.2098, + "step": 4816 + }, + { + "epoch": 13.455307262569832, + "grad_norm": 0.8099170327186584, + "learning_rate": 1.613296089385475e-05, + "loss": 0.1692, + "step": 4817 + }, + { + "epoch": 13.458100558659218, + "grad_norm": 0.7161393165588379, + "learning_rate": 1.613631284916201e-05, + "loss": 0.2217, + "step": 4818 + }, + { + "epoch": 13.460893854748603, + "grad_norm": 0.8384212255477905, + "learning_rate": 1.6139664804469273e-05, + "loss": 0.1677, + "step": 4819 + }, + { + "epoch": 13.46368715083799, + "grad_norm": 1.2560189962387085, + "learning_rate": 1.6143016759776538e-05, + "loss": 0.2361, + "step": 4820 + }, + { + "epoch": 13.466480446927374, + "grad_norm": 1.0565204620361328, + "learning_rate": 1.61463687150838e-05, + "loss": 0.1782, + "step": 4821 + }, + { + "epoch": 13.46927374301676, + "grad_norm": 1.1636459827423096, + "learning_rate": 1.614972067039106e-05, + "loss": 0.1849, + "step": 4822 + }, + { + "epoch": 13.472067039106145, + "grad_norm": 5.4785475730896, + "learning_rate": 1.6153072625698325e-05, + "loss": 0.2105, + "step": 4823 + }, + { + "epoch": 13.474860335195531, + "grad_norm": 1.934780240058899, + "learning_rate": 1.6156424581005586e-05, + "loss": 0.2155, + "step": 4824 + }, + { + "epoch": 13.477653631284916, + "grad_norm": 16.415307998657227, + "learning_rate": 1.6159776536312848e-05, + "loss": 0.2043, + "step": 4825 + }, + { + "epoch": 13.480446927374302, + "grad_norm": 1.3127055168151855, + "learning_rate": 1.6163128491620112e-05, + "loss": 0.2106, + "step": 4826 + }, + { + "epoch": 13.483240223463687, + "grad_norm": 2.5339457988739014, + "learning_rate": 1.6166480446927374e-05, + "loss": 0.2526, + "step": 4827 + }, + { + "epoch": 13.486033519553073, + "grad_norm": 1.4295283555984497, + "learning_rate": 1.616983240223464e-05, + "loss": 0.2283, + "step": 4828 + }, + { + "epoch": 13.488826815642458, + "grad_norm": 5.241528034210205, + "learning_rate": 1.61731843575419e-05, + "loss": 0.3069, + "step": 4829 + }, + { + "epoch": 13.491620111731844, + "grad_norm": 0.7426085472106934, + "learning_rate": 1.617653631284916e-05, + "loss": 0.2137, + "step": 4830 + }, + { + "epoch": 13.494413407821229, + "grad_norm": 0.6801235675811768, + "learning_rate": 1.6179888268156426e-05, + "loss": 0.1828, + "step": 4831 + }, + { + "epoch": 13.497206703910614, + "grad_norm": 0.698849081993103, + "learning_rate": 1.6183240223463687e-05, + "loss": 0.2021, + "step": 4832 + }, + { + "epoch": 13.5, + "grad_norm": 0.5751041173934937, + "learning_rate": 1.618659217877095e-05, + "loss": 0.1948, + "step": 4833 + }, + { + "epoch": 13.502793296089386, + "grad_norm": 0.6604710221290588, + "learning_rate": 1.6189944134078213e-05, + "loss": 0.202, + "step": 4834 + }, + { + "epoch": 13.505586592178771, + "grad_norm": 0.7131868004798889, + "learning_rate": 1.6193296089385475e-05, + "loss": 0.2146, + "step": 4835 + }, + { + "epoch": 13.508379888268156, + "grad_norm": 0.6764258742332458, + "learning_rate": 1.6196648044692736e-05, + "loss": 0.1831, + "step": 4836 + }, + { + "epoch": 13.511173184357542, + "grad_norm": 0.6770328283309937, + "learning_rate": 1.62e-05, + "loss": 0.2175, + "step": 4837 + }, + { + "epoch": 13.513966480446927, + "grad_norm": 0.6424245238304138, + "learning_rate": 1.6203351955307262e-05, + "loss": 0.1869, + "step": 4838 + }, + { + "epoch": 13.516759776536313, + "grad_norm": 0.9169637560844421, + "learning_rate": 1.6206703910614523e-05, + "loss": 0.183, + "step": 4839 + }, + { + "epoch": 13.519553072625698, + "grad_norm": 1.0951793193817139, + "learning_rate": 1.6210055865921788e-05, + "loss": 0.2162, + "step": 4840 + }, + { + "epoch": 13.522346368715084, + "grad_norm": 1.0545357465744019, + "learning_rate": 1.621340782122905e-05, + "loss": 0.1699, + "step": 4841 + }, + { + "epoch": 13.525139664804469, + "grad_norm": 4.02010440826416, + "learning_rate": 1.621675977653631e-05, + "loss": 0.1971, + "step": 4842 + }, + { + "epoch": 13.527932960893855, + "grad_norm": 0.9859873056411743, + "learning_rate": 1.6220111731843576e-05, + "loss": 0.1683, + "step": 4843 + }, + { + "epoch": 13.53072625698324, + "grad_norm": 1.2520396709442139, + "learning_rate": 1.6223463687150837e-05, + "loss": 0.1893, + "step": 4844 + }, + { + "epoch": 13.533519553072626, + "grad_norm": 1.1282094717025757, + "learning_rate": 1.6226815642458098e-05, + "loss": 0.1816, + "step": 4845 + }, + { + "epoch": 13.53631284916201, + "grad_norm": 1.0683798789978027, + "learning_rate": 1.6230167597765363e-05, + "loss": 0.1762, + "step": 4846 + }, + { + "epoch": 13.539106145251397, + "grad_norm": 0.9781039357185364, + "learning_rate": 1.6233519553072624e-05, + "loss": 0.1943, + "step": 4847 + }, + { + "epoch": 13.541899441340782, + "grad_norm": 1.167076587677002, + "learning_rate": 1.623687150837989e-05, + "loss": 0.1757, + "step": 4848 + }, + { + "epoch": 13.544692737430168, + "grad_norm": 1.464657187461853, + "learning_rate": 1.624022346368715e-05, + "loss": 0.2422, + "step": 4849 + }, + { + "epoch": 13.547486033519553, + "grad_norm": 1.3794893026351929, + "learning_rate": 1.6243575418994412e-05, + "loss": 0.1925, + "step": 4850 + }, + { + "epoch": 13.550279329608939, + "grad_norm": 1.1808971166610718, + "learning_rate": 1.6246927374301677e-05, + "loss": 0.2084, + "step": 4851 + }, + { + "epoch": 13.553072625698324, + "grad_norm": 1.163001537322998, + "learning_rate": 1.6250279329608938e-05, + "loss": 0.1992, + "step": 4852 + }, + { + "epoch": 13.55586592178771, + "grad_norm": 1.3631139993667603, + "learning_rate": 1.62536312849162e-05, + "loss": 0.2115, + "step": 4853 + }, + { + "epoch": 13.558659217877095, + "grad_norm": 1.9578940868377686, + "learning_rate": 1.6256983240223464e-05, + "loss": 0.2671, + "step": 4854 + }, + { + "epoch": 13.561452513966481, + "grad_norm": 0.4682580828666687, + "learning_rate": 1.6260335195530725e-05, + "loss": 0.2056, + "step": 4855 + }, + { + "epoch": 13.564245810055866, + "grad_norm": 0.5711252093315125, + "learning_rate": 1.6263687150837987e-05, + "loss": 0.1985, + "step": 4856 + }, + { + "epoch": 13.567039106145252, + "grad_norm": 0.5827006697654724, + "learning_rate": 1.626703910614525e-05, + "loss": 0.1892, + "step": 4857 + }, + { + "epoch": 13.569832402234637, + "grad_norm": 0.5313262939453125, + "learning_rate": 1.6270391061452513e-05, + "loss": 0.2005, + "step": 4858 + }, + { + "epoch": 13.572625698324023, + "grad_norm": 1.1151374578475952, + "learning_rate": 1.6273743016759774e-05, + "loss": 0.2016, + "step": 4859 + }, + { + "epoch": 13.575418994413408, + "grad_norm": 0.5088601112365723, + "learning_rate": 1.627709497206704e-05, + "loss": 0.2051, + "step": 4860 + }, + { + "epoch": 13.578212290502794, + "grad_norm": 0.6617470979690552, + "learning_rate": 1.62804469273743e-05, + "loss": 0.1646, + "step": 4861 + }, + { + "epoch": 13.581005586592179, + "grad_norm": 1.0858641862869263, + "learning_rate": 1.628379888268156e-05, + "loss": 0.2063, + "step": 4862 + }, + { + "epoch": 13.583798882681565, + "grad_norm": 1.821784257888794, + "learning_rate": 1.6287150837988826e-05, + "loss": 0.174, + "step": 4863 + }, + { + "epoch": 13.58659217877095, + "grad_norm": 0.616023063659668, + "learning_rate": 1.629050279329609e-05, + "loss": 0.1841, + "step": 4864 + }, + { + "epoch": 13.589385474860336, + "grad_norm": 0.9288156032562256, + "learning_rate": 1.6293854748603356e-05, + "loss": 0.2015, + "step": 4865 + }, + { + "epoch": 13.59217877094972, + "grad_norm": 0.8742575645446777, + "learning_rate": 1.6297206703910617e-05, + "loss": 0.1972, + "step": 4866 + }, + { + "epoch": 13.594972067039105, + "grad_norm": 0.6511961817741394, + "learning_rate": 1.630055865921788e-05, + "loss": 0.1855, + "step": 4867 + }, + { + "epoch": 13.597765363128492, + "grad_norm": 0.6911728382110596, + "learning_rate": 1.6303910614525143e-05, + "loss": 0.1782, + "step": 4868 + }, + { + "epoch": 13.600558659217878, + "grad_norm": 1.5647207498550415, + "learning_rate": 1.6307262569832404e-05, + "loss": 0.1997, + "step": 4869 + }, + { + "epoch": 13.603351955307263, + "grad_norm": 0.9513446092605591, + "learning_rate": 1.6310614525139666e-05, + "loss": 0.1882, + "step": 4870 + }, + { + "epoch": 13.606145251396647, + "grad_norm": 0.7969967126846313, + "learning_rate": 1.631396648044693e-05, + "loss": 0.1859, + "step": 4871 + }, + { + "epoch": 13.608938547486034, + "grad_norm": 1.9807301759719849, + "learning_rate": 1.6317318435754192e-05, + "loss": 0.1689, + "step": 4872 + }, + { + "epoch": 13.611731843575418, + "grad_norm": 0.8081663846969604, + "learning_rate": 1.6320670391061453e-05, + "loss": 0.194, + "step": 4873 + }, + { + "epoch": 13.614525139664805, + "grad_norm": 1.022426724433899, + "learning_rate": 1.6324022346368718e-05, + "loss": 0.1557, + "step": 4874 + }, + { + "epoch": 13.61731843575419, + "grad_norm": 1.4856237173080444, + "learning_rate": 1.632737430167598e-05, + "loss": 0.2026, + "step": 4875 + }, + { + "epoch": 13.620111731843576, + "grad_norm": 1.0040868520736694, + "learning_rate": 1.633072625698324e-05, + "loss": 0.1999, + "step": 4876 + }, + { + "epoch": 13.62290502793296, + "grad_norm": 1.0290107727050781, + "learning_rate": 1.6334078212290505e-05, + "loss": 0.2125, + "step": 4877 + }, + { + "epoch": 13.625698324022347, + "grad_norm": 2.036759614944458, + "learning_rate": 1.6337430167597767e-05, + "loss": 0.2678, + "step": 4878 + }, + { + "epoch": 13.628491620111731, + "grad_norm": 4.732199192047119, + "learning_rate": 1.6340782122905028e-05, + "loss": 0.2988, + "step": 4879 + }, + { + "epoch": 13.631284916201118, + "grad_norm": 0.6978581547737122, + "learning_rate": 1.6344134078212293e-05, + "loss": 0.2142, + "step": 4880 + }, + { + "epoch": 13.634078212290502, + "grad_norm": 0.9870418310165405, + "learning_rate": 1.6347486033519554e-05, + "loss": 0.2559, + "step": 4881 + }, + { + "epoch": 13.636871508379889, + "grad_norm": 0.6430525183677673, + "learning_rate": 1.6350837988826815e-05, + "loss": 0.1838, + "step": 4882 + }, + { + "epoch": 13.639664804469273, + "grad_norm": 0.6024045944213867, + "learning_rate": 1.635418994413408e-05, + "loss": 0.1885, + "step": 4883 + }, + { + "epoch": 13.64245810055866, + "grad_norm": 1.2551088333129883, + "learning_rate": 1.635754189944134e-05, + "loss": 0.1913, + "step": 4884 + }, + { + "epoch": 13.645251396648044, + "grad_norm": 0.6535409092903137, + "learning_rate": 1.6360893854748606e-05, + "loss": 0.1762, + "step": 4885 + }, + { + "epoch": 13.64804469273743, + "grad_norm": 0.6879927515983582, + "learning_rate": 1.6364245810055868e-05, + "loss": 0.1991, + "step": 4886 + }, + { + "epoch": 13.650837988826815, + "grad_norm": 1.6056405305862427, + "learning_rate": 1.636759776536313e-05, + "loss": 0.1846, + "step": 4887 + }, + { + "epoch": 13.653631284916202, + "grad_norm": 0.8268870115280151, + "learning_rate": 1.6370949720670394e-05, + "loss": 0.1978, + "step": 4888 + }, + { + "epoch": 13.656424581005586, + "grad_norm": 0.7376096248626709, + "learning_rate": 1.6374301675977655e-05, + "loss": 0.2203, + "step": 4889 + }, + { + "epoch": 13.659217877094973, + "grad_norm": 0.8710257411003113, + "learning_rate": 1.6377653631284916e-05, + "loss": 0.212, + "step": 4890 + }, + { + "epoch": 13.662011173184357, + "grad_norm": 0.8297674059867859, + "learning_rate": 1.638100558659218e-05, + "loss": 0.1954, + "step": 4891 + }, + { + "epoch": 13.664804469273744, + "grad_norm": 0.8233268857002258, + "learning_rate": 1.6384357541899443e-05, + "loss": 0.2002, + "step": 4892 + }, + { + "epoch": 13.667597765363128, + "grad_norm": 1.1252233982086182, + "learning_rate": 1.6387709497206704e-05, + "loss": 0.1992, + "step": 4893 + }, + { + "epoch": 13.670391061452515, + "grad_norm": 0.8685394525527954, + "learning_rate": 1.639106145251397e-05, + "loss": 0.1895, + "step": 4894 + }, + { + "epoch": 13.6731843575419, + "grad_norm": 0.9333240985870361, + "learning_rate": 1.639441340782123e-05, + "loss": 0.165, + "step": 4895 + }, + { + "epoch": 13.675977653631286, + "grad_norm": 1.3548628091812134, + "learning_rate": 1.639776536312849e-05, + "loss": 0.1885, + "step": 4896 + }, + { + "epoch": 13.67877094972067, + "grad_norm": 0.82316654920578, + "learning_rate": 1.6401117318435756e-05, + "loss": 0.1786, + "step": 4897 + }, + { + "epoch": 13.681564245810057, + "grad_norm": 2.630763292312622, + "learning_rate": 1.6404469273743017e-05, + "loss": 0.2105, + "step": 4898 + }, + { + "epoch": 13.684357541899441, + "grad_norm": 0.9692097306251526, + "learning_rate": 1.640782122905028e-05, + "loss": 0.2051, + "step": 4899 + }, + { + "epoch": 13.687150837988828, + "grad_norm": 0.9836537837982178, + "learning_rate": 1.6411173184357543e-05, + "loss": 0.1801, + "step": 4900 + }, + { + "epoch": 13.689944134078212, + "grad_norm": 1.5052658319473267, + "learning_rate": 1.6414525139664805e-05, + "loss": 0.2185, + "step": 4901 + }, + { + "epoch": 13.692737430167599, + "grad_norm": 0.994295060634613, + "learning_rate": 1.6417877094972066e-05, + "loss": 0.2015, + "step": 4902 + }, + { + "epoch": 13.695530726256983, + "grad_norm": 1.3510743379592896, + "learning_rate": 1.642122905027933e-05, + "loss": 0.2138, + "step": 4903 + }, + { + "epoch": 13.69832402234637, + "grad_norm": 3.8053367137908936, + "learning_rate": 1.6424581005586592e-05, + "loss": 0.2527, + "step": 4904 + }, + { + "epoch": 13.701117318435754, + "grad_norm": 0.5208854675292969, + "learning_rate": 1.6427932960893857e-05, + "loss": 0.1889, + "step": 4905 + }, + { + "epoch": 13.703910614525139, + "grad_norm": 0.6959810853004456, + "learning_rate": 1.6431284916201118e-05, + "loss": 0.2075, + "step": 4906 + }, + { + "epoch": 13.706703910614525, + "grad_norm": 0.5112706422805786, + "learning_rate": 1.643463687150838e-05, + "loss": 0.1901, + "step": 4907 + }, + { + "epoch": 13.70949720670391, + "grad_norm": 0.6710006594657898, + "learning_rate": 1.6437988826815644e-05, + "loss": 0.1939, + "step": 4908 + }, + { + "epoch": 13.712290502793296, + "grad_norm": 1.6936928033828735, + "learning_rate": 1.6441340782122906e-05, + "loss": 0.2284, + "step": 4909 + }, + { + "epoch": 13.71508379888268, + "grad_norm": 1.1848633289337158, + "learning_rate": 1.6444692737430167e-05, + "loss": 0.2218, + "step": 4910 + }, + { + "epoch": 13.717877094972067, + "grad_norm": 0.5186358690261841, + "learning_rate": 1.6448044692737432e-05, + "loss": 0.2111, + "step": 4911 + }, + { + "epoch": 13.720670391061452, + "grad_norm": 0.5416008234024048, + "learning_rate": 1.6451396648044693e-05, + "loss": 0.1782, + "step": 4912 + }, + { + "epoch": 13.723463687150838, + "grad_norm": 0.6545098423957825, + "learning_rate": 1.6454748603351954e-05, + "loss": 0.1915, + "step": 4913 + }, + { + "epoch": 13.726256983240223, + "grad_norm": 0.741942822933197, + "learning_rate": 1.645810055865922e-05, + "loss": 0.225, + "step": 4914 + }, + { + "epoch": 13.72905027932961, + "grad_norm": 0.7312400341033936, + "learning_rate": 1.646145251396648e-05, + "loss": 0.1798, + "step": 4915 + }, + { + "epoch": 13.731843575418994, + "grad_norm": 0.8509039878845215, + "learning_rate": 1.6464804469273742e-05, + "loss": 0.2012, + "step": 4916 + }, + { + "epoch": 13.73463687150838, + "grad_norm": 3.8216049671173096, + "learning_rate": 1.6468156424581007e-05, + "loss": 0.2362, + "step": 4917 + }, + { + "epoch": 13.737430167597765, + "grad_norm": 1.0112452507019043, + "learning_rate": 1.6471508379888268e-05, + "loss": 0.1891, + "step": 4918 + }, + { + "epoch": 13.740223463687151, + "grad_norm": 0.8338775634765625, + "learning_rate": 1.647486033519553e-05, + "loss": 0.1876, + "step": 4919 + }, + { + "epoch": 13.743016759776536, + "grad_norm": 0.8817514181137085, + "learning_rate": 1.6478212290502794e-05, + "loss": 0.1779, + "step": 4920 + }, + { + "epoch": 13.745810055865922, + "grad_norm": 0.7120217084884644, + "learning_rate": 1.6481564245810055e-05, + "loss": 0.2086, + "step": 4921 + }, + { + "epoch": 13.748603351955307, + "grad_norm": 1.0177205801010132, + "learning_rate": 1.648491620111732e-05, + "loss": 0.2163, + "step": 4922 + }, + { + "epoch": 13.751396648044693, + "grad_norm": 1.0130175352096558, + "learning_rate": 1.648826815642458e-05, + "loss": 0.2147, + "step": 4923 + }, + { + "epoch": 13.754189944134078, + "grad_norm": 0.8456910252571106, + "learning_rate": 1.6491620111731843e-05, + "loss": 0.2181, + "step": 4924 + }, + { + "epoch": 13.756983240223464, + "grad_norm": 0.880809485912323, + "learning_rate": 1.6494972067039108e-05, + "loss": 0.1533, + "step": 4925 + }, + { + "epoch": 13.759776536312849, + "grad_norm": 1.1533148288726807, + "learning_rate": 1.649832402234637e-05, + "loss": 0.2184, + "step": 4926 + }, + { + "epoch": 13.762569832402235, + "grad_norm": 1.6389262676239014, + "learning_rate": 1.650167597765363e-05, + "loss": 0.2401, + "step": 4927 + }, + { + "epoch": 13.76536312849162, + "grad_norm": 1.2300353050231934, + "learning_rate": 1.6505027932960895e-05, + "loss": 0.2129, + "step": 4928 + }, + { + "epoch": 13.768156424581006, + "grad_norm": 2.1579983234405518, + "learning_rate": 1.6508379888268156e-05, + "loss": 0.2963, + "step": 4929 + }, + { + "epoch": 13.77094972067039, + "grad_norm": 0.6384286284446716, + "learning_rate": 1.6511731843575418e-05, + "loss": 0.2434, + "step": 4930 + }, + { + "epoch": 13.773743016759777, + "grad_norm": 0.580607533454895, + "learning_rate": 1.6515083798882682e-05, + "loss": 0.1681, + "step": 4931 + }, + { + "epoch": 13.776536312849162, + "grad_norm": 0.6481080055236816, + "learning_rate": 1.6518435754189944e-05, + "loss": 0.1997, + "step": 4932 + }, + { + "epoch": 13.779329608938548, + "grad_norm": 1.208233118057251, + "learning_rate": 1.6521787709497205e-05, + "loss": 0.2171, + "step": 4933 + }, + { + "epoch": 13.782122905027933, + "grad_norm": 0.9955776333808899, + "learning_rate": 1.652513966480447e-05, + "loss": 0.2453, + "step": 4934 + }, + { + "epoch": 13.78491620111732, + "grad_norm": 0.6483702063560486, + "learning_rate": 1.652849162011173e-05, + "loss": 0.2134, + "step": 4935 + }, + { + "epoch": 13.787709497206704, + "grad_norm": 0.7597767114639282, + "learning_rate": 1.6531843575418993e-05, + "loss": 0.2407, + "step": 4936 + }, + { + "epoch": 13.79050279329609, + "grad_norm": 0.7257301211357117, + "learning_rate": 1.6535195530726257e-05, + "loss": 0.2035, + "step": 4937 + }, + { + "epoch": 13.793296089385475, + "grad_norm": 0.7295578122138977, + "learning_rate": 1.653854748603352e-05, + "loss": 0.1954, + "step": 4938 + }, + { + "epoch": 13.796089385474861, + "grad_norm": 1.3227779865264893, + "learning_rate": 1.654189944134078e-05, + "loss": 0.2072, + "step": 4939 + }, + { + "epoch": 13.798882681564246, + "grad_norm": 0.8837338089942932, + "learning_rate": 1.6545251396648045e-05, + "loss": 0.1934, + "step": 4940 + }, + { + "epoch": 13.80167597765363, + "grad_norm": 0.6040142774581909, + "learning_rate": 1.6548603351955306e-05, + "loss": 0.2021, + "step": 4941 + }, + { + "epoch": 13.804469273743017, + "grad_norm": 0.8035222291946411, + "learning_rate": 1.655195530726257e-05, + "loss": 0.2252, + "step": 4942 + }, + { + "epoch": 13.807262569832401, + "grad_norm": 1.0336552858352661, + "learning_rate": 1.6555307262569832e-05, + "loss": 0.2302, + "step": 4943 + }, + { + "epoch": 13.810055865921788, + "grad_norm": 0.7583746910095215, + "learning_rate": 1.6558659217877093e-05, + "loss": 0.1879, + "step": 4944 + }, + { + "epoch": 13.812849162011172, + "grad_norm": 1.197224736213684, + "learning_rate": 1.6562011173184358e-05, + "loss": 0.2063, + "step": 4945 + }, + { + "epoch": 13.815642458100559, + "grad_norm": 1.1123013496398926, + "learning_rate": 1.656536312849162e-05, + "loss": 0.2147, + "step": 4946 + }, + { + "epoch": 13.818435754189943, + "grad_norm": 0.7925002574920654, + "learning_rate": 1.656871508379888e-05, + "loss": 0.1849, + "step": 4947 + }, + { + "epoch": 13.82122905027933, + "grad_norm": 1.0876836776733398, + "learning_rate": 1.6572067039106146e-05, + "loss": 0.2306, + "step": 4948 + }, + { + "epoch": 13.824022346368714, + "grad_norm": 0.8712457418441772, + "learning_rate": 1.6575418994413407e-05, + "loss": 0.1887, + "step": 4949 + }, + { + "epoch": 13.8268156424581, + "grad_norm": 0.9821040630340576, + "learning_rate": 1.6578770949720668e-05, + "loss": 0.1864, + "step": 4950 + }, + { + "epoch": 13.829608938547485, + "grad_norm": 1.0772448778152466, + "learning_rate": 1.6582122905027933e-05, + "loss": 0.1923, + "step": 4951 + }, + { + "epoch": 13.832402234636872, + "grad_norm": 1.8208863735198975, + "learning_rate": 1.6585474860335194e-05, + "loss": 0.2256, + "step": 4952 + }, + { + "epoch": 13.835195530726256, + "grad_norm": 1.6711926460266113, + "learning_rate": 1.6588826815642456e-05, + "loss": 0.2451, + "step": 4953 + }, + { + "epoch": 13.837988826815643, + "grad_norm": 2.1026315689086914, + "learning_rate": 1.659217877094972e-05, + "loss": 0.3117, + "step": 4954 + }, + { + "epoch": 13.840782122905027, + "grad_norm": 0.5733019709587097, + "learning_rate": 1.6595530726256982e-05, + "loss": 0.2003, + "step": 4955 + }, + { + "epoch": 13.843575418994414, + "grad_norm": 0.7182850241661072, + "learning_rate": 1.6598882681564243e-05, + "loss": 0.202, + "step": 4956 + }, + { + "epoch": 13.846368715083798, + "grad_norm": 0.6279488801956177, + "learning_rate": 1.6602234636871508e-05, + "loss": 0.1965, + "step": 4957 + }, + { + "epoch": 13.849162011173185, + "grad_norm": 0.7642987966537476, + "learning_rate": 1.660558659217877e-05, + "loss": 0.1833, + "step": 4958 + }, + { + "epoch": 13.85195530726257, + "grad_norm": 1.2714903354644775, + "learning_rate": 1.6608938547486034e-05, + "loss": 0.2025, + "step": 4959 + }, + { + "epoch": 13.854748603351956, + "grad_norm": 0.9914622902870178, + "learning_rate": 1.6612290502793295e-05, + "loss": 0.2345, + "step": 4960 + }, + { + "epoch": 13.85754189944134, + "grad_norm": 0.7452812790870667, + "learning_rate": 1.661564245810056e-05, + "loss": 0.1884, + "step": 4961 + }, + { + "epoch": 13.860335195530727, + "grad_norm": 1.7915602922439575, + "learning_rate": 1.6618994413407825e-05, + "loss": 0.1547, + "step": 4962 + }, + { + "epoch": 13.863128491620111, + "grad_norm": 0.9747093915939331, + "learning_rate": 1.6622346368715086e-05, + "loss": 0.1838, + "step": 4963 + }, + { + "epoch": 13.865921787709498, + "grad_norm": 0.6860989332199097, + "learning_rate": 1.6625698324022347e-05, + "loss": 0.2344, + "step": 4964 + }, + { + "epoch": 13.868715083798882, + "grad_norm": 0.9014551043510437, + "learning_rate": 1.6629050279329612e-05, + "loss": 0.1874, + "step": 4965 + }, + { + "epoch": 13.871508379888269, + "grad_norm": 0.7122955918312073, + "learning_rate": 1.6632402234636874e-05, + "loss": 0.1913, + "step": 4966 + }, + { + "epoch": 13.874301675977653, + "grad_norm": 0.6558124423027039, + "learning_rate": 1.6635754189944135e-05, + "loss": 0.1871, + "step": 4967 + }, + { + "epoch": 13.87709497206704, + "grad_norm": 0.6551256775856018, + "learning_rate": 1.66391061452514e-05, + "loss": 0.1847, + "step": 4968 + }, + { + "epoch": 13.879888268156424, + "grad_norm": 1.1940463781356812, + "learning_rate": 1.664245810055866e-05, + "loss": 0.2137, + "step": 4969 + }, + { + "epoch": 13.88268156424581, + "grad_norm": 1.1009800434112549, + "learning_rate": 1.6645810055865922e-05, + "loss": 0.2189, + "step": 4970 + }, + { + "epoch": 13.885474860335195, + "grad_norm": 0.8737493753433228, + "learning_rate": 1.6649162011173187e-05, + "loss": 0.195, + "step": 4971 + }, + { + "epoch": 13.888268156424582, + "grad_norm": 0.7803341746330261, + "learning_rate": 1.665251396648045e-05, + "loss": 0.1895, + "step": 4972 + }, + { + "epoch": 13.891061452513966, + "grad_norm": 1.1278537511825562, + "learning_rate": 1.665586592178771e-05, + "loss": 0.1773, + "step": 4973 + }, + { + "epoch": 13.893854748603353, + "grad_norm": 2.391446352005005, + "learning_rate": 1.6659217877094974e-05, + "loss": 0.2132, + "step": 4974 + }, + { + "epoch": 13.896648044692737, + "grad_norm": 1.000569462776184, + "learning_rate": 1.6662569832402236e-05, + "loss": 0.2048, + "step": 4975 + }, + { + "epoch": 13.899441340782122, + "grad_norm": 1.2575429677963257, + "learning_rate": 1.6665921787709497e-05, + "loss": 0.2342, + "step": 4976 + }, + { + "epoch": 13.902234636871508, + "grad_norm": 2.1158783435821533, + "learning_rate": 1.6669273743016762e-05, + "loss": 0.2332, + "step": 4977 + }, + { + "epoch": 13.905027932960895, + "grad_norm": 1.8064324855804443, + "learning_rate": 1.6672625698324023e-05, + "loss": 0.2188, + "step": 4978 + }, + { + "epoch": 13.90782122905028, + "grad_norm": 2.236698865890503, + "learning_rate": 1.6675977653631288e-05, + "loss": 0.2731, + "step": 4979 + }, + { + "epoch": 13.910614525139664, + "grad_norm": 0.5054540634155273, + "learning_rate": 1.667932960893855e-05, + "loss": 0.2009, + "step": 4980 + }, + { + "epoch": 13.91340782122905, + "grad_norm": 0.6906905770301819, + "learning_rate": 1.668268156424581e-05, + "loss": 0.2328, + "step": 4981 + }, + { + "epoch": 13.916201117318435, + "grad_norm": 0.8248416781425476, + "learning_rate": 1.6686033519553075e-05, + "loss": 0.1949, + "step": 4982 + }, + { + "epoch": 13.918994413407821, + "grad_norm": 0.7354496121406555, + "learning_rate": 1.6689385474860337e-05, + "loss": 0.1883, + "step": 4983 + }, + { + "epoch": 13.921787709497206, + "grad_norm": 0.7645291686058044, + "learning_rate": 1.6692737430167598e-05, + "loss": 0.2092, + "step": 4984 + }, + { + "epoch": 13.924581005586592, + "grad_norm": 0.8879695534706116, + "learning_rate": 1.6696089385474863e-05, + "loss": 0.2205, + "step": 4985 + }, + { + "epoch": 13.927374301675977, + "grad_norm": 0.6354501843452454, + "learning_rate": 1.6699441340782124e-05, + "loss": 0.1774, + "step": 4986 + }, + { + "epoch": 13.930167597765363, + "grad_norm": 0.6581356525421143, + "learning_rate": 1.6702793296089385e-05, + "loss": 0.1834, + "step": 4987 + }, + { + "epoch": 13.932960893854748, + "grad_norm": 0.7097819447517395, + "learning_rate": 1.670614525139665e-05, + "loss": 0.1772, + "step": 4988 + }, + { + "epoch": 13.935754189944134, + "grad_norm": 0.5648629665374756, + "learning_rate": 1.670949720670391e-05, + "loss": 0.2005, + "step": 4989 + }, + { + "epoch": 13.938547486033519, + "grad_norm": 0.5968577265739441, + "learning_rate": 1.6712849162011173e-05, + "loss": 0.1926, + "step": 4990 + }, + { + "epoch": 13.941340782122905, + "grad_norm": 0.7153709530830383, + "learning_rate": 1.6716201117318438e-05, + "loss": 0.1951, + "step": 4991 + }, + { + "epoch": 13.94413407821229, + "grad_norm": 1.1575407981872559, + "learning_rate": 1.67195530726257e-05, + "loss": 0.2011, + "step": 4992 + }, + { + "epoch": 13.946927374301676, + "grad_norm": 0.8208765983581543, + "learning_rate": 1.672290502793296e-05, + "loss": 0.1784, + "step": 4993 + }, + { + "epoch": 13.949720670391061, + "grad_norm": 0.8400183916091919, + "learning_rate": 1.6726256983240225e-05, + "loss": 0.1797, + "step": 4994 + }, + { + "epoch": 13.952513966480447, + "grad_norm": 0.9737865924835205, + "learning_rate": 1.6729608938547486e-05, + "loss": 0.1636, + "step": 4995 + }, + { + "epoch": 13.955307262569832, + "grad_norm": 0.7989575266838074, + "learning_rate": 1.6732960893854748e-05, + "loss": 0.2285, + "step": 4996 + }, + { + "epoch": 13.958100558659218, + "grad_norm": 2.215975046157837, + "learning_rate": 1.6736312849162013e-05, + "loss": 0.1927, + "step": 4997 + }, + { + "epoch": 13.960893854748603, + "grad_norm": 1.1134378910064697, + "learning_rate": 1.6739664804469274e-05, + "loss": 0.1882, + "step": 4998 + }, + { + "epoch": 13.96368715083799, + "grad_norm": 1.008774995803833, + "learning_rate": 1.674301675977654e-05, + "loss": 0.2051, + "step": 4999 + }, + { + "epoch": 13.966480446927374, + "grad_norm": 0.9099233150482178, + "learning_rate": 1.67463687150838e-05, + "loss": 0.2338, + "step": 5000 + }, + { + "epoch": 13.966480446927374, + "eval_cer": 0.19290694248960755, + "eval_loss": 0.2707229554653168, + "eval_runtime": 23.744, + "eval_samples_per_second": 191.122, + "eval_steps_per_second": 0.632, + "eval_wer": 0.343007178570431, + "step": 5000 + }, + { + "epoch": 13.96927374301676, + "grad_norm": 1.2281434535980225, + "learning_rate": 1.674972067039106e-05, + "loss": 0.2115, + "step": 5001 + }, + { + "epoch": 13.972067039106145, + "grad_norm": 1.0482540130615234, + "learning_rate": 1.6753072625698326e-05, + "loss": 0.1916, + "step": 5002 + }, + { + "epoch": 13.974860335195531, + "grad_norm": 1.0724822282791138, + "learning_rate": 1.6756424581005587e-05, + "loss": 0.1907, + "step": 5003 + }, + { + "epoch": 13.977653631284916, + "grad_norm": 1.938852310180664, + "learning_rate": 1.675977653631285e-05, + "loss": 0.3177, + "step": 5004 + }, + { + "epoch": 13.980446927374302, + "grad_norm": 0.5940600633621216, + "learning_rate": 1.6763128491620113e-05, + "loss": 0.1988, + "step": 5005 + }, + { + "epoch": 13.983240223463687, + "grad_norm": 0.5772649049758911, + "learning_rate": 1.6766480446927375e-05, + "loss": 0.1643, + "step": 5006 + }, + { + "epoch": 13.986033519553073, + "grad_norm": 0.8421567678451538, + "learning_rate": 1.6769832402234636e-05, + "loss": 0.2061, + "step": 5007 + }, + { + "epoch": 13.988826815642458, + "grad_norm": 1.641265869140625, + "learning_rate": 1.67731843575419e-05, + "loss": 0.1998, + "step": 5008 + }, + { + "epoch": 13.991620111731844, + "grad_norm": 0.6688755750656128, + "learning_rate": 1.6776536312849162e-05, + "loss": 0.2009, + "step": 5009 + }, + { + "epoch": 13.994413407821229, + "grad_norm": 0.686514139175415, + "learning_rate": 1.6779888268156424e-05, + "loss": 0.2043, + "step": 5010 + }, + { + "epoch": 13.997206703910614, + "grad_norm": 5.479888439178467, + "learning_rate": 1.6783240223463688e-05, + "loss": 0.1841, + "step": 5011 + }, + { + "epoch": 14.0, + "grad_norm": 2.936248302459717, + "learning_rate": 1.678659217877095e-05, + "loss": 0.3088, + "step": 5012 + }, + { + "epoch": 14.002793296089385, + "grad_norm": 0.5097810626029968, + "learning_rate": 1.678994413407821e-05, + "loss": 0.1653, + "step": 5013 + }, + { + "epoch": 14.005586592178771, + "grad_norm": 0.5686606764793396, + "learning_rate": 1.6793296089385476e-05, + "loss": 0.2245, + "step": 5014 + }, + { + "epoch": 14.008379888268156, + "grad_norm": 0.5669854879379272, + "learning_rate": 1.6796648044692737e-05, + "loss": 0.1846, + "step": 5015 + }, + { + "epoch": 14.011173184357542, + "grad_norm": 0.5723404288291931, + "learning_rate": 1.6800000000000002e-05, + "loss": 0.2217, + "step": 5016 + }, + { + "epoch": 14.013966480446927, + "grad_norm": 0.5100845098495483, + "learning_rate": 1.6803351955307263e-05, + "loss": 0.1787, + "step": 5017 + }, + { + "epoch": 14.016759776536313, + "grad_norm": 3.2940311431884766, + "learning_rate": 1.6806703910614524e-05, + "loss": 0.176, + "step": 5018 + }, + { + "epoch": 14.019553072625698, + "grad_norm": 0.5647488832473755, + "learning_rate": 1.681005586592179e-05, + "loss": 0.184, + "step": 5019 + }, + { + "epoch": 14.022346368715084, + "grad_norm": 0.7633018493652344, + "learning_rate": 1.681340782122905e-05, + "loss": 0.1868, + "step": 5020 + }, + { + "epoch": 14.025139664804469, + "grad_norm": 0.659299373626709, + "learning_rate": 1.6816759776536312e-05, + "loss": 0.1835, + "step": 5021 + }, + { + "epoch": 14.027932960893855, + "grad_norm": 0.6121032238006592, + "learning_rate": 1.6820111731843577e-05, + "loss": 0.1868, + "step": 5022 + }, + { + "epoch": 14.03072625698324, + "grad_norm": 0.7961002588272095, + "learning_rate": 1.6823463687150838e-05, + "loss": 0.1859, + "step": 5023 + }, + { + "epoch": 14.033519553072626, + "grad_norm": 0.5954718589782715, + "learning_rate": 1.68268156424581e-05, + "loss": 0.146, + "step": 5024 + }, + { + "epoch": 14.03631284916201, + "grad_norm": 0.856962263584137, + "learning_rate": 1.6830167597765364e-05, + "loss": 0.1712, + "step": 5025 + }, + { + "epoch": 14.039106145251397, + "grad_norm": 0.9839105606079102, + "learning_rate": 1.6833519553072625e-05, + "loss": 0.1911, + "step": 5026 + }, + { + "epoch": 14.041899441340782, + "grad_norm": 0.9186225533485413, + "learning_rate": 1.6836871508379887e-05, + "loss": 0.2068, + "step": 5027 + }, + { + "epoch": 14.044692737430168, + "grad_norm": 0.76984703540802, + "learning_rate": 1.684022346368715e-05, + "loss": 0.158, + "step": 5028 + }, + { + "epoch": 14.047486033519553, + "grad_norm": 1.5826243162155151, + "learning_rate": 1.6843575418994413e-05, + "loss": 0.179, + "step": 5029 + }, + { + "epoch": 14.050279329608939, + "grad_norm": 0.9263628721237183, + "learning_rate": 1.6846927374301674e-05, + "loss": 0.177, + "step": 5030 + }, + { + "epoch": 14.053072625698324, + "grad_norm": 0.8350117206573486, + "learning_rate": 1.685027932960894e-05, + "loss": 0.1802, + "step": 5031 + }, + { + "epoch": 14.05586592178771, + "grad_norm": 1.2179635763168335, + "learning_rate": 1.68536312849162e-05, + "loss": 0.2009, + "step": 5032 + }, + { + "epoch": 14.058659217877095, + "grad_norm": 0.7923583388328552, + "learning_rate": 1.685698324022346e-05, + "loss": 0.1949, + "step": 5033 + }, + { + "epoch": 14.061452513966481, + "grad_norm": 1.061172366142273, + "learning_rate": 1.6860335195530726e-05, + "loss": 0.1892, + "step": 5034 + }, + { + "epoch": 14.064245810055866, + "grad_norm": 1.9076520204544067, + "learning_rate": 1.6863687150837988e-05, + "loss": 0.2181, + "step": 5035 + }, + { + "epoch": 14.067039106145252, + "grad_norm": 1.1644617319107056, + "learning_rate": 1.6867039106145252e-05, + "loss": 0.2024, + "step": 5036 + }, + { + "epoch": 14.069832402234637, + "grad_norm": 1.5275007486343384, + "learning_rate": 1.6870391061452514e-05, + "loss": 0.2472, + "step": 5037 + }, + { + "epoch": 14.072625698324023, + "grad_norm": 0.4860062301158905, + "learning_rate": 1.6873743016759775e-05, + "loss": 0.2044, + "step": 5038 + }, + { + "epoch": 14.075418994413408, + "grad_norm": 0.682094156742096, + "learning_rate": 1.687709497206704e-05, + "loss": 0.1627, + "step": 5039 + }, + { + "epoch": 14.078212290502794, + "grad_norm": 0.8239700198173523, + "learning_rate": 1.68804469273743e-05, + "loss": 0.2385, + "step": 5040 + }, + { + "epoch": 14.081005586592179, + "grad_norm": 0.5241459608078003, + "learning_rate": 1.6883798882681563e-05, + "loss": 0.1964, + "step": 5041 + }, + { + "epoch": 14.083798882681565, + "grad_norm": 0.5357932448387146, + "learning_rate": 1.6887150837988827e-05, + "loss": 0.2119, + "step": 5042 + }, + { + "epoch": 14.08659217877095, + "grad_norm": 0.8274497389793396, + "learning_rate": 1.689050279329609e-05, + "loss": 0.1998, + "step": 5043 + }, + { + "epoch": 14.089385474860336, + "grad_norm": 0.7429519891738892, + "learning_rate": 1.689385474860335e-05, + "loss": 0.186, + "step": 5044 + }, + { + "epoch": 14.09217877094972, + "grad_norm": 0.7212750315666199, + "learning_rate": 1.6897206703910615e-05, + "loss": 0.1973, + "step": 5045 + }, + { + "epoch": 14.094972067039107, + "grad_norm": 0.6184440851211548, + "learning_rate": 1.6900558659217876e-05, + "loss": 0.2, + "step": 5046 + }, + { + "epoch": 14.097765363128492, + "grad_norm": 0.9133778214454651, + "learning_rate": 1.6903910614525137e-05, + "loss": 0.1724, + "step": 5047 + }, + { + "epoch": 14.100558659217878, + "grad_norm": 0.7479596734046936, + "learning_rate": 1.6907262569832402e-05, + "loss": 0.229, + "step": 5048 + }, + { + "epoch": 14.103351955307263, + "grad_norm": 0.8044571876525879, + "learning_rate": 1.6910614525139663e-05, + "loss": 0.1993, + "step": 5049 + }, + { + "epoch": 14.106145251396647, + "grad_norm": 0.9547159671783447, + "learning_rate": 1.6913966480446925e-05, + "loss": 0.2241, + "step": 5050 + }, + { + "epoch": 14.108938547486034, + "grad_norm": 0.8705384135246277, + "learning_rate": 1.691731843575419e-05, + "loss": 0.1848, + "step": 5051 + }, + { + "epoch": 14.111731843575418, + "grad_norm": 0.8476273417472839, + "learning_rate": 1.692067039106145e-05, + "loss": 0.2051, + "step": 5052 + }, + { + "epoch": 14.114525139664805, + "grad_norm": 1.4904192686080933, + "learning_rate": 1.6924022346368716e-05, + "loss": 0.168, + "step": 5053 + }, + { + "epoch": 14.11731843575419, + "grad_norm": 0.5912845730781555, + "learning_rate": 1.6927374301675977e-05, + "loss": 0.1855, + "step": 5054 + }, + { + "epoch": 14.120111731843576, + "grad_norm": 0.9858276844024658, + "learning_rate": 1.6930726256983238e-05, + "loss": 0.2092, + "step": 5055 + }, + { + "epoch": 14.12290502793296, + "grad_norm": 0.8868897557258606, + "learning_rate": 1.6934078212290503e-05, + "loss": 0.2108, + "step": 5056 + }, + { + "epoch": 14.125698324022347, + "grad_norm": 0.8934311270713806, + "learning_rate": 1.6937430167597768e-05, + "loss": 0.1963, + "step": 5057 + }, + { + "epoch": 14.128491620111731, + "grad_norm": 1.2979662418365479, + "learning_rate": 1.694078212290503e-05, + "loss": 0.2268, + "step": 5058 + }, + { + "epoch": 14.131284916201118, + "grad_norm": 1.6866134405136108, + "learning_rate": 1.6944134078212294e-05, + "loss": 0.2209, + "step": 5059 + }, + { + "epoch": 14.134078212290502, + "grad_norm": 1.8963457345962524, + "learning_rate": 1.6947486033519555e-05, + "loss": 0.2128, + "step": 5060 + }, + { + "epoch": 14.136871508379889, + "grad_norm": 1.5267311334609985, + "learning_rate": 1.6950837988826817e-05, + "loss": 0.2761, + "step": 5061 + }, + { + "epoch": 14.139664804469273, + "grad_norm": 3.736767530441284, + "learning_rate": 1.695418994413408e-05, + "loss": 0.2831, + "step": 5062 + }, + { + "epoch": 14.14245810055866, + "grad_norm": 0.6089104413986206, + "learning_rate": 1.6957541899441343e-05, + "loss": 0.2364, + "step": 5063 + }, + { + "epoch": 14.145251396648044, + "grad_norm": 0.6711127161979675, + "learning_rate": 1.6960893854748604e-05, + "loss": 0.194, + "step": 5064 + }, + { + "epoch": 14.14804469273743, + "grad_norm": 0.6882439851760864, + "learning_rate": 1.696424581005587e-05, + "loss": 0.1906, + "step": 5065 + }, + { + "epoch": 14.150837988826815, + "grad_norm": 0.65103679895401, + "learning_rate": 1.696759776536313e-05, + "loss": 0.1771, + "step": 5066 + }, + { + "epoch": 14.153631284916202, + "grad_norm": 0.8668402433395386, + "learning_rate": 1.697094972067039e-05, + "loss": 0.1944, + "step": 5067 + }, + { + "epoch": 14.156424581005586, + "grad_norm": 0.6067304015159607, + "learning_rate": 1.6974301675977656e-05, + "loss": 0.1847, + "step": 5068 + }, + { + "epoch": 14.159217877094973, + "grad_norm": 0.8462134003639221, + "learning_rate": 1.6977653631284917e-05, + "loss": 0.2164, + "step": 5069 + }, + { + "epoch": 14.162011173184357, + "grad_norm": 0.7238622307777405, + "learning_rate": 1.698100558659218e-05, + "loss": 0.1686, + "step": 5070 + }, + { + "epoch": 14.164804469273744, + "grad_norm": 0.6397600173950195, + "learning_rate": 1.6984357541899444e-05, + "loss": 0.1573, + "step": 5071 + }, + { + "epoch": 14.167597765363128, + "grad_norm": 0.5528873801231384, + "learning_rate": 1.6987709497206705e-05, + "loss": 0.1733, + "step": 5072 + }, + { + "epoch": 14.170391061452515, + "grad_norm": 0.663000226020813, + "learning_rate": 1.699106145251397e-05, + "loss": 0.2187, + "step": 5073 + }, + { + "epoch": 14.1731843575419, + "grad_norm": 1.3018397092819214, + "learning_rate": 1.699441340782123e-05, + "loss": 0.206, + "step": 5074 + }, + { + "epoch": 14.175977653631286, + "grad_norm": 0.8535932302474976, + "learning_rate": 1.6997765363128492e-05, + "loss": 0.1946, + "step": 5075 + }, + { + "epoch": 14.17877094972067, + "grad_norm": 1.747796893119812, + "learning_rate": 1.7001117318435757e-05, + "loss": 0.2128, + "step": 5076 + }, + { + "epoch": 14.181564245810057, + "grad_norm": 0.8595774173736572, + "learning_rate": 1.700446927374302e-05, + "loss": 0.1928, + "step": 5077 + }, + { + "epoch": 14.184357541899441, + "grad_norm": 0.7101022005081177, + "learning_rate": 1.700782122905028e-05, + "loss": 0.1651, + "step": 5078 + }, + { + "epoch": 14.187150837988828, + "grad_norm": 0.7740656733512878, + "learning_rate": 1.7011173184357544e-05, + "loss": 0.2104, + "step": 5079 + }, + { + "epoch": 14.189944134078212, + "grad_norm": 1.1501387357711792, + "learning_rate": 1.7014525139664806e-05, + "loss": 0.2225, + "step": 5080 + }, + { + "epoch": 14.192737430167599, + "grad_norm": 1.0636407136917114, + "learning_rate": 1.7017877094972067e-05, + "loss": 0.2495, + "step": 5081 + }, + { + "epoch": 14.195530726256983, + "grad_norm": 1.193702220916748, + "learning_rate": 1.7021229050279332e-05, + "loss": 0.2053, + "step": 5082 + }, + { + "epoch": 14.19832402234637, + "grad_norm": 1.2703653573989868, + "learning_rate": 1.7024581005586593e-05, + "loss": 0.1907, + "step": 5083 + }, + { + "epoch": 14.201117318435754, + "grad_norm": 1.0085948705673218, + "learning_rate": 1.7027932960893855e-05, + "loss": 0.1603, + "step": 5084 + }, + { + "epoch": 14.203910614525139, + "grad_norm": 1.1704822778701782, + "learning_rate": 1.703128491620112e-05, + "loss": 0.2075, + "step": 5085 + }, + { + "epoch": 14.206703910614525, + "grad_norm": 1.3956891298294067, + "learning_rate": 1.703463687150838e-05, + "loss": 0.206, + "step": 5086 + }, + { + "epoch": 14.20949720670391, + "grad_norm": 5.456483364105225, + "learning_rate": 1.7037988826815642e-05, + "loss": 0.2821, + "step": 5087 + }, + { + "epoch": 14.212290502793296, + "grad_norm": 0.7300254702568054, + "learning_rate": 1.7041340782122907e-05, + "loss": 0.1978, + "step": 5088 + }, + { + "epoch": 14.21508379888268, + "grad_norm": 0.47508758306503296, + "learning_rate": 1.7044692737430168e-05, + "loss": 0.1826, + "step": 5089 + }, + { + "epoch": 14.217877094972067, + "grad_norm": 0.7738616466522217, + "learning_rate": 1.704804469273743e-05, + "loss": 0.1843, + "step": 5090 + }, + { + "epoch": 14.220670391061452, + "grad_norm": 0.8398893475532532, + "learning_rate": 1.7051396648044694e-05, + "loss": 0.1965, + "step": 5091 + }, + { + "epoch": 14.223463687150838, + "grad_norm": 0.5727632641792297, + "learning_rate": 1.7054748603351956e-05, + "loss": 0.1848, + "step": 5092 + }, + { + "epoch": 14.226256983240223, + "grad_norm": 0.7477567195892334, + "learning_rate": 1.705810055865922e-05, + "loss": 0.1748, + "step": 5093 + }, + { + "epoch": 14.22905027932961, + "grad_norm": 0.5726558566093445, + "learning_rate": 1.706145251396648e-05, + "loss": 0.2086, + "step": 5094 + }, + { + "epoch": 14.231843575418994, + "grad_norm": 0.6315433382987976, + "learning_rate": 1.7064804469273743e-05, + "loss": 0.2073, + "step": 5095 + }, + { + "epoch": 14.23463687150838, + "grad_norm": 2.0871787071228027, + "learning_rate": 1.7068156424581008e-05, + "loss": 0.1814, + "step": 5096 + }, + { + "epoch": 14.237430167597765, + "grad_norm": 0.7200970649719238, + "learning_rate": 1.707150837988827e-05, + "loss": 0.1871, + "step": 5097 + }, + { + "epoch": 14.240223463687151, + "grad_norm": 0.9858104586601257, + "learning_rate": 1.707486033519553e-05, + "loss": 0.1945, + "step": 5098 + }, + { + "epoch": 14.243016759776536, + "grad_norm": 1.06559419631958, + "learning_rate": 1.7078212290502795e-05, + "loss": 0.1801, + "step": 5099 + }, + { + "epoch": 14.245810055865922, + "grad_norm": 0.7114586234092712, + "learning_rate": 1.7081564245810056e-05, + "loss": 0.2027, + "step": 5100 + }, + { + "epoch": 14.248603351955307, + "grad_norm": 0.7421502470970154, + "learning_rate": 1.7084916201117318e-05, + "loss": 0.1675, + "step": 5101 + }, + { + "epoch": 14.251396648044693, + "grad_norm": 0.6904516816139221, + "learning_rate": 1.7088268156424583e-05, + "loss": 0.1673, + "step": 5102 + }, + { + "epoch": 14.254189944134078, + "grad_norm": 0.8257575035095215, + "learning_rate": 1.7091620111731844e-05, + "loss": 0.1649, + "step": 5103 + }, + { + "epoch": 14.256983240223464, + "grad_norm": 1.2610681056976318, + "learning_rate": 1.7094972067039105e-05, + "loss": 0.2283, + "step": 5104 + }, + { + "epoch": 14.259776536312849, + "grad_norm": 0.7267904281616211, + "learning_rate": 1.709832402234637e-05, + "loss": 0.1949, + "step": 5105 + }, + { + "epoch": 14.262569832402235, + "grad_norm": 0.9099175333976746, + "learning_rate": 1.710167597765363e-05, + "loss": 0.1977, + "step": 5106 + }, + { + "epoch": 14.26536312849162, + "grad_norm": 1.0154900550842285, + "learning_rate": 1.7105027932960893e-05, + "loss": 0.1823, + "step": 5107 + }, + { + "epoch": 14.268156424581006, + "grad_norm": 1.0052292346954346, + "learning_rate": 1.7108379888268157e-05, + "loss": 0.1879, + "step": 5108 + }, + { + "epoch": 14.27094972067039, + "grad_norm": 1.1330411434173584, + "learning_rate": 1.711173184357542e-05, + "loss": 0.2108, + "step": 5109 + }, + { + "epoch": 14.273743016759777, + "grad_norm": 1.3329631090164185, + "learning_rate": 1.7115083798882683e-05, + "loss": 0.2399, + "step": 5110 + }, + { + "epoch": 14.276536312849162, + "grad_norm": 1.4698855876922607, + "learning_rate": 1.7118435754189945e-05, + "loss": 0.2218, + "step": 5111 + }, + { + "epoch": 14.279329608938548, + "grad_norm": 2.375162124633789, + "learning_rate": 1.7121787709497206e-05, + "loss": 0.258, + "step": 5112 + }, + { + "epoch": 14.282122905027933, + "grad_norm": 0.5957632064819336, + "learning_rate": 1.712513966480447e-05, + "loss": 0.1808, + "step": 5113 + }, + { + "epoch": 14.28491620111732, + "grad_norm": 0.9819090962409973, + "learning_rate": 1.7128491620111732e-05, + "loss": 0.2273, + "step": 5114 + }, + { + "epoch": 14.287709497206704, + "grad_norm": 0.5800276398658752, + "learning_rate": 1.7131843575418994e-05, + "loss": 0.194, + "step": 5115 + }, + { + "epoch": 14.29050279329609, + "grad_norm": 0.9035034775733948, + "learning_rate": 1.7135195530726258e-05, + "loss": 0.1873, + "step": 5116 + }, + { + "epoch": 14.293296089385475, + "grad_norm": 0.6345489025115967, + "learning_rate": 1.713854748603352e-05, + "loss": 0.2026, + "step": 5117 + }, + { + "epoch": 14.296089385474861, + "grad_norm": 0.9345027208328247, + "learning_rate": 1.714189944134078e-05, + "loss": 0.1928, + "step": 5118 + }, + { + "epoch": 14.298882681564246, + "grad_norm": 0.6476150751113892, + "learning_rate": 1.7145251396648046e-05, + "loss": 0.1856, + "step": 5119 + }, + { + "epoch": 14.30167597765363, + "grad_norm": 0.6979705691337585, + "learning_rate": 1.7148603351955307e-05, + "loss": 0.2271, + "step": 5120 + }, + { + "epoch": 14.304469273743017, + "grad_norm": 0.7362495064735413, + "learning_rate": 1.715195530726257e-05, + "loss": 0.2029, + "step": 5121 + }, + { + "epoch": 14.307262569832401, + "grad_norm": 0.7219176888465881, + "learning_rate": 1.7155307262569833e-05, + "loss": 0.182, + "step": 5122 + }, + { + "epoch": 14.310055865921788, + "grad_norm": 0.775141716003418, + "learning_rate": 1.7158659217877094e-05, + "loss": 0.1984, + "step": 5123 + }, + { + "epoch": 14.312849162011172, + "grad_norm": 0.8255548477172852, + "learning_rate": 1.7162011173184356e-05, + "loss": 0.1835, + "step": 5124 + }, + { + "epoch": 14.315642458100559, + "grad_norm": 1.0451346635818481, + "learning_rate": 1.716536312849162e-05, + "loss": 0.1634, + "step": 5125 + }, + { + "epoch": 14.318435754189943, + "grad_norm": 0.9542490243911743, + "learning_rate": 1.7168715083798882e-05, + "loss": 0.1851, + "step": 5126 + }, + { + "epoch": 14.32122905027933, + "grad_norm": 0.6556843519210815, + "learning_rate": 1.7172067039106143e-05, + "loss": 0.1573, + "step": 5127 + }, + { + "epoch": 14.324022346368714, + "grad_norm": 0.9137011170387268, + "learning_rate": 1.7175418994413408e-05, + "loss": 0.1856, + "step": 5128 + }, + { + "epoch": 14.3268156424581, + "grad_norm": 1.1841174364089966, + "learning_rate": 1.717877094972067e-05, + "loss": 0.1825, + "step": 5129 + }, + { + "epoch": 14.329608938547485, + "grad_norm": 1.0993934869766235, + "learning_rate": 1.7182122905027934e-05, + "loss": 0.197, + "step": 5130 + }, + { + "epoch": 14.332402234636872, + "grad_norm": 0.8640457391738892, + "learning_rate": 1.7185474860335195e-05, + "loss": 0.1815, + "step": 5131 + }, + { + "epoch": 14.335195530726256, + "grad_norm": 1.1170774698257446, + "learning_rate": 1.7188826815642457e-05, + "loss": 0.197, + "step": 5132 + }, + { + "epoch": 14.337988826815643, + "grad_norm": 1.4414327144622803, + "learning_rate": 1.719217877094972e-05, + "loss": 0.191, + "step": 5133 + }, + { + "epoch": 14.340782122905027, + "grad_norm": 2.0535037517547607, + "learning_rate": 1.7195530726256983e-05, + "loss": 0.2058, + "step": 5134 + }, + { + "epoch": 14.343575418994414, + "grad_norm": 1.3121263980865479, + "learning_rate": 1.7198882681564244e-05, + "loss": 0.1978, + "step": 5135 + }, + { + "epoch": 14.346368715083798, + "grad_norm": 1.7094604969024658, + "learning_rate": 1.720223463687151e-05, + "loss": 0.275, + "step": 5136 + }, + { + "epoch": 14.349162011173185, + "grad_norm": 2.0520589351654053, + "learning_rate": 1.720558659217877e-05, + "loss": 0.2919, + "step": 5137 + }, + { + "epoch": 14.35195530726257, + "grad_norm": 0.5974161028862, + "learning_rate": 1.720893854748603e-05, + "loss": 0.2214, + "step": 5138 + }, + { + "epoch": 14.354748603351956, + "grad_norm": 0.5717990398406982, + "learning_rate": 1.7212290502793296e-05, + "loss": 0.1985, + "step": 5139 + }, + { + "epoch": 14.35754189944134, + "grad_norm": 0.6809808611869812, + "learning_rate": 1.7215642458100558e-05, + "loss": 0.2082, + "step": 5140 + }, + { + "epoch": 14.360335195530727, + "grad_norm": 0.7521038055419922, + "learning_rate": 1.721899441340782e-05, + "loss": 0.2145, + "step": 5141 + }, + { + "epoch": 14.363128491620111, + "grad_norm": 0.6871249675750732, + "learning_rate": 1.7222346368715084e-05, + "loss": 0.2199, + "step": 5142 + }, + { + "epoch": 14.365921787709498, + "grad_norm": 0.8222054839134216, + "learning_rate": 1.7225698324022345e-05, + "loss": 0.1819, + "step": 5143 + }, + { + "epoch": 14.368715083798882, + "grad_norm": 0.6354455351829529, + "learning_rate": 1.7229050279329606e-05, + "loss": 0.2169, + "step": 5144 + }, + { + "epoch": 14.371508379888269, + "grad_norm": 1.147626519203186, + "learning_rate": 1.723240223463687e-05, + "loss": 0.1979, + "step": 5145 + }, + { + "epoch": 14.374301675977653, + "grad_norm": 0.6658684015274048, + "learning_rate": 1.7235754189944133e-05, + "loss": 0.1662, + "step": 5146 + }, + { + "epoch": 14.37709497206704, + "grad_norm": 1.1231590509414673, + "learning_rate": 1.7239106145251397e-05, + "loss": 0.2041, + "step": 5147 + }, + { + "epoch": 14.379888268156424, + "grad_norm": 0.6021534204483032, + "learning_rate": 1.724245810055866e-05, + "loss": 0.1683, + "step": 5148 + }, + { + "epoch": 14.38268156424581, + "grad_norm": 0.8212235569953918, + "learning_rate": 1.724581005586592e-05, + "loss": 0.1861, + "step": 5149 + }, + { + "epoch": 14.385474860335195, + "grad_norm": 0.6855723261833191, + "learning_rate": 1.7249162011173185e-05, + "loss": 0.1597, + "step": 5150 + }, + { + "epoch": 14.388268156424582, + "grad_norm": 0.6801570057868958, + "learning_rate": 1.7252513966480446e-05, + "loss": 0.1776, + "step": 5151 + }, + { + "epoch": 14.391061452513966, + "grad_norm": 0.833120584487915, + "learning_rate": 1.7255865921787707e-05, + "loss": 0.212, + "step": 5152 + }, + { + "epoch": 14.393854748603353, + "grad_norm": 0.8944430351257324, + "learning_rate": 1.7259217877094972e-05, + "loss": 0.1754, + "step": 5153 + }, + { + "epoch": 14.396648044692737, + "grad_norm": 0.7718711495399475, + "learning_rate": 1.7262569832402237e-05, + "loss": 0.1942, + "step": 5154 + }, + { + "epoch": 14.399441340782122, + "grad_norm": 1.3760921955108643, + "learning_rate": 1.7265921787709498e-05, + "loss": 0.1597, + "step": 5155 + }, + { + "epoch": 14.402234636871508, + "grad_norm": 1.226315975189209, + "learning_rate": 1.7269273743016763e-05, + "loss": 0.188, + "step": 5156 + }, + { + "epoch": 14.405027932960893, + "grad_norm": 1.3422685861587524, + "learning_rate": 1.7272625698324024e-05, + "loss": 0.2425, + "step": 5157 + }, + { + "epoch": 14.40782122905028, + "grad_norm": 1.1732748746871948, + "learning_rate": 1.7275977653631286e-05, + "loss": 0.1787, + "step": 5158 + }, + { + "epoch": 14.410614525139664, + "grad_norm": 1.3538166284561157, + "learning_rate": 1.727932960893855e-05, + "loss": 0.2029, + "step": 5159 + }, + { + "epoch": 14.41340782122905, + "grad_norm": 2.400662660598755, + "learning_rate": 1.728268156424581e-05, + "loss": 0.189, + "step": 5160 + }, + { + "epoch": 14.416201117318435, + "grad_norm": 1.4506539106369019, + "learning_rate": 1.7286033519553073e-05, + "loss": 0.2299, + "step": 5161 + }, + { + "epoch": 14.418994413407821, + "grad_norm": 3.467122793197632, + "learning_rate": 1.7289385474860338e-05, + "loss": 0.2414, + "step": 5162 + }, + { + "epoch": 14.421787709497206, + "grad_norm": 0.6172969937324524, + "learning_rate": 1.72927374301676e-05, + "loss": 0.1984, + "step": 5163 + }, + { + "epoch": 14.424581005586592, + "grad_norm": 0.4902450144290924, + "learning_rate": 1.729608938547486e-05, + "loss": 0.1713, + "step": 5164 + }, + { + "epoch": 14.427374301675977, + "grad_norm": 0.4806141257286072, + "learning_rate": 1.7299441340782125e-05, + "loss": 0.223, + "step": 5165 + }, + { + "epoch": 14.430167597765363, + "grad_norm": 0.6677011847496033, + "learning_rate": 1.7302793296089387e-05, + "loss": 0.1803, + "step": 5166 + }, + { + "epoch": 14.432960893854748, + "grad_norm": 1.0174788236618042, + "learning_rate": 1.730614525139665e-05, + "loss": 0.2153, + "step": 5167 + }, + { + "epoch": 14.435754189944134, + "grad_norm": 0.5995932221412659, + "learning_rate": 1.7309497206703913e-05, + "loss": 0.1874, + "step": 5168 + }, + { + "epoch": 14.438547486033519, + "grad_norm": 0.5318551659584045, + "learning_rate": 1.7312849162011174e-05, + "loss": 0.1898, + "step": 5169 + }, + { + "epoch": 14.441340782122905, + "grad_norm": 0.7481793165206909, + "learning_rate": 1.731620111731844e-05, + "loss": 0.1766, + "step": 5170 + }, + { + "epoch": 14.44413407821229, + "grad_norm": 0.56796795129776, + "learning_rate": 1.73195530726257e-05, + "loss": 0.1967, + "step": 5171 + }, + { + "epoch": 14.446927374301676, + "grad_norm": 0.773189902305603, + "learning_rate": 1.732290502793296e-05, + "loss": 0.1865, + "step": 5172 + }, + { + "epoch": 14.449720670391061, + "grad_norm": 0.578679621219635, + "learning_rate": 1.7326256983240226e-05, + "loss": 0.162, + "step": 5173 + }, + { + "epoch": 14.452513966480447, + "grad_norm": 0.8440898656845093, + "learning_rate": 1.7329608938547487e-05, + "loss": 0.1783, + "step": 5174 + }, + { + "epoch": 14.455307262569832, + "grad_norm": 0.7501804828643799, + "learning_rate": 1.733296089385475e-05, + "loss": 0.1994, + "step": 5175 + }, + { + "epoch": 14.458100558659218, + "grad_norm": 1.013456106185913, + "learning_rate": 1.7336312849162014e-05, + "loss": 0.1985, + "step": 5176 + }, + { + "epoch": 14.460893854748603, + "grad_norm": 0.854615330696106, + "learning_rate": 1.7339664804469275e-05, + "loss": 0.1603, + "step": 5177 + }, + { + "epoch": 14.46368715083799, + "grad_norm": 0.6173352003097534, + "learning_rate": 1.7343016759776536e-05, + "loss": 0.1747, + "step": 5178 + }, + { + "epoch": 14.466480446927374, + "grad_norm": 0.8713481426239014, + "learning_rate": 1.73463687150838e-05, + "loss": 0.2325, + "step": 5179 + }, + { + "epoch": 14.46927374301676, + "grad_norm": 2.031515598297119, + "learning_rate": 1.7349720670391062e-05, + "loss": 0.1858, + "step": 5180 + }, + { + "epoch": 14.472067039106145, + "grad_norm": 1.1038472652435303, + "learning_rate": 1.7353072625698324e-05, + "loss": 0.2047, + "step": 5181 + }, + { + "epoch": 14.474860335195531, + "grad_norm": 0.8233575820922852, + "learning_rate": 1.735642458100559e-05, + "loss": 0.2003, + "step": 5182 + }, + { + "epoch": 14.477653631284916, + "grad_norm": 2.4570469856262207, + "learning_rate": 1.735977653631285e-05, + "loss": 0.2198, + "step": 5183 + }, + { + "epoch": 14.480446927374302, + "grad_norm": 1.2360502481460571, + "learning_rate": 1.736312849162011e-05, + "loss": 0.186, + "step": 5184 + }, + { + "epoch": 14.483240223463687, + "grad_norm": 1.156750202178955, + "learning_rate": 1.7366480446927376e-05, + "loss": 0.2197, + "step": 5185 + }, + { + "epoch": 14.486033519553073, + "grad_norm": 1.4431462287902832, + "learning_rate": 1.7369832402234637e-05, + "loss": 0.228, + "step": 5186 + }, + { + "epoch": 14.488826815642458, + "grad_norm": 5.256291389465332, + "learning_rate": 1.7373184357541902e-05, + "loss": 0.3127, + "step": 5187 + }, + { + "epoch": 14.491620111731844, + "grad_norm": 0.5623120069503784, + "learning_rate": 1.7376536312849163e-05, + "loss": 0.1808, + "step": 5188 + }, + { + "epoch": 14.494413407821229, + "grad_norm": 0.806419849395752, + "learning_rate": 1.7379888268156425e-05, + "loss": 0.1988, + "step": 5189 + }, + { + "epoch": 14.497206703910614, + "grad_norm": 0.8473981022834778, + "learning_rate": 1.738324022346369e-05, + "loss": 0.1845, + "step": 5190 + }, + { + "epoch": 14.5, + "grad_norm": 0.5002455115318298, + "learning_rate": 1.738659217877095e-05, + "loss": 0.1997, + "step": 5191 + }, + { + "epoch": 14.502793296089386, + "grad_norm": 0.745423436164856, + "learning_rate": 1.7389944134078212e-05, + "loss": 0.2158, + "step": 5192 + }, + { + "epoch": 14.505586592178771, + "grad_norm": 0.5932745933532715, + "learning_rate": 1.7393296089385477e-05, + "loss": 0.2058, + "step": 5193 + }, + { + "epoch": 14.508379888268156, + "grad_norm": 0.5666309595108032, + "learning_rate": 1.7396648044692738e-05, + "loss": 0.1819, + "step": 5194 + }, + { + "epoch": 14.511173184357542, + "grad_norm": 0.6354514956474304, + "learning_rate": 1.74e-05, + "loss": 0.182, + "step": 5195 + }, + { + "epoch": 14.513966480446927, + "grad_norm": 1.1739368438720703, + "learning_rate": 1.7403351955307264e-05, + "loss": 0.171, + "step": 5196 + }, + { + "epoch": 14.516759776536313, + "grad_norm": 0.6206181645393372, + "learning_rate": 1.7406703910614526e-05, + "loss": 0.1724, + "step": 5197 + }, + { + "epoch": 14.519553072625698, + "grad_norm": 0.692364513874054, + "learning_rate": 1.7410055865921787e-05, + "loss": 0.2003, + "step": 5198 + }, + { + "epoch": 14.522346368715084, + "grad_norm": 0.8339012265205383, + "learning_rate": 1.741340782122905e-05, + "loss": 0.2115, + "step": 5199 + }, + { + "epoch": 14.525139664804469, + "grad_norm": 1.1273937225341797, + "learning_rate": 1.7416759776536313e-05, + "loss": 0.1718, + "step": 5200 + }, + { + "epoch": 14.527932960893855, + "grad_norm": 0.9194111227989197, + "learning_rate": 1.7420111731843574e-05, + "loss": 0.1787, + "step": 5201 + }, + { + "epoch": 14.53072625698324, + "grad_norm": 0.824435830116272, + "learning_rate": 1.742346368715084e-05, + "loss": 0.2242, + "step": 5202 + }, + { + "epoch": 14.533519553072626, + "grad_norm": 1.0002049207687378, + "learning_rate": 1.74268156424581e-05, + "loss": 0.2063, + "step": 5203 + }, + { + "epoch": 14.53631284916201, + "grad_norm": 0.9718796014785767, + "learning_rate": 1.7430167597765365e-05, + "loss": 0.1914, + "step": 5204 + }, + { + "epoch": 14.539106145251397, + "grad_norm": 0.7251378297805786, + "learning_rate": 1.7433519553072626e-05, + "loss": 0.1687, + "step": 5205 + }, + { + "epoch": 14.541899441340782, + "grad_norm": 1.1350210905075073, + "learning_rate": 1.7436871508379888e-05, + "loss": 0.1492, + "step": 5206 + }, + { + "epoch": 14.544692737430168, + "grad_norm": 2.7378242015838623, + "learning_rate": 1.7440223463687153e-05, + "loss": 0.2067, + "step": 5207 + }, + { + "epoch": 14.547486033519553, + "grad_norm": 1.3967255353927612, + "learning_rate": 1.7443575418994414e-05, + "loss": 0.1751, + "step": 5208 + }, + { + "epoch": 14.550279329608939, + "grad_norm": 0.9531848430633545, + "learning_rate": 1.7446927374301675e-05, + "loss": 0.1809, + "step": 5209 + }, + { + "epoch": 14.553072625698324, + "grad_norm": 1.3670423030853271, + "learning_rate": 1.745027932960894e-05, + "loss": 0.2541, + "step": 5210 + }, + { + "epoch": 14.55586592178771, + "grad_norm": 1.4940139055252075, + "learning_rate": 1.74536312849162e-05, + "loss": 0.21, + "step": 5211 + }, + { + "epoch": 14.558659217877095, + "grad_norm": 1.858693242073059, + "learning_rate": 1.7456983240223463e-05, + "loss": 0.2916, + "step": 5212 + }, + { + "epoch": 14.561452513966481, + "grad_norm": 0.6553477048873901, + "learning_rate": 1.7460335195530727e-05, + "loss": 0.2443, + "step": 5213 + }, + { + "epoch": 14.564245810055866, + "grad_norm": 0.6276983618736267, + "learning_rate": 1.746368715083799e-05, + "loss": 0.1867, + "step": 5214 + }, + { + "epoch": 14.567039106145252, + "grad_norm": 1.446393370628357, + "learning_rate": 1.746703910614525e-05, + "loss": 0.1849, + "step": 5215 + }, + { + "epoch": 14.569832402234637, + "grad_norm": 0.5356502532958984, + "learning_rate": 1.7470391061452515e-05, + "loss": 0.1636, + "step": 5216 + }, + { + "epoch": 14.572625698324023, + "grad_norm": 0.6465040445327759, + "learning_rate": 1.7473743016759776e-05, + "loss": 0.1768, + "step": 5217 + }, + { + "epoch": 14.575418994413408, + "grad_norm": 0.558599054813385, + "learning_rate": 1.7477094972067037e-05, + "loss": 0.1838, + "step": 5218 + }, + { + "epoch": 14.578212290502794, + "grad_norm": 0.5643178224563599, + "learning_rate": 1.7480446927374302e-05, + "loss": 0.1736, + "step": 5219 + }, + { + "epoch": 14.581005586592179, + "grad_norm": 0.7047673463821411, + "learning_rate": 1.7483798882681564e-05, + "loss": 0.1735, + "step": 5220 + }, + { + "epoch": 14.583798882681565, + "grad_norm": 0.849492609500885, + "learning_rate": 1.7487150837988825e-05, + "loss": 0.1784, + "step": 5221 + }, + { + "epoch": 14.58659217877095, + "grad_norm": 0.7889730334281921, + "learning_rate": 1.749050279329609e-05, + "loss": 0.1813, + "step": 5222 + }, + { + "epoch": 14.589385474860336, + "grad_norm": 0.7525766491889954, + "learning_rate": 1.749385474860335e-05, + "loss": 0.1962, + "step": 5223 + }, + { + "epoch": 14.59217877094972, + "grad_norm": 0.9174278378486633, + "learning_rate": 1.7497206703910616e-05, + "loss": 0.1953, + "step": 5224 + }, + { + "epoch": 14.594972067039105, + "grad_norm": 0.9273908734321594, + "learning_rate": 1.7500558659217877e-05, + "loss": 0.1753, + "step": 5225 + }, + { + "epoch": 14.597765363128492, + "grad_norm": 1.0077521800994873, + "learning_rate": 1.750391061452514e-05, + "loss": 0.1914, + "step": 5226 + }, + { + "epoch": 14.600558659217878, + "grad_norm": 0.9167560338973999, + "learning_rate": 1.7507262569832403e-05, + "loss": 0.1619, + "step": 5227 + }, + { + "epoch": 14.603351955307263, + "grad_norm": 0.6834773421287537, + "learning_rate": 1.7510614525139664e-05, + "loss": 0.1712, + "step": 5228 + }, + { + "epoch": 14.606145251396647, + "grad_norm": 0.8407741189002991, + "learning_rate": 1.7513966480446926e-05, + "loss": 0.1917, + "step": 5229 + }, + { + "epoch": 14.608938547486034, + "grad_norm": 1.2240240573883057, + "learning_rate": 1.751731843575419e-05, + "loss": 0.1918, + "step": 5230 + }, + { + "epoch": 14.611731843575418, + "grad_norm": 1.3187683820724487, + "learning_rate": 1.7520670391061452e-05, + "loss": 0.2006, + "step": 5231 + }, + { + "epoch": 14.614525139664805, + "grad_norm": 0.9397504329681396, + "learning_rate": 1.7524022346368713e-05, + "loss": 0.1773, + "step": 5232 + }, + { + "epoch": 14.61731843575419, + "grad_norm": 1.2782046794891357, + "learning_rate": 1.7527374301675978e-05, + "loss": 0.1858, + "step": 5233 + }, + { + "epoch": 14.620111731843576, + "grad_norm": 0.9210992455482483, + "learning_rate": 1.753072625698324e-05, + "loss": 0.2005, + "step": 5234 + }, + { + "epoch": 14.62290502793296, + "grad_norm": 0.9138185381889343, + "learning_rate": 1.75340782122905e-05, + "loss": 0.176, + "step": 5235 + }, + { + "epoch": 14.625698324022347, + "grad_norm": 1.362288236618042, + "learning_rate": 1.7537430167597765e-05, + "loss": 0.2946, + "step": 5236 + }, + { + "epoch": 14.628491620111731, + "grad_norm": 2.314206838607788, + "learning_rate": 1.7540782122905027e-05, + "loss": 0.3049, + "step": 5237 + }, + { + "epoch": 14.631284916201118, + "grad_norm": 0.5205069184303284, + "learning_rate": 1.7544134078212288e-05, + "loss": 0.2015, + "step": 5238 + }, + { + "epoch": 14.634078212290502, + "grad_norm": 0.8590618371963501, + "learning_rate": 1.7547486033519553e-05, + "loss": 0.1916, + "step": 5239 + }, + { + "epoch": 14.636871508379889, + "grad_norm": 0.6126362085342407, + "learning_rate": 1.7550837988826814e-05, + "loss": 0.2064, + "step": 5240 + }, + { + "epoch": 14.639664804469273, + "grad_norm": 0.8078811764717102, + "learning_rate": 1.755418994413408e-05, + "loss": 0.1463, + "step": 5241 + }, + { + "epoch": 14.64245810055866, + "grad_norm": 0.6788820624351501, + "learning_rate": 1.755754189944134e-05, + "loss": 0.2102, + "step": 5242 + }, + { + "epoch": 14.645251396648044, + "grad_norm": 0.46861425042152405, + "learning_rate": 1.75608938547486e-05, + "loss": 0.1809, + "step": 5243 + }, + { + "epoch": 14.64804469273743, + "grad_norm": 0.6321874856948853, + "learning_rate": 1.7564245810055866e-05, + "loss": 0.1921, + "step": 5244 + }, + { + "epoch": 14.650837988826815, + "grad_norm": 0.6293928027153015, + "learning_rate": 1.7567597765363128e-05, + "loss": 0.1978, + "step": 5245 + }, + { + "epoch": 14.653631284916202, + "grad_norm": 0.6335256099700928, + "learning_rate": 1.757094972067039e-05, + "loss": 0.1695, + "step": 5246 + }, + { + "epoch": 14.656424581005586, + "grad_norm": 0.8716128468513489, + "learning_rate": 1.7574301675977654e-05, + "loss": 0.2272, + "step": 5247 + }, + { + "epoch": 14.659217877094973, + "grad_norm": 0.6366837620735168, + "learning_rate": 1.7577653631284915e-05, + "loss": 0.2021, + "step": 5248 + }, + { + "epoch": 14.662011173184357, + "grad_norm": 0.9979801774024963, + "learning_rate": 1.758100558659218e-05, + "loss": 0.1683, + "step": 5249 + }, + { + "epoch": 14.664804469273744, + "grad_norm": 0.7747973203659058, + "learning_rate": 1.7584357541899445e-05, + "loss": 0.1824, + "step": 5250 + }, + { + "epoch": 14.667597765363128, + "grad_norm": 0.9300662875175476, + "learning_rate": 1.7587709497206706e-05, + "loss": 0.1739, + "step": 5251 + }, + { + "epoch": 14.670391061452515, + "grad_norm": 1.4098864793777466, + "learning_rate": 1.7591061452513967e-05, + "loss": 0.1643, + "step": 5252 + }, + { + "epoch": 14.6731843575419, + "grad_norm": 0.8078183531761169, + "learning_rate": 1.7594413407821232e-05, + "loss": 0.1875, + "step": 5253 + }, + { + "epoch": 14.675977653631286, + "grad_norm": 1.5789486169815063, + "learning_rate": 1.7597765363128493e-05, + "loss": 0.1872, + "step": 5254 + }, + { + "epoch": 14.67877094972067, + "grad_norm": 0.9943745732307434, + "learning_rate": 1.7601117318435755e-05, + "loss": 0.1745, + "step": 5255 + }, + { + "epoch": 14.681564245810057, + "grad_norm": 1.0800831317901611, + "learning_rate": 1.760446927374302e-05, + "loss": 0.1697, + "step": 5256 + }, + { + "epoch": 14.684357541899441, + "grad_norm": 1.1433217525482178, + "learning_rate": 1.760782122905028e-05, + "loss": 0.189, + "step": 5257 + }, + { + "epoch": 14.687150837988828, + "grad_norm": 1.2423213720321655, + "learning_rate": 1.7611173184357542e-05, + "loss": 0.2277, + "step": 5258 + }, + { + "epoch": 14.689944134078212, + "grad_norm": 2.280787706375122, + "learning_rate": 1.7614525139664807e-05, + "loss": 0.2149, + "step": 5259 + }, + { + "epoch": 14.692737430167599, + "grad_norm": 1.0065503120422363, + "learning_rate": 1.7617877094972068e-05, + "loss": 0.1697, + "step": 5260 + }, + { + "epoch": 14.695530726256983, + "grad_norm": 1.883996844291687, + "learning_rate": 1.7621229050279333e-05, + "loss": 0.2483, + "step": 5261 + }, + { + "epoch": 14.69832402234637, + "grad_norm": 5.347286701202393, + "learning_rate": 1.7624581005586594e-05, + "loss": 0.3412, + "step": 5262 + }, + { + "epoch": 14.701117318435754, + "grad_norm": 0.4326632618904114, + "learning_rate": 1.7627932960893856e-05, + "loss": 0.2074, + "step": 5263 + }, + { + "epoch": 14.703910614525139, + "grad_norm": 0.5143949389457703, + "learning_rate": 1.763128491620112e-05, + "loss": 0.1936, + "step": 5264 + }, + { + "epoch": 14.706703910614525, + "grad_norm": 0.7839817404747009, + "learning_rate": 1.763463687150838e-05, + "loss": 0.1766, + "step": 5265 + }, + { + "epoch": 14.70949720670391, + "grad_norm": 0.5075476765632629, + "learning_rate": 1.7637988826815643e-05, + "loss": 0.1984, + "step": 5266 + }, + { + "epoch": 14.712290502793296, + "grad_norm": 1.1026802062988281, + "learning_rate": 1.7641340782122908e-05, + "loss": 0.1824, + "step": 5267 + }, + { + "epoch": 14.71508379888268, + "grad_norm": 0.5609515905380249, + "learning_rate": 1.764469273743017e-05, + "loss": 0.2027, + "step": 5268 + }, + { + "epoch": 14.717877094972067, + "grad_norm": 0.9849872589111328, + "learning_rate": 1.764804469273743e-05, + "loss": 0.2195, + "step": 5269 + }, + { + "epoch": 14.720670391061452, + "grad_norm": 0.6054843068122864, + "learning_rate": 1.7651396648044695e-05, + "loss": 0.1854, + "step": 5270 + }, + { + "epoch": 14.723463687150838, + "grad_norm": 0.7682620286941528, + "learning_rate": 1.7654748603351957e-05, + "loss": 0.1595, + "step": 5271 + }, + { + "epoch": 14.726256983240223, + "grad_norm": 0.858316957950592, + "learning_rate": 1.7658100558659218e-05, + "loss": 0.202, + "step": 5272 + }, + { + "epoch": 14.72905027932961, + "grad_norm": 0.7501189112663269, + "learning_rate": 1.7661452513966483e-05, + "loss": 0.1729, + "step": 5273 + }, + { + "epoch": 14.731843575418994, + "grad_norm": 0.9036000370979309, + "learning_rate": 1.7664804469273744e-05, + "loss": 0.1831, + "step": 5274 + }, + { + "epoch": 14.73463687150838, + "grad_norm": 0.8348864912986755, + "learning_rate": 1.7668156424581005e-05, + "loss": 0.1941, + "step": 5275 + }, + { + "epoch": 14.737430167597765, + "grad_norm": 1.1426677703857422, + "learning_rate": 1.767150837988827e-05, + "loss": 0.1839, + "step": 5276 + }, + { + "epoch": 14.740223463687151, + "grad_norm": 0.6874288320541382, + "learning_rate": 1.767486033519553e-05, + "loss": 0.1819, + "step": 5277 + }, + { + "epoch": 14.743016759776536, + "grad_norm": 1.0388387441635132, + "learning_rate": 1.7678212290502793e-05, + "loss": 0.1941, + "step": 5278 + }, + { + "epoch": 14.745810055865922, + "grad_norm": 0.7658501863479614, + "learning_rate": 1.7681564245810057e-05, + "loss": 0.1822, + "step": 5279 + }, + { + "epoch": 14.748603351955307, + "grad_norm": 0.7673095464706421, + "learning_rate": 1.768491620111732e-05, + "loss": 0.18, + "step": 5280 + }, + { + "epoch": 14.751396648044693, + "grad_norm": 0.9403380751609802, + "learning_rate": 1.7688268156424584e-05, + "loss": 0.1679, + "step": 5281 + }, + { + "epoch": 14.754189944134078, + "grad_norm": 0.8547789454460144, + "learning_rate": 1.7691620111731845e-05, + "loss": 0.1701, + "step": 5282 + }, + { + "epoch": 14.756983240223464, + "grad_norm": 0.7148831486701965, + "learning_rate": 1.7694972067039106e-05, + "loss": 0.1715, + "step": 5283 + }, + { + "epoch": 14.759776536312849, + "grad_norm": 1.1654236316680908, + "learning_rate": 1.769832402234637e-05, + "loss": 0.2011, + "step": 5284 + }, + { + "epoch": 14.762569832402235, + "grad_norm": 1.4934712648391724, + "learning_rate": 1.7701675977653632e-05, + "loss": 0.2128, + "step": 5285 + }, + { + "epoch": 14.76536312849162, + "grad_norm": 1.1211246252059937, + "learning_rate": 1.7705027932960894e-05, + "loss": 0.1946, + "step": 5286 + }, + { + "epoch": 14.768156424581006, + "grad_norm": 2.0585808753967285, + "learning_rate": 1.770837988826816e-05, + "loss": 0.2602, + "step": 5287 + }, + { + "epoch": 14.77094972067039, + "grad_norm": 0.5379813313484192, + "learning_rate": 1.771173184357542e-05, + "loss": 0.1757, + "step": 5288 + }, + { + "epoch": 14.773743016759777, + "grad_norm": 0.6689831614494324, + "learning_rate": 1.771508379888268e-05, + "loss": 0.196, + "step": 5289 + }, + { + "epoch": 14.776536312849162, + "grad_norm": 0.7049437761306763, + "learning_rate": 1.7718435754189946e-05, + "loss": 0.173, + "step": 5290 + }, + { + "epoch": 14.779329608938548, + "grad_norm": 1.0554572343826294, + "learning_rate": 1.7721787709497207e-05, + "loss": 0.2334, + "step": 5291 + }, + { + "epoch": 14.782122905027933, + "grad_norm": 0.6565403342247009, + "learning_rate": 1.772513966480447e-05, + "loss": 0.1923, + "step": 5292 + }, + { + "epoch": 14.78491620111732, + "grad_norm": 0.5655161738395691, + "learning_rate": 1.7728491620111733e-05, + "loss": 0.1775, + "step": 5293 + }, + { + "epoch": 14.787709497206704, + "grad_norm": 0.5910884737968445, + "learning_rate": 1.7731843575418995e-05, + "loss": 0.2072, + "step": 5294 + }, + { + "epoch": 14.79050279329609, + "grad_norm": 0.5990604162216187, + "learning_rate": 1.7735195530726256e-05, + "loss": 0.1872, + "step": 5295 + }, + { + "epoch": 14.793296089385475, + "grad_norm": 0.7197539806365967, + "learning_rate": 1.773854748603352e-05, + "loss": 0.1966, + "step": 5296 + }, + { + "epoch": 14.796089385474861, + "grad_norm": 0.6353759765625, + "learning_rate": 1.7741899441340782e-05, + "loss": 0.2022, + "step": 5297 + }, + { + "epoch": 14.798882681564246, + "grad_norm": 0.534000813961029, + "learning_rate": 1.7745251396648047e-05, + "loss": 0.1969, + "step": 5298 + }, + { + "epoch": 14.80167597765363, + "grad_norm": 0.5878714919090271, + "learning_rate": 1.7748603351955308e-05, + "loss": 0.1829, + "step": 5299 + }, + { + "epoch": 14.804469273743017, + "grad_norm": 0.7076089978218079, + "learning_rate": 1.775195530726257e-05, + "loss": 0.1706, + "step": 5300 + }, + { + "epoch": 14.807262569832401, + "grad_norm": 0.6261314153671265, + "learning_rate": 1.7755307262569834e-05, + "loss": 0.1741, + "step": 5301 + }, + { + "epoch": 14.810055865921788, + "grad_norm": 1.824934720993042, + "learning_rate": 1.7758659217877096e-05, + "loss": 0.1882, + "step": 5302 + }, + { + "epoch": 14.812849162011172, + "grad_norm": 0.7575598955154419, + "learning_rate": 1.7762011173184357e-05, + "loss": 0.1897, + "step": 5303 + }, + { + "epoch": 14.815642458100559, + "grad_norm": 0.9772080779075623, + "learning_rate": 1.776536312849162e-05, + "loss": 0.1703, + "step": 5304 + }, + { + "epoch": 14.818435754189943, + "grad_norm": 0.8602588176727295, + "learning_rate": 1.7768715083798883e-05, + "loss": 0.1419, + "step": 5305 + }, + { + "epoch": 14.82122905027933, + "grad_norm": 0.756101131439209, + "learning_rate": 1.7772067039106144e-05, + "loss": 0.1773, + "step": 5306 + }, + { + "epoch": 14.824022346368714, + "grad_norm": 1.1944538354873657, + "learning_rate": 1.777541899441341e-05, + "loss": 0.193, + "step": 5307 + }, + { + "epoch": 14.8268156424581, + "grad_norm": 1.3971936702728271, + "learning_rate": 1.777877094972067e-05, + "loss": 0.2099, + "step": 5308 + }, + { + "epoch": 14.829608938547485, + "grad_norm": 1.6767597198486328, + "learning_rate": 1.7782122905027932e-05, + "loss": 0.1981, + "step": 5309 + }, + { + "epoch": 14.832402234636872, + "grad_norm": 1.2538063526153564, + "learning_rate": 1.7785474860335196e-05, + "loss": 0.221, + "step": 5310 + }, + { + "epoch": 14.835195530726256, + "grad_norm": 1.7076013088226318, + "learning_rate": 1.7788826815642458e-05, + "loss": 0.2156, + "step": 5311 + }, + { + "epoch": 14.837988826815643, + "grad_norm": 3.8120012283325195, + "learning_rate": 1.779217877094972e-05, + "loss": 0.2837, + "step": 5312 + }, + { + "epoch": 14.840782122905027, + "grad_norm": 1.2786812782287598, + "learning_rate": 1.7795530726256984e-05, + "loss": 0.1784, + "step": 5313 + }, + { + "epoch": 14.843575418994414, + "grad_norm": 0.8257588148117065, + "learning_rate": 1.7798882681564245e-05, + "loss": 0.1632, + "step": 5314 + }, + { + "epoch": 14.846368715083798, + "grad_norm": 0.9777702689170837, + "learning_rate": 1.7802234636871507e-05, + "loss": 0.1587, + "step": 5315 + }, + { + "epoch": 14.849162011173185, + "grad_norm": 0.8988518714904785, + "learning_rate": 1.780558659217877e-05, + "loss": 0.2423, + "step": 5316 + }, + { + "epoch": 14.85195530726257, + "grad_norm": 1.1572656631469727, + "learning_rate": 1.7808938547486033e-05, + "loss": 0.2357, + "step": 5317 + }, + { + "epoch": 14.854748603351956, + "grad_norm": 0.9496991038322449, + "learning_rate": 1.7812290502793297e-05, + "loss": 0.24, + "step": 5318 + }, + { + "epoch": 14.85754189944134, + "grad_norm": 1.9591820240020752, + "learning_rate": 1.781564245810056e-05, + "loss": 0.1899, + "step": 5319 + }, + { + "epoch": 14.860335195530727, + "grad_norm": 1.52360200881958, + "learning_rate": 1.781899441340782e-05, + "loss": 0.1979, + "step": 5320 + }, + { + "epoch": 14.863128491620111, + "grad_norm": 0.727249801158905, + "learning_rate": 1.7822346368715085e-05, + "loss": 0.1906, + "step": 5321 + }, + { + "epoch": 14.865921787709498, + "grad_norm": 0.772865891456604, + "learning_rate": 1.7825698324022346e-05, + "loss": 0.1921, + "step": 5322 + }, + { + "epoch": 14.868715083798882, + "grad_norm": 0.9101760387420654, + "learning_rate": 1.7829050279329607e-05, + "loss": 0.2, + "step": 5323 + }, + { + "epoch": 14.871508379888269, + "grad_norm": 0.7353464365005493, + "learning_rate": 1.7832402234636872e-05, + "loss": 0.1943, + "step": 5324 + }, + { + "epoch": 14.874301675977653, + "grad_norm": 1.0054141283035278, + "learning_rate": 1.7835754189944134e-05, + "loss": 0.1874, + "step": 5325 + }, + { + "epoch": 14.87709497206704, + "grad_norm": 1.0295093059539795, + "learning_rate": 1.7839106145251395e-05, + "loss": 0.2133, + "step": 5326 + }, + { + "epoch": 14.879888268156424, + "grad_norm": 2.113431215286255, + "learning_rate": 1.784245810055866e-05, + "loss": 0.1532, + "step": 5327 + }, + { + "epoch": 14.88268156424581, + "grad_norm": 0.8499321341514587, + "learning_rate": 1.784581005586592e-05, + "loss": 0.1967, + "step": 5328 + }, + { + "epoch": 14.885474860335195, + "grad_norm": 0.8761413097381592, + "learning_rate": 1.7849162011173182e-05, + "loss": 0.1545, + "step": 5329 + }, + { + "epoch": 14.888268156424582, + "grad_norm": 0.9777705669403076, + "learning_rate": 1.7852513966480447e-05, + "loss": 0.1856, + "step": 5330 + }, + { + "epoch": 14.891061452513966, + "grad_norm": 2.137449264526367, + "learning_rate": 1.785586592178771e-05, + "loss": 0.1638, + "step": 5331 + }, + { + "epoch": 14.893854748603353, + "grad_norm": 0.9079502820968628, + "learning_rate": 1.785921787709497e-05, + "loss": 0.1993, + "step": 5332 + }, + { + "epoch": 14.896648044692737, + "grad_norm": 1.6817976236343384, + "learning_rate": 1.7862569832402234e-05, + "loss": 0.1725, + "step": 5333 + }, + { + "epoch": 14.899441340782122, + "grad_norm": 1.249648094177246, + "learning_rate": 1.7865921787709496e-05, + "loss": 0.2322, + "step": 5334 + }, + { + "epoch": 14.902234636871508, + "grad_norm": 1.7914098501205444, + "learning_rate": 1.786927374301676e-05, + "loss": 0.2342, + "step": 5335 + }, + { + "epoch": 14.905027932960895, + "grad_norm": 2.1256749629974365, + "learning_rate": 1.7872625698324022e-05, + "loss": 0.2285, + "step": 5336 + }, + { + "epoch": 14.90782122905028, + "grad_norm": 1.5839430093765259, + "learning_rate": 1.7875977653631283e-05, + "loss": 0.3354, + "step": 5337 + }, + { + "epoch": 14.910614525139664, + "grad_norm": 1.4942432641983032, + "learning_rate": 1.7879329608938548e-05, + "loss": 0.228, + "step": 5338 + }, + { + "epoch": 14.91340782122905, + "grad_norm": 0.6231188178062439, + "learning_rate": 1.788268156424581e-05, + "loss": 0.1908, + "step": 5339 + }, + { + "epoch": 14.916201117318435, + "grad_norm": 1.4801113605499268, + "learning_rate": 1.788603351955307e-05, + "loss": 0.2113, + "step": 5340 + }, + { + "epoch": 14.918994413407821, + "grad_norm": 0.6909085512161255, + "learning_rate": 1.7889385474860335e-05, + "loss": 0.1874, + "step": 5341 + }, + { + "epoch": 14.921787709497206, + "grad_norm": 0.7906087040901184, + "learning_rate": 1.7892737430167597e-05, + "loss": 0.192, + "step": 5342 + }, + { + "epoch": 14.924581005586592, + "grad_norm": 0.5631007552146912, + "learning_rate": 1.7896089385474858e-05, + "loss": 0.1902, + "step": 5343 + }, + { + "epoch": 14.927374301675977, + "grad_norm": 0.9749992489814758, + "learning_rate": 1.7899441340782123e-05, + "loss": 0.1565, + "step": 5344 + }, + { + "epoch": 14.930167597765363, + "grad_norm": 1.1353789567947388, + "learning_rate": 1.7902793296089384e-05, + "loss": 0.1986, + "step": 5345 + }, + { + "epoch": 14.932960893854748, + "grad_norm": 0.9097453951835632, + "learning_rate": 1.790614525139665e-05, + "loss": 0.1653, + "step": 5346 + }, + { + "epoch": 14.935754189944134, + "grad_norm": 0.9892898201942444, + "learning_rate": 1.7909497206703914e-05, + "loss": 0.2061, + "step": 5347 + }, + { + "epoch": 14.938547486033519, + "grad_norm": 0.9853661060333252, + "learning_rate": 1.7912849162011175e-05, + "loss": 0.2245, + "step": 5348 + }, + { + "epoch": 14.941340782122905, + "grad_norm": 1.1538175344467163, + "learning_rate": 1.7916201117318436e-05, + "loss": 0.1947, + "step": 5349 + }, + { + "epoch": 14.94413407821229, + "grad_norm": 0.7679402232170105, + "learning_rate": 1.79195530726257e-05, + "loss": 0.1859, + "step": 5350 + }, + { + "epoch": 14.946927374301676, + "grad_norm": 1.9291404485702515, + "learning_rate": 1.7922905027932962e-05, + "loss": 0.2037, + "step": 5351 + }, + { + "epoch": 14.949720670391061, + "grad_norm": 0.8362616300582886, + "learning_rate": 1.7926256983240224e-05, + "loss": 0.2148, + "step": 5352 + }, + { + "epoch": 14.952513966480447, + "grad_norm": 0.9848764538764954, + "learning_rate": 1.792960893854749e-05, + "loss": 0.2276, + "step": 5353 + }, + { + "epoch": 14.955307262569832, + "grad_norm": 0.9925324320793152, + "learning_rate": 1.793296089385475e-05, + "loss": 0.2469, + "step": 5354 + }, + { + "epoch": 14.958100558659218, + "grad_norm": 0.9262887835502625, + "learning_rate": 1.7936312849162015e-05, + "loss": 0.2065, + "step": 5355 + }, + { + "epoch": 14.960893854748603, + "grad_norm": 1.0662223100662231, + "learning_rate": 1.7939664804469276e-05, + "loss": 0.218, + "step": 5356 + }, + { + "epoch": 14.96368715083799, + "grad_norm": 1.7205967903137207, + "learning_rate": 1.7943016759776537e-05, + "loss": 0.2593, + "step": 5357 + }, + { + "epoch": 14.966480446927374, + "grad_norm": 0.8366590142250061, + "learning_rate": 1.7946368715083802e-05, + "loss": 0.196, + "step": 5358 + }, + { + "epoch": 14.96927374301676, + "grad_norm": 1.2276511192321777, + "learning_rate": 1.7949720670391063e-05, + "loss": 0.2009, + "step": 5359 + }, + { + "epoch": 14.972067039106145, + "grad_norm": 8.529196739196777, + "learning_rate": 1.7953072625698325e-05, + "loss": 0.163, + "step": 5360 + }, + { + "epoch": 14.974860335195531, + "grad_norm": 1.5483835935592651, + "learning_rate": 1.795642458100559e-05, + "loss": 0.2215, + "step": 5361 + }, + { + "epoch": 14.977653631284916, + "grad_norm": 2.0050134658813477, + "learning_rate": 1.795977653631285e-05, + "loss": 0.2862, + "step": 5362 + }, + { + "epoch": 14.980446927374302, + "grad_norm": 0.5706510543823242, + "learning_rate": 1.7963128491620112e-05, + "loss": 0.1899, + "step": 5363 + }, + { + "epoch": 14.983240223463687, + "grad_norm": 1.3359463214874268, + "learning_rate": 1.7966480446927377e-05, + "loss": 0.1848, + "step": 5364 + }, + { + "epoch": 14.986033519553073, + "grad_norm": 0.620718777179718, + "learning_rate": 1.7969832402234638e-05, + "loss": 0.2062, + "step": 5365 + }, + { + "epoch": 14.988826815642458, + "grad_norm": 1.3796255588531494, + "learning_rate": 1.79731843575419e-05, + "loss": 0.2279, + "step": 5366 + }, + { + "epoch": 14.991620111731844, + "grad_norm": 0.8691429495811462, + "learning_rate": 1.7976536312849164e-05, + "loss": 0.2016, + "step": 5367 + }, + { + "epoch": 14.994413407821229, + "grad_norm": 0.6881114840507507, + "learning_rate": 1.7979888268156426e-05, + "loss": 0.1846, + "step": 5368 + }, + { + "epoch": 14.997206703910614, + "grad_norm": 1.40552818775177, + "learning_rate": 1.7983240223463687e-05, + "loss": 0.1905, + "step": 5369 + }, + { + "epoch": 15.0, + "grad_norm": 2.631906032562256, + "learning_rate": 1.798659217877095e-05, + "loss": 0.2423, + "step": 5370 + }, + { + "epoch": 15.002793296089385, + "grad_norm": 0.5308542251586914, + "learning_rate": 1.7989944134078213e-05, + "loss": 0.1802, + "step": 5371 + }, + { + "epoch": 15.005586592178771, + "grad_norm": 0.5721537470817566, + "learning_rate": 1.7993296089385474e-05, + "loss": 0.2032, + "step": 5372 + }, + { + "epoch": 15.008379888268156, + "grad_norm": 0.7849298119544983, + "learning_rate": 1.799664804469274e-05, + "loss": 0.1786, + "step": 5373 + }, + { + "epoch": 15.011173184357542, + "grad_norm": 0.6399182081222534, + "learning_rate": 1.8e-05, + "loss": 0.1948, + "step": 5374 + }, + { + "epoch": 15.013966480446927, + "grad_norm": 1.9500199556350708, + "learning_rate": 1.8003351955307265e-05, + "loss": 0.1796, + "step": 5375 + }, + { + "epoch": 15.016759776536313, + "grad_norm": 0.7855203151702881, + "learning_rate": 1.8006703910614527e-05, + "loss": 0.1857, + "step": 5376 + }, + { + "epoch": 15.019553072625698, + "grad_norm": 0.7264398336410522, + "learning_rate": 1.8010055865921788e-05, + "loss": 0.1903, + "step": 5377 + }, + { + "epoch": 15.022346368715084, + "grad_norm": 0.787211000919342, + "learning_rate": 1.8013407821229053e-05, + "loss": 0.1646, + "step": 5378 + }, + { + "epoch": 15.025139664804469, + "grad_norm": 2.773928642272949, + "learning_rate": 1.8016759776536314e-05, + "loss": 0.1498, + "step": 5379 + }, + { + "epoch": 15.027932960893855, + "grad_norm": 1.348749041557312, + "learning_rate": 1.8020111731843575e-05, + "loss": 0.2034, + "step": 5380 + }, + { + "epoch": 15.03072625698324, + "grad_norm": 0.8884660601615906, + "learning_rate": 1.802346368715084e-05, + "loss": 0.1859, + "step": 5381 + }, + { + "epoch": 15.033519553072626, + "grad_norm": 0.7512675523757935, + "learning_rate": 1.80268156424581e-05, + "loss": 0.1734, + "step": 5382 + }, + { + "epoch": 15.03631284916201, + "grad_norm": 1.231083631515503, + "learning_rate": 1.8030167597765363e-05, + "loss": 0.1825, + "step": 5383 + }, + { + "epoch": 15.039106145251397, + "grad_norm": 0.7027963399887085, + "learning_rate": 1.8033519553072627e-05, + "loss": 0.1606, + "step": 5384 + }, + { + "epoch": 15.041899441340782, + "grad_norm": 0.7291545867919922, + "learning_rate": 1.803687150837989e-05, + "loss": 0.1353, + "step": 5385 + }, + { + "epoch": 15.044692737430168, + "grad_norm": 0.97948157787323, + "learning_rate": 1.804022346368715e-05, + "loss": 0.1639, + "step": 5386 + }, + { + "epoch": 15.047486033519553, + "grad_norm": 0.7278848886489868, + "learning_rate": 1.8043575418994415e-05, + "loss": 0.2067, + "step": 5387 + }, + { + "epoch": 15.050279329608939, + "grad_norm": 0.7209067344665527, + "learning_rate": 1.8046927374301676e-05, + "loss": 0.1779, + "step": 5388 + }, + { + "epoch": 15.053072625698324, + "grad_norm": 0.8602139353752136, + "learning_rate": 1.8050279329608938e-05, + "loss": 0.1971, + "step": 5389 + }, + { + "epoch": 15.05586592178771, + "grad_norm": 1.2403377294540405, + "learning_rate": 1.8053631284916202e-05, + "loss": 0.1895, + "step": 5390 + }, + { + "epoch": 15.058659217877095, + "grad_norm": 0.9090428948402405, + "learning_rate": 1.8056983240223464e-05, + "loss": 0.2019, + "step": 5391 + }, + { + "epoch": 15.061452513966481, + "grad_norm": 1.5280709266662598, + "learning_rate": 1.806033519553073e-05, + "loss": 0.1996, + "step": 5392 + }, + { + "epoch": 15.064245810055866, + "grad_norm": 1.1440225839614868, + "learning_rate": 1.806368715083799e-05, + "loss": 0.1929, + "step": 5393 + }, + { + "epoch": 15.067039106145252, + "grad_norm": 1.418965458869934, + "learning_rate": 1.806703910614525e-05, + "loss": 0.1883, + "step": 5394 + }, + { + "epoch": 15.069832402234637, + "grad_norm": 1.5868462324142456, + "learning_rate": 1.8070391061452516e-05, + "loss": 0.253, + "step": 5395 + }, + { + "epoch": 15.072625698324023, + "grad_norm": 0.40300965309143066, + "learning_rate": 1.8073743016759777e-05, + "loss": 0.1796, + "step": 5396 + }, + { + "epoch": 15.075418994413408, + "grad_norm": 0.6423327326774597, + "learning_rate": 1.807709497206704e-05, + "loss": 0.1932, + "step": 5397 + }, + { + "epoch": 15.078212290502794, + "grad_norm": 1.6926259994506836, + "learning_rate": 1.8080446927374303e-05, + "loss": 0.1932, + "step": 5398 + }, + { + "epoch": 15.081005586592179, + "grad_norm": 0.9475044012069702, + "learning_rate": 1.8083798882681565e-05, + "loss": 0.1987, + "step": 5399 + }, + { + "epoch": 15.083798882681565, + "grad_norm": 0.6857426166534424, + "learning_rate": 1.8087150837988826e-05, + "loss": 0.2019, + "step": 5400 + }, + { + "epoch": 15.08659217877095, + "grad_norm": 0.6484384536743164, + "learning_rate": 1.809050279329609e-05, + "loss": 0.1682, + "step": 5401 + }, + { + "epoch": 15.089385474860336, + "grad_norm": 0.5256891846656799, + "learning_rate": 1.8093854748603352e-05, + "loss": 0.1798, + "step": 5402 + }, + { + "epoch": 15.09217877094972, + "grad_norm": 1.1072635650634766, + "learning_rate": 1.8097206703910613e-05, + "loss": 0.1917, + "step": 5403 + }, + { + "epoch": 15.094972067039107, + "grad_norm": 0.7472240924835205, + "learning_rate": 1.8100558659217878e-05, + "loss": 0.2024, + "step": 5404 + }, + { + "epoch": 15.097765363128492, + "grad_norm": 0.7861294150352478, + "learning_rate": 1.810391061452514e-05, + "loss": 0.1967, + "step": 5405 + }, + { + "epoch": 15.100558659217878, + "grad_norm": 1.0035423040390015, + "learning_rate": 1.81072625698324e-05, + "loss": 0.1722, + "step": 5406 + }, + { + "epoch": 15.103351955307263, + "grad_norm": 0.8701342344284058, + "learning_rate": 1.8110614525139666e-05, + "loss": 0.1857, + "step": 5407 + }, + { + "epoch": 15.106145251396647, + "grad_norm": 0.7325621843338013, + "learning_rate": 1.8113966480446927e-05, + "loss": 0.1728, + "step": 5408 + }, + { + "epoch": 15.108938547486034, + "grad_norm": 3.470393419265747, + "learning_rate": 1.8117318435754188e-05, + "loss": 0.1926, + "step": 5409 + }, + { + "epoch": 15.111731843575418, + "grad_norm": 0.8216171264648438, + "learning_rate": 1.8120670391061453e-05, + "loss": 0.1574, + "step": 5410 + }, + { + "epoch": 15.114525139664805, + "grad_norm": 0.7630009055137634, + "learning_rate": 1.8124022346368714e-05, + "loss": 0.1778, + "step": 5411 + }, + { + "epoch": 15.11731843575419, + "grad_norm": 1.076061725616455, + "learning_rate": 1.812737430167598e-05, + "loss": 0.1918, + "step": 5412 + }, + { + "epoch": 15.120111731843576, + "grad_norm": 0.9492613077163696, + "learning_rate": 1.813072625698324e-05, + "loss": 0.1671, + "step": 5413 + }, + { + "epoch": 15.12290502793296, + "grad_norm": 0.9484812021255493, + "learning_rate": 1.8134078212290502e-05, + "loss": 0.1875, + "step": 5414 + }, + { + "epoch": 15.125698324022347, + "grad_norm": 1.6282764673233032, + "learning_rate": 1.8137430167597766e-05, + "loss": 0.1735, + "step": 5415 + }, + { + "epoch": 15.128491620111731, + "grad_norm": 1.3577158451080322, + "learning_rate": 1.8140782122905028e-05, + "loss": 0.1813, + "step": 5416 + }, + { + "epoch": 15.131284916201118, + "grad_norm": 1.1550276279449463, + "learning_rate": 1.814413407821229e-05, + "loss": 0.2115, + "step": 5417 + }, + { + "epoch": 15.134078212290502, + "grad_norm": 1.0994523763656616, + "learning_rate": 1.8147486033519554e-05, + "loss": 0.2137, + "step": 5418 + }, + { + "epoch": 15.136871508379889, + "grad_norm": 1.0957247018814087, + "learning_rate": 1.8150837988826815e-05, + "loss": 0.2262, + "step": 5419 + }, + { + "epoch": 15.139664804469273, + "grad_norm": 2.317291021347046, + "learning_rate": 1.8154189944134077e-05, + "loss": 0.2595, + "step": 5420 + }, + { + "epoch": 15.14245810055866, + "grad_norm": 0.5051766633987427, + "learning_rate": 1.815754189944134e-05, + "loss": 0.2026, + "step": 5421 + }, + { + "epoch": 15.145251396648044, + "grad_norm": 0.511879026889801, + "learning_rate": 1.8160893854748603e-05, + "loss": 0.1515, + "step": 5422 + }, + { + "epoch": 15.14804469273743, + "grad_norm": 0.48539412021636963, + "learning_rate": 1.8164245810055864e-05, + "loss": 0.1599, + "step": 5423 + }, + { + "epoch": 15.150837988826815, + "grad_norm": 0.6394636034965515, + "learning_rate": 1.816759776536313e-05, + "loss": 0.1765, + "step": 5424 + }, + { + "epoch": 15.153631284916202, + "grad_norm": 1.4280989170074463, + "learning_rate": 1.817094972067039e-05, + "loss": 0.2373, + "step": 5425 + }, + { + "epoch": 15.156424581005586, + "grad_norm": 0.6656328439712524, + "learning_rate": 1.817430167597765e-05, + "loss": 0.2005, + "step": 5426 + }, + { + "epoch": 15.159217877094973, + "grad_norm": 3.7945075035095215, + "learning_rate": 1.8177653631284916e-05, + "loss": 0.1991, + "step": 5427 + }, + { + "epoch": 15.162011173184357, + "grad_norm": 0.735578179359436, + "learning_rate": 1.8181005586592177e-05, + "loss": 0.1648, + "step": 5428 + }, + { + "epoch": 15.164804469273744, + "grad_norm": 1.4410291910171509, + "learning_rate": 1.8184357541899442e-05, + "loss": 0.176, + "step": 5429 + }, + { + "epoch": 15.167597765363128, + "grad_norm": 0.544895350933075, + "learning_rate": 1.8187709497206704e-05, + "loss": 0.1933, + "step": 5430 + }, + { + "epoch": 15.170391061452515, + "grad_norm": 0.8057953715324402, + "learning_rate": 1.8191061452513965e-05, + "loss": 0.1918, + "step": 5431 + }, + { + "epoch": 15.1731843575419, + "grad_norm": 0.9053359627723694, + "learning_rate": 1.819441340782123e-05, + "loss": 0.2115, + "step": 5432 + }, + { + "epoch": 15.175977653631286, + "grad_norm": 0.9749478101730347, + "learning_rate": 1.819776536312849e-05, + "loss": 0.162, + "step": 5433 + }, + { + "epoch": 15.17877094972067, + "grad_norm": 0.6820585131645203, + "learning_rate": 1.8201117318435752e-05, + "loss": 0.1827, + "step": 5434 + }, + { + "epoch": 15.181564245810057, + "grad_norm": 0.7356789112091064, + "learning_rate": 1.8204469273743017e-05, + "loss": 0.2073, + "step": 5435 + }, + { + "epoch": 15.184357541899441, + "grad_norm": 0.9775916934013367, + "learning_rate": 1.820782122905028e-05, + "loss": 0.1758, + "step": 5436 + }, + { + "epoch": 15.187150837988828, + "grad_norm": 1.0525225400924683, + "learning_rate": 1.821117318435754e-05, + "loss": 0.2049, + "step": 5437 + }, + { + "epoch": 15.189944134078212, + "grad_norm": 2.5947039127349854, + "learning_rate": 1.8214525139664804e-05, + "loss": 0.1867, + "step": 5438 + }, + { + "epoch": 15.192737430167599, + "grad_norm": 0.9716818332672119, + "learning_rate": 1.8217877094972066e-05, + "loss": 0.2203, + "step": 5439 + }, + { + "epoch": 15.195530726256983, + "grad_norm": 1.1465822458267212, + "learning_rate": 1.8221229050279327e-05, + "loss": 0.1858, + "step": 5440 + }, + { + "epoch": 15.19832402234637, + "grad_norm": 1.1874370574951172, + "learning_rate": 1.8224581005586592e-05, + "loss": 0.1763, + "step": 5441 + }, + { + "epoch": 15.201117318435754, + "grad_norm": 4.733560085296631, + "learning_rate": 1.8227932960893857e-05, + "loss": 0.2193, + "step": 5442 + }, + { + "epoch": 15.203910614525139, + "grad_norm": 1.1498299837112427, + "learning_rate": 1.8231284916201118e-05, + "loss": 0.2046, + "step": 5443 + }, + { + "epoch": 15.206703910614525, + "grad_norm": 1.0900349617004395, + "learning_rate": 1.8234636871508383e-05, + "loss": 0.2107, + "step": 5444 + }, + { + "epoch": 15.20949720670391, + "grad_norm": 6.262004852294922, + "learning_rate": 1.8237988826815644e-05, + "loss": 0.2911, + "step": 5445 + }, + { + "epoch": 15.212290502793296, + "grad_norm": 0.5850556492805481, + "learning_rate": 1.8241340782122905e-05, + "loss": 0.1888, + "step": 5446 + }, + { + "epoch": 15.21508379888268, + "grad_norm": 0.5652551054954529, + "learning_rate": 1.824469273743017e-05, + "loss": 0.1784, + "step": 5447 + }, + { + "epoch": 15.217877094972067, + "grad_norm": 0.7192457318305969, + "learning_rate": 1.824804469273743e-05, + "loss": 0.1735, + "step": 5448 + }, + { + "epoch": 15.220670391061452, + "grad_norm": 0.7188121676445007, + "learning_rate": 1.8251396648044696e-05, + "loss": 0.2095, + "step": 5449 + }, + { + "epoch": 15.223463687150838, + "grad_norm": 1.2235159873962402, + "learning_rate": 1.8254748603351958e-05, + "loss": 0.2059, + "step": 5450 + }, + { + "epoch": 15.226256983240223, + "grad_norm": 0.5067228078842163, + "learning_rate": 1.825810055865922e-05, + "loss": 0.1782, + "step": 5451 + }, + { + "epoch": 15.22905027932961, + "grad_norm": 0.9458021521568298, + "learning_rate": 1.8261452513966484e-05, + "loss": 0.2072, + "step": 5452 + }, + { + "epoch": 15.231843575418994, + "grad_norm": 1.037036657333374, + "learning_rate": 1.8264804469273745e-05, + "loss": 0.1553, + "step": 5453 + }, + { + "epoch": 15.23463687150838, + "grad_norm": 0.9713526368141174, + "learning_rate": 1.8268156424581006e-05, + "loss": 0.185, + "step": 5454 + }, + { + "epoch": 15.237430167597765, + "grad_norm": 0.617790699005127, + "learning_rate": 1.827150837988827e-05, + "loss": 0.1759, + "step": 5455 + }, + { + "epoch": 15.240223463687151, + "grad_norm": 0.8572084903717041, + "learning_rate": 1.8274860335195532e-05, + "loss": 0.1705, + "step": 5456 + }, + { + "epoch": 15.243016759776536, + "grad_norm": 0.6129744648933411, + "learning_rate": 1.8278212290502794e-05, + "loss": 0.1502, + "step": 5457 + }, + { + "epoch": 15.245810055865922, + "grad_norm": 0.7205598950386047, + "learning_rate": 1.828156424581006e-05, + "loss": 0.1842, + "step": 5458 + }, + { + "epoch": 15.248603351955307, + "grad_norm": 0.6877211928367615, + "learning_rate": 1.828491620111732e-05, + "loss": 0.1729, + "step": 5459 + }, + { + "epoch": 15.251396648044693, + "grad_norm": 0.9356571435928345, + "learning_rate": 1.828826815642458e-05, + "loss": 0.1679, + "step": 5460 + }, + { + "epoch": 15.254189944134078, + "grad_norm": 1.0339823961257935, + "learning_rate": 1.8291620111731846e-05, + "loss": 0.205, + "step": 5461 + }, + { + "epoch": 15.256983240223464, + "grad_norm": 0.887626051902771, + "learning_rate": 1.8294972067039107e-05, + "loss": 0.1499, + "step": 5462 + }, + { + "epoch": 15.259776536312849, + "grad_norm": 1.4485483169555664, + "learning_rate": 1.829832402234637e-05, + "loss": 0.1712, + "step": 5463 + }, + { + "epoch": 15.262569832402235, + "grad_norm": 1.066403865814209, + "learning_rate": 1.8301675977653633e-05, + "loss": 0.1915, + "step": 5464 + }, + { + "epoch": 15.26536312849162, + "grad_norm": 1.1943721771240234, + "learning_rate": 1.8305027932960895e-05, + "loss": 0.2154, + "step": 5465 + }, + { + "epoch": 15.268156424581006, + "grad_norm": 1.1577739715576172, + "learning_rate": 1.8308379888268156e-05, + "loss": 0.2067, + "step": 5466 + }, + { + "epoch": 15.27094972067039, + "grad_norm": 2.0230374336242676, + "learning_rate": 1.831173184357542e-05, + "loss": 0.1781, + "step": 5467 + }, + { + "epoch": 15.273743016759777, + "grad_norm": 1.2303073406219482, + "learning_rate": 1.8315083798882682e-05, + "loss": 0.2252, + "step": 5468 + }, + { + "epoch": 15.276536312849162, + "grad_norm": 1.1662582159042358, + "learning_rate": 1.8318435754189947e-05, + "loss": 0.1984, + "step": 5469 + }, + { + "epoch": 15.279329608938548, + "grad_norm": 2.26593279838562, + "learning_rate": 1.8321787709497208e-05, + "loss": 0.2861, + "step": 5470 + }, + { + "epoch": 15.282122905027933, + "grad_norm": 0.7635455131530762, + "learning_rate": 1.832513966480447e-05, + "loss": 0.1986, + "step": 5471 + }, + { + "epoch": 15.28491620111732, + "grad_norm": 0.5377920866012573, + "learning_rate": 1.8328491620111734e-05, + "loss": 0.1919, + "step": 5472 + }, + { + "epoch": 15.287709497206704, + "grad_norm": 0.6691102385520935, + "learning_rate": 1.8331843575418996e-05, + "loss": 0.2073, + "step": 5473 + }, + { + "epoch": 15.29050279329609, + "grad_norm": 0.8559274077415466, + "learning_rate": 1.8335195530726257e-05, + "loss": 0.1915, + "step": 5474 + }, + { + "epoch": 15.293296089385475, + "grad_norm": 1.0646377801895142, + "learning_rate": 1.833854748603352e-05, + "loss": 0.2303, + "step": 5475 + }, + { + "epoch": 15.296089385474861, + "grad_norm": 0.8533020615577698, + "learning_rate": 1.8341899441340783e-05, + "loss": 0.2035, + "step": 5476 + }, + { + "epoch": 15.298882681564246, + "grad_norm": 0.7001170516014099, + "learning_rate": 1.8345251396648044e-05, + "loss": 0.1868, + "step": 5477 + }, + { + "epoch": 15.30167597765363, + "grad_norm": 0.7070342302322388, + "learning_rate": 1.834860335195531e-05, + "loss": 0.1911, + "step": 5478 + }, + { + "epoch": 15.304469273743017, + "grad_norm": 0.8622980713844299, + "learning_rate": 1.835195530726257e-05, + "loss": 0.1592, + "step": 5479 + }, + { + "epoch": 15.307262569832401, + "grad_norm": 0.8719980716705322, + "learning_rate": 1.8355307262569832e-05, + "loss": 0.2009, + "step": 5480 + }, + { + "epoch": 15.310055865921788, + "grad_norm": 0.745566189289093, + "learning_rate": 1.8358659217877097e-05, + "loss": 0.1875, + "step": 5481 + }, + { + "epoch": 15.312849162011172, + "grad_norm": 0.9547754526138306, + "learning_rate": 1.8362011173184358e-05, + "loss": 0.1711, + "step": 5482 + }, + { + "epoch": 15.315642458100559, + "grad_norm": 0.9322498440742493, + "learning_rate": 1.836536312849162e-05, + "loss": 0.208, + "step": 5483 + }, + { + "epoch": 15.318435754189943, + "grad_norm": 1.2503743171691895, + "learning_rate": 1.8368715083798884e-05, + "loss": 0.1838, + "step": 5484 + }, + { + "epoch": 15.32122905027933, + "grad_norm": 0.6744921803474426, + "learning_rate": 1.8372067039106145e-05, + "loss": 0.1854, + "step": 5485 + }, + { + "epoch": 15.324022346368714, + "grad_norm": 1.1249297857284546, + "learning_rate": 1.837541899441341e-05, + "loss": 0.1883, + "step": 5486 + }, + { + "epoch": 15.3268156424581, + "grad_norm": 1.1251689195632935, + "learning_rate": 1.837877094972067e-05, + "loss": 0.1806, + "step": 5487 + }, + { + "epoch": 15.329608938547485, + "grad_norm": 1.284307599067688, + "learning_rate": 1.8382122905027933e-05, + "loss": 0.2154, + "step": 5488 + }, + { + "epoch": 15.332402234636872, + "grad_norm": 1.2216941118240356, + "learning_rate": 1.8385474860335197e-05, + "loss": 0.1823, + "step": 5489 + }, + { + "epoch": 15.335195530726256, + "grad_norm": 0.9642823934555054, + "learning_rate": 1.838882681564246e-05, + "loss": 0.1755, + "step": 5490 + }, + { + "epoch": 15.337988826815643, + "grad_norm": 1.2617616653442383, + "learning_rate": 1.839217877094972e-05, + "loss": 0.183, + "step": 5491 + }, + { + "epoch": 15.340782122905027, + "grad_norm": 3.3879714012145996, + "learning_rate": 1.8395530726256985e-05, + "loss": 0.1741, + "step": 5492 + }, + { + "epoch": 15.343575418994414, + "grad_norm": 1.0110509395599365, + "learning_rate": 1.8398882681564246e-05, + "loss": 0.1674, + "step": 5493 + }, + { + "epoch": 15.346368715083798, + "grad_norm": 1.3865855932235718, + "learning_rate": 1.8402234636871508e-05, + "loss": 0.1953, + "step": 5494 + }, + { + "epoch": 15.349162011173185, + "grad_norm": 3.7922306060791016, + "learning_rate": 1.8405586592178772e-05, + "loss": 0.3002, + "step": 5495 + }, + { + "epoch": 15.35195530726257, + "grad_norm": 0.9885417819023132, + "learning_rate": 1.8408938547486034e-05, + "loss": 0.2242, + "step": 5496 + }, + { + "epoch": 15.354748603351956, + "grad_norm": 0.5790302753448486, + "learning_rate": 1.8412290502793295e-05, + "loss": 0.1825, + "step": 5497 + }, + { + "epoch": 15.35754189944134, + "grad_norm": 0.6260929703712463, + "learning_rate": 1.841564245810056e-05, + "loss": 0.1981, + "step": 5498 + }, + { + "epoch": 15.360335195530727, + "grad_norm": 0.8677269220352173, + "learning_rate": 1.841899441340782e-05, + "loss": 0.2172, + "step": 5499 + }, + { + "epoch": 15.363128491620111, + "grad_norm": 0.7061383128166199, + "learning_rate": 1.8422346368715082e-05, + "loss": 0.1781, + "step": 5500 + }, + { + "epoch": 15.365921787709498, + "grad_norm": 0.7074316740036011, + "learning_rate": 1.8425698324022347e-05, + "loss": 0.1702, + "step": 5501 + }, + { + "epoch": 15.368715083798882, + "grad_norm": 0.6627846360206604, + "learning_rate": 1.842905027932961e-05, + "loss": 0.1747, + "step": 5502 + }, + { + "epoch": 15.371508379888269, + "grad_norm": 1.631402611732483, + "learning_rate": 1.843240223463687e-05, + "loss": 0.1931, + "step": 5503 + }, + { + "epoch": 15.374301675977653, + "grad_norm": 0.6822512745857239, + "learning_rate": 1.8435754189944135e-05, + "loss": 0.1742, + "step": 5504 + }, + { + "epoch": 15.37709497206704, + "grad_norm": 0.9383516311645508, + "learning_rate": 1.8439106145251396e-05, + "loss": 0.2042, + "step": 5505 + }, + { + "epoch": 15.379888268156424, + "grad_norm": 1.140408992767334, + "learning_rate": 1.844245810055866e-05, + "loss": 0.1822, + "step": 5506 + }, + { + "epoch": 15.38268156424581, + "grad_norm": 0.7207919359207153, + "learning_rate": 1.8445810055865922e-05, + "loss": 0.1531, + "step": 5507 + }, + { + "epoch": 15.385474860335195, + "grad_norm": 1.45779550075531, + "learning_rate": 1.8449162011173183e-05, + "loss": 0.1741, + "step": 5508 + }, + { + "epoch": 15.388268156424582, + "grad_norm": 0.8823292255401611, + "learning_rate": 1.8452513966480448e-05, + "loss": 0.23, + "step": 5509 + }, + { + "epoch": 15.391061452513966, + "grad_norm": 1.0888893604278564, + "learning_rate": 1.845586592178771e-05, + "loss": 0.1701, + "step": 5510 + }, + { + "epoch": 15.393854748603353, + "grad_norm": 1.133079171180725, + "learning_rate": 1.845921787709497e-05, + "loss": 0.1604, + "step": 5511 + }, + { + "epoch": 15.396648044692737, + "grad_norm": 0.8637135028839111, + "learning_rate": 1.8462569832402236e-05, + "loss": 0.1761, + "step": 5512 + }, + { + "epoch": 15.399441340782122, + "grad_norm": 1.0721571445465088, + "learning_rate": 1.8465921787709497e-05, + "loss": 0.1729, + "step": 5513 + }, + { + "epoch": 15.402234636871508, + "grad_norm": 0.9079535603523254, + "learning_rate": 1.8469273743016758e-05, + "loss": 0.1637, + "step": 5514 + }, + { + "epoch": 15.405027932960893, + "grad_norm": 0.8566271662712097, + "learning_rate": 1.8472625698324023e-05, + "loss": 0.1738, + "step": 5515 + }, + { + "epoch": 15.40782122905028, + "grad_norm": 3.9447529315948486, + "learning_rate": 1.8475977653631284e-05, + "loss": 0.1581, + "step": 5516 + }, + { + "epoch": 15.410614525139664, + "grad_norm": 1.6298120021820068, + "learning_rate": 1.8479329608938546e-05, + "loss": 0.2042, + "step": 5517 + }, + { + "epoch": 15.41340782122905, + "grad_norm": 1.416832447052002, + "learning_rate": 1.848268156424581e-05, + "loss": 0.2186, + "step": 5518 + }, + { + "epoch": 15.416201117318435, + "grad_norm": 2.0731589794158936, + "learning_rate": 1.8486033519553072e-05, + "loss": 0.2916, + "step": 5519 + }, + { + "epoch": 15.418994413407821, + "grad_norm": 3.6909964084625244, + "learning_rate": 1.8489385474860333e-05, + "loss": 0.2751, + "step": 5520 + }, + { + "epoch": 15.421787709497206, + "grad_norm": 0.6786301732063293, + "learning_rate": 1.8492737430167598e-05, + "loss": 0.1902, + "step": 5521 + }, + { + "epoch": 15.424581005586592, + "grad_norm": 0.7497189044952393, + "learning_rate": 1.849608938547486e-05, + "loss": 0.216, + "step": 5522 + }, + { + "epoch": 15.427374301675977, + "grad_norm": 0.6891225576400757, + "learning_rate": 1.8499441340782124e-05, + "loss": 0.1909, + "step": 5523 + }, + { + "epoch": 15.430167597765363, + "grad_norm": 0.6230394840240479, + "learning_rate": 1.8502793296089385e-05, + "loss": 0.2236, + "step": 5524 + }, + { + "epoch": 15.432960893854748, + "grad_norm": 0.7640438675880432, + "learning_rate": 1.8506145251396647e-05, + "loss": 0.2001, + "step": 5525 + }, + { + "epoch": 15.435754189944134, + "grad_norm": 0.7718290090560913, + "learning_rate": 1.850949720670391e-05, + "loss": 0.1803, + "step": 5526 + }, + { + "epoch": 15.438547486033519, + "grad_norm": 0.8632411956787109, + "learning_rate": 1.8512849162011173e-05, + "loss": 0.1622, + "step": 5527 + }, + { + "epoch": 15.441340782122905, + "grad_norm": 0.6519826650619507, + "learning_rate": 1.8516201117318434e-05, + "loss": 0.2014, + "step": 5528 + }, + { + "epoch": 15.44413407821229, + "grad_norm": 0.6199743151664734, + "learning_rate": 1.85195530726257e-05, + "loss": 0.1952, + "step": 5529 + }, + { + "epoch": 15.446927374301676, + "grad_norm": 0.9065355062484741, + "learning_rate": 1.852290502793296e-05, + "loss": 0.196, + "step": 5530 + }, + { + "epoch": 15.449720670391061, + "grad_norm": 0.876612663269043, + "learning_rate": 1.852625698324022e-05, + "loss": 0.1782, + "step": 5531 + }, + { + "epoch": 15.452513966480447, + "grad_norm": 0.7265151739120483, + "learning_rate": 1.8529608938547486e-05, + "loss": 0.1853, + "step": 5532 + }, + { + "epoch": 15.455307262569832, + "grad_norm": 1.0503625869750977, + "learning_rate": 1.8532960893854747e-05, + "loss": 0.1767, + "step": 5533 + }, + { + "epoch": 15.458100558659218, + "grad_norm": 0.7195703983306885, + "learning_rate": 1.853631284916201e-05, + "loss": 0.2021, + "step": 5534 + }, + { + "epoch": 15.460893854748603, + "grad_norm": 0.9026429057121277, + "learning_rate": 1.8539664804469274e-05, + "loss": 0.217, + "step": 5535 + }, + { + "epoch": 15.46368715083799, + "grad_norm": 0.8191381692886353, + "learning_rate": 1.8543016759776535e-05, + "loss": 0.2032, + "step": 5536 + }, + { + "epoch": 15.466480446927374, + "grad_norm": 0.6931017637252808, + "learning_rate": 1.85463687150838e-05, + "loss": 0.2064, + "step": 5537 + }, + { + "epoch": 15.46927374301676, + "grad_norm": 0.7283341884613037, + "learning_rate": 1.8549720670391064e-05, + "loss": 0.1749, + "step": 5538 + }, + { + "epoch": 15.472067039106145, + "grad_norm": 1.5280698537826538, + "learning_rate": 1.8553072625698326e-05, + "loss": 0.1714, + "step": 5539 + }, + { + "epoch": 15.474860335195531, + "grad_norm": 2.1137075424194336, + "learning_rate": 1.8556424581005587e-05, + "loss": 0.1975, + "step": 5540 + }, + { + "epoch": 15.477653631284916, + "grad_norm": 0.717642068862915, + "learning_rate": 1.8559776536312852e-05, + "loss": 0.1955, + "step": 5541 + }, + { + "epoch": 15.480446927374302, + "grad_norm": 2.4102370738983154, + "learning_rate": 1.8563128491620113e-05, + "loss": 0.2087, + "step": 5542 + }, + { + "epoch": 15.483240223463687, + "grad_norm": 1.8126271963119507, + "learning_rate": 1.8566480446927378e-05, + "loss": 0.202, + "step": 5543 + }, + { + "epoch": 15.486033519553073, + "grad_norm": 1.205686092376709, + "learning_rate": 1.856983240223464e-05, + "loss": 0.1978, + "step": 5544 + }, + { + "epoch": 15.488826815642458, + "grad_norm": 3.7466113567352295, + "learning_rate": 1.85731843575419e-05, + "loss": 0.2494, + "step": 5545 + }, + { + "epoch": 15.491620111731844, + "grad_norm": 0.5545952916145325, + "learning_rate": 1.8576536312849165e-05, + "loss": 0.1807, + "step": 5546 + }, + { + "epoch": 15.494413407821229, + "grad_norm": 0.5834225416183472, + "learning_rate": 1.8579888268156427e-05, + "loss": 0.218, + "step": 5547 + }, + { + "epoch": 15.497206703910614, + "grad_norm": 1.1797261238098145, + "learning_rate": 1.8583240223463688e-05, + "loss": 0.1818, + "step": 5548 + }, + { + "epoch": 15.5, + "grad_norm": 0.4823102056980133, + "learning_rate": 1.8586592178770953e-05, + "loss": 0.1558, + "step": 5549 + }, + { + "epoch": 15.502793296089386, + "grad_norm": 0.5513525605201721, + "learning_rate": 1.8589944134078214e-05, + "loss": 0.1887, + "step": 5550 + }, + { + "epoch": 15.505586592178771, + "grad_norm": 0.4745798110961914, + "learning_rate": 1.8593296089385475e-05, + "loss": 0.203, + "step": 5551 + }, + { + "epoch": 15.508379888268156, + "grad_norm": 0.6844167709350586, + "learning_rate": 1.859664804469274e-05, + "loss": 0.2113, + "step": 5552 + }, + { + "epoch": 15.511173184357542, + "grad_norm": 0.7085041403770447, + "learning_rate": 1.86e-05, + "loss": 0.2075, + "step": 5553 + }, + { + "epoch": 15.513966480446927, + "grad_norm": 0.6208303570747375, + "learning_rate": 1.8603351955307263e-05, + "loss": 0.1613, + "step": 5554 + }, + { + "epoch": 15.516759776536313, + "grad_norm": 0.6693980693817139, + "learning_rate": 1.8606703910614528e-05, + "loss": 0.1574, + "step": 5555 + }, + { + "epoch": 15.519553072625698, + "grad_norm": 0.8531777262687683, + "learning_rate": 1.861005586592179e-05, + "loss": 0.1822, + "step": 5556 + }, + { + "epoch": 15.522346368715084, + "grad_norm": 0.6678202748298645, + "learning_rate": 1.861340782122905e-05, + "loss": 0.2098, + "step": 5557 + }, + { + "epoch": 15.525139664804469, + "grad_norm": 0.9600443840026855, + "learning_rate": 1.8616759776536315e-05, + "loss": 0.1801, + "step": 5558 + }, + { + "epoch": 15.527932960893855, + "grad_norm": 0.9724751114845276, + "learning_rate": 1.8620111731843576e-05, + "loss": 0.178, + "step": 5559 + }, + { + "epoch": 15.53072625698324, + "grad_norm": 0.7174497246742249, + "learning_rate": 1.8623463687150838e-05, + "loss": 0.1606, + "step": 5560 + }, + { + "epoch": 15.533519553072626, + "grad_norm": 0.5869032144546509, + "learning_rate": 1.8626815642458102e-05, + "loss": 0.1464, + "step": 5561 + }, + { + "epoch": 15.53631284916201, + "grad_norm": 1.2017145156860352, + "learning_rate": 1.8630167597765364e-05, + "loss": 0.2116, + "step": 5562 + }, + { + "epoch": 15.539106145251397, + "grad_norm": 0.7745858430862427, + "learning_rate": 1.863351955307263e-05, + "loss": 0.1847, + "step": 5563 + }, + { + "epoch": 15.541899441340782, + "grad_norm": 3.0528323650360107, + "learning_rate": 1.863687150837989e-05, + "loss": 0.1909, + "step": 5564 + }, + { + "epoch": 15.544692737430168, + "grad_norm": 1.2136220932006836, + "learning_rate": 1.864022346368715e-05, + "loss": 0.2095, + "step": 5565 + }, + { + "epoch": 15.547486033519553, + "grad_norm": 1.3780443668365479, + "learning_rate": 1.8643575418994416e-05, + "loss": 0.1904, + "step": 5566 + }, + { + "epoch": 15.550279329608939, + "grad_norm": 0.9672235250473022, + "learning_rate": 1.8646927374301677e-05, + "loss": 0.1805, + "step": 5567 + }, + { + "epoch": 15.553072625698324, + "grad_norm": 2.1274423599243164, + "learning_rate": 1.865027932960894e-05, + "loss": 0.2017, + "step": 5568 + }, + { + "epoch": 15.55586592178771, + "grad_norm": 2.4930734634399414, + "learning_rate": 1.8653631284916203e-05, + "loss": 0.2518, + "step": 5569 + }, + { + "epoch": 15.558659217877095, + "grad_norm": 4.786435604095459, + "learning_rate": 1.8656983240223465e-05, + "loss": 0.3068, + "step": 5570 + }, + { + "epoch": 15.561452513966481, + "grad_norm": 0.8834460377693176, + "learning_rate": 1.8660335195530726e-05, + "loss": 0.2366, + "step": 5571 + }, + { + "epoch": 15.564245810055866, + "grad_norm": 0.6717724800109863, + "learning_rate": 1.866368715083799e-05, + "loss": 0.1951, + "step": 5572 + }, + { + "epoch": 15.567039106145252, + "grad_norm": 0.8890587091445923, + "learning_rate": 1.8667039106145252e-05, + "loss": 0.1738, + "step": 5573 + }, + { + "epoch": 15.569832402234637, + "grad_norm": 0.6614871025085449, + "learning_rate": 1.8670391061452513e-05, + "loss": 0.2178, + "step": 5574 + }, + { + "epoch": 15.572625698324023, + "grad_norm": 0.6587030291557312, + "learning_rate": 1.8673743016759778e-05, + "loss": 0.1983, + "step": 5575 + }, + { + "epoch": 15.575418994413408, + "grad_norm": 0.8449211716651917, + "learning_rate": 1.867709497206704e-05, + "loss": 0.1977, + "step": 5576 + }, + { + "epoch": 15.578212290502794, + "grad_norm": 0.7420019507408142, + "learning_rate": 1.86804469273743e-05, + "loss": 0.2128, + "step": 5577 + }, + { + "epoch": 15.581005586592179, + "grad_norm": 0.8243714570999146, + "learning_rate": 1.8683798882681566e-05, + "loss": 0.1862, + "step": 5578 + }, + { + "epoch": 15.583798882681565, + "grad_norm": 0.6111689805984497, + "learning_rate": 1.8687150837988827e-05, + "loss": 0.1668, + "step": 5579 + }, + { + "epoch": 15.58659217877095, + "grad_norm": 0.6741356253623962, + "learning_rate": 1.869050279329609e-05, + "loss": 0.1823, + "step": 5580 + }, + { + "epoch": 15.589385474860336, + "grad_norm": 1.5238929986953735, + "learning_rate": 1.8693854748603353e-05, + "loss": 0.1726, + "step": 5581 + }, + { + "epoch": 15.59217877094972, + "grad_norm": 2.5792007446289062, + "learning_rate": 1.8697206703910614e-05, + "loss": 0.1739, + "step": 5582 + }, + { + "epoch": 15.594972067039105, + "grad_norm": 0.6909595131874084, + "learning_rate": 1.870055865921788e-05, + "loss": 0.1655, + "step": 5583 + }, + { + "epoch": 15.597765363128492, + "grad_norm": 1.0268527269363403, + "learning_rate": 1.870391061452514e-05, + "loss": 0.1891, + "step": 5584 + }, + { + "epoch": 15.600558659217878, + "grad_norm": 1.1560301780700684, + "learning_rate": 1.8707262569832402e-05, + "loss": 0.1812, + "step": 5585 + }, + { + "epoch": 15.603351955307263, + "grad_norm": 0.9321133494377136, + "learning_rate": 1.8710614525139667e-05, + "loss": 0.1572, + "step": 5586 + }, + { + "epoch": 15.606145251396647, + "grad_norm": 0.7644237279891968, + "learning_rate": 1.8713966480446928e-05, + "loss": 0.1821, + "step": 5587 + }, + { + "epoch": 15.608938547486034, + "grad_norm": 0.7476121187210083, + "learning_rate": 1.871731843575419e-05, + "loss": 0.1735, + "step": 5588 + }, + { + "epoch": 15.611731843575418, + "grad_norm": 1.0261495113372803, + "learning_rate": 1.8720670391061454e-05, + "loss": 0.1787, + "step": 5589 + }, + { + "epoch": 15.614525139664805, + "grad_norm": 1.6085379123687744, + "learning_rate": 1.8724022346368715e-05, + "loss": 0.2235, + "step": 5590 + }, + { + "epoch": 15.61731843575419, + "grad_norm": 0.8130521774291992, + "learning_rate": 1.8727374301675977e-05, + "loss": 0.1741, + "step": 5591 + }, + { + "epoch": 15.620111731843576, + "grad_norm": 1.2220783233642578, + "learning_rate": 1.873072625698324e-05, + "loss": 0.1703, + "step": 5592 + }, + { + "epoch": 15.62290502793296, + "grad_norm": 2.862027645111084, + "learning_rate": 1.8734078212290503e-05, + "loss": 0.205, + "step": 5593 + }, + { + "epoch": 15.625698324022347, + "grad_norm": 3.4298853874206543, + "learning_rate": 1.8737430167597764e-05, + "loss": 0.2396, + "step": 5594 + }, + { + "epoch": 15.628491620111731, + "grad_norm": 1.3519026041030884, + "learning_rate": 1.874078212290503e-05, + "loss": 0.2664, + "step": 5595 + }, + { + "epoch": 15.631284916201118, + "grad_norm": 0.6109738945960999, + "learning_rate": 1.874413407821229e-05, + "loss": 0.2183, + "step": 5596 + }, + { + "epoch": 15.634078212290502, + "grad_norm": 0.6795789003372192, + "learning_rate": 1.874748603351955e-05, + "loss": 0.1885, + "step": 5597 + }, + { + "epoch": 15.636871508379889, + "grad_norm": 1.08349609375, + "learning_rate": 1.8750837988826816e-05, + "loss": 0.1797, + "step": 5598 + }, + { + "epoch": 15.639664804469273, + "grad_norm": 0.5678629875183105, + "learning_rate": 1.8754189944134078e-05, + "loss": 0.1655, + "step": 5599 + }, + { + "epoch": 15.64245810055866, + "grad_norm": 1.0905065536499023, + "learning_rate": 1.8757541899441342e-05, + "loss": 0.1871, + "step": 5600 + }, + { + "epoch": 15.645251396648044, + "grad_norm": 1.2619456052780151, + "learning_rate": 1.8760893854748604e-05, + "loss": 0.2009, + "step": 5601 + }, + { + "epoch": 15.64804469273743, + "grad_norm": 0.674611508846283, + "learning_rate": 1.8764245810055865e-05, + "loss": 0.1792, + "step": 5602 + }, + { + "epoch": 15.650837988826815, + "grad_norm": 0.8432426452636719, + "learning_rate": 1.876759776536313e-05, + "loss": 0.177, + "step": 5603 + }, + { + "epoch": 15.653631284916202, + "grad_norm": 0.6227656006813049, + "learning_rate": 1.877094972067039e-05, + "loss": 0.1704, + "step": 5604 + }, + { + "epoch": 15.656424581005586, + "grad_norm": 0.8813119530677795, + "learning_rate": 1.8774301675977652e-05, + "loss": 0.1779, + "step": 5605 + }, + { + "epoch": 15.659217877094973, + "grad_norm": 0.8616018891334534, + "learning_rate": 1.8777653631284917e-05, + "loss": 0.2312, + "step": 5606 + }, + { + "epoch": 15.662011173184357, + "grad_norm": 0.7511171102523804, + "learning_rate": 1.878100558659218e-05, + "loss": 0.1608, + "step": 5607 + }, + { + "epoch": 15.664804469273744, + "grad_norm": 2.769911527633667, + "learning_rate": 1.878435754189944e-05, + "loss": 0.1685, + "step": 5608 + }, + { + "epoch": 15.667597765363128, + "grad_norm": 0.701017439365387, + "learning_rate": 1.8787709497206705e-05, + "loss": 0.1872, + "step": 5609 + }, + { + "epoch": 15.670391061452515, + "grad_norm": 2.462855339050293, + "learning_rate": 1.8791061452513966e-05, + "loss": 0.2047, + "step": 5610 + }, + { + "epoch": 15.6731843575419, + "grad_norm": 1.3393617868423462, + "learning_rate": 1.8794413407821227e-05, + "loss": 0.2164, + "step": 5611 + }, + { + "epoch": 15.675977653631286, + "grad_norm": 0.9499475359916687, + "learning_rate": 1.8797765363128492e-05, + "loss": 0.1888, + "step": 5612 + }, + { + "epoch": 15.67877094972067, + "grad_norm": 1.7902191877365112, + "learning_rate": 1.8801117318435753e-05, + "loss": 0.2123, + "step": 5613 + }, + { + "epoch": 15.681564245810057, + "grad_norm": 1.1599262952804565, + "learning_rate": 1.8804469273743015e-05, + "loss": 0.1714, + "step": 5614 + }, + { + "epoch": 15.684357541899441, + "grad_norm": 1.1424977779388428, + "learning_rate": 1.880782122905028e-05, + "loss": 0.2056, + "step": 5615 + }, + { + "epoch": 15.687150837988828, + "grad_norm": 1.8632616996765137, + "learning_rate": 1.881117318435754e-05, + "loss": 0.1955, + "step": 5616 + }, + { + "epoch": 15.689944134078212, + "grad_norm": 1.2803829908370972, + "learning_rate": 1.8814525139664806e-05, + "loss": 0.2011, + "step": 5617 + }, + { + "epoch": 15.692737430167599, + "grad_norm": 1.2109342813491821, + "learning_rate": 1.8817877094972067e-05, + "loss": 0.213, + "step": 5618 + }, + { + "epoch": 15.695530726256983, + "grad_norm": 1.330417275428772, + "learning_rate": 1.8821229050279328e-05, + "loss": 0.1954, + "step": 5619 + }, + { + "epoch": 15.69832402234637, + "grad_norm": 3.471757173538208, + "learning_rate": 1.8824581005586593e-05, + "loss": 0.3511, + "step": 5620 + }, + { + "epoch": 15.701117318435754, + "grad_norm": 1.3445507287979126, + "learning_rate": 1.8827932960893854e-05, + "loss": 0.2034, + "step": 5621 + }, + { + "epoch": 15.703910614525139, + "grad_norm": 0.5125529766082764, + "learning_rate": 1.8831284916201116e-05, + "loss": 0.1701, + "step": 5622 + }, + { + "epoch": 15.706703910614525, + "grad_norm": 0.6577967405319214, + "learning_rate": 1.883463687150838e-05, + "loss": 0.231, + "step": 5623 + }, + { + "epoch": 15.70949720670391, + "grad_norm": 0.453673779964447, + "learning_rate": 1.8837988826815642e-05, + "loss": 0.1721, + "step": 5624 + }, + { + "epoch": 15.712290502793296, + "grad_norm": 0.9899497628211975, + "learning_rate": 1.8841340782122903e-05, + "loss": 0.195, + "step": 5625 + }, + { + "epoch": 15.71508379888268, + "grad_norm": 0.6947128176689148, + "learning_rate": 1.8844692737430168e-05, + "loss": 0.2125, + "step": 5626 + }, + { + "epoch": 15.717877094972067, + "grad_norm": 0.9814050197601318, + "learning_rate": 1.884804469273743e-05, + "loss": 0.2173, + "step": 5627 + }, + { + "epoch": 15.720670391061452, + "grad_norm": 0.5328679084777832, + "learning_rate": 1.885139664804469e-05, + "loss": 0.1719, + "step": 5628 + }, + { + "epoch": 15.723463687150838, + "grad_norm": 1.4907890558242798, + "learning_rate": 1.8854748603351955e-05, + "loss": 0.1763, + "step": 5629 + }, + { + "epoch": 15.726256983240223, + "grad_norm": 0.7542809844017029, + "learning_rate": 1.8858100558659217e-05, + "loss": 0.2416, + "step": 5630 + }, + { + "epoch": 15.72905027932961, + "grad_norm": 2.2017555236816406, + "learning_rate": 1.8861452513966478e-05, + "loss": 0.1778, + "step": 5631 + }, + { + "epoch": 15.731843575418994, + "grad_norm": 1.3984143733978271, + "learning_rate": 1.8864804469273743e-05, + "loss": 0.1846, + "step": 5632 + }, + { + "epoch": 15.73463687150838, + "grad_norm": 0.8952956795692444, + "learning_rate": 1.8868156424581004e-05, + "loss": 0.2077, + "step": 5633 + }, + { + "epoch": 15.737430167597765, + "grad_norm": 0.9598250985145569, + "learning_rate": 1.887150837988827e-05, + "loss": 0.2267, + "step": 5634 + }, + { + "epoch": 15.740223463687151, + "grad_norm": 0.5869441628456116, + "learning_rate": 1.8874860335195533e-05, + "loss": 0.1726, + "step": 5635 + }, + { + "epoch": 15.743016759776536, + "grad_norm": 0.8967300057411194, + "learning_rate": 1.8878212290502795e-05, + "loss": 0.179, + "step": 5636 + }, + { + "epoch": 15.745810055865922, + "grad_norm": 0.6964938044548035, + "learning_rate": 1.888156424581006e-05, + "loss": 0.1719, + "step": 5637 + }, + { + "epoch": 15.748603351955307, + "grad_norm": 1.0068968534469604, + "learning_rate": 1.888491620111732e-05, + "loss": 0.1791, + "step": 5638 + }, + { + "epoch": 15.751396648044693, + "grad_norm": 1.8715153932571411, + "learning_rate": 1.8888268156424582e-05, + "loss": 0.2045, + "step": 5639 + }, + { + "epoch": 15.754189944134078, + "grad_norm": 0.706832766532898, + "learning_rate": 1.8891620111731847e-05, + "loss": 0.1808, + "step": 5640 + }, + { + "epoch": 15.756983240223464, + "grad_norm": 1.6431876420974731, + "learning_rate": 1.8894972067039108e-05, + "loss": 0.1661, + "step": 5641 + }, + { + "epoch": 15.759776536312849, + "grad_norm": 1.4632755517959595, + "learning_rate": 1.889832402234637e-05, + "loss": 0.169, + "step": 5642 + }, + { + "epoch": 15.762569832402235, + "grad_norm": 0.9360888600349426, + "learning_rate": 1.8901675977653634e-05, + "loss": 0.1778, + "step": 5643 + }, + { + "epoch": 15.76536312849162, + "grad_norm": 1.5189536809921265, + "learning_rate": 1.8905027932960896e-05, + "loss": 0.2606, + "step": 5644 + }, + { + "epoch": 15.768156424581006, + "grad_norm": 3.374814033508301, + "learning_rate": 1.8908379888268157e-05, + "loss": 0.2566, + "step": 5645 + }, + { + "epoch": 15.77094972067039, + "grad_norm": 0.631033182144165, + "learning_rate": 1.8911731843575422e-05, + "loss": 0.2108, + "step": 5646 + }, + { + "epoch": 15.773743016759777, + "grad_norm": 0.832180380821228, + "learning_rate": 1.8915083798882683e-05, + "loss": 0.1671, + "step": 5647 + }, + { + "epoch": 15.776536312849162, + "grad_norm": 0.7850459218025208, + "learning_rate": 1.8918435754189944e-05, + "loss": 0.2248, + "step": 5648 + }, + { + "epoch": 15.779329608938548, + "grad_norm": 1.3332706689834595, + "learning_rate": 1.892178770949721e-05, + "loss": 0.1927, + "step": 5649 + }, + { + "epoch": 15.782122905027933, + "grad_norm": 0.726493239402771, + "learning_rate": 1.892513966480447e-05, + "loss": 0.1632, + "step": 5650 + }, + { + "epoch": 15.78491620111732, + "grad_norm": 0.7987567186355591, + "learning_rate": 1.8928491620111732e-05, + "loss": 0.1913, + "step": 5651 + }, + { + "epoch": 15.787709497206704, + "grad_norm": 0.7305020093917847, + "learning_rate": 1.8931843575418997e-05, + "loss": 0.2234, + "step": 5652 + }, + { + "epoch": 15.79050279329609, + "grad_norm": 1.2187258005142212, + "learning_rate": 1.8935195530726258e-05, + "loss": 0.2042, + "step": 5653 + }, + { + "epoch": 15.793296089385475, + "grad_norm": 0.9080995321273804, + "learning_rate": 1.893854748603352e-05, + "loss": 0.1899, + "step": 5654 + }, + { + "epoch": 15.796089385474861, + "grad_norm": 0.8676538467407227, + "learning_rate": 1.8941899441340784e-05, + "loss": 0.1875, + "step": 5655 + }, + { + "epoch": 15.798882681564246, + "grad_norm": 0.7441958785057068, + "learning_rate": 1.8945251396648045e-05, + "loss": 0.2121, + "step": 5656 + }, + { + "epoch": 15.80167597765363, + "grad_norm": 0.6223496198654175, + "learning_rate": 1.894860335195531e-05, + "loss": 0.1856, + "step": 5657 + }, + { + "epoch": 15.804469273743017, + "grad_norm": 0.984860897064209, + "learning_rate": 1.895195530726257e-05, + "loss": 0.1884, + "step": 5658 + }, + { + "epoch": 15.807262569832401, + "grad_norm": 0.7685349583625793, + "learning_rate": 1.8955307262569833e-05, + "loss": 0.1507, + "step": 5659 + }, + { + "epoch": 15.810055865921788, + "grad_norm": 0.8875697255134583, + "learning_rate": 1.8958659217877098e-05, + "loss": 0.193, + "step": 5660 + }, + { + "epoch": 15.812849162011172, + "grad_norm": 2.3011958599090576, + "learning_rate": 1.896201117318436e-05, + "loss": 0.1739, + "step": 5661 + }, + { + "epoch": 15.815642458100559, + "grad_norm": 1.4168593883514404, + "learning_rate": 1.896536312849162e-05, + "loss": 0.2065, + "step": 5662 + }, + { + "epoch": 15.818435754189943, + "grad_norm": 0.9812067747116089, + "learning_rate": 1.8968715083798885e-05, + "loss": 0.1655, + "step": 5663 + }, + { + "epoch": 15.82122905027933, + "grad_norm": 0.7923923134803772, + "learning_rate": 1.8972067039106146e-05, + "loss": 0.1772, + "step": 5664 + }, + { + "epoch": 15.824022346368714, + "grad_norm": 1.2436976432800293, + "learning_rate": 1.8975418994413408e-05, + "loss": 0.1859, + "step": 5665 + }, + { + "epoch": 15.8268156424581, + "grad_norm": 1.4572898149490356, + "learning_rate": 1.8978770949720672e-05, + "loss": 0.2577, + "step": 5666 + }, + { + "epoch": 15.829608938547485, + "grad_norm": 1.2125049829483032, + "learning_rate": 1.8982122905027934e-05, + "loss": 0.1969, + "step": 5667 + }, + { + "epoch": 15.832402234636872, + "grad_norm": 1.9431543350219727, + "learning_rate": 1.8985474860335195e-05, + "loss": 0.221, + "step": 5668 + }, + { + "epoch": 15.835195530726256, + "grad_norm": 1.5414739847183228, + "learning_rate": 1.898882681564246e-05, + "loss": 0.2308, + "step": 5669 + }, + { + "epoch": 15.837988826815643, + "grad_norm": 3.3252339363098145, + "learning_rate": 1.899217877094972e-05, + "loss": 0.2249, + "step": 5670 + }, + { + "epoch": 15.840782122905027, + "grad_norm": 0.47312262654304504, + "learning_rate": 1.8995530726256983e-05, + "loss": 0.1943, + "step": 5671 + }, + { + "epoch": 15.843575418994414, + "grad_norm": 0.6095109581947327, + "learning_rate": 1.8998882681564247e-05, + "loss": 0.1956, + "step": 5672 + }, + { + "epoch": 15.846368715083798, + "grad_norm": 0.546805202960968, + "learning_rate": 1.900223463687151e-05, + "loss": 0.193, + "step": 5673 + }, + { + "epoch": 15.849162011173185, + "grad_norm": 0.47789573669433594, + "learning_rate": 1.9005586592178773e-05, + "loss": 0.1683, + "step": 5674 + }, + { + "epoch": 15.85195530726257, + "grad_norm": 0.5789629817008972, + "learning_rate": 1.9008938547486035e-05, + "loss": 0.2008, + "step": 5675 + }, + { + "epoch": 15.854748603351956, + "grad_norm": 0.5238616466522217, + "learning_rate": 1.9012290502793296e-05, + "loss": 0.1724, + "step": 5676 + }, + { + "epoch": 15.85754189944134, + "grad_norm": 0.5786634683609009, + "learning_rate": 1.901564245810056e-05, + "loss": 0.1729, + "step": 5677 + }, + { + "epoch": 15.860335195530727, + "grad_norm": 0.8175095915794373, + "learning_rate": 1.9018994413407822e-05, + "loss": 0.1868, + "step": 5678 + }, + { + "epoch": 15.863128491620111, + "grad_norm": 0.7352153658866882, + "learning_rate": 1.9022346368715083e-05, + "loss": 0.1849, + "step": 5679 + }, + { + "epoch": 15.865921787709498, + "grad_norm": 0.7510939240455627, + "learning_rate": 1.9025698324022348e-05, + "loss": 0.2219, + "step": 5680 + }, + { + "epoch": 15.868715083798882, + "grad_norm": 0.623010516166687, + "learning_rate": 1.902905027932961e-05, + "loss": 0.165, + "step": 5681 + }, + { + "epoch": 15.871508379888269, + "grad_norm": 0.836879312992096, + "learning_rate": 1.903240223463687e-05, + "loss": 0.1829, + "step": 5682 + }, + { + "epoch": 15.874301675977653, + "grad_norm": 1.2018545866012573, + "learning_rate": 1.9035754189944136e-05, + "loss": 0.1793, + "step": 5683 + }, + { + "epoch": 15.87709497206704, + "grad_norm": 0.7207831144332886, + "learning_rate": 1.9039106145251397e-05, + "loss": 0.1849, + "step": 5684 + }, + { + "epoch": 15.879888268156424, + "grad_norm": 1.6215265989303589, + "learning_rate": 1.9042458100558658e-05, + "loss": 0.203, + "step": 5685 + }, + { + "epoch": 15.88268156424581, + "grad_norm": 1.235607624053955, + "learning_rate": 1.9045810055865923e-05, + "loss": 0.1701, + "step": 5686 + }, + { + "epoch": 15.885474860335195, + "grad_norm": 0.8148708343505859, + "learning_rate": 1.9049162011173184e-05, + "loss": 0.1745, + "step": 5687 + }, + { + "epoch": 15.888268156424582, + "grad_norm": 0.8926193118095398, + "learning_rate": 1.9052513966480446e-05, + "loss": 0.1869, + "step": 5688 + }, + { + "epoch": 15.891061452513966, + "grad_norm": 0.7312188744544983, + "learning_rate": 1.905586592178771e-05, + "loss": 0.1633, + "step": 5689 + }, + { + "epoch": 15.893854748603353, + "grad_norm": 3.0810177326202393, + "learning_rate": 1.9059217877094972e-05, + "loss": 0.2172, + "step": 5690 + }, + { + "epoch": 15.896648044692737, + "grad_norm": 2.0184378623962402, + "learning_rate": 1.9062569832402233e-05, + "loss": 0.2264, + "step": 5691 + }, + { + "epoch": 15.899441340782122, + "grad_norm": 0.9839572906494141, + "learning_rate": 1.9065921787709498e-05, + "loss": 0.1884, + "step": 5692 + }, + { + "epoch": 15.902234636871508, + "grad_norm": 1.3782804012298584, + "learning_rate": 1.906927374301676e-05, + "loss": 0.2069, + "step": 5693 + }, + { + "epoch": 15.905027932960895, + "grad_norm": 1.2050318717956543, + "learning_rate": 1.9072625698324024e-05, + "loss": 0.217, + "step": 5694 + }, + { + "epoch": 15.90782122905028, + "grad_norm": 4.225982666015625, + "learning_rate": 1.9075977653631285e-05, + "loss": 0.2902, + "step": 5695 + }, + { + "epoch": 15.910614525139664, + "grad_norm": 0.45696184039115906, + "learning_rate": 1.9079329608938547e-05, + "loss": 0.1831, + "step": 5696 + }, + { + "epoch": 15.91340782122905, + "grad_norm": 0.6545003652572632, + "learning_rate": 1.908268156424581e-05, + "loss": 0.1831, + "step": 5697 + }, + { + "epoch": 15.916201117318435, + "grad_norm": 0.6415268182754517, + "learning_rate": 1.9086033519553073e-05, + "loss": 0.226, + "step": 5698 + }, + { + "epoch": 15.918994413407821, + "grad_norm": 0.6492296457290649, + "learning_rate": 1.9089385474860334e-05, + "loss": 0.191, + "step": 5699 + }, + { + "epoch": 15.921787709497206, + "grad_norm": 0.49898919463157654, + "learning_rate": 1.90927374301676e-05, + "loss": 0.1825, + "step": 5700 + }, + { + "epoch": 15.924581005586592, + "grad_norm": 0.7702747583389282, + "learning_rate": 1.909608938547486e-05, + "loss": 0.1803, + "step": 5701 + }, + { + "epoch": 15.927374301675977, + "grad_norm": 2.6749520301818848, + "learning_rate": 1.909944134078212e-05, + "loss": 0.1666, + "step": 5702 + }, + { + "epoch": 15.930167597765363, + "grad_norm": 1.1427894830703735, + "learning_rate": 1.9102793296089386e-05, + "loss": 0.2147, + "step": 5703 + }, + { + "epoch": 15.932960893854748, + "grad_norm": 0.6308075189590454, + "learning_rate": 1.9106145251396648e-05, + "loss": 0.1639, + "step": 5704 + }, + { + "epoch": 15.935754189944134, + "grad_norm": 0.4957255423069, + "learning_rate": 1.910949720670391e-05, + "loss": 0.1892, + "step": 5705 + }, + { + "epoch": 15.938547486033519, + "grad_norm": 0.5553117394447327, + "learning_rate": 1.9112849162011174e-05, + "loss": 0.1841, + "step": 5706 + }, + { + "epoch": 15.941340782122905, + "grad_norm": 0.8668846487998962, + "learning_rate": 1.9116201117318435e-05, + "loss": 0.2099, + "step": 5707 + }, + { + "epoch": 15.94413407821229, + "grad_norm": 1.3744827508926392, + "learning_rate": 1.9119553072625696e-05, + "loss": 0.1756, + "step": 5708 + }, + { + "epoch": 15.946927374301676, + "grad_norm": 0.8144103288650513, + "learning_rate": 1.912290502793296e-05, + "loss": 0.1702, + "step": 5709 + }, + { + "epoch": 15.949720670391061, + "grad_norm": 0.695149302482605, + "learning_rate": 1.9126256983240222e-05, + "loss": 0.1611, + "step": 5710 + }, + { + "epoch": 15.952513966480447, + "grad_norm": 2.62773060798645, + "learning_rate": 1.9129608938547487e-05, + "loss": 0.2053, + "step": 5711 + }, + { + "epoch": 15.955307262569832, + "grad_norm": 1.2282671928405762, + "learning_rate": 1.913296089385475e-05, + "loss": 0.1929, + "step": 5712 + }, + { + "epoch": 15.958100558659218, + "grad_norm": 0.7325605154037476, + "learning_rate": 1.913631284916201e-05, + "loss": 0.1881, + "step": 5713 + }, + { + "epoch": 15.960893854748603, + "grad_norm": 1.5962591171264648, + "learning_rate": 1.9139664804469275e-05, + "loss": 0.1843, + "step": 5714 + }, + { + "epoch": 15.96368715083799, + "grad_norm": 1.1042903661727905, + "learning_rate": 1.9143016759776536e-05, + "loss": 0.1605, + "step": 5715 + }, + { + "epoch": 15.966480446927374, + "grad_norm": 1.0347912311553955, + "learning_rate": 1.9146368715083797e-05, + "loss": 0.1905, + "step": 5716 + }, + { + "epoch": 15.96927374301676, + "grad_norm": 2.876072406768799, + "learning_rate": 1.9149720670391062e-05, + "loss": 0.1964, + "step": 5717 + }, + { + "epoch": 15.972067039106145, + "grad_norm": 1.0975346565246582, + "learning_rate": 1.9153072625698323e-05, + "loss": 0.2059, + "step": 5718 + }, + { + "epoch": 15.974860335195531, + "grad_norm": 2.0710854530334473, + "learning_rate": 1.9156424581005585e-05, + "loss": 0.1955, + "step": 5719 + }, + { + "epoch": 15.977653631284916, + "grad_norm": 2.6176798343658447, + "learning_rate": 1.915977653631285e-05, + "loss": 0.2935, + "step": 5720 + }, + { + "epoch": 15.980446927374302, + "grad_norm": 1.3346441984176636, + "learning_rate": 1.916312849162011e-05, + "loss": 0.2067, + "step": 5721 + }, + { + "epoch": 15.983240223463687, + "grad_norm": 0.6005075573921204, + "learning_rate": 1.9166480446927372e-05, + "loss": 0.18, + "step": 5722 + }, + { + "epoch": 15.986033519553073, + "grad_norm": 0.8145706653594971, + "learning_rate": 1.9169832402234637e-05, + "loss": 0.1993, + "step": 5723 + }, + { + "epoch": 15.988826815642458, + "grad_norm": 1.2051368951797485, + "learning_rate": 1.9173184357541898e-05, + "loss": 0.1714, + "step": 5724 + }, + { + "epoch": 15.991620111731844, + "grad_norm": 1.2883059978485107, + "learning_rate": 1.917653631284916e-05, + "loss": 0.1701, + "step": 5725 + }, + { + "epoch": 15.994413407821229, + "grad_norm": 0.6192737817764282, + "learning_rate": 1.9179888268156424e-05, + "loss": 0.1745, + "step": 5726 + }, + { + "epoch": 15.997206703910614, + "grad_norm": 0.9873599410057068, + "learning_rate": 1.9183240223463686e-05, + "loss": 0.2016, + "step": 5727 + }, + { + "epoch": 16.0, + "grad_norm": 1.0467407703399658, + "learning_rate": 1.9186592178770947e-05, + "loss": 0.2052, + "step": 5728 + }, + { + "epoch": 16.002793296089386, + "grad_norm": 0.6876585483551025, + "learning_rate": 1.9189944134078212e-05, + "loss": 0.2323, + "step": 5729 + }, + { + "epoch": 16.00558659217877, + "grad_norm": 0.7050426006317139, + "learning_rate": 1.9193296089385476e-05, + "loss": 0.1967, + "step": 5730 + }, + { + "epoch": 16.008379888268156, + "grad_norm": 0.751594066619873, + "learning_rate": 1.919664804469274e-05, + "loss": 0.1583, + "step": 5731 + }, + { + "epoch": 16.011173184357542, + "grad_norm": 0.9258936643600464, + "learning_rate": 1.9200000000000003e-05, + "loss": 0.1848, + "step": 5732 + }, + { + "epoch": 16.01396648044693, + "grad_norm": 0.7250924110412598, + "learning_rate": 1.9203351955307264e-05, + "loss": 0.1885, + "step": 5733 + }, + { + "epoch": 16.01675977653631, + "grad_norm": 0.6125675439834595, + "learning_rate": 1.920670391061453e-05, + "loss": 0.1703, + "step": 5734 + }, + { + "epoch": 16.019553072625698, + "grad_norm": 1.0594289302825928, + "learning_rate": 1.921005586592179e-05, + "loss": 0.2, + "step": 5735 + }, + { + "epoch": 16.022346368715084, + "grad_norm": 5.310128688812256, + "learning_rate": 1.921340782122905e-05, + "loss": 0.1686, + "step": 5736 + }, + { + "epoch": 16.02513966480447, + "grad_norm": 1.215975284576416, + "learning_rate": 1.9216759776536316e-05, + "loss": 0.1817, + "step": 5737 + }, + { + "epoch": 16.027932960893853, + "grad_norm": 1.0459468364715576, + "learning_rate": 1.9220111731843577e-05, + "loss": 0.1998, + "step": 5738 + }, + { + "epoch": 16.03072625698324, + "grad_norm": 1.3041032552719116, + "learning_rate": 1.922346368715084e-05, + "loss": 0.1573, + "step": 5739 + }, + { + "epoch": 16.033519553072626, + "grad_norm": 1.313170313835144, + "learning_rate": 1.9226815642458103e-05, + "loss": 0.1812, + "step": 5740 + }, + { + "epoch": 16.036312849162012, + "grad_norm": 0.6014472246170044, + "learning_rate": 1.9230167597765365e-05, + "loss": 0.1616, + "step": 5741 + }, + { + "epoch": 16.039106145251395, + "grad_norm": 0.8557477593421936, + "learning_rate": 1.9233519553072626e-05, + "loss": 0.208, + "step": 5742 + }, + { + "epoch": 16.04189944134078, + "grad_norm": 0.9072015285491943, + "learning_rate": 1.923687150837989e-05, + "loss": 0.1513, + "step": 5743 + }, + { + "epoch": 16.044692737430168, + "grad_norm": 0.8945603966712952, + "learning_rate": 1.9240223463687152e-05, + "loss": 0.1795, + "step": 5744 + }, + { + "epoch": 16.047486033519554, + "grad_norm": 1.8702211380004883, + "learning_rate": 1.9243575418994414e-05, + "loss": 0.1835, + "step": 5745 + }, + { + "epoch": 16.050279329608937, + "grad_norm": 0.7017609477043152, + "learning_rate": 1.9246927374301678e-05, + "loss": 0.16, + "step": 5746 + }, + { + "epoch": 16.053072625698324, + "grad_norm": 0.9926375150680542, + "learning_rate": 1.925027932960894e-05, + "loss": 0.1876, + "step": 5747 + }, + { + "epoch": 16.05586592178771, + "grad_norm": 1.6543422937393188, + "learning_rate": 1.92536312849162e-05, + "loss": 0.1971, + "step": 5748 + }, + { + "epoch": 16.058659217877096, + "grad_norm": 1.0083043575286865, + "learning_rate": 1.9256983240223466e-05, + "loss": 0.1906, + "step": 5749 + }, + { + "epoch": 16.06145251396648, + "grad_norm": 1.0376359224319458, + "learning_rate": 1.9260335195530727e-05, + "loss": 0.1755, + "step": 5750 + }, + { + "epoch": 16.064245810055866, + "grad_norm": 1.469549298286438, + "learning_rate": 1.9263687150837992e-05, + "loss": 0.2063, + "step": 5751 + }, + { + "epoch": 16.067039106145252, + "grad_norm": 1.122471570968628, + "learning_rate": 1.9267039106145253e-05, + "loss": 0.2006, + "step": 5752 + }, + { + "epoch": 16.06983240223464, + "grad_norm": 4.555715560913086, + "learning_rate": 1.9270391061452514e-05, + "loss": 0.2847, + "step": 5753 + }, + { + "epoch": 16.07262569832402, + "grad_norm": 0.6928714513778687, + "learning_rate": 1.927374301675978e-05, + "loss": 0.208, + "step": 5754 + }, + { + "epoch": 16.075418994413408, + "grad_norm": 0.6966777443885803, + "learning_rate": 1.927709497206704e-05, + "loss": 0.1802, + "step": 5755 + }, + { + "epoch": 16.078212290502794, + "grad_norm": 0.5766451358795166, + "learning_rate": 1.9280446927374302e-05, + "loss": 0.1967, + "step": 5756 + }, + { + "epoch": 16.08100558659218, + "grad_norm": 0.6066396236419678, + "learning_rate": 1.9283798882681567e-05, + "loss": 0.185, + "step": 5757 + }, + { + "epoch": 16.083798882681563, + "grad_norm": 0.7514068484306335, + "learning_rate": 1.9287150837988828e-05, + "loss": 0.1781, + "step": 5758 + }, + { + "epoch": 16.08659217877095, + "grad_norm": 0.7468105554580688, + "learning_rate": 1.929050279329609e-05, + "loss": 0.1915, + "step": 5759 + }, + { + "epoch": 16.089385474860336, + "grad_norm": 0.8405874967575073, + "learning_rate": 1.9293854748603354e-05, + "loss": 0.1724, + "step": 5760 + }, + { + "epoch": 16.092178770949722, + "grad_norm": 3.32767915725708, + "learning_rate": 1.9297206703910615e-05, + "loss": 0.1541, + "step": 5761 + }, + { + "epoch": 16.094972067039105, + "grad_norm": 1.1103161573410034, + "learning_rate": 1.9300558659217877e-05, + "loss": 0.1727, + "step": 5762 + }, + { + "epoch": 16.09776536312849, + "grad_norm": 0.956878125667572, + "learning_rate": 1.930391061452514e-05, + "loss": 0.1575, + "step": 5763 + }, + { + "epoch": 16.100558659217878, + "grad_norm": 0.8224087357521057, + "learning_rate": 1.9307262569832403e-05, + "loss": 0.1748, + "step": 5764 + }, + { + "epoch": 16.10335195530726, + "grad_norm": 0.8303358554840088, + "learning_rate": 1.9310614525139664e-05, + "loss": 0.1827, + "step": 5765 + }, + { + "epoch": 16.106145251396647, + "grad_norm": 0.7121956944465637, + "learning_rate": 1.931396648044693e-05, + "loss": 0.211, + "step": 5766 + }, + { + "epoch": 16.108938547486034, + "grad_norm": 1.3952330350875854, + "learning_rate": 1.931731843575419e-05, + "loss": 0.2416, + "step": 5767 + }, + { + "epoch": 16.11173184357542, + "grad_norm": 1.7143166065216064, + "learning_rate": 1.9320670391061455e-05, + "loss": 0.1489, + "step": 5768 + }, + { + "epoch": 16.114525139664803, + "grad_norm": 1.4829065799713135, + "learning_rate": 1.9324022346368716e-05, + "loss": 0.1783, + "step": 5769 + }, + { + "epoch": 16.11731843575419, + "grad_norm": 0.9336336255073547, + "learning_rate": 1.9327374301675978e-05, + "loss": 0.1862, + "step": 5770 + }, + { + "epoch": 16.120111731843576, + "grad_norm": 1.1770561933517456, + "learning_rate": 1.9330726256983242e-05, + "loss": 0.1628, + "step": 5771 + }, + { + "epoch": 16.122905027932962, + "grad_norm": 1.8483673334121704, + "learning_rate": 1.9334078212290504e-05, + "loss": 0.1446, + "step": 5772 + }, + { + "epoch": 16.125698324022345, + "grad_norm": 1.4235873222351074, + "learning_rate": 1.9337430167597765e-05, + "loss": 0.1481, + "step": 5773 + }, + { + "epoch": 16.12849162011173, + "grad_norm": 1.4147595167160034, + "learning_rate": 1.934078212290503e-05, + "loss": 0.2482, + "step": 5774 + }, + { + "epoch": 16.131284916201118, + "grad_norm": 1.2481216192245483, + "learning_rate": 1.934413407821229e-05, + "loss": 0.1739, + "step": 5775 + }, + { + "epoch": 16.134078212290504, + "grad_norm": 1.7846935987472534, + "learning_rate": 1.9347486033519553e-05, + "loss": 0.2136, + "step": 5776 + }, + { + "epoch": 16.136871508379887, + "grad_norm": 1.429471731185913, + "learning_rate": 1.9350837988826817e-05, + "loss": 0.2428, + "step": 5777 + }, + { + "epoch": 16.139664804469273, + "grad_norm": 1.6725866794586182, + "learning_rate": 1.935418994413408e-05, + "loss": 0.2784, + "step": 5778 + }, + { + "epoch": 16.14245810055866, + "grad_norm": 0.6412115097045898, + "learning_rate": 1.935754189944134e-05, + "loss": 0.177, + "step": 5779 + }, + { + "epoch": 16.145251396648046, + "grad_norm": 1.0823171138763428, + "learning_rate": 1.9360893854748605e-05, + "loss": 0.1992, + "step": 5780 + }, + { + "epoch": 16.14804469273743, + "grad_norm": 1.9317622184753418, + "learning_rate": 1.9364245810055866e-05, + "loss": 0.1816, + "step": 5781 + }, + { + "epoch": 16.150837988826815, + "grad_norm": 1.2754995822906494, + "learning_rate": 1.9367597765363127e-05, + "loss": 0.1889, + "step": 5782 + }, + { + "epoch": 16.1536312849162, + "grad_norm": 0.72211092710495, + "learning_rate": 1.9370949720670392e-05, + "loss": 0.1945, + "step": 5783 + }, + { + "epoch": 16.156424581005588, + "grad_norm": 1.8432763814926147, + "learning_rate": 1.9374301675977653e-05, + "loss": 0.1979, + "step": 5784 + }, + { + "epoch": 16.15921787709497, + "grad_norm": 1.0330169200897217, + "learning_rate": 1.9377653631284915e-05, + "loss": 0.2011, + "step": 5785 + }, + { + "epoch": 16.162011173184357, + "grad_norm": 0.7949180603027344, + "learning_rate": 1.938100558659218e-05, + "loss": 0.1853, + "step": 5786 + }, + { + "epoch": 16.164804469273744, + "grad_norm": 0.767631471157074, + "learning_rate": 1.938435754189944e-05, + "loss": 0.158, + "step": 5787 + }, + { + "epoch": 16.16759776536313, + "grad_norm": 1.087835431098938, + "learning_rate": 1.9387709497206706e-05, + "loss": 0.1535, + "step": 5788 + }, + { + "epoch": 16.170391061452513, + "grad_norm": 0.5491178631782532, + "learning_rate": 1.9391061452513967e-05, + "loss": 0.1923, + "step": 5789 + }, + { + "epoch": 16.1731843575419, + "grad_norm": 1.3717846870422363, + "learning_rate": 1.939441340782123e-05, + "loss": 0.1976, + "step": 5790 + }, + { + "epoch": 16.175977653631286, + "grad_norm": 0.9666928648948669, + "learning_rate": 1.9397765363128493e-05, + "loss": 0.1756, + "step": 5791 + }, + { + "epoch": 16.178770949720672, + "grad_norm": 0.6771291494369507, + "learning_rate": 1.9401117318435754e-05, + "loss": 0.1983, + "step": 5792 + }, + { + "epoch": 16.181564245810055, + "grad_norm": 1.165290117263794, + "learning_rate": 1.9404469273743016e-05, + "loss": 0.1577, + "step": 5793 + }, + { + "epoch": 16.18435754189944, + "grad_norm": 0.6461641788482666, + "learning_rate": 1.940782122905028e-05, + "loss": 0.1581, + "step": 5794 + }, + { + "epoch": 16.187150837988828, + "grad_norm": 1.6318038702011108, + "learning_rate": 1.9411173184357542e-05, + "loss": 0.1895, + "step": 5795 + }, + { + "epoch": 16.189944134078214, + "grad_norm": 1.0697968006134033, + "learning_rate": 1.9414525139664803e-05, + "loss": 0.2075, + "step": 5796 + }, + { + "epoch": 16.192737430167597, + "grad_norm": 0.8332831859588623, + "learning_rate": 1.9417877094972068e-05, + "loss": 0.1882, + "step": 5797 + }, + { + "epoch": 16.195530726256983, + "grad_norm": 0.9994171857833862, + "learning_rate": 1.942122905027933e-05, + "loss": 0.1597, + "step": 5798 + }, + { + "epoch": 16.19832402234637, + "grad_norm": 1.2603111267089844, + "learning_rate": 1.942458100558659e-05, + "loss": 0.1735, + "step": 5799 + }, + { + "epoch": 16.201117318435756, + "grad_norm": 1.0342991352081299, + "learning_rate": 1.9427932960893855e-05, + "loss": 0.1661, + "step": 5800 + }, + { + "epoch": 16.20391061452514, + "grad_norm": 7.820026397705078, + "learning_rate": 1.9431284916201117e-05, + "loss": 0.1659, + "step": 5801 + }, + { + "epoch": 16.206703910614525, + "grad_norm": 1.555130958557129, + "learning_rate": 1.9434636871508378e-05, + "loss": 0.2305, + "step": 5802 + }, + { + "epoch": 16.20949720670391, + "grad_norm": 2.1637556552886963, + "learning_rate": 1.9437988826815643e-05, + "loss": 0.3246, + "step": 5803 + }, + { + "epoch": 16.212290502793294, + "grad_norm": 0.8095383644104004, + "learning_rate": 1.9441340782122904e-05, + "loss": 0.2142, + "step": 5804 + }, + { + "epoch": 16.21508379888268, + "grad_norm": 0.5543080568313599, + "learning_rate": 1.944469273743017e-05, + "loss": 0.1841, + "step": 5805 + }, + { + "epoch": 16.217877094972067, + "grad_norm": 0.7974898815155029, + "learning_rate": 1.944804469273743e-05, + "loss": 0.1673, + "step": 5806 + }, + { + "epoch": 16.220670391061454, + "grad_norm": 0.5193461179733276, + "learning_rate": 1.945139664804469e-05, + "loss": 0.1714, + "step": 5807 + }, + { + "epoch": 16.223463687150836, + "grad_norm": 0.6429254412651062, + "learning_rate": 1.9454748603351956e-05, + "loss": 0.2001, + "step": 5808 + }, + { + "epoch": 16.226256983240223, + "grad_norm": 1.3456521034240723, + "learning_rate": 1.9458100558659218e-05, + "loss": 0.1792, + "step": 5809 + }, + { + "epoch": 16.22905027932961, + "grad_norm": 0.7646437883377075, + "learning_rate": 1.946145251396648e-05, + "loss": 0.2142, + "step": 5810 + }, + { + "epoch": 16.231843575418996, + "grad_norm": 0.8739559650421143, + "learning_rate": 1.9464804469273744e-05, + "loss": 0.1884, + "step": 5811 + }, + { + "epoch": 16.23463687150838, + "grad_norm": 0.6744294166564941, + "learning_rate": 1.9468156424581005e-05, + "loss": 0.1636, + "step": 5812 + }, + { + "epoch": 16.237430167597765, + "grad_norm": 1.0414643287658691, + "learning_rate": 1.9471508379888266e-05, + "loss": 0.1894, + "step": 5813 + }, + { + "epoch": 16.24022346368715, + "grad_norm": 0.6037180423736572, + "learning_rate": 1.947486033519553e-05, + "loss": 0.1822, + "step": 5814 + }, + { + "epoch": 16.243016759776538, + "grad_norm": 1.0475513935089111, + "learning_rate": 1.9478212290502792e-05, + "loss": 0.2096, + "step": 5815 + }, + { + "epoch": 16.24581005586592, + "grad_norm": 0.9979348182678223, + "learning_rate": 1.9481564245810054e-05, + "loss": 0.177, + "step": 5816 + }, + { + "epoch": 16.248603351955307, + "grad_norm": 0.5982809066772461, + "learning_rate": 1.948491620111732e-05, + "loss": 0.1707, + "step": 5817 + }, + { + "epoch": 16.251396648044693, + "grad_norm": 1.1567827463150024, + "learning_rate": 1.948826815642458e-05, + "loss": 0.1785, + "step": 5818 + }, + { + "epoch": 16.25418994413408, + "grad_norm": 0.965018093585968, + "learning_rate": 1.949162011173184e-05, + "loss": 0.1735, + "step": 5819 + }, + { + "epoch": 16.256983240223462, + "grad_norm": 0.9561141729354858, + "learning_rate": 1.9494972067039106e-05, + "loss": 0.1591, + "step": 5820 + }, + { + "epoch": 16.25977653631285, + "grad_norm": 1.0379170179367065, + "learning_rate": 1.9498324022346367e-05, + "loss": 0.1751, + "step": 5821 + }, + { + "epoch": 16.262569832402235, + "grad_norm": 1.4235342741012573, + "learning_rate": 1.950167597765363e-05, + "loss": 0.1971, + "step": 5822 + }, + { + "epoch": 16.26536312849162, + "grad_norm": 2.1355271339416504, + "learning_rate": 1.9505027932960893e-05, + "loss": 0.2043, + "step": 5823 + }, + { + "epoch": 16.268156424581004, + "grad_norm": 1.0181339979171753, + "learning_rate": 1.9508379888268155e-05, + "loss": 0.2221, + "step": 5824 + }, + { + "epoch": 16.27094972067039, + "grad_norm": 1.3124529123306274, + "learning_rate": 1.951173184357542e-05, + "loss": 0.2042, + "step": 5825 + }, + { + "epoch": 16.273743016759777, + "grad_norm": 1.5319230556488037, + "learning_rate": 1.951508379888268e-05, + "loss": 0.1955, + "step": 5826 + }, + { + "epoch": 16.276536312849164, + "grad_norm": 1.4179824590682983, + "learning_rate": 1.9518435754189946e-05, + "loss": 0.1934, + "step": 5827 + }, + { + "epoch": 16.279329608938546, + "grad_norm": 2.1724700927734375, + "learning_rate": 1.952178770949721e-05, + "loss": 0.3223, + "step": 5828 + }, + { + "epoch": 16.282122905027933, + "grad_norm": 0.5374197959899902, + "learning_rate": 1.952513966480447e-05, + "loss": 0.1808, + "step": 5829 + }, + { + "epoch": 16.28491620111732, + "grad_norm": 0.7843204736709595, + "learning_rate": 1.9528491620111733e-05, + "loss": 0.1884, + "step": 5830 + }, + { + "epoch": 16.287709497206706, + "grad_norm": 0.6064081788063049, + "learning_rate": 1.9531843575418998e-05, + "loss": 0.2349, + "step": 5831 + }, + { + "epoch": 16.29050279329609, + "grad_norm": 0.6208264827728271, + "learning_rate": 1.953519553072626e-05, + "loss": 0.1608, + "step": 5832 + }, + { + "epoch": 16.293296089385475, + "grad_norm": 0.7612592577934265, + "learning_rate": 1.953854748603352e-05, + "loss": 0.2222, + "step": 5833 + }, + { + "epoch": 16.29608938547486, + "grad_norm": 0.5522499680519104, + "learning_rate": 1.9541899441340785e-05, + "loss": 0.1633, + "step": 5834 + }, + { + "epoch": 16.298882681564244, + "grad_norm": 0.733447790145874, + "learning_rate": 1.9545251396648046e-05, + "loss": 0.1814, + "step": 5835 + }, + { + "epoch": 16.30167597765363, + "grad_norm": 0.8245477080345154, + "learning_rate": 1.9548603351955308e-05, + "loss": 0.2154, + "step": 5836 + }, + { + "epoch": 16.304469273743017, + "grad_norm": 1.5482478141784668, + "learning_rate": 1.9551955307262573e-05, + "loss": 0.1869, + "step": 5837 + }, + { + "epoch": 16.307262569832403, + "grad_norm": 1.0440372228622437, + "learning_rate": 1.9555307262569834e-05, + "loss": 0.205, + "step": 5838 + }, + { + "epoch": 16.310055865921786, + "grad_norm": 0.556326150894165, + "learning_rate": 1.9558659217877095e-05, + "loss": 0.1626, + "step": 5839 + }, + { + "epoch": 16.312849162011172, + "grad_norm": 0.5407446622848511, + "learning_rate": 1.956201117318436e-05, + "loss": 0.188, + "step": 5840 + }, + { + "epoch": 16.31564245810056, + "grad_norm": 1.573560357093811, + "learning_rate": 1.956536312849162e-05, + "loss": 0.192, + "step": 5841 + }, + { + "epoch": 16.318435754189945, + "grad_norm": 0.886385440826416, + "learning_rate": 1.9568715083798883e-05, + "loss": 0.1749, + "step": 5842 + }, + { + "epoch": 16.321229050279328, + "grad_norm": 0.6377646923065186, + "learning_rate": 1.9572067039106147e-05, + "loss": 0.1836, + "step": 5843 + }, + { + "epoch": 16.324022346368714, + "grad_norm": 1.0486475229263306, + "learning_rate": 1.957541899441341e-05, + "loss": 0.192, + "step": 5844 + }, + { + "epoch": 16.3268156424581, + "grad_norm": 1.4788267612457275, + "learning_rate": 1.9578770949720673e-05, + "loss": 0.1526, + "step": 5845 + }, + { + "epoch": 16.329608938547487, + "grad_norm": 1.389596700668335, + "learning_rate": 1.9582122905027935e-05, + "loss": 0.189, + "step": 5846 + }, + { + "epoch": 16.33240223463687, + "grad_norm": 0.8937526345252991, + "learning_rate": 1.9585474860335196e-05, + "loss": 0.2119, + "step": 5847 + }, + { + "epoch": 16.335195530726256, + "grad_norm": 0.7266169190406799, + "learning_rate": 1.958882681564246e-05, + "loss": 0.1516, + "step": 5848 + }, + { + "epoch": 16.337988826815643, + "grad_norm": 0.7091546058654785, + "learning_rate": 1.9592178770949722e-05, + "loss": 0.1718, + "step": 5849 + }, + { + "epoch": 16.34078212290503, + "grad_norm": 1.0242987871170044, + "learning_rate": 1.9595530726256984e-05, + "loss": 0.1698, + "step": 5850 + }, + { + "epoch": 16.343575418994412, + "grad_norm": 1.158544659614563, + "learning_rate": 1.9598882681564248e-05, + "loss": 0.1937, + "step": 5851 + }, + { + "epoch": 16.3463687150838, + "grad_norm": 1.6920698881149292, + "learning_rate": 1.960223463687151e-05, + "loss": 0.1963, + "step": 5852 + }, + { + "epoch": 16.349162011173185, + "grad_norm": 1.7908693552017212, + "learning_rate": 1.960558659217877e-05, + "loss": 0.2654, + "step": 5853 + }, + { + "epoch": 16.35195530726257, + "grad_norm": 0.5307396650314331, + "learning_rate": 1.9608938547486036e-05, + "loss": 0.2051, + "step": 5854 + }, + { + "epoch": 16.354748603351954, + "grad_norm": 0.5588791966438293, + "learning_rate": 1.9612290502793297e-05, + "loss": 0.2011, + "step": 5855 + }, + { + "epoch": 16.35754189944134, + "grad_norm": 0.43220266699790955, + "learning_rate": 1.961564245810056e-05, + "loss": 0.1581, + "step": 5856 + }, + { + "epoch": 16.360335195530727, + "grad_norm": 1.078705906867981, + "learning_rate": 1.9618994413407823e-05, + "loss": 0.1921, + "step": 5857 + }, + { + "epoch": 16.363128491620113, + "grad_norm": 0.5344694256782532, + "learning_rate": 1.9622346368715084e-05, + "loss": 0.1949, + "step": 5858 + }, + { + "epoch": 16.365921787709496, + "grad_norm": 0.715362548828125, + "learning_rate": 1.9625698324022346e-05, + "loss": 0.2004, + "step": 5859 + }, + { + "epoch": 16.368715083798882, + "grad_norm": 0.6082919836044312, + "learning_rate": 1.962905027932961e-05, + "loss": 0.1562, + "step": 5860 + }, + { + "epoch": 16.37150837988827, + "grad_norm": 0.4872756004333496, + "learning_rate": 1.9632402234636872e-05, + "loss": 0.1523, + "step": 5861 + }, + { + "epoch": 16.374301675977655, + "grad_norm": 0.7324002981185913, + "learning_rate": 1.9635754189944137e-05, + "loss": 0.1748, + "step": 5862 + }, + { + "epoch": 16.377094972067038, + "grad_norm": 0.7190079092979431, + "learning_rate": 1.9639106145251398e-05, + "loss": 0.1785, + "step": 5863 + }, + { + "epoch": 16.379888268156424, + "grad_norm": 0.8159694671630859, + "learning_rate": 1.964245810055866e-05, + "loss": 0.1899, + "step": 5864 + }, + { + "epoch": 16.38268156424581, + "grad_norm": 0.5700684189796448, + "learning_rate": 1.9645810055865924e-05, + "loss": 0.1636, + "step": 5865 + }, + { + "epoch": 16.385474860335197, + "grad_norm": 1.2027521133422852, + "learning_rate": 1.9649162011173185e-05, + "loss": 0.1571, + "step": 5866 + }, + { + "epoch": 16.38826815642458, + "grad_norm": 0.7358332276344299, + "learning_rate": 1.9652513966480447e-05, + "loss": 0.165, + "step": 5867 + }, + { + "epoch": 16.391061452513966, + "grad_norm": 0.9313184022903442, + "learning_rate": 1.965586592178771e-05, + "loss": 0.1603, + "step": 5868 + }, + { + "epoch": 16.393854748603353, + "grad_norm": 1.0991677045822144, + "learning_rate": 1.9659217877094973e-05, + "loss": 0.1894, + "step": 5869 + }, + { + "epoch": 16.39664804469274, + "grad_norm": 1.5144155025482178, + "learning_rate": 1.9662569832402234e-05, + "loss": 0.1983, + "step": 5870 + }, + { + "epoch": 16.399441340782122, + "grad_norm": 1.135430097579956, + "learning_rate": 1.96659217877095e-05, + "loss": 0.2054, + "step": 5871 + }, + { + "epoch": 16.40223463687151, + "grad_norm": 1.2212462425231934, + "learning_rate": 1.966927374301676e-05, + "loss": 0.1933, + "step": 5872 + }, + { + "epoch": 16.405027932960895, + "grad_norm": 1.1437976360321045, + "learning_rate": 1.967262569832402e-05, + "loss": 0.1967, + "step": 5873 + }, + { + "epoch": 16.407821229050278, + "grad_norm": 1.6805050373077393, + "learning_rate": 1.9675977653631286e-05, + "loss": 0.1976, + "step": 5874 + }, + { + "epoch": 16.410614525139664, + "grad_norm": 1.3702940940856934, + "learning_rate": 1.9679329608938548e-05, + "loss": 0.2001, + "step": 5875 + }, + { + "epoch": 16.41340782122905, + "grad_norm": 1.684321641921997, + "learning_rate": 1.968268156424581e-05, + "loss": 0.1957, + "step": 5876 + }, + { + "epoch": 16.416201117318437, + "grad_norm": 1.121792197227478, + "learning_rate": 1.9686033519553074e-05, + "loss": 0.1849, + "step": 5877 + }, + { + "epoch": 16.41899441340782, + "grad_norm": 1.8655019998550415, + "learning_rate": 1.9689385474860335e-05, + "loss": 0.2274, + "step": 5878 + }, + { + "epoch": 16.421787709497206, + "grad_norm": 0.5348700881004333, + "learning_rate": 1.9692737430167596e-05, + "loss": 0.1632, + "step": 5879 + }, + { + "epoch": 16.424581005586592, + "grad_norm": 0.7120680809020996, + "learning_rate": 1.969608938547486e-05, + "loss": 0.2058, + "step": 5880 + }, + { + "epoch": 16.42737430167598, + "grad_norm": 0.4995569586753845, + "learning_rate": 1.9699441340782123e-05, + "loss": 0.1673, + "step": 5881 + }, + { + "epoch": 16.43016759776536, + "grad_norm": 0.5837088227272034, + "learning_rate": 1.9702793296089387e-05, + "loss": 0.183, + "step": 5882 + }, + { + "epoch": 16.432960893854748, + "grad_norm": 0.7866269946098328, + "learning_rate": 1.970614525139665e-05, + "loss": 0.1684, + "step": 5883 + }, + { + "epoch": 16.435754189944134, + "grad_norm": 0.689897358417511, + "learning_rate": 1.970949720670391e-05, + "loss": 0.1903, + "step": 5884 + }, + { + "epoch": 16.43854748603352, + "grad_norm": 0.85750812292099, + "learning_rate": 1.9712849162011175e-05, + "loss": 0.2193, + "step": 5885 + }, + { + "epoch": 16.441340782122904, + "grad_norm": 0.6550633311271667, + "learning_rate": 1.9716201117318436e-05, + "loss": 0.1637, + "step": 5886 + }, + { + "epoch": 16.44413407821229, + "grad_norm": 0.7614025473594666, + "learning_rate": 1.9719553072625697e-05, + "loss": 0.186, + "step": 5887 + }, + { + "epoch": 16.446927374301676, + "grad_norm": 0.9328618049621582, + "learning_rate": 1.9722905027932962e-05, + "loss": 0.1943, + "step": 5888 + }, + { + "epoch": 16.449720670391063, + "grad_norm": 0.8918062448501587, + "learning_rate": 1.9726256983240223e-05, + "loss": 0.2183, + "step": 5889 + }, + { + "epoch": 16.452513966480446, + "grad_norm": 1.024586796760559, + "learning_rate": 1.9729608938547485e-05, + "loss": 0.2149, + "step": 5890 + }, + { + "epoch": 16.455307262569832, + "grad_norm": 0.7969521284103394, + "learning_rate": 1.973296089385475e-05, + "loss": 0.1722, + "step": 5891 + }, + { + "epoch": 16.45810055865922, + "grad_norm": 0.974992573261261, + "learning_rate": 1.973631284916201e-05, + "loss": 0.1553, + "step": 5892 + }, + { + "epoch": 16.460893854748605, + "grad_norm": 1.0363444089889526, + "learning_rate": 1.9739664804469272e-05, + "loss": 0.1536, + "step": 5893 + }, + { + "epoch": 16.463687150837988, + "grad_norm": 2.8143651485443115, + "learning_rate": 1.9743016759776537e-05, + "loss": 0.1772, + "step": 5894 + }, + { + "epoch": 16.466480446927374, + "grad_norm": 1.1301976442337036, + "learning_rate": 1.97463687150838e-05, + "loss": 0.1867, + "step": 5895 + }, + { + "epoch": 16.46927374301676, + "grad_norm": 1.5248794555664062, + "learning_rate": 1.974972067039106e-05, + "loss": 0.168, + "step": 5896 + }, + { + "epoch": 16.472067039106147, + "grad_norm": 0.8021107316017151, + "learning_rate": 1.9753072625698324e-05, + "loss": 0.18, + "step": 5897 + }, + { + "epoch": 16.47486033519553, + "grad_norm": 1.9845572710037231, + "learning_rate": 1.9756424581005586e-05, + "loss": 0.2042, + "step": 5898 + }, + { + "epoch": 16.477653631284916, + "grad_norm": 1.6193259954452515, + "learning_rate": 1.975977653631285e-05, + "loss": 0.1965, + "step": 5899 + }, + { + "epoch": 16.480446927374302, + "grad_norm": 1.8381743431091309, + "learning_rate": 1.9763128491620112e-05, + "loss": 0.2216, + "step": 5900 + }, + { + "epoch": 16.48324022346369, + "grad_norm": 1.6792151927947998, + "learning_rate": 1.9766480446927373e-05, + "loss": 0.1729, + "step": 5901 + }, + { + "epoch": 16.48603351955307, + "grad_norm": 2.8398356437683105, + "learning_rate": 1.9769832402234638e-05, + "loss": 0.2013, + "step": 5902 + }, + { + "epoch": 16.488826815642458, + "grad_norm": 2.429287910461426, + "learning_rate": 1.97731843575419e-05, + "loss": 0.2571, + "step": 5903 + }, + { + "epoch": 16.491620111731844, + "grad_norm": 0.7441657185554504, + "learning_rate": 1.977653631284916e-05, + "loss": 0.193, + "step": 5904 + }, + { + "epoch": 16.49441340782123, + "grad_norm": 0.5730157494544983, + "learning_rate": 1.9779888268156425e-05, + "loss": 0.1852, + "step": 5905 + }, + { + "epoch": 16.497206703910614, + "grad_norm": 0.6026260852813721, + "learning_rate": 1.9783240223463687e-05, + "loss": 0.196, + "step": 5906 + }, + { + "epoch": 16.5, + "grad_norm": 0.8806973695755005, + "learning_rate": 1.9786592178770948e-05, + "loss": 0.1957, + "step": 5907 + }, + { + "epoch": 16.502793296089386, + "grad_norm": 0.7619754672050476, + "learning_rate": 1.9789944134078213e-05, + "loss": 0.1787, + "step": 5908 + }, + { + "epoch": 16.505586592178773, + "grad_norm": 1.0333995819091797, + "learning_rate": 1.9793296089385474e-05, + "loss": 0.192, + "step": 5909 + }, + { + "epoch": 16.508379888268156, + "grad_norm": 0.9905533790588379, + "learning_rate": 1.9796648044692735e-05, + "loss": 0.2102, + "step": 5910 + }, + { + "epoch": 16.511173184357542, + "grad_norm": 0.7787652015686035, + "learning_rate": 1.98e-05, + "loss": 0.2068, + "step": 5911 + }, + { + "epoch": 16.51396648044693, + "grad_norm": 0.8018123507499695, + "learning_rate": 1.980335195530726e-05, + "loss": 0.1828, + "step": 5912 + }, + { + "epoch": 16.51675977653631, + "grad_norm": 0.6556683778762817, + "learning_rate": 1.9806703910614523e-05, + "loss": 0.193, + "step": 5913 + }, + { + "epoch": 16.519553072625698, + "grad_norm": 0.7198950052261353, + "learning_rate": 1.9810055865921788e-05, + "loss": 0.1842, + "step": 5914 + }, + { + "epoch": 16.522346368715084, + "grad_norm": 0.7620674967765808, + "learning_rate": 1.981340782122905e-05, + "loss": 0.1809, + "step": 5915 + }, + { + "epoch": 16.52513966480447, + "grad_norm": 0.8521338701248169, + "learning_rate": 1.981675977653631e-05, + "loss": 0.1875, + "step": 5916 + }, + { + "epoch": 16.527932960893853, + "grad_norm": 0.6633079648017883, + "learning_rate": 1.9820111731843575e-05, + "loss": 0.1834, + "step": 5917 + }, + { + "epoch": 16.53072625698324, + "grad_norm": 0.9306690692901611, + "learning_rate": 1.9823463687150836e-05, + "loss": 0.1627, + "step": 5918 + }, + { + "epoch": 16.533519553072626, + "grad_norm": 0.7355237603187561, + "learning_rate": 1.98268156424581e-05, + "loss": 0.1747, + "step": 5919 + }, + { + "epoch": 16.536312849162012, + "grad_norm": 1.089219331741333, + "learning_rate": 1.9830167597765362e-05, + "loss": 0.215, + "step": 5920 + }, + { + "epoch": 16.539106145251395, + "grad_norm": 0.8674601316452026, + "learning_rate": 1.9833519553072624e-05, + "loss": 0.1807, + "step": 5921 + }, + { + "epoch": 16.54189944134078, + "grad_norm": 1.201404333114624, + "learning_rate": 1.983687150837989e-05, + "loss": 0.2, + "step": 5922 + }, + { + "epoch": 16.544692737430168, + "grad_norm": 0.8389508724212646, + "learning_rate": 1.9840223463687153e-05, + "loss": 0.1635, + "step": 5923 + }, + { + "epoch": 16.547486033519554, + "grad_norm": 0.9775781035423279, + "learning_rate": 1.9843575418994415e-05, + "loss": 0.169, + "step": 5924 + }, + { + "epoch": 16.550279329608937, + "grad_norm": 1.1684406995773315, + "learning_rate": 1.984692737430168e-05, + "loss": 0.1791, + "step": 5925 + }, + { + "epoch": 16.553072625698324, + "grad_norm": 1.5287076234817505, + "learning_rate": 1.985027932960894e-05, + "loss": 0.1742, + "step": 5926 + }, + { + "epoch": 16.55586592178771, + "grad_norm": 1.7475539445877075, + "learning_rate": 1.9853631284916202e-05, + "loss": 0.1904, + "step": 5927 + }, + { + "epoch": 16.558659217877096, + "grad_norm": 2.1542162895202637, + "learning_rate": 1.9856983240223467e-05, + "loss": 0.2508, + "step": 5928 + }, + { + "epoch": 16.56145251396648, + "grad_norm": 0.8869678378105164, + "learning_rate": 1.9860335195530728e-05, + "loss": 0.1997, + "step": 5929 + }, + { + "epoch": 16.564245810055866, + "grad_norm": 0.7166237831115723, + "learning_rate": 1.986368715083799e-05, + "loss": 0.1884, + "step": 5930 + }, + { + "epoch": 16.567039106145252, + "grad_norm": 1.0796215534210205, + "learning_rate": 1.9867039106145254e-05, + "loss": 0.212, + "step": 5931 + }, + { + "epoch": 16.56983240223464, + "grad_norm": 1.5622729063034058, + "learning_rate": 1.9870391061452516e-05, + "loss": 0.2371, + "step": 5932 + }, + { + "epoch": 16.57262569832402, + "grad_norm": 0.669573962688446, + "learning_rate": 1.9873743016759777e-05, + "loss": 0.2126, + "step": 5933 + }, + { + "epoch": 16.575418994413408, + "grad_norm": 0.6535643935203552, + "learning_rate": 1.987709497206704e-05, + "loss": 0.2015, + "step": 5934 + }, + { + "epoch": 16.578212290502794, + "grad_norm": 1.4564166069030762, + "learning_rate": 1.9880446927374303e-05, + "loss": 0.179, + "step": 5935 + }, + { + "epoch": 16.58100558659218, + "grad_norm": 0.9758133292198181, + "learning_rate": 1.9883798882681564e-05, + "loss": 0.211, + "step": 5936 + }, + { + "epoch": 16.583798882681563, + "grad_norm": 0.67426997423172, + "learning_rate": 1.988715083798883e-05, + "loss": 0.1704, + "step": 5937 + }, + { + "epoch": 16.58659217877095, + "grad_norm": 0.6225221753120422, + "learning_rate": 1.989050279329609e-05, + "loss": 0.1679, + "step": 5938 + }, + { + "epoch": 16.589385474860336, + "grad_norm": 0.8609606027603149, + "learning_rate": 1.9893854748603355e-05, + "loss": 0.2077, + "step": 5939 + }, + { + "epoch": 16.592178770949722, + "grad_norm": 0.8343431949615479, + "learning_rate": 1.9897206703910616e-05, + "loss": 0.1816, + "step": 5940 + }, + { + "epoch": 16.594972067039105, + "grad_norm": 1.132385015487671, + "learning_rate": 1.9900558659217878e-05, + "loss": 0.1537, + "step": 5941 + }, + { + "epoch": 16.59776536312849, + "grad_norm": 0.9977555274963379, + "learning_rate": 1.9903910614525143e-05, + "loss": 0.1744, + "step": 5942 + }, + { + "epoch": 16.600558659217878, + "grad_norm": 1.2667436599731445, + "learning_rate": 1.9907262569832404e-05, + "loss": 0.1971, + "step": 5943 + }, + { + "epoch": 16.60335195530726, + "grad_norm": 0.7418209314346313, + "learning_rate": 1.9910614525139665e-05, + "loss": 0.1503, + "step": 5944 + }, + { + "epoch": 16.606145251396647, + "grad_norm": 1.2106512784957886, + "learning_rate": 1.991396648044693e-05, + "loss": 0.2075, + "step": 5945 + }, + { + "epoch": 16.608938547486034, + "grad_norm": 0.8580309748649597, + "learning_rate": 1.991731843575419e-05, + "loss": 0.1619, + "step": 5946 + }, + { + "epoch": 16.61173184357542, + "grad_norm": 2.243924856185913, + "learning_rate": 1.9920670391061453e-05, + "loss": 0.1859, + "step": 5947 + }, + { + "epoch": 16.614525139664803, + "grad_norm": 1.4864705801010132, + "learning_rate": 1.9924022346368717e-05, + "loss": 0.1664, + "step": 5948 + }, + { + "epoch": 16.61731843575419, + "grad_norm": 1.057424783706665, + "learning_rate": 1.992737430167598e-05, + "loss": 0.1699, + "step": 5949 + }, + { + "epoch": 16.620111731843576, + "grad_norm": 0.8248178362846375, + "learning_rate": 1.993072625698324e-05, + "loss": 0.2302, + "step": 5950 + }, + { + "epoch": 16.622905027932962, + "grad_norm": 0.905230700969696, + "learning_rate": 1.9934078212290505e-05, + "loss": 0.2005, + "step": 5951 + }, + { + "epoch": 16.625698324022345, + "grad_norm": 1.4510602951049805, + "learning_rate": 1.9937430167597766e-05, + "loss": 0.2385, + "step": 5952 + }, + { + "epoch": 16.62849162011173, + "grad_norm": 1.6224864721298218, + "learning_rate": 1.9940782122905027e-05, + "loss": 0.2341, + "step": 5953 + }, + { + "epoch": 16.631284916201118, + "grad_norm": 0.6296976804733276, + "learning_rate": 1.9944134078212292e-05, + "loss": 0.1959, + "step": 5954 + }, + { + "epoch": 16.634078212290504, + "grad_norm": 0.48394206166267395, + "learning_rate": 1.9947486033519554e-05, + "loss": 0.195, + "step": 5955 + }, + { + "epoch": 16.636871508379887, + "grad_norm": 1.3960762023925781, + "learning_rate": 1.9950837988826818e-05, + "loss": 0.1779, + "step": 5956 + }, + { + "epoch": 16.639664804469273, + "grad_norm": 0.6206721067428589, + "learning_rate": 1.995418994413408e-05, + "loss": 0.1712, + "step": 5957 + }, + { + "epoch": 16.64245810055866, + "grad_norm": 1.0659148693084717, + "learning_rate": 1.995754189944134e-05, + "loss": 0.206, + "step": 5958 + }, + { + "epoch": 16.645251396648046, + "grad_norm": 0.6170654892921448, + "learning_rate": 1.9960893854748606e-05, + "loss": 0.1665, + "step": 5959 + }, + { + "epoch": 16.64804469273743, + "grad_norm": 0.5795278549194336, + "learning_rate": 1.9964245810055867e-05, + "loss": 0.1739, + "step": 5960 + }, + { + "epoch": 16.650837988826815, + "grad_norm": 0.8916805386543274, + "learning_rate": 1.996759776536313e-05, + "loss": 0.1856, + "step": 5961 + }, + { + "epoch": 16.6536312849162, + "grad_norm": 0.7777912616729736, + "learning_rate": 1.9970949720670393e-05, + "loss": 0.1751, + "step": 5962 + }, + { + "epoch": 16.656424581005588, + "grad_norm": 0.6202384233474731, + "learning_rate": 1.9974301675977654e-05, + "loss": 0.1872, + "step": 5963 + }, + { + "epoch": 16.65921787709497, + "grad_norm": 1.1392936706542969, + "learning_rate": 1.9977653631284916e-05, + "loss": 0.1822, + "step": 5964 + }, + { + "epoch": 16.662011173184357, + "grad_norm": 1.09144127368927, + "learning_rate": 1.998100558659218e-05, + "loss": 0.1796, + "step": 5965 + }, + { + "epoch": 16.664804469273744, + "grad_norm": 0.5765389800071716, + "learning_rate": 1.9984357541899442e-05, + "loss": 0.1746, + "step": 5966 + }, + { + "epoch": 16.66759776536313, + "grad_norm": 0.6020392179489136, + "learning_rate": 1.9987709497206703e-05, + "loss": 0.1441, + "step": 5967 + }, + { + "epoch": 16.670391061452513, + "grad_norm": 0.9503408670425415, + "learning_rate": 1.9991061452513968e-05, + "loss": 0.2021, + "step": 5968 + }, + { + "epoch": 16.6731843575419, + "grad_norm": 1.060049295425415, + "learning_rate": 1.999441340782123e-05, + "loss": 0.1777, + "step": 5969 + }, + { + "epoch": 16.675977653631286, + "grad_norm": 1.1033780574798584, + "learning_rate": 1.999776536312849e-05, + "loss": 0.1858, + "step": 5970 + }, + { + "epoch": 16.678770949720672, + "grad_norm": 1.086956262588501, + "learning_rate": 2.0001117318435755e-05, + "loss": 0.1828, + "step": 5971 + }, + { + "epoch": 16.681564245810055, + "grad_norm": 0.8564766049385071, + "learning_rate": 2.0004469273743017e-05, + "loss": 0.21, + "step": 5972 + }, + { + "epoch": 16.68435754189944, + "grad_norm": 1.1257343292236328, + "learning_rate": 2.0007821229050278e-05, + "loss": 0.1959, + "step": 5973 + }, + { + "epoch": 16.687150837988828, + "grad_norm": 1.650771141052246, + "learning_rate": 2.0011173184357543e-05, + "loss": 0.211, + "step": 5974 + }, + { + "epoch": 16.689944134078214, + "grad_norm": 1.0730290412902832, + "learning_rate": 2.0014525139664804e-05, + "loss": 0.1818, + "step": 5975 + }, + { + "epoch": 16.692737430167597, + "grad_norm": 1.0560986995697021, + "learning_rate": 2.001787709497207e-05, + "loss": 0.2151, + "step": 5976 + }, + { + "epoch": 16.695530726256983, + "grad_norm": 1.6965817213058472, + "learning_rate": 2.002122905027933e-05, + "loss": 0.2086, + "step": 5977 + }, + { + "epoch": 16.69832402234637, + "grad_norm": 1.7079346179962158, + "learning_rate": 2.002458100558659e-05, + "loss": 0.2463, + "step": 5978 + }, + { + "epoch": 16.701117318435756, + "grad_norm": 0.934985876083374, + "learning_rate": 2.0027932960893856e-05, + "loss": 0.2163, + "step": 5979 + }, + { + "epoch": 16.70391061452514, + "grad_norm": 0.8519138097763062, + "learning_rate": 2.0031284916201118e-05, + "loss": 0.1921, + "step": 5980 + }, + { + "epoch": 16.706703910614525, + "grad_norm": 0.850037693977356, + "learning_rate": 2.003463687150838e-05, + "loss": 0.201, + "step": 5981 + }, + { + "epoch": 16.70949720670391, + "grad_norm": 0.8497642874717712, + "learning_rate": 2.0037988826815644e-05, + "loss": 0.1947, + "step": 5982 + }, + { + "epoch": 16.712290502793294, + "grad_norm": 0.6012024879455566, + "learning_rate": 2.0041340782122905e-05, + "loss": 0.209, + "step": 5983 + }, + { + "epoch": 16.71508379888268, + "grad_norm": 0.8282449841499329, + "learning_rate": 2.0044692737430166e-05, + "loss": 0.2213, + "step": 5984 + }, + { + "epoch": 16.717877094972067, + "grad_norm": 2.4088659286499023, + "learning_rate": 2.004804469273743e-05, + "loss": 0.172, + "step": 5985 + }, + { + "epoch": 16.720670391061454, + "grad_norm": 1.5182751417160034, + "learning_rate": 2.0051396648044693e-05, + "loss": 0.1661, + "step": 5986 + }, + { + "epoch": 16.723463687150836, + "grad_norm": 0.4818273186683655, + "learning_rate": 2.0054748603351954e-05, + "loss": 0.1738, + "step": 5987 + }, + { + "epoch": 16.726256983240223, + "grad_norm": 0.8011188507080078, + "learning_rate": 2.005810055865922e-05, + "loss": 0.2265, + "step": 5988 + }, + { + "epoch": 16.72905027932961, + "grad_norm": 0.7228988409042358, + "learning_rate": 2.006145251396648e-05, + "loss": 0.1788, + "step": 5989 + }, + { + "epoch": 16.731843575418996, + "grad_norm": 1.2500355243682861, + "learning_rate": 2.006480446927374e-05, + "loss": 0.1937, + "step": 5990 + }, + { + "epoch": 16.73463687150838, + "grad_norm": 0.7818806171417236, + "learning_rate": 2.0068156424581006e-05, + "loss": 0.1828, + "step": 5991 + }, + { + "epoch": 16.737430167597765, + "grad_norm": 1.0143556594848633, + "learning_rate": 2.0071508379888267e-05, + "loss": 0.1751, + "step": 5992 + }, + { + "epoch": 16.74022346368715, + "grad_norm": 0.8912897109985352, + "learning_rate": 2.0074860335195532e-05, + "loss": 0.1666, + "step": 5993 + }, + { + "epoch": 16.743016759776538, + "grad_norm": 0.7487048506736755, + "learning_rate": 2.0078212290502793e-05, + "loss": 0.173, + "step": 5994 + }, + { + "epoch": 16.74581005586592, + "grad_norm": 3.133702278137207, + "learning_rate": 2.0081564245810055e-05, + "loss": 0.2251, + "step": 5995 + }, + { + "epoch": 16.748603351955307, + "grad_norm": 1.85382878780365, + "learning_rate": 2.008491620111732e-05, + "loss": 0.1556, + "step": 5996 + }, + { + "epoch": 16.751396648044693, + "grad_norm": 1.2456825971603394, + "learning_rate": 2.008826815642458e-05, + "loss": 0.1546, + "step": 5997 + }, + { + "epoch": 16.75418994413408, + "grad_norm": 0.8036320209503174, + "learning_rate": 2.0091620111731842e-05, + "loss": 0.1777, + "step": 5998 + }, + { + "epoch": 16.756983240223462, + "grad_norm": 1.0817253589630127, + "learning_rate": 2.0094972067039107e-05, + "loss": 0.1674, + "step": 5999 + }, + { + "epoch": 16.75977653631285, + "grad_norm": 1.298591136932373, + "learning_rate": 2.009832402234637e-05, + "loss": 0.2105, + "step": 6000 + }, + { + "epoch": 16.75977653631285, + "eval_cer": 0.19165220996584945, + "eval_loss": 0.2686631977558136, + "eval_runtime": 23.7558, + "eval_samples_per_second": 191.027, + "eval_steps_per_second": 0.631, + "eval_wer": 0.33890114801262533, + "step": 6000 + }, + { + "epoch": 16.762569832402235, + "grad_norm": 1.5015164613723755, + "learning_rate": 2.010167597765363e-05, + "loss": 0.2431, + "step": 6001 + }, + { + "epoch": 16.76536312849162, + "grad_norm": 1.3272188901901245, + "learning_rate": 2.0105027932960894e-05, + "loss": 0.2158, + "step": 6002 + }, + { + "epoch": 16.768156424581004, + "grad_norm": 1.2525218725204468, + "learning_rate": 2.0108379888268156e-05, + "loss": 0.2512, + "step": 6003 + }, + { + "epoch": 16.77094972067039, + "grad_norm": 0.7218210101127625, + "learning_rate": 2.0111731843575417e-05, + "loss": 0.1906, + "step": 6004 + }, + { + "epoch": 16.773743016759777, + "grad_norm": 0.49930325150489807, + "learning_rate": 2.0115083798882682e-05, + "loss": 0.1596, + "step": 6005 + }, + { + "epoch": 16.776536312849164, + "grad_norm": 0.9455858469009399, + "learning_rate": 2.0118435754189943e-05, + "loss": 0.159, + "step": 6006 + }, + { + "epoch": 16.779329608938546, + "grad_norm": 0.8473210334777832, + "learning_rate": 2.0121787709497205e-05, + "loss": 0.1964, + "step": 6007 + }, + { + "epoch": 16.782122905027933, + "grad_norm": 0.7438949346542358, + "learning_rate": 2.012513966480447e-05, + "loss": 0.1546, + "step": 6008 + }, + { + "epoch": 16.78491620111732, + "grad_norm": 0.9767144918441772, + "learning_rate": 2.012849162011173e-05, + "loss": 0.1732, + "step": 6009 + }, + { + "epoch": 16.787709497206706, + "grad_norm": 0.6159979701042175, + "learning_rate": 2.0131843575418992e-05, + "loss": 0.1962, + "step": 6010 + }, + { + "epoch": 16.79050279329609, + "grad_norm": 0.8341845870018005, + "learning_rate": 2.0135195530726257e-05, + "loss": 0.213, + "step": 6011 + }, + { + "epoch": 16.793296089385475, + "grad_norm": 0.6529709696769714, + "learning_rate": 2.0138547486033518e-05, + "loss": 0.1658, + "step": 6012 + }, + { + "epoch": 16.79608938547486, + "grad_norm": 0.925045907497406, + "learning_rate": 2.0141899441340783e-05, + "loss": 0.2051, + "step": 6013 + }, + { + "epoch": 16.798882681564244, + "grad_norm": 1.1142586469650269, + "learning_rate": 2.0145251396648044e-05, + "loss": 0.1922, + "step": 6014 + }, + { + "epoch": 16.80167597765363, + "grad_norm": 0.6049912571907043, + "learning_rate": 2.0148603351955305e-05, + "loss": 0.1768, + "step": 6015 + }, + { + "epoch": 16.804469273743017, + "grad_norm": 0.6410613656044006, + "learning_rate": 2.015195530726257e-05, + "loss": 0.1802, + "step": 6016 + }, + { + "epoch": 16.807262569832403, + "grad_norm": 0.7171903252601624, + "learning_rate": 2.015530726256983e-05, + "loss": 0.1798, + "step": 6017 + }, + { + "epoch": 16.810055865921786, + "grad_norm": 0.708947479724884, + "learning_rate": 2.0158659217877093e-05, + "loss": 0.2041, + "step": 6018 + }, + { + "epoch": 16.812849162011172, + "grad_norm": 0.6932775974273682, + "learning_rate": 2.0162011173184358e-05, + "loss": 0.1593, + "step": 6019 + }, + { + "epoch": 16.81564245810056, + "grad_norm": 0.7445617318153381, + "learning_rate": 2.0165363128491622e-05, + "loss": 0.1757, + "step": 6020 + }, + { + "epoch": 16.818435754189945, + "grad_norm": 1.065722942352295, + "learning_rate": 2.0168715083798884e-05, + "loss": 0.2075, + "step": 6021 + }, + { + "epoch": 16.821229050279328, + "grad_norm": 0.7429558038711548, + "learning_rate": 2.017206703910615e-05, + "loss": 0.1576, + "step": 6022 + }, + { + "epoch": 16.824022346368714, + "grad_norm": 0.9666042923927307, + "learning_rate": 2.017541899441341e-05, + "loss": 0.1845, + "step": 6023 + }, + { + "epoch": 16.8268156424581, + "grad_norm": 0.7710238695144653, + "learning_rate": 2.017877094972067e-05, + "loss": 0.2085, + "step": 6024 + }, + { + "epoch": 16.829608938547487, + "grad_norm": 1.0698350667953491, + "learning_rate": 2.0182122905027936e-05, + "loss": 0.2028, + "step": 6025 + }, + { + "epoch": 16.83240223463687, + "grad_norm": 1.1429990530014038, + "learning_rate": 2.0185474860335197e-05, + "loss": 0.2186, + "step": 6026 + }, + { + "epoch": 16.835195530726256, + "grad_norm": 1.0522236824035645, + "learning_rate": 2.018882681564246e-05, + "loss": 0.2198, + "step": 6027 + }, + { + "epoch": 16.837988826815643, + "grad_norm": 1.6316841840744019, + "learning_rate": 2.0192178770949723e-05, + "loss": 0.3053, + "step": 6028 + }, + { + "epoch": 16.84078212290503, + "grad_norm": 0.5452074408531189, + "learning_rate": 2.0195530726256985e-05, + "loss": 0.1811, + "step": 6029 + }, + { + "epoch": 16.843575418994412, + "grad_norm": 0.6944013237953186, + "learning_rate": 2.0198882681564246e-05, + "loss": 0.1829, + "step": 6030 + }, + { + "epoch": 16.8463687150838, + "grad_norm": 0.7985291481018066, + "learning_rate": 2.020223463687151e-05, + "loss": 0.1784, + "step": 6031 + }, + { + "epoch": 16.849162011173185, + "grad_norm": 0.5377930402755737, + "learning_rate": 2.0205586592178772e-05, + "loss": 0.1881, + "step": 6032 + }, + { + "epoch": 16.85195530726257, + "grad_norm": 0.6478134393692017, + "learning_rate": 2.0208938547486037e-05, + "loss": 0.174, + "step": 6033 + }, + { + "epoch": 16.854748603351954, + "grad_norm": 0.5892351269721985, + "learning_rate": 2.0212290502793298e-05, + "loss": 0.1627, + "step": 6034 + }, + { + "epoch": 16.85754189944134, + "grad_norm": 0.78426194190979, + "learning_rate": 2.021564245810056e-05, + "loss": 0.1692, + "step": 6035 + }, + { + "epoch": 16.860335195530727, + "grad_norm": 0.9105619788169861, + "learning_rate": 2.0218994413407824e-05, + "loss": 0.2159, + "step": 6036 + }, + { + "epoch": 16.863128491620113, + "grad_norm": 0.568096399307251, + "learning_rate": 2.0222346368715086e-05, + "loss": 0.1618, + "step": 6037 + }, + { + "epoch": 16.865921787709496, + "grad_norm": 0.6662490367889404, + "learning_rate": 2.0225698324022347e-05, + "loss": 0.194, + "step": 6038 + }, + { + "epoch": 16.868715083798882, + "grad_norm": 0.7135036587715149, + "learning_rate": 2.022905027932961e-05, + "loss": 0.1721, + "step": 6039 + }, + { + "epoch": 16.87150837988827, + "grad_norm": 0.65996253490448, + "learning_rate": 2.0232402234636873e-05, + "loss": 0.1564, + "step": 6040 + }, + { + "epoch": 16.874301675977655, + "grad_norm": 3.8892197608947754, + "learning_rate": 2.0235754189944134e-05, + "loss": 0.1638, + "step": 6041 + }, + { + "epoch": 16.877094972067038, + "grad_norm": 1.240559458732605, + "learning_rate": 2.02391061452514e-05, + "loss": 0.1745, + "step": 6042 + }, + { + "epoch": 16.879888268156424, + "grad_norm": 0.7238781452178955, + "learning_rate": 2.024245810055866e-05, + "loss": 0.2056, + "step": 6043 + }, + { + "epoch": 16.88268156424581, + "grad_norm": 1.5841728448867798, + "learning_rate": 2.0245810055865922e-05, + "loss": 0.1858, + "step": 6044 + }, + { + "epoch": 16.885474860335197, + "grad_norm": 1.5160754919052124, + "learning_rate": 2.0249162011173186e-05, + "loss": 0.1721, + "step": 6045 + }, + { + "epoch": 16.88826815642458, + "grad_norm": 3.4320528507232666, + "learning_rate": 2.0252513966480448e-05, + "loss": 0.1727, + "step": 6046 + }, + { + "epoch": 16.891061452513966, + "grad_norm": 2.4956600666046143, + "learning_rate": 2.025586592178771e-05, + "loss": 0.1773, + "step": 6047 + }, + { + "epoch": 16.893854748603353, + "grad_norm": 1.225408673286438, + "learning_rate": 2.0259217877094974e-05, + "loss": 0.1637, + "step": 6048 + }, + { + "epoch": 16.89664804469274, + "grad_norm": 1.3635185956954956, + "learning_rate": 2.0262569832402235e-05, + "loss": 0.1832, + "step": 6049 + }, + { + "epoch": 16.899441340782122, + "grad_norm": 1.222791314125061, + "learning_rate": 2.02659217877095e-05, + "loss": 0.182, + "step": 6050 + }, + { + "epoch": 16.90223463687151, + "grad_norm": 1.1559112071990967, + "learning_rate": 2.026927374301676e-05, + "loss": 0.1772, + "step": 6051 + }, + { + "epoch": 16.905027932960895, + "grad_norm": 1.237762451171875, + "learning_rate": 2.0272625698324023e-05, + "loss": 0.2264, + "step": 6052 + }, + { + "epoch": 16.907821229050278, + "grad_norm": 3.4707272052764893, + "learning_rate": 2.0275977653631287e-05, + "loss": 0.2205, + "step": 6053 + }, + { + "epoch": 16.910614525139664, + "grad_norm": 1.7735618352890015, + "learning_rate": 2.027932960893855e-05, + "loss": 0.1913, + "step": 6054 + }, + { + "epoch": 16.91340782122905, + "grad_norm": 0.8020482063293457, + "learning_rate": 2.028268156424581e-05, + "loss": 0.1854, + "step": 6055 + }, + { + "epoch": 16.916201117318437, + "grad_norm": 1.4416261911392212, + "learning_rate": 2.0286033519553075e-05, + "loss": 0.1868, + "step": 6056 + }, + { + "epoch": 16.91899441340782, + "grad_norm": 1.0427569150924683, + "learning_rate": 2.0289385474860336e-05, + "loss": 0.1697, + "step": 6057 + }, + { + "epoch": 16.921787709497206, + "grad_norm": 0.7009031772613525, + "learning_rate": 2.0292737430167597e-05, + "loss": 0.1839, + "step": 6058 + }, + { + "epoch": 16.924581005586592, + "grad_norm": 1.2043408155441284, + "learning_rate": 2.0296089385474862e-05, + "loss": 0.1814, + "step": 6059 + }, + { + "epoch": 16.92737430167598, + "grad_norm": 0.6934646368026733, + "learning_rate": 2.0299441340782124e-05, + "loss": 0.1616, + "step": 6060 + }, + { + "epoch": 16.93016759776536, + "grad_norm": 0.9731733202934265, + "learning_rate": 2.0302793296089385e-05, + "loss": 0.171, + "step": 6061 + }, + { + "epoch": 16.932960893854748, + "grad_norm": 1.0356038808822632, + "learning_rate": 2.030614525139665e-05, + "loss": 0.1575, + "step": 6062 + }, + { + "epoch": 16.935754189944134, + "grad_norm": 1.6282119750976562, + "learning_rate": 2.030949720670391e-05, + "loss": 0.1814, + "step": 6063 + }, + { + "epoch": 16.93854748603352, + "grad_norm": 0.9591637253761292, + "learning_rate": 2.0312849162011172e-05, + "loss": 0.1925, + "step": 6064 + }, + { + "epoch": 16.941340782122904, + "grad_norm": 0.6405372023582458, + "learning_rate": 2.0316201117318437e-05, + "loss": 0.1433, + "step": 6065 + }, + { + "epoch": 16.94413407821229, + "grad_norm": 1.7418839931488037, + "learning_rate": 2.03195530726257e-05, + "loss": 0.1657, + "step": 6066 + }, + { + "epoch": 16.946927374301676, + "grad_norm": 0.8597262501716614, + "learning_rate": 2.032290502793296e-05, + "loss": 0.1866, + "step": 6067 + }, + { + "epoch": 16.949720670391063, + "grad_norm": 1.099050521850586, + "learning_rate": 2.0326256983240224e-05, + "loss": 0.204, + "step": 6068 + }, + { + "epoch": 16.952513966480446, + "grad_norm": 0.880308210849762, + "learning_rate": 2.0329608938547486e-05, + "loss": 0.1466, + "step": 6069 + }, + { + "epoch": 16.955307262569832, + "grad_norm": 1.1370768547058105, + "learning_rate": 2.033296089385475e-05, + "loss": 0.1701, + "step": 6070 + }, + { + "epoch": 16.95810055865922, + "grad_norm": 0.6207517981529236, + "learning_rate": 2.0336312849162012e-05, + "loss": 0.1651, + "step": 6071 + }, + { + "epoch": 16.960893854748605, + "grad_norm": 0.9535444378852844, + "learning_rate": 2.0339664804469273e-05, + "loss": 0.1626, + "step": 6072 + }, + { + "epoch": 16.963687150837988, + "grad_norm": 1.3601596355438232, + "learning_rate": 2.0343016759776538e-05, + "loss": 0.1892, + "step": 6073 + }, + { + "epoch": 16.966480446927374, + "grad_norm": 1.471065878868103, + "learning_rate": 2.03463687150838e-05, + "loss": 0.1869, + "step": 6074 + }, + { + "epoch": 16.96927374301676, + "grad_norm": 1.9066492319107056, + "learning_rate": 2.034972067039106e-05, + "loss": 0.187, + "step": 6075 + }, + { + "epoch": 16.972067039106147, + "grad_norm": 1.2765334844589233, + "learning_rate": 2.0353072625698325e-05, + "loss": 0.2068, + "step": 6076 + }, + { + "epoch": 16.97486033519553, + "grad_norm": 2.8864009380340576, + "learning_rate": 2.0356424581005587e-05, + "loss": 0.2569, + "step": 6077 + }, + { + "epoch": 16.977653631284916, + "grad_norm": 1.6386641263961792, + "learning_rate": 2.0359776536312848e-05, + "loss": 0.299, + "step": 6078 + }, + { + "epoch": 16.980446927374302, + "grad_norm": 0.6520002484321594, + "learning_rate": 2.0363128491620113e-05, + "loss": 0.1817, + "step": 6079 + }, + { + "epoch": 16.98324022346369, + "grad_norm": 0.784160315990448, + "learning_rate": 2.0366480446927374e-05, + "loss": 0.1655, + "step": 6080 + }, + { + "epoch": 16.98603351955307, + "grad_norm": 0.7592374086380005, + "learning_rate": 2.0369832402234636e-05, + "loss": 0.1731, + "step": 6081 + }, + { + "epoch": 16.988826815642458, + "grad_norm": 0.7578975558280945, + "learning_rate": 2.03731843575419e-05, + "loss": 0.1858, + "step": 6082 + }, + { + "epoch": 16.991620111731844, + "grad_norm": 1.3751890659332275, + "learning_rate": 2.037653631284916e-05, + "loss": 0.1719, + "step": 6083 + }, + { + "epoch": 16.994413407821227, + "grad_norm": 1.1944993734359741, + "learning_rate": 2.0379888268156423e-05, + "loss": 0.1925, + "step": 6084 + }, + { + "epoch": 16.997206703910614, + "grad_norm": 1.0912871360778809, + "learning_rate": 2.0383240223463688e-05, + "loss": 0.1703, + "step": 6085 + }, + { + "epoch": 17.0, + "grad_norm": 2.5676326751708984, + "learning_rate": 2.038659217877095e-05, + "loss": 0.2104, + "step": 6086 + }, + { + "epoch": 17.002793296089386, + "grad_norm": 0.7419584393501282, + "learning_rate": 2.0389944134078214e-05, + "loss": 0.2052, + "step": 6087 + }, + { + "epoch": 17.00558659217877, + "grad_norm": 0.8119620084762573, + "learning_rate": 2.0393296089385475e-05, + "loss": 0.1574, + "step": 6088 + }, + { + "epoch": 17.008379888268156, + "grad_norm": 1.0576229095458984, + "learning_rate": 2.0396648044692736e-05, + "loss": 0.1905, + "step": 6089 + }, + { + "epoch": 17.011173184357542, + "grad_norm": 1.4724912643432617, + "learning_rate": 2.04e-05, + "loss": 0.196, + "step": 6090 + }, + { + "epoch": 17.01396648044693, + "grad_norm": 0.7458210587501526, + "learning_rate": 2.0403351955307263e-05, + "loss": 0.2077, + "step": 6091 + }, + { + "epoch": 17.01675977653631, + "grad_norm": 0.7087650299072266, + "learning_rate": 2.0406703910614524e-05, + "loss": 0.1507, + "step": 6092 + }, + { + "epoch": 17.019553072625698, + "grad_norm": 0.8739767074584961, + "learning_rate": 2.041005586592179e-05, + "loss": 0.19, + "step": 6093 + }, + { + "epoch": 17.022346368715084, + "grad_norm": 0.6435679197311401, + "learning_rate": 2.041340782122905e-05, + "loss": 0.145, + "step": 6094 + }, + { + "epoch": 17.02513966480447, + "grad_norm": 0.6908910274505615, + "learning_rate": 2.041675977653631e-05, + "loss": 0.1587, + "step": 6095 + }, + { + "epoch": 17.027932960893853, + "grad_norm": 0.6849102973937988, + "learning_rate": 2.0420111731843576e-05, + "loss": 0.1587, + "step": 6096 + }, + { + "epoch": 17.03072625698324, + "grad_norm": 1.0070867538452148, + "learning_rate": 2.0423463687150837e-05, + "loss": 0.1805, + "step": 6097 + }, + { + "epoch": 17.033519553072626, + "grad_norm": 0.5021836757659912, + "learning_rate": 2.04268156424581e-05, + "loss": 0.1462, + "step": 6098 + }, + { + "epoch": 17.036312849162012, + "grad_norm": 0.6580798625946045, + "learning_rate": 2.0430167597765363e-05, + "loss": 0.1777, + "step": 6099 + }, + { + "epoch": 17.039106145251395, + "grad_norm": 0.9105710387229919, + "learning_rate": 2.0433519553072625e-05, + "loss": 0.1711, + "step": 6100 + }, + { + "epoch": 17.04189944134078, + "grad_norm": 1.2565882205963135, + "learning_rate": 2.0436871508379886e-05, + "loss": 0.1818, + "step": 6101 + }, + { + "epoch": 17.044692737430168, + "grad_norm": 0.7418020963668823, + "learning_rate": 2.044022346368715e-05, + "loss": 0.1582, + "step": 6102 + }, + { + "epoch": 17.047486033519554, + "grad_norm": 1.4182056188583374, + "learning_rate": 2.0443575418994412e-05, + "loss": 0.1713, + "step": 6103 + }, + { + "epoch": 17.050279329608937, + "grad_norm": 0.9723373055458069, + "learning_rate": 2.0446927374301674e-05, + "loss": 0.1687, + "step": 6104 + }, + { + "epoch": 17.053072625698324, + "grad_norm": 0.9693424105644226, + "learning_rate": 2.045027932960894e-05, + "loss": 0.1738, + "step": 6105 + }, + { + "epoch": 17.05586592178771, + "grad_norm": 1.2427780628204346, + "learning_rate": 2.04536312849162e-05, + "loss": 0.1824, + "step": 6106 + }, + { + "epoch": 17.058659217877096, + "grad_norm": 0.7813154458999634, + "learning_rate": 2.0456983240223464e-05, + "loss": 0.1647, + "step": 6107 + }, + { + "epoch": 17.06145251396648, + "grad_norm": 1.061123013496399, + "learning_rate": 2.0460335195530726e-05, + "loss": 0.1851, + "step": 6108 + }, + { + "epoch": 17.064245810055866, + "grad_norm": 0.857311487197876, + "learning_rate": 2.0463687150837987e-05, + "loss": 0.1666, + "step": 6109 + }, + { + "epoch": 17.067039106145252, + "grad_norm": 1.1863481998443604, + "learning_rate": 2.0467039106145252e-05, + "loss": 0.1643, + "step": 6110 + }, + { + "epoch": 17.06983240223464, + "grad_norm": 4.14269495010376, + "learning_rate": 2.0470391061452513e-05, + "loss": 0.2481, + "step": 6111 + }, + { + "epoch": 17.07262569832402, + "grad_norm": 1.260615348815918, + "learning_rate": 2.0473743016759775e-05, + "loss": 0.164, + "step": 6112 + }, + { + "epoch": 17.075418994413408, + "grad_norm": 0.5938602685928345, + "learning_rate": 2.047709497206704e-05, + "loss": 0.2078, + "step": 6113 + }, + { + "epoch": 17.078212290502794, + "grad_norm": 0.6340175867080688, + "learning_rate": 2.04804469273743e-05, + "loss": 0.1917, + "step": 6114 + }, + { + "epoch": 17.08100558659218, + "grad_norm": 0.6992279887199402, + "learning_rate": 2.0483798882681565e-05, + "loss": 0.201, + "step": 6115 + }, + { + "epoch": 17.083798882681563, + "grad_norm": 0.6323086619377136, + "learning_rate": 2.048715083798883e-05, + "loss": 0.1678, + "step": 6116 + }, + { + "epoch": 17.08659217877095, + "grad_norm": 0.5380739569664001, + "learning_rate": 2.049050279329609e-05, + "loss": 0.1347, + "step": 6117 + }, + { + "epoch": 17.089385474860336, + "grad_norm": 0.7691534161567688, + "learning_rate": 2.0493854748603353e-05, + "loss": 0.1592, + "step": 6118 + }, + { + "epoch": 17.092178770949722, + "grad_norm": 3.65378737449646, + "learning_rate": 2.0497206703910617e-05, + "loss": 0.1631, + "step": 6119 + }, + { + "epoch": 17.094972067039105, + "grad_norm": 0.5914955139160156, + "learning_rate": 2.050055865921788e-05, + "loss": 0.1791, + "step": 6120 + }, + { + "epoch": 17.09776536312849, + "grad_norm": 0.7525936365127563, + "learning_rate": 2.050391061452514e-05, + "loss": 0.1772, + "step": 6121 + }, + { + "epoch": 17.100558659217878, + "grad_norm": 0.9493700861930847, + "learning_rate": 2.0507262569832405e-05, + "loss": 0.1877, + "step": 6122 + }, + { + "epoch": 17.10335195530726, + "grad_norm": 3.9888107776641846, + "learning_rate": 2.0510614525139666e-05, + "loss": 0.2048, + "step": 6123 + }, + { + "epoch": 17.106145251396647, + "grad_norm": 1.340187430381775, + "learning_rate": 2.0513966480446928e-05, + "loss": 0.1977, + "step": 6124 + }, + { + "epoch": 17.108938547486034, + "grad_norm": 1.4611738920211792, + "learning_rate": 2.0517318435754192e-05, + "loss": 0.1823, + "step": 6125 + }, + { + "epoch": 17.11173184357542, + "grad_norm": 1.2073115110397339, + "learning_rate": 2.0520670391061454e-05, + "loss": 0.1592, + "step": 6126 + }, + { + "epoch": 17.114525139664803, + "grad_norm": 0.8448192477226257, + "learning_rate": 2.052402234636872e-05, + "loss": 0.1914, + "step": 6127 + }, + { + "epoch": 17.11731843575419, + "grad_norm": 3.041191339492798, + "learning_rate": 2.052737430167598e-05, + "loss": 0.1825, + "step": 6128 + }, + { + "epoch": 17.120111731843576, + "grad_norm": 5.5335211753845215, + "learning_rate": 2.053072625698324e-05, + "loss": 0.1541, + "step": 6129 + }, + { + "epoch": 17.122905027932962, + "grad_norm": 1.2664488554000854, + "learning_rate": 2.0534078212290506e-05, + "loss": 0.1834, + "step": 6130 + }, + { + "epoch": 17.125698324022345, + "grad_norm": 1.440117597579956, + "learning_rate": 2.0537430167597767e-05, + "loss": 0.2009, + "step": 6131 + }, + { + "epoch": 17.12849162011173, + "grad_norm": 1.2936142683029175, + "learning_rate": 2.054078212290503e-05, + "loss": 0.1891, + "step": 6132 + }, + { + "epoch": 17.131284916201118, + "grad_norm": 1.251384973526001, + "learning_rate": 2.0544134078212293e-05, + "loss": 0.2341, + "step": 6133 + }, + { + "epoch": 17.134078212290504, + "grad_norm": 1.1947511434555054, + "learning_rate": 2.0547486033519555e-05, + "loss": 0.1885, + "step": 6134 + }, + { + "epoch": 17.136871508379887, + "grad_norm": 1.3119763135910034, + "learning_rate": 2.0550837988826816e-05, + "loss": 0.2449, + "step": 6135 + }, + { + "epoch": 17.139664804469273, + "grad_norm": 1.4627337455749512, + "learning_rate": 2.055418994413408e-05, + "loss": 0.2611, + "step": 6136 + }, + { + "epoch": 17.14245810055866, + "grad_norm": 0.7942196130752563, + "learning_rate": 2.0557541899441342e-05, + "loss": 0.1691, + "step": 6137 + }, + { + "epoch": 17.145251396648046, + "grad_norm": 0.5605328679084778, + "learning_rate": 2.0560893854748603e-05, + "loss": 0.1759, + "step": 6138 + }, + { + "epoch": 17.14804469273743, + "grad_norm": 0.5733950138092041, + "learning_rate": 2.0564245810055868e-05, + "loss": 0.1699, + "step": 6139 + }, + { + "epoch": 17.150837988826815, + "grad_norm": 0.5309332013130188, + "learning_rate": 2.056759776536313e-05, + "loss": 0.1813, + "step": 6140 + }, + { + "epoch": 17.1536312849162, + "grad_norm": 0.5456206202507019, + "learning_rate": 2.057094972067039e-05, + "loss": 0.1617, + "step": 6141 + }, + { + "epoch": 17.156424581005588, + "grad_norm": 0.7394863367080688, + "learning_rate": 2.0574301675977656e-05, + "loss": 0.1652, + "step": 6142 + }, + { + "epoch": 17.15921787709497, + "grad_norm": 1.4748857021331787, + "learning_rate": 2.0577653631284917e-05, + "loss": 0.1608, + "step": 6143 + }, + { + "epoch": 17.162011173184357, + "grad_norm": 0.6724356412887573, + "learning_rate": 2.058100558659218e-05, + "loss": 0.1554, + "step": 6144 + }, + { + "epoch": 17.164804469273744, + "grad_norm": 0.5589832067489624, + "learning_rate": 2.0584357541899443e-05, + "loss": 0.1547, + "step": 6145 + }, + { + "epoch": 17.16759776536313, + "grad_norm": 0.7538211941719055, + "learning_rate": 2.0587709497206704e-05, + "loss": 0.2371, + "step": 6146 + }, + { + "epoch": 17.170391061452513, + "grad_norm": 1.116688847541809, + "learning_rate": 2.059106145251397e-05, + "loss": 0.1679, + "step": 6147 + }, + { + "epoch": 17.1731843575419, + "grad_norm": 0.7323755025863647, + "learning_rate": 2.059441340782123e-05, + "loss": 0.1772, + "step": 6148 + }, + { + "epoch": 17.175977653631286, + "grad_norm": 0.833161473274231, + "learning_rate": 2.0597765363128492e-05, + "loss": 0.1743, + "step": 6149 + }, + { + "epoch": 17.178770949720672, + "grad_norm": 0.8149371147155762, + "learning_rate": 2.0601117318435756e-05, + "loss": 0.1582, + "step": 6150 + }, + { + "epoch": 17.181564245810055, + "grad_norm": 0.7853180766105652, + "learning_rate": 2.0604469273743018e-05, + "loss": 0.1983, + "step": 6151 + }, + { + "epoch": 17.18435754189944, + "grad_norm": 1.207940697669983, + "learning_rate": 2.060782122905028e-05, + "loss": 0.1808, + "step": 6152 + }, + { + "epoch": 17.187150837988828, + "grad_norm": 1.7862366437911987, + "learning_rate": 2.0611173184357544e-05, + "loss": 0.1809, + "step": 6153 + }, + { + "epoch": 17.189944134078214, + "grad_norm": 1.2146223783493042, + "learning_rate": 2.0614525139664805e-05, + "loss": 0.1568, + "step": 6154 + }, + { + "epoch": 17.192737430167597, + "grad_norm": 0.929789125919342, + "learning_rate": 2.0617877094972067e-05, + "loss": 0.1799, + "step": 6155 + }, + { + "epoch": 17.195530726256983, + "grad_norm": 0.9499701261520386, + "learning_rate": 2.062122905027933e-05, + "loss": 0.1867, + "step": 6156 + }, + { + "epoch": 17.19832402234637, + "grad_norm": 1.4057719707489014, + "learning_rate": 2.0624581005586593e-05, + "loss": 0.1982, + "step": 6157 + }, + { + "epoch": 17.201117318435756, + "grad_norm": 1.2080193758010864, + "learning_rate": 2.0627932960893854e-05, + "loss": 0.1776, + "step": 6158 + }, + { + "epoch": 17.20391061452514, + "grad_norm": 2.0756421089172363, + "learning_rate": 2.063128491620112e-05, + "loss": 0.1872, + "step": 6159 + }, + { + "epoch": 17.206703910614525, + "grad_norm": 1.7939406633377075, + "learning_rate": 2.063463687150838e-05, + "loss": 0.1852, + "step": 6160 + }, + { + "epoch": 17.20949720670391, + "grad_norm": 1.463123083114624, + "learning_rate": 2.063798882681564e-05, + "loss": 0.2333, + "step": 6161 + }, + { + "epoch": 17.212290502793294, + "grad_norm": 0.6609507203102112, + "learning_rate": 2.0641340782122906e-05, + "loss": 0.1865, + "step": 6162 + }, + { + "epoch": 17.21508379888268, + "grad_norm": 0.5193392038345337, + "learning_rate": 2.0644692737430167e-05, + "loss": 0.1829, + "step": 6163 + }, + { + "epoch": 17.217877094972067, + "grad_norm": 0.7989371418952942, + "learning_rate": 2.0648044692737432e-05, + "loss": 0.1745, + "step": 6164 + }, + { + "epoch": 17.220670391061454, + "grad_norm": 1.2597942352294922, + "learning_rate": 2.0651396648044694e-05, + "loss": 0.1821, + "step": 6165 + }, + { + "epoch": 17.223463687150836, + "grad_norm": 0.625677764415741, + "learning_rate": 2.0654748603351955e-05, + "loss": 0.1707, + "step": 6166 + }, + { + "epoch": 17.226256983240223, + "grad_norm": 0.9276540279388428, + "learning_rate": 2.065810055865922e-05, + "loss": 0.2023, + "step": 6167 + }, + { + "epoch": 17.22905027932961, + "grad_norm": 0.6214410662651062, + "learning_rate": 2.066145251396648e-05, + "loss": 0.1765, + "step": 6168 + }, + { + "epoch": 17.231843575418996, + "grad_norm": 0.8957575559616089, + "learning_rate": 2.0664804469273742e-05, + "loss": 0.1606, + "step": 6169 + }, + { + "epoch": 17.23463687150838, + "grad_norm": 0.5936989784240723, + "learning_rate": 2.0668156424581007e-05, + "loss": 0.1679, + "step": 6170 + }, + { + "epoch": 17.237430167597765, + "grad_norm": 0.6989136934280396, + "learning_rate": 2.067150837988827e-05, + "loss": 0.1905, + "step": 6171 + }, + { + "epoch": 17.24022346368715, + "grad_norm": 0.7486265897750854, + "learning_rate": 2.067486033519553e-05, + "loss": 0.1594, + "step": 6172 + }, + { + "epoch": 17.243016759776538, + "grad_norm": 0.9789102673530579, + "learning_rate": 2.0678212290502794e-05, + "loss": 0.1703, + "step": 6173 + }, + { + "epoch": 17.24581005586592, + "grad_norm": 1.1164276599884033, + "learning_rate": 2.0681564245810056e-05, + "loss": 0.1635, + "step": 6174 + }, + { + "epoch": 17.248603351955307, + "grad_norm": 0.6672331094741821, + "learning_rate": 2.0684916201117317e-05, + "loss": 0.1753, + "step": 6175 + }, + { + "epoch": 17.251396648044693, + "grad_norm": 0.729102611541748, + "learning_rate": 2.0688268156424582e-05, + "loss": 0.1467, + "step": 6176 + }, + { + "epoch": 17.25418994413408, + "grad_norm": 1.085320234298706, + "learning_rate": 2.0691620111731843e-05, + "loss": 0.2042, + "step": 6177 + }, + { + "epoch": 17.256983240223462, + "grad_norm": 0.7337324619293213, + "learning_rate": 2.0694972067039105e-05, + "loss": 0.1525, + "step": 6178 + }, + { + "epoch": 17.25977653631285, + "grad_norm": 3.495490074157715, + "learning_rate": 2.069832402234637e-05, + "loss": 0.176, + "step": 6179 + }, + { + "epoch": 17.262569832402235, + "grad_norm": 1.3160804510116577, + "learning_rate": 2.070167597765363e-05, + "loss": 0.174, + "step": 6180 + }, + { + "epoch": 17.26536312849162, + "grad_norm": 1.6641128063201904, + "learning_rate": 2.0705027932960895e-05, + "loss": 0.2036, + "step": 6181 + }, + { + "epoch": 17.268156424581004, + "grad_norm": 1.6539602279663086, + "learning_rate": 2.0708379888268157e-05, + "loss": 0.2061, + "step": 6182 + }, + { + "epoch": 17.27094972067039, + "grad_norm": 1.6503206491470337, + "learning_rate": 2.0711731843575418e-05, + "loss": 0.1775, + "step": 6183 + }, + { + "epoch": 17.273743016759777, + "grad_norm": 1.2897707223892212, + "learning_rate": 2.0715083798882683e-05, + "loss": 0.2169, + "step": 6184 + }, + { + "epoch": 17.276536312849164, + "grad_norm": 1.4945714473724365, + "learning_rate": 2.0718435754189944e-05, + "loss": 0.2136, + "step": 6185 + }, + { + "epoch": 17.279329608938546, + "grad_norm": 2.0337021350860596, + "learning_rate": 2.0721787709497206e-05, + "loss": 0.2718, + "step": 6186 + }, + { + "epoch": 17.282122905027933, + "grad_norm": 0.6006680130958557, + "learning_rate": 2.072513966480447e-05, + "loss": 0.2008, + "step": 6187 + }, + { + "epoch": 17.28491620111732, + "grad_norm": 0.85647052526474, + "learning_rate": 2.072849162011173e-05, + "loss": 0.2002, + "step": 6188 + }, + { + "epoch": 17.287709497206706, + "grad_norm": 0.5388543605804443, + "learning_rate": 2.0731843575418993e-05, + "loss": 0.1968, + "step": 6189 + }, + { + "epoch": 17.29050279329609, + "grad_norm": 0.6081468462944031, + "learning_rate": 2.0735195530726258e-05, + "loss": 0.1931, + "step": 6190 + }, + { + "epoch": 17.293296089385475, + "grad_norm": 0.8678581714630127, + "learning_rate": 2.073854748603352e-05, + "loss": 0.2028, + "step": 6191 + }, + { + "epoch": 17.29608938547486, + "grad_norm": 1.2055460214614868, + "learning_rate": 2.074189944134078e-05, + "loss": 0.2043, + "step": 6192 + }, + { + "epoch": 17.298882681564244, + "grad_norm": 0.6281092166900635, + "learning_rate": 2.0745251396648045e-05, + "loss": 0.1964, + "step": 6193 + }, + { + "epoch": 17.30167597765363, + "grad_norm": 0.5454573631286621, + "learning_rate": 2.0748603351955306e-05, + "loss": 0.1621, + "step": 6194 + }, + { + "epoch": 17.304469273743017, + "grad_norm": 0.6036502718925476, + "learning_rate": 2.0751955307262568e-05, + "loss": 0.1736, + "step": 6195 + }, + { + "epoch": 17.307262569832403, + "grad_norm": 0.5458258390426636, + "learning_rate": 2.0755307262569833e-05, + "loss": 0.1718, + "step": 6196 + }, + { + "epoch": 17.310055865921786, + "grad_norm": 1.1945610046386719, + "learning_rate": 2.0758659217877094e-05, + "loss": 0.1825, + "step": 6197 + }, + { + "epoch": 17.312849162011172, + "grad_norm": 1.2657074928283691, + "learning_rate": 2.0762011173184355e-05, + "loss": 0.1707, + "step": 6198 + }, + { + "epoch": 17.31564245810056, + "grad_norm": 0.948506772518158, + "learning_rate": 2.076536312849162e-05, + "loss": 0.1631, + "step": 6199 + }, + { + "epoch": 17.318435754189945, + "grad_norm": 0.7397724986076355, + "learning_rate": 2.076871508379888e-05, + "loss": 0.1726, + "step": 6200 + }, + { + "epoch": 17.321229050279328, + "grad_norm": 0.9546582698822021, + "learning_rate": 2.0772067039106146e-05, + "loss": 0.1525, + "step": 6201 + }, + { + "epoch": 17.324022346368714, + "grad_norm": 2.128571033477783, + "learning_rate": 2.0775418994413407e-05, + "loss": 0.1629, + "step": 6202 + }, + { + "epoch": 17.3268156424581, + "grad_norm": 0.9295694828033447, + "learning_rate": 2.077877094972067e-05, + "loss": 0.1742, + "step": 6203 + }, + { + "epoch": 17.329608938547487, + "grad_norm": 2.133315086364746, + "learning_rate": 2.0782122905027933e-05, + "loss": 0.2203, + "step": 6204 + }, + { + "epoch": 17.33240223463687, + "grad_norm": 1.247321367263794, + "learning_rate": 2.0785474860335195e-05, + "loss": 0.1475, + "step": 6205 + }, + { + "epoch": 17.335195530726256, + "grad_norm": 2.042318105697632, + "learning_rate": 2.0788826815642456e-05, + "loss": 0.1951, + "step": 6206 + }, + { + "epoch": 17.337988826815643, + "grad_norm": 1.155673861503601, + "learning_rate": 2.079217877094972e-05, + "loss": 0.187, + "step": 6207 + }, + { + "epoch": 17.34078212290503, + "grad_norm": 0.9616581797599792, + "learning_rate": 2.0795530726256982e-05, + "loss": 0.175, + "step": 6208 + }, + { + "epoch": 17.343575418994412, + "grad_norm": 1.487188458442688, + "learning_rate": 2.0798882681564244e-05, + "loss": 0.1952, + "step": 6209 + }, + { + "epoch": 17.3463687150838, + "grad_norm": 2.5579276084899902, + "learning_rate": 2.080223463687151e-05, + "loss": 0.2121, + "step": 6210 + }, + { + "epoch": 17.349162011173185, + "grad_norm": 3.8243026733398438, + "learning_rate": 2.080558659217877e-05, + "loss": 0.2416, + "step": 6211 + }, + { + "epoch": 17.35195530726257, + "grad_norm": 0.5964899659156799, + "learning_rate": 2.0808938547486034e-05, + "loss": 0.1838, + "step": 6212 + }, + { + "epoch": 17.354748603351954, + "grad_norm": 0.8702559471130371, + "learning_rate": 2.08122905027933e-05, + "loss": 0.1843, + "step": 6213 + }, + { + "epoch": 17.35754189944134, + "grad_norm": 0.8856706619262695, + "learning_rate": 2.081564245810056e-05, + "loss": 0.2147, + "step": 6214 + }, + { + "epoch": 17.360335195530727, + "grad_norm": 0.58980393409729, + "learning_rate": 2.0818994413407822e-05, + "loss": 0.1821, + "step": 6215 + }, + { + "epoch": 17.363128491620113, + "grad_norm": 0.7314609885215759, + "learning_rate": 2.0822346368715087e-05, + "loss": 0.1719, + "step": 6216 + }, + { + "epoch": 17.365921787709496, + "grad_norm": 3.8556668758392334, + "learning_rate": 2.0825698324022348e-05, + "loss": 0.1909, + "step": 6217 + }, + { + "epoch": 17.368715083798882, + "grad_norm": 0.6443886160850525, + "learning_rate": 2.082905027932961e-05, + "loss": 0.2192, + "step": 6218 + }, + { + "epoch": 17.37150837988827, + "grad_norm": 0.955737829208374, + "learning_rate": 2.0832402234636874e-05, + "loss": 0.1706, + "step": 6219 + }, + { + "epoch": 17.374301675977655, + "grad_norm": 0.8394348621368408, + "learning_rate": 2.0835754189944135e-05, + "loss": 0.1822, + "step": 6220 + }, + { + "epoch": 17.377094972067038, + "grad_norm": 0.8398982286453247, + "learning_rate": 2.08391061452514e-05, + "loss": 0.1777, + "step": 6221 + }, + { + "epoch": 17.379888268156424, + "grad_norm": 0.8126246333122253, + "learning_rate": 2.084245810055866e-05, + "loss": 0.1765, + "step": 6222 + }, + { + "epoch": 17.38268156424581, + "grad_norm": 1.0272585153579712, + "learning_rate": 2.0845810055865923e-05, + "loss": 0.1839, + "step": 6223 + }, + { + "epoch": 17.385474860335197, + "grad_norm": 0.9674957990646362, + "learning_rate": 2.0849162011173187e-05, + "loss": 0.1746, + "step": 6224 + }, + { + "epoch": 17.38826815642458, + "grad_norm": 0.7997081875801086, + "learning_rate": 2.085251396648045e-05, + "loss": 0.1783, + "step": 6225 + }, + { + "epoch": 17.391061452513966, + "grad_norm": 1.0567878484725952, + "learning_rate": 2.085586592178771e-05, + "loss": 0.1763, + "step": 6226 + }, + { + "epoch": 17.393854748603353, + "grad_norm": 1.2645505666732788, + "learning_rate": 2.0859217877094975e-05, + "loss": 0.1618, + "step": 6227 + }, + { + "epoch": 17.39664804469274, + "grad_norm": 0.8732265830039978, + "learning_rate": 2.0862569832402236e-05, + "loss": 0.1678, + "step": 6228 + }, + { + "epoch": 17.399441340782122, + "grad_norm": 1.0609465837478638, + "learning_rate": 2.0865921787709498e-05, + "loss": 0.1627, + "step": 6229 + }, + { + "epoch": 17.40223463687151, + "grad_norm": 0.7626755237579346, + "learning_rate": 2.0869273743016762e-05, + "loss": 0.1616, + "step": 6230 + }, + { + "epoch": 17.405027932960895, + "grad_norm": 1.0164796113967896, + "learning_rate": 2.0872625698324024e-05, + "loss": 0.1978, + "step": 6231 + }, + { + "epoch": 17.407821229050278, + "grad_norm": 0.884812593460083, + "learning_rate": 2.0875977653631285e-05, + "loss": 0.1756, + "step": 6232 + }, + { + "epoch": 17.410614525139664, + "grad_norm": 1.4557331800460815, + "learning_rate": 2.087932960893855e-05, + "loss": 0.2265, + "step": 6233 + }, + { + "epoch": 17.41340782122905, + "grad_norm": 1.4996304512023926, + "learning_rate": 2.088268156424581e-05, + "loss": 0.1927, + "step": 6234 + }, + { + "epoch": 17.416201117318437, + "grad_norm": 2.5791945457458496, + "learning_rate": 2.0886033519553072e-05, + "loss": 0.2036, + "step": 6235 + }, + { + "epoch": 17.41899441340782, + "grad_norm": 1.9572151899337769, + "learning_rate": 2.0889385474860337e-05, + "loss": 0.272, + "step": 6236 + }, + { + "epoch": 17.421787709497206, + "grad_norm": 0.6069589853286743, + "learning_rate": 2.08927374301676e-05, + "loss": 0.2353, + "step": 6237 + }, + { + "epoch": 17.424581005586592, + "grad_norm": 0.7402726411819458, + "learning_rate": 2.0896089385474863e-05, + "loss": 0.1919, + "step": 6238 + }, + { + "epoch": 17.42737430167598, + "grad_norm": 0.7997162938117981, + "learning_rate": 2.0899441340782125e-05, + "loss": 0.1814, + "step": 6239 + }, + { + "epoch": 17.43016759776536, + "grad_norm": 0.8955235481262207, + "learning_rate": 2.0902793296089386e-05, + "loss": 0.1786, + "step": 6240 + }, + { + "epoch": 17.432960893854748, + "grad_norm": 1.3312714099884033, + "learning_rate": 2.090614525139665e-05, + "loss": 0.1769, + "step": 6241 + }, + { + "epoch": 17.435754189944134, + "grad_norm": 0.9765978455543518, + "learning_rate": 2.0909497206703912e-05, + "loss": 0.1835, + "step": 6242 + }, + { + "epoch": 17.43854748603352, + "grad_norm": 0.6378364562988281, + "learning_rate": 2.0912849162011173e-05, + "loss": 0.1859, + "step": 6243 + }, + { + "epoch": 17.441340782122904, + "grad_norm": 0.6843480467796326, + "learning_rate": 2.0916201117318438e-05, + "loss": 0.2048, + "step": 6244 + }, + { + "epoch": 17.44413407821229, + "grad_norm": 0.4966743588447571, + "learning_rate": 2.09195530726257e-05, + "loss": 0.1459, + "step": 6245 + }, + { + "epoch": 17.446927374301676, + "grad_norm": 0.9907164573669434, + "learning_rate": 2.092290502793296e-05, + "loss": 0.1793, + "step": 6246 + }, + { + "epoch": 17.449720670391063, + "grad_norm": 1.0266932249069214, + "learning_rate": 2.0926256983240226e-05, + "loss": 0.1657, + "step": 6247 + }, + { + "epoch": 17.452513966480446, + "grad_norm": 0.971979558467865, + "learning_rate": 2.0929608938547487e-05, + "loss": 0.1861, + "step": 6248 + }, + { + "epoch": 17.455307262569832, + "grad_norm": 1.1176115274429321, + "learning_rate": 2.0932960893854748e-05, + "loss": 0.16, + "step": 6249 + }, + { + "epoch": 17.45810055865922, + "grad_norm": 1.5883920192718506, + "learning_rate": 2.0936312849162013e-05, + "loss": 0.1808, + "step": 6250 + }, + { + "epoch": 17.460893854748605, + "grad_norm": 0.6574868559837341, + "learning_rate": 2.0939664804469274e-05, + "loss": 0.1531, + "step": 6251 + }, + { + "epoch": 17.463687150837988, + "grad_norm": 1.2130508422851562, + "learning_rate": 2.0943016759776536e-05, + "loss": 0.151, + "step": 6252 + }, + { + "epoch": 17.466480446927374, + "grad_norm": 0.7255468368530273, + "learning_rate": 2.09463687150838e-05, + "loss": 0.192, + "step": 6253 + }, + { + "epoch": 17.46927374301676, + "grad_norm": 1.092699408531189, + "learning_rate": 2.0949720670391062e-05, + "loss": 0.1822, + "step": 6254 + }, + { + "epoch": 17.472067039106147, + "grad_norm": 1.3975285291671753, + "learning_rate": 2.0953072625698323e-05, + "loss": 0.1801, + "step": 6255 + }, + { + "epoch": 17.47486033519553, + "grad_norm": 2.3122506141662598, + "learning_rate": 2.0956424581005588e-05, + "loss": 0.1715, + "step": 6256 + }, + { + "epoch": 17.477653631284916, + "grad_norm": 1.1680665016174316, + "learning_rate": 2.095977653631285e-05, + "loss": 0.1611, + "step": 6257 + }, + { + "epoch": 17.480446927374302, + "grad_norm": 0.8010864853858948, + "learning_rate": 2.0963128491620114e-05, + "loss": 0.1875, + "step": 6258 + }, + { + "epoch": 17.48324022346369, + "grad_norm": 1.0084494352340698, + "learning_rate": 2.0966480446927375e-05, + "loss": 0.181, + "step": 6259 + }, + { + "epoch": 17.48603351955307, + "grad_norm": 2.3053722381591797, + "learning_rate": 2.0969832402234637e-05, + "loss": 0.2124, + "step": 6260 + }, + { + "epoch": 17.488826815642458, + "grad_norm": 1.363692283630371, + "learning_rate": 2.09731843575419e-05, + "loss": 0.245, + "step": 6261 + }, + { + "epoch": 17.491620111731844, + "grad_norm": 0.6865903735160828, + "learning_rate": 2.0976536312849163e-05, + "loss": 0.2051, + "step": 6262 + }, + { + "epoch": 17.49441340782123, + "grad_norm": 0.5113147497177124, + "learning_rate": 2.0979888268156424e-05, + "loss": 0.1983, + "step": 6263 + }, + { + "epoch": 17.497206703910614, + "grad_norm": 0.8279721736907959, + "learning_rate": 2.098324022346369e-05, + "loss": 0.2069, + "step": 6264 + }, + { + "epoch": 17.5, + "grad_norm": 0.7479648590087891, + "learning_rate": 2.098659217877095e-05, + "loss": 0.1857, + "step": 6265 + }, + { + "epoch": 17.502793296089386, + "grad_norm": 0.6877180933952332, + "learning_rate": 2.098994413407821e-05, + "loss": 0.1767, + "step": 6266 + }, + { + "epoch": 17.505586592178773, + "grad_norm": 0.5920787453651428, + "learning_rate": 2.0993296089385476e-05, + "loss": 0.179, + "step": 6267 + }, + { + "epoch": 17.508379888268156, + "grad_norm": 0.9232597947120667, + "learning_rate": 2.0996648044692737e-05, + "loss": 0.1668, + "step": 6268 + }, + { + "epoch": 17.511173184357542, + "grad_norm": 0.8566027879714966, + "learning_rate": 2.1e-05, + "loss": 0.1877, + "step": 6269 + }, + { + "epoch": 17.51396648044693, + "grad_norm": 0.8429160714149475, + "learning_rate": 2.1003351955307264e-05, + "loss": 0.1765, + "step": 6270 + }, + { + "epoch": 17.51675977653631, + "grad_norm": 0.6731369495391846, + "learning_rate": 2.1006703910614525e-05, + "loss": 0.1564, + "step": 6271 + }, + { + "epoch": 17.519553072625698, + "grad_norm": 0.8928357362747192, + "learning_rate": 2.1010055865921786e-05, + "loss": 0.1844, + "step": 6272 + }, + { + "epoch": 17.522346368715084, + "grad_norm": 0.9538990259170532, + "learning_rate": 2.101340782122905e-05, + "loss": 0.1658, + "step": 6273 + }, + { + "epoch": 17.52513966480447, + "grad_norm": 1.1164343357086182, + "learning_rate": 2.1016759776536312e-05, + "loss": 0.144, + "step": 6274 + }, + { + "epoch": 17.527932960893853, + "grad_norm": 1.5288039445877075, + "learning_rate": 2.1020111731843577e-05, + "loss": 0.1803, + "step": 6275 + }, + { + "epoch": 17.53072625698324, + "grad_norm": 1.5509990453720093, + "learning_rate": 2.102346368715084e-05, + "loss": 0.1708, + "step": 6276 + }, + { + "epoch": 17.533519553072626, + "grad_norm": 2.07979154586792, + "learning_rate": 2.10268156424581e-05, + "loss": 0.1713, + "step": 6277 + }, + { + "epoch": 17.536312849162012, + "grad_norm": 1.0653797388076782, + "learning_rate": 2.1030167597765364e-05, + "loss": 0.1824, + "step": 6278 + }, + { + "epoch": 17.539106145251395, + "grad_norm": 1.36439847946167, + "learning_rate": 2.1033519553072626e-05, + "loss": 0.1528, + "step": 6279 + }, + { + "epoch": 17.54189944134078, + "grad_norm": 2.338469982147217, + "learning_rate": 2.1036871508379887e-05, + "loss": 0.1799, + "step": 6280 + }, + { + "epoch": 17.544692737430168, + "grad_norm": 1.1485556364059448, + "learning_rate": 2.1040223463687152e-05, + "loss": 0.1537, + "step": 6281 + }, + { + "epoch": 17.547486033519554, + "grad_norm": 1.1405435800552368, + "learning_rate": 2.1043575418994413e-05, + "loss": 0.2102, + "step": 6282 + }, + { + "epoch": 17.550279329608937, + "grad_norm": 1.3352869749069214, + "learning_rate": 2.1046927374301675e-05, + "loss": 0.1927, + "step": 6283 + }, + { + "epoch": 17.553072625698324, + "grad_norm": 1.1354867219924927, + "learning_rate": 2.105027932960894e-05, + "loss": 0.1769, + "step": 6284 + }, + { + "epoch": 17.55586592178771, + "grad_norm": 1.3513585329055786, + "learning_rate": 2.10536312849162e-05, + "loss": 0.1943, + "step": 6285 + }, + { + "epoch": 17.558659217877096, + "grad_norm": 2.353461742401123, + "learning_rate": 2.1056983240223462e-05, + "loss": 0.222, + "step": 6286 + }, + { + "epoch": 17.56145251396648, + "grad_norm": 1.0757452249526978, + "learning_rate": 2.1060335195530727e-05, + "loss": 0.1847, + "step": 6287 + }, + { + "epoch": 17.564245810055866, + "grad_norm": 0.9010648727416992, + "learning_rate": 2.1063687150837988e-05, + "loss": 0.1589, + "step": 6288 + }, + { + "epoch": 17.567039106145252, + "grad_norm": 0.6671091318130493, + "learning_rate": 2.106703910614525e-05, + "loss": 0.1738, + "step": 6289 + }, + { + "epoch": 17.56983240223464, + "grad_norm": 0.6769937872886658, + "learning_rate": 2.1070391061452514e-05, + "loss": 0.1748, + "step": 6290 + }, + { + "epoch": 17.57262569832402, + "grad_norm": 3.79670786857605, + "learning_rate": 2.1073743016759776e-05, + "loss": 0.2204, + "step": 6291 + }, + { + "epoch": 17.575418994413408, + "grad_norm": 1.0408565998077393, + "learning_rate": 2.1077094972067037e-05, + "loss": 0.1887, + "step": 6292 + }, + { + "epoch": 17.578212290502794, + "grad_norm": 2.675527811050415, + "learning_rate": 2.10804469273743e-05, + "loss": 0.1843, + "step": 6293 + }, + { + "epoch": 17.58100558659218, + "grad_norm": 1.5849608182907104, + "learning_rate": 2.1083798882681563e-05, + "loss": 0.212, + "step": 6294 + }, + { + "epoch": 17.583798882681563, + "grad_norm": 0.5914669632911682, + "learning_rate": 2.1087150837988828e-05, + "loss": 0.1788, + "step": 6295 + }, + { + "epoch": 17.58659217877095, + "grad_norm": 2.6292507648468018, + "learning_rate": 2.109050279329609e-05, + "loss": 0.1965, + "step": 6296 + }, + { + "epoch": 17.589385474860336, + "grad_norm": 1.2523884773254395, + "learning_rate": 2.109385474860335e-05, + "loss": 0.2156, + "step": 6297 + }, + { + "epoch": 17.592178770949722, + "grad_norm": 1.9870976209640503, + "learning_rate": 2.1097206703910615e-05, + "loss": 0.1614, + "step": 6298 + }, + { + "epoch": 17.594972067039105, + "grad_norm": 1.3748586177825928, + "learning_rate": 2.1100558659217876e-05, + "loss": 0.2079, + "step": 6299 + }, + { + "epoch": 17.59776536312849, + "grad_norm": 4.6869730949401855, + "learning_rate": 2.1103910614525138e-05, + "loss": 0.1566, + "step": 6300 + }, + { + "epoch": 17.600558659217878, + "grad_norm": 1.1724293231964111, + "learning_rate": 2.1107262569832403e-05, + "loss": 0.1782, + "step": 6301 + }, + { + "epoch": 17.60335195530726, + "grad_norm": 0.770584225654602, + "learning_rate": 2.1110614525139664e-05, + "loss": 0.168, + "step": 6302 + }, + { + "epoch": 17.606145251396647, + "grad_norm": 1.2818559408187866, + "learning_rate": 2.1113966480446925e-05, + "loss": 0.1701, + "step": 6303 + }, + { + "epoch": 17.608938547486034, + "grad_norm": 2.9428024291992188, + "learning_rate": 2.111731843575419e-05, + "loss": 0.1718, + "step": 6304 + }, + { + "epoch": 17.61173184357542, + "grad_norm": 1.198986530303955, + "learning_rate": 2.112067039106145e-05, + "loss": 0.1527, + "step": 6305 + }, + { + "epoch": 17.614525139664803, + "grad_norm": 0.9388255476951599, + "learning_rate": 2.1124022346368713e-05, + "loss": 0.1789, + "step": 6306 + }, + { + "epoch": 17.61731843575419, + "grad_norm": 5.511944770812988, + "learning_rate": 2.1127374301675977e-05, + "loss": 0.1611, + "step": 6307 + }, + { + "epoch": 17.620111731843576, + "grad_norm": 1.5841012001037598, + "learning_rate": 2.1130726256983242e-05, + "loss": 0.2021, + "step": 6308 + }, + { + "epoch": 17.622905027932962, + "grad_norm": 2.2797157764434814, + "learning_rate": 2.1134078212290503e-05, + "loss": 0.2258, + "step": 6309 + }, + { + "epoch": 17.625698324022345, + "grad_norm": 2.888566732406616, + "learning_rate": 2.1137430167597768e-05, + "loss": 0.2352, + "step": 6310 + }, + { + "epoch": 17.62849162011173, + "grad_norm": 2.6833248138427734, + "learning_rate": 2.114078212290503e-05, + "loss": 0.3307, + "step": 6311 + }, + { + "epoch": 17.631284916201118, + "grad_norm": 0.7389935255050659, + "learning_rate": 2.114413407821229e-05, + "loss": 0.2027, + "step": 6312 + }, + { + "epoch": 17.634078212290504, + "grad_norm": 0.6562895774841309, + "learning_rate": 2.1147486033519556e-05, + "loss": 0.1633, + "step": 6313 + }, + { + "epoch": 17.636871508379887, + "grad_norm": 1.0681227445602417, + "learning_rate": 2.1150837988826817e-05, + "loss": 0.1814, + "step": 6314 + }, + { + "epoch": 17.639664804469273, + "grad_norm": 1.0457907915115356, + "learning_rate": 2.1154189944134082e-05, + "loss": 0.1901, + "step": 6315 + }, + { + "epoch": 17.64245810055866, + "grad_norm": 0.7678982019424438, + "learning_rate": 2.1157541899441343e-05, + "loss": 0.2123, + "step": 6316 + }, + { + "epoch": 17.645251396648046, + "grad_norm": 0.930755615234375, + "learning_rate": 2.1160893854748604e-05, + "loss": 0.159, + "step": 6317 + }, + { + "epoch": 17.64804469273743, + "grad_norm": 0.6003066301345825, + "learning_rate": 2.116424581005587e-05, + "loss": 0.188, + "step": 6318 + }, + { + "epoch": 17.650837988826815, + "grad_norm": 0.8669813871383667, + "learning_rate": 2.116759776536313e-05, + "loss": 0.1885, + "step": 6319 + }, + { + "epoch": 17.6536312849162, + "grad_norm": 0.9442107081413269, + "learning_rate": 2.1170949720670392e-05, + "loss": 0.16, + "step": 6320 + }, + { + "epoch": 17.656424581005588, + "grad_norm": 0.7868989109992981, + "learning_rate": 2.1174301675977657e-05, + "loss": 0.1716, + "step": 6321 + }, + { + "epoch": 17.65921787709497, + "grad_norm": 0.6792812943458557, + "learning_rate": 2.1177653631284918e-05, + "loss": 0.1625, + "step": 6322 + }, + { + "epoch": 17.662011173184357, + "grad_norm": 1.1901204586029053, + "learning_rate": 2.118100558659218e-05, + "loss": 0.1947, + "step": 6323 + }, + { + "epoch": 17.664804469273744, + "grad_norm": 1.1718626022338867, + "learning_rate": 2.1184357541899444e-05, + "loss": 0.1887, + "step": 6324 + }, + { + "epoch": 17.66759776536313, + "grad_norm": 0.7319594025611877, + "learning_rate": 2.1187709497206705e-05, + "loss": 0.1819, + "step": 6325 + }, + { + "epoch": 17.670391061452513, + "grad_norm": 0.7918857932090759, + "learning_rate": 2.1191061452513967e-05, + "loss": 0.1528, + "step": 6326 + }, + { + "epoch": 17.6731843575419, + "grad_norm": 0.8012604713439941, + "learning_rate": 2.119441340782123e-05, + "loss": 0.1976, + "step": 6327 + }, + { + "epoch": 17.675977653631286, + "grad_norm": 1.804272174835205, + "learning_rate": 2.1197765363128493e-05, + "loss": 0.1826, + "step": 6328 + }, + { + "epoch": 17.678770949720672, + "grad_norm": 0.8945886492729187, + "learning_rate": 2.1201117318435754e-05, + "loss": 0.1598, + "step": 6329 + }, + { + "epoch": 17.681564245810055, + "grad_norm": 0.6596807837486267, + "learning_rate": 2.120446927374302e-05, + "loss": 0.1557, + "step": 6330 + }, + { + "epoch": 17.68435754189944, + "grad_norm": 0.957363486289978, + "learning_rate": 2.120782122905028e-05, + "loss": 0.181, + "step": 6331 + }, + { + "epoch": 17.687150837988828, + "grad_norm": 4.753033638000488, + "learning_rate": 2.1211173184357545e-05, + "loss": 0.1779, + "step": 6332 + }, + { + "epoch": 17.689944134078214, + "grad_norm": 1.8500969409942627, + "learning_rate": 2.1214525139664806e-05, + "loss": 0.2023, + "step": 6333 + }, + { + "epoch": 17.692737430167597, + "grad_norm": 0.987273633480072, + "learning_rate": 2.1217877094972068e-05, + "loss": 0.1706, + "step": 6334 + }, + { + "epoch": 17.695530726256983, + "grad_norm": 2.1105189323425293, + "learning_rate": 2.1221229050279332e-05, + "loss": 0.237, + "step": 6335 + }, + { + "epoch": 17.69832402234637, + "grad_norm": 3.4068174362182617, + "learning_rate": 2.1224581005586594e-05, + "loss": 0.2628, + "step": 6336 + }, + { + "epoch": 17.701117318435756, + "grad_norm": 0.5312089920043945, + "learning_rate": 2.1227932960893855e-05, + "loss": 0.1698, + "step": 6337 + }, + { + "epoch": 17.70391061452514, + "grad_norm": 0.948086142539978, + "learning_rate": 2.123128491620112e-05, + "loss": 0.2128, + "step": 6338 + }, + { + "epoch": 17.706703910614525, + "grad_norm": 0.6946269869804382, + "learning_rate": 2.123463687150838e-05, + "loss": 0.1824, + "step": 6339 + }, + { + "epoch": 17.70949720670391, + "grad_norm": 0.6276617050170898, + "learning_rate": 2.1237988826815642e-05, + "loss": 0.2015, + "step": 6340 + }, + { + "epoch": 17.712290502793294, + "grad_norm": 2.1627752780914307, + "learning_rate": 2.1241340782122907e-05, + "loss": 0.1713, + "step": 6341 + }, + { + "epoch": 17.71508379888268, + "grad_norm": 0.7968697547912598, + "learning_rate": 2.124469273743017e-05, + "loss": 0.1763, + "step": 6342 + }, + { + "epoch": 17.717877094972067, + "grad_norm": 1.6451997756958008, + "learning_rate": 2.124804469273743e-05, + "loss": 0.1895, + "step": 6343 + }, + { + "epoch": 17.720670391061454, + "grad_norm": 1.1933448314666748, + "learning_rate": 2.1251396648044695e-05, + "loss": 0.1515, + "step": 6344 + }, + { + "epoch": 17.723463687150836, + "grad_norm": 1.5117391347885132, + "learning_rate": 2.1254748603351956e-05, + "loss": 0.1701, + "step": 6345 + }, + { + "epoch": 17.726256983240223, + "grad_norm": 0.9739192724227905, + "learning_rate": 2.1258100558659217e-05, + "loss": 0.1939, + "step": 6346 + }, + { + "epoch": 17.72905027932961, + "grad_norm": 0.8233645558357239, + "learning_rate": 2.1261452513966482e-05, + "loss": 0.1932, + "step": 6347 + }, + { + "epoch": 17.731843575418996, + "grad_norm": 1.2804356813430786, + "learning_rate": 2.1264804469273743e-05, + "loss": 0.1396, + "step": 6348 + }, + { + "epoch": 17.73463687150838, + "grad_norm": 1.3807358741760254, + "learning_rate": 2.1268156424581005e-05, + "loss": 0.1668, + "step": 6349 + }, + { + "epoch": 17.737430167597765, + "grad_norm": 2.954758405685425, + "learning_rate": 2.127150837988827e-05, + "loss": 0.1745, + "step": 6350 + }, + { + "epoch": 17.74022346368715, + "grad_norm": 0.8845977187156677, + "learning_rate": 2.127486033519553e-05, + "loss": 0.1527, + "step": 6351 + }, + { + "epoch": 17.743016759776538, + "grad_norm": 1.0305477380752563, + "learning_rate": 2.1278212290502796e-05, + "loss": 0.1895, + "step": 6352 + }, + { + "epoch": 17.74581005586592, + "grad_norm": 1.6552091836929321, + "learning_rate": 2.1281564245810057e-05, + "loss": 0.1766, + "step": 6353 + }, + { + "epoch": 17.748603351955307, + "grad_norm": 0.9653846621513367, + "learning_rate": 2.1284916201117318e-05, + "loss": 0.1758, + "step": 6354 + }, + { + "epoch": 17.751396648044693, + "grad_norm": 1.362411379814148, + "learning_rate": 2.1288268156424583e-05, + "loss": 0.1756, + "step": 6355 + }, + { + "epoch": 17.75418994413408, + "grad_norm": 2.053684949874878, + "learning_rate": 2.1291620111731844e-05, + "loss": 0.1947, + "step": 6356 + }, + { + "epoch": 17.756983240223462, + "grad_norm": 0.9923363327980042, + "learning_rate": 2.1294972067039106e-05, + "loss": 0.1842, + "step": 6357 + }, + { + "epoch": 17.75977653631285, + "grad_norm": 1.1346365213394165, + "learning_rate": 2.129832402234637e-05, + "loss": 0.1909, + "step": 6358 + }, + { + "epoch": 17.762569832402235, + "grad_norm": 1.6032673120498657, + "learning_rate": 2.1301675977653632e-05, + "loss": 0.2028, + "step": 6359 + }, + { + "epoch": 17.76536312849162, + "grad_norm": 4.6361589431762695, + "learning_rate": 2.1305027932960893e-05, + "loss": 0.1992, + "step": 6360 + }, + { + "epoch": 17.768156424581004, + "grad_norm": 2.8720407485961914, + "learning_rate": 2.1308379888268158e-05, + "loss": 0.2761, + "step": 6361 + }, + { + "epoch": 17.77094972067039, + "grad_norm": 1.1659364700317383, + "learning_rate": 2.131173184357542e-05, + "loss": 0.1878, + "step": 6362 + }, + { + "epoch": 17.773743016759777, + "grad_norm": 0.4952990412712097, + "learning_rate": 2.131508379888268e-05, + "loss": 0.1632, + "step": 6363 + }, + { + "epoch": 17.776536312849164, + "grad_norm": 0.5296598076820374, + "learning_rate": 2.1318435754189945e-05, + "loss": 0.1577, + "step": 6364 + }, + { + "epoch": 17.779329608938546, + "grad_norm": 0.6613145470619202, + "learning_rate": 2.1321787709497207e-05, + "loss": 0.1646, + "step": 6365 + }, + { + "epoch": 17.782122905027933, + "grad_norm": 0.9957204461097717, + "learning_rate": 2.1325139664804468e-05, + "loss": 0.2107, + "step": 6366 + }, + { + "epoch": 17.78491620111732, + "grad_norm": 0.9180287718772888, + "learning_rate": 2.1328491620111733e-05, + "loss": 0.1907, + "step": 6367 + }, + { + "epoch": 17.787709497206706, + "grad_norm": 0.530378520488739, + "learning_rate": 2.1331843575418994e-05, + "loss": 0.1884, + "step": 6368 + }, + { + "epoch": 17.79050279329609, + "grad_norm": 0.6046866774559021, + "learning_rate": 2.133519553072626e-05, + "loss": 0.1649, + "step": 6369 + }, + { + "epoch": 17.793296089385475, + "grad_norm": 0.6931436657905579, + "learning_rate": 2.133854748603352e-05, + "loss": 0.1767, + "step": 6370 + }, + { + "epoch": 17.79608938547486, + "grad_norm": 0.6749007701873779, + "learning_rate": 2.134189944134078e-05, + "loss": 0.1999, + "step": 6371 + }, + { + "epoch": 17.798882681564244, + "grad_norm": 1.1788979768753052, + "learning_rate": 2.1345251396648046e-05, + "loss": 0.1939, + "step": 6372 + }, + { + "epoch": 17.80167597765363, + "grad_norm": 1.1068018674850464, + "learning_rate": 2.1348603351955307e-05, + "loss": 0.1849, + "step": 6373 + }, + { + "epoch": 17.804469273743017, + "grad_norm": 0.7316251397132874, + "learning_rate": 2.135195530726257e-05, + "loss": 0.1686, + "step": 6374 + }, + { + "epoch": 17.807262569832403, + "grad_norm": 1.2378922700881958, + "learning_rate": 2.1355307262569834e-05, + "loss": 0.1602, + "step": 6375 + }, + { + "epoch": 17.810055865921786, + "grad_norm": 0.9503644704818726, + "learning_rate": 2.1358659217877095e-05, + "loss": 0.1659, + "step": 6376 + }, + { + "epoch": 17.812849162011172, + "grad_norm": 0.8158159852027893, + "learning_rate": 2.1362011173184356e-05, + "loss": 0.167, + "step": 6377 + }, + { + "epoch": 17.81564245810056, + "grad_norm": 1.09507155418396, + "learning_rate": 2.136536312849162e-05, + "loss": 0.1848, + "step": 6378 + }, + { + "epoch": 17.818435754189945, + "grad_norm": 1.2421398162841797, + "learning_rate": 2.1368715083798882e-05, + "loss": 0.1859, + "step": 6379 + }, + { + "epoch": 17.821229050279328, + "grad_norm": 0.781992495059967, + "learning_rate": 2.1372067039106144e-05, + "loss": 0.1564, + "step": 6380 + }, + { + "epoch": 17.824022346368714, + "grad_norm": 2.3897008895874023, + "learning_rate": 2.137541899441341e-05, + "loss": 0.1872, + "step": 6381 + }, + { + "epoch": 17.8268156424581, + "grad_norm": 0.9674078226089478, + "learning_rate": 2.137877094972067e-05, + "loss": 0.1873, + "step": 6382 + }, + { + "epoch": 17.829608938547487, + "grad_norm": 2.0902562141418457, + "learning_rate": 2.138212290502793e-05, + "loss": 0.1667, + "step": 6383 + }, + { + "epoch": 17.83240223463687, + "grad_norm": 1.1688218116760254, + "learning_rate": 2.1385474860335196e-05, + "loss": 0.2116, + "step": 6384 + }, + { + "epoch": 17.835195530726256, + "grad_norm": 1.4278844594955444, + "learning_rate": 2.1388826815642457e-05, + "loss": 0.2086, + "step": 6385 + }, + { + "epoch": 17.837988826815643, + "grad_norm": 2.2551801204681396, + "learning_rate": 2.139217877094972e-05, + "loss": 0.2312, + "step": 6386 + }, + { + "epoch": 17.84078212290503, + "grad_norm": 1.9209116697311401, + "learning_rate": 2.1395530726256983e-05, + "loss": 0.1905, + "step": 6387 + }, + { + "epoch": 17.843575418994412, + "grad_norm": 1.2401552200317383, + "learning_rate": 2.1398882681564245e-05, + "loss": 0.1886, + "step": 6388 + }, + { + "epoch": 17.8463687150838, + "grad_norm": 0.7493698000907898, + "learning_rate": 2.140223463687151e-05, + "loss": 0.1729, + "step": 6389 + }, + { + "epoch": 17.849162011173185, + "grad_norm": 0.5474507808685303, + "learning_rate": 2.140558659217877e-05, + "loss": 0.1759, + "step": 6390 + }, + { + "epoch": 17.85195530726257, + "grad_norm": 0.9039734601974487, + "learning_rate": 2.1408938547486032e-05, + "loss": 0.1943, + "step": 6391 + }, + { + "epoch": 17.854748603351954, + "grad_norm": 1.637297511100769, + "learning_rate": 2.1412290502793297e-05, + "loss": 0.1946, + "step": 6392 + }, + { + "epoch": 17.85754189944134, + "grad_norm": 1.3480383157730103, + "learning_rate": 2.1415642458100558e-05, + "loss": 0.2042, + "step": 6393 + }, + { + "epoch": 17.860335195530727, + "grad_norm": 0.6761502027511597, + "learning_rate": 2.141899441340782e-05, + "loss": 0.221, + "step": 6394 + }, + { + "epoch": 17.863128491620113, + "grad_norm": 0.643243670463562, + "learning_rate": 2.1422346368715084e-05, + "loss": 0.193, + "step": 6395 + }, + { + "epoch": 17.865921787709496, + "grad_norm": 0.5942966938018799, + "learning_rate": 2.1425698324022346e-05, + "loss": 0.1656, + "step": 6396 + }, + { + "epoch": 17.868715083798882, + "grad_norm": 0.831484854221344, + "learning_rate": 2.1429050279329607e-05, + "loss": 0.1848, + "step": 6397 + }, + { + "epoch": 17.87150837988827, + "grad_norm": 0.7980980277061462, + "learning_rate": 2.143240223463687e-05, + "loss": 0.1716, + "step": 6398 + }, + { + "epoch": 17.874301675977655, + "grad_norm": 0.8499917387962341, + "learning_rate": 2.1435754189944133e-05, + "loss": 0.1676, + "step": 6399 + }, + { + "epoch": 17.877094972067038, + "grad_norm": 0.8423599004745483, + "learning_rate": 2.1439106145251394e-05, + "loss": 0.1682, + "step": 6400 + }, + { + "epoch": 17.879888268156424, + "grad_norm": 0.6008381843566895, + "learning_rate": 2.144245810055866e-05, + "loss": 0.1644, + "step": 6401 + }, + { + "epoch": 17.88268156424581, + "grad_norm": 0.683998167514801, + "learning_rate": 2.144581005586592e-05, + "loss": 0.165, + "step": 6402 + }, + { + "epoch": 17.885474860335197, + "grad_norm": 1.4266544580459595, + "learning_rate": 2.1449162011173182e-05, + "loss": 0.1893, + "step": 6403 + }, + { + "epoch": 17.88826815642458, + "grad_norm": 0.872183620929718, + "learning_rate": 2.1452513966480446e-05, + "loss": 0.164, + "step": 6404 + }, + { + "epoch": 17.891061452513966, + "grad_norm": 1.0124653577804565, + "learning_rate": 2.145586592178771e-05, + "loss": 0.1993, + "step": 6405 + }, + { + "epoch": 17.893854748603353, + "grad_norm": 2.0978105068206787, + "learning_rate": 2.1459217877094973e-05, + "loss": 0.1564, + "step": 6406 + }, + { + "epoch": 17.89664804469274, + "grad_norm": 1.639642596244812, + "learning_rate": 2.1462569832402237e-05, + "loss": 0.1534, + "step": 6407 + }, + { + "epoch": 17.899441340782122, + "grad_norm": 1.2808945178985596, + "learning_rate": 2.14659217877095e-05, + "loss": 0.1635, + "step": 6408 + }, + { + "epoch": 17.90223463687151, + "grad_norm": 1.7703239917755127, + "learning_rate": 2.1469273743016763e-05, + "loss": 0.1863, + "step": 6409 + }, + { + "epoch": 17.905027932960895, + "grad_norm": 1.1202274560928345, + "learning_rate": 2.1472625698324025e-05, + "loss": 0.1761, + "step": 6410 + }, + { + "epoch": 17.907821229050278, + "grad_norm": 1.5194226503372192, + "learning_rate": 2.1475977653631286e-05, + "loss": 0.2777, + "step": 6411 + }, + { + "epoch": 17.910614525139664, + "grad_norm": 0.6041330695152283, + "learning_rate": 2.147932960893855e-05, + "loss": 0.2022, + "step": 6412 + }, + { + "epoch": 17.91340782122905, + "grad_norm": 0.684644877910614, + "learning_rate": 2.1482681564245812e-05, + "loss": 0.1874, + "step": 6413 + }, + { + "epoch": 17.916201117318437, + "grad_norm": 0.6737584471702576, + "learning_rate": 2.1486033519553073e-05, + "loss": 0.1607, + "step": 6414 + }, + { + "epoch": 17.91899441340782, + "grad_norm": 0.8411276340484619, + "learning_rate": 2.1489385474860338e-05, + "loss": 0.1686, + "step": 6415 + }, + { + "epoch": 17.921787709497206, + "grad_norm": 0.9113268852233887, + "learning_rate": 2.14927374301676e-05, + "loss": 0.2181, + "step": 6416 + }, + { + "epoch": 17.924581005586592, + "grad_norm": 0.5599209666252136, + "learning_rate": 2.149608938547486e-05, + "loss": 0.1935, + "step": 6417 + }, + { + "epoch": 17.92737430167598, + "grad_norm": 0.7919377684593201, + "learning_rate": 2.1499441340782126e-05, + "loss": 0.1991, + "step": 6418 + }, + { + "epoch": 17.93016759776536, + "grad_norm": 0.8976686000823975, + "learning_rate": 2.1502793296089387e-05, + "loss": 0.1907, + "step": 6419 + }, + { + "epoch": 17.932960893854748, + "grad_norm": 1.4403138160705566, + "learning_rate": 2.150614525139665e-05, + "loss": 0.1738, + "step": 6420 + }, + { + "epoch": 17.935754189944134, + "grad_norm": 0.9853764176368713, + "learning_rate": 2.1509497206703913e-05, + "loss": 0.1789, + "step": 6421 + }, + { + "epoch": 17.93854748603352, + "grad_norm": 0.5838185548782349, + "learning_rate": 2.1512849162011174e-05, + "loss": 0.1666, + "step": 6422 + }, + { + "epoch": 17.941340782122904, + "grad_norm": 0.632374107837677, + "learning_rate": 2.1516201117318436e-05, + "loss": 0.201, + "step": 6423 + }, + { + "epoch": 17.94413407821229, + "grad_norm": 1.3562716245651245, + "learning_rate": 2.15195530726257e-05, + "loss": 0.1886, + "step": 6424 + }, + { + "epoch": 17.946927374301676, + "grad_norm": 1.9985496997833252, + "learning_rate": 2.1522905027932962e-05, + "loss": 0.2028, + "step": 6425 + }, + { + "epoch": 17.949720670391063, + "grad_norm": 1.3077149391174316, + "learning_rate": 2.1526256983240227e-05, + "loss": 0.2049, + "step": 6426 + }, + { + "epoch": 17.952513966480446, + "grad_norm": 0.7740367650985718, + "learning_rate": 2.1529608938547488e-05, + "loss": 0.1902, + "step": 6427 + }, + { + "epoch": 17.955307262569832, + "grad_norm": 0.8100991249084473, + "learning_rate": 2.153296089385475e-05, + "loss": 0.1511, + "step": 6428 + }, + { + "epoch": 17.95810055865922, + "grad_norm": 1.3421355485916138, + "learning_rate": 2.1536312849162014e-05, + "loss": 0.1761, + "step": 6429 + }, + { + "epoch": 17.960893854748605, + "grad_norm": 0.7710348963737488, + "learning_rate": 2.1539664804469275e-05, + "loss": 0.205, + "step": 6430 + }, + { + "epoch": 17.963687150837988, + "grad_norm": 1.1426604986190796, + "learning_rate": 2.1543016759776537e-05, + "loss": 0.1943, + "step": 6431 + }, + { + "epoch": 17.966480446927374, + "grad_norm": 3.0819990634918213, + "learning_rate": 2.15463687150838e-05, + "loss": 0.1459, + "step": 6432 + }, + { + "epoch": 17.96927374301676, + "grad_norm": 1.0667576789855957, + "learning_rate": 2.1549720670391063e-05, + "loss": 0.1712, + "step": 6433 + }, + { + "epoch": 17.972067039106147, + "grad_norm": 2.601743221282959, + "learning_rate": 2.1553072625698324e-05, + "loss": 0.1963, + "step": 6434 + }, + { + "epoch": 17.97486033519553, + "grad_norm": 0.8258443474769592, + "learning_rate": 2.155642458100559e-05, + "loss": 0.195, + "step": 6435 + }, + { + "epoch": 17.977653631284916, + "grad_norm": 1.3864750862121582, + "learning_rate": 2.155977653631285e-05, + "loss": 0.2226, + "step": 6436 + }, + { + "epoch": 17.980446927374302, + "grad_norm": 0.6586910486221313, + "learning_rate": 2.156312849162011e-05, + "loss": 0.1792, + "step": 6437 + }, + { + "epoch": 17.98324022346369, + "grad_norm": 0.6658897995948792, + "learning_rate": 2.1566480446927376e-05, + "loss": 0.1862, + "step": 6438 + }, + { + "epoch": 17.98603351955307, + "grad_norm": 0.7029942870140076, + "learning_rate": 2.1569832402234638e-05, + "loss": 0.1904, + "step": 6439 + }, + { + "epoch": 17.988826815642458, + "grad_norm": 1.2114858627319336, + "learning_rate": 2.15731843575419e-05, + "loss": 0.2152, + "step": 6440 + }, + { + "epoch": 17.991620111731844, + "grad_norm": 0.6276467442512512, + "learning_rate": 2.1576536312849164e-05, + "loss": 0.1572, + "step": 6441 + }, + { + "epoch": 17.994413407821227, + "grad_norm": 1.0446746349334717, + "learning_rate": 2.1579888268156425e-05, + "loss": 0.1735, + "step": 6442 + }, + { + "epoch": 17.997206703910614, + "grad_norm": 1.1247352361679077, + "learning_rate": 2.1583240223463686e-05, + "loss": 0.1854, + "step": 6443 + }, + { + "epoch": 18.0, + "grad_norm": 1.687760353088379, + "learning_rate": 2.158659217877095e-05, + "loss": 0.2286, + "step": 6444 + }, + { + "epoch": 18.002793296089386, + "grad_norm": 0.9982057213783264, + "learning_rate": 2.1589944134078212e-05, + "loss": 0.2368, + "step": 6445 + }, + { + "epoch": 18.00558659217877, + "grad_norm": 0.6658401489257812, + "learning_rate": 2.1593296089385477e-05, + "loss": 0.1856, + "step": 6446 + }, + { + "epoch": 18.008379888268156, + "grad_norm": 0.5766686797142029, + "learning_rate": 2.159664804469274e-05, + "loss": 0.1852, + "step": 6447 + }, + { + "epoch": 18.011173184357542, + "grad_norm": 0.71714848279953, + "learning_rate": 2.16e-05, + "loss": 0.1659, + "step": 6448 + }, + { + "epoch": 18.01396648044693, + "grad_norm": 1.0198194980621338, + "learning_rate": 2.1603351955307265e-05, + "loss": 0.1721, + "step": 6449 + }, + { + "epoch": 18.01675977653631, + "grad_norm": 1.226637601852417, + "learning_rate": 2.1606703910614526e-05, + "loss": 0.1851, + "step": 6450 + }, + { + "epoch": 18.019553072625698, + "grad_norm": 1.6701922416687012, + "learning_rate": 2.1610055865921787e-05, + "loss": 0.2285, + "step": 6451 + }, + { + "epoch": 18.022346368715084, + "grad_norm": 0.9297319650650024, + "learning_rate": 2.1613407821229052e-05, + "loss": 0.1679, + "step": 6452 + }, + { + "epoch": 18.02513966480447, + "grad_norm": 1.376517415046692, + "learning_rate": 2.1616759776536313e-05, + "loss": 0.1925, + "step": 6453 + }, + { + "epoch": 18.027932960893853, + "grad_norm": 1.0973191261291504, + "learning_rate": 2.1620111731843575e-05, + "loss": 0.1662, + "step": 6454 + }, + { + "epoch": 18.03072625698324, + "grad_norm": 1.5733816623687744, + "learning_rate": 2.162346368715084e-05, + "loss": 0.171, + "step": 6455 + }, + { + "epoch": 18.033519553072626, + "grad_norm": 0.9926033616065979, + "learning_rate": 2.16268156424581e-05, + "loss": 0.1805, + "step": 6456 + }, + { + "epoch": 18.036312849162012, + "grad_norm": 1.119775652885437, + "learning_rate": 2.1630167597765362e-05, + "loss": 0.1952, + "step": 6457 + }, + { + "epoch": 18.039106145251395, + "grad_norm": 0.8606629967689514, + "learning_rate": 2.1633519553072627e-05, + "loss": 0.1669, + "step": 6458 + }, + { + "epoch": 18.04189944134078, + "grad_norm": 0.8384382724761963, + "learning_rate": 2.1636871508379888e-05, + "loss": 0.1767, + "step": 6459 + }, + { + "epoch": 18.044692737430168, + "grad_norm": 1.0589582920074463, + "learning_rate": 2.164022346368715e-05, + "loss": 0.1789, + "step": 6460 + }, + { + "epoch": 18.047486033519554, + "grad_norm": 2.482854127883911, + "learning_rate": 2.1643575418994414e-05, + "loss": 0.1537, + "step": 6461 + }, + { + "epoch": 18.050279329608937, + "grad_norm": 1.1355960369110107, + "learning_rate": 2.1646927374301676e-05, + "loss": 0.1941, + "step": 6462 + }, + { + "epoch": 18.053072625698324, + "grad_norm": 1.3371891975402832, + "learning_rate": 2.165027932960894e-05, + "loss": 0.1739, + "step": 6463 + }, + { + "epoch": 18.05586592178771, + "grad_norm": 1.1095671653747559, + "learning_rate": 2.1653631284916202e-05, + "loss": 0.1935, + "step": 6464 + }, + { + "epoch": 18.058659217877096, + "grad_norm": 1.361465573310852, + "learning_rate": 2.1656983240223463e-05, + "loss": 0.1769, + "step": 6465 + }, + { + "epoch": 18.06145251396648, + "grad_norm": 1.036400556564331, + "learning_rate": 2.1660335195530728e-05, + "loss": 0.1717, + "step": 6466 + }, + { + "epoch": 18.064245810055866, + "grad_norm": 1.7252082824707031, + "learning_rate": 2.166368715083799e-05, + "loss": 0.2232, + "step": 6467 + }, + { + "epoch": 18.067039106145252, + "grad_norm": 1.3320544958114624, + "learning_rate": 2.166703910614525e-05, + "loss": 0.212, + "step": 6468 + }, + { + "epoch": 18.06983240223464, + "grad_norm": 1.8421549797058105, + "learning_rate": 2.1670391061452515e-05, + "loss": 0.2641, + "step": 6469 + }, + { + "epoch": 18.07262569832402, + "grad_norm": 0.7527721524238586, + "learning_rate": 2.1673743016759777e-05, + "loss": 0.1895, + "step": 6470 + }, + { + "epoch": 18.075418994413408, + "grad_norm": 0.5827402472496033, + "learning_rate": 2.1677094972067038e-05, + "loss": 0.19, + "step": 6471 + }, + { + "epoch": 18.078212290502794, + "grad_norm": 1.2014063596725464, + "learning_rate": 2.1680446927374303e-05, + "loss": 0.1731, + "step": 6472 + }, + { + "epoch": 18.08100558659218, + "grad_norm": 1.6046949625015259, + "learning_rate": 2.1683798882681564e-05, + "loss": 0.1898, + "step": 6473 + }, + { + "epoch": 18.083798882681563, + "grad_norm": 1.8158786296844482, + "learning_rate": 2.1687150837988825e-05, + "loss": 0.1737, + "step": 6474 + }, + { + "epoch": 18.08659217877095, + "grad_norm": 0.7260918617248535, + "learning_rate": 2.169050279329609e-05, + "loss": 0.1845, + "step": 6475 + }, + { + "epoch": 18.089385474860336, + "grad_norm": 0.5897423624992371, + "learning_rate": 2.169385474860335e-05, + "loss": 0.1817, + "step": 6476 + }, + { + "epoch": 18.092178770949722, + "grad_norm": 0.5267654061317444, + "learning_rate": 2.1697206703910613e-05, + "loss": 0.1811, + "step": 6477 + }, + { + "epoch": 18.094972067039105, + "grad_norm": 1.8073841333389282, + "learning_rate": 2.1700558659217877e-05, + "loss": 0.1714, + "step": 6478 + }, + { + "epoch": 18.09776536312849, + "grad_norm": 0.7667608857154846, + "learning_rate": 2.170391061452514e-05, + "loss": 0.1569, + "step": 6479 + }, + { + "epoch": 18.100558659217878, + "grad_norm": 1.4188144207000732, + "learning_rate": 2.17072625698324e-05, + "loss": 0.15, + "step": 6480 + }, + { + "epoch": 18.10335195530726, + "grad_norm": 0.6091414093971252, + "learning_rate": 2.1710614525139665e-05, + "loss": 0.1673, + "step": 6481 + }, + { + "epoch": 18.106145251396647, + "grad_norm": 0.8405781388282776, + "learning_rate": 2.1713966480446926e-05, + "loss": 0.1758, + "step": 6482 + }, + { + "epoch": 18.108938547486034, + "grad_norm": 1.2117761373519897, + "learning_rate": 2.171731843575419e-05, + "loss": 0.1923, + "step": 6483 + }, + { + "epoch": 18.11173184357542, + "grad_norm": 0.8737962245941162, + "learning_rate": 2.1720670391061452e-05, + "loss": 0.1799, + "step": 6484 + }, + { + "epoch": 18.114525139664803, + "grad_norm": 0.5819265246391296, + "learning_rate": 2.1724022346368714e-05, + "loss": 0.1925, + "step": 6485 + }, + { + "epoch": 18.11731843575419, + "grad_norm": 1.767973780632019, + "learning_rate": 2.172737430167598e-05, + "loss": 0.1587, + "step": 6486 + }, + { + "epoch": 18.120111731843576, + "grad_norm": 1.2421834468841553, + "learning_rate": 2.173072625698324e-05, + "loss": 0.1827, + "step": 6487 + }, + { + "epoch": 18.122905027932962, + "grad_norm": 0.8495721817016602, + "learning_rate": 2.17340782122905e-05, + "loss": 0.1815, + "step": 6488 + }, + { + "epoch": 18.125698324022345, + "grad_norm": 1.2869646549224854, + "learning_rate": 2.1737430167597766e-05, + "loss": 0.1813, + "step": 6489 + }, + { + "epoch": 18.12849162011173, + "grad_norm": 1.1360547542572021, + "learning_rate": 2.1740782122905027e-05, + "loss": 0.1932, + "step": 6490 + }, + { + "epoch": 18.131284916201118, + "grad_norm": 1.3709173202514648, + "learning_rate": 2.174413407821229e-05, + "loss": 0.1616, + "step": 6491 + }, + { + "epoch": 18.134078212290504, + "grad_norm": 1.5549310445785522, + "learning_rate": 2.1747486033519553e-05, + "loss": 0.1842, + "step": 6492 + }, + { + "epoch": 18.136871508379887, + "grad_norm": 1.2440437078475952, + "learning_rate": 2.1750837988826815e-05, + "loss": 0.169, + "step": 6493 + }, + { + "epoch": 18.139664804469273, + "grad_norm": 1.7598258256912231, + "learning_rate": 2.1754189944134076e-05, + "loss": 0.2448, + "step": 6494 + }, + { + "epoch": 18.14245810055866, + "grad_norm": 0.7422120571136475, + "learning_rate": 2.175754189944134e-05, + "loss": 0.1709, + "step": 6495 + }, + { + "epoch": 18.145251396648046, + "grad_norm": 0.6887905597686768, + "learning_rate": 2.1760893854748602e-05, + "loss": 0.1701, + "step": 6496 + }, + { + "epoch": 18.14804469273743, + "grad_norm": 4.78468132019043, + "learning_rate": 2.1764245810055863e-05, + "loss": 0.186, + "step": 6497 + }, + { + "epoch": 18.150837988826815, + "grad_norm": 0.7779973149299622, + "learning_rate": 2.1767597765363128e-05, + "loss": 0.2149, + "step": 6498 + }, + { + "epoch": 18.1536312849162, + "grad_norm": 2.3326730728149414, + "learning_rate": 2.177094972067039e-05, + "loss": 0.2037, + "step": 6499 + }, + { + "epoch": 18.156424581005588, + "grad_norm": 0.7523485422134399, + "learning_rate": 2.1774301675977654e-05, + "loss": 0.1671, + "step": 6500 + }, + { + "epoch": 18.15921787709497, + "grad_norm": 1.0710772275924683, + "learning_rate": 2.177765363128492e-05, + "loss": 0.1796, + "step": 6501 + }, + { + "epoch": 18.162011173184357, + "grad_norm": 1.649715542793274, + "learning_rate": 2.178100558659218e-05, + "loss": 0.1801, + "step": 6502 + }, + { + "epoch": 18.164804469273744, + "grad_norm": 0.653749406337738, + "learning_rate": 2.1784357541899445e-05, + "loss": 0.1685, + "step": 6503 + }, + { + "epoch": 18.16759776536313, + "grad_norm": 4.158608913421631, + "learning_rate": 2.1787709497206706e-05, + "loss": 0.1782, + "step": 6504 + }, + { + "epoch": 18.170391061452513, + "grad_norm": 0.8472782969474792, + "learning_rate": 2.1791061452513968e-05, + "loss": 0.1834, + "step": 6505 + }, + { + "epoch": 18.1731843575419, + "grad_norm": 0.9562689661979675, + "learning_rate": 2.1794413407821232e-05, + "loss": 0.1762, + "step": 6506 + }, + { + "epoch": 18.175977653631286, + "grad_norm": 0.48943156003952026, + "learning_rate": 2.1797765363128494e-05, + "loss": 0.1596, + "step": 6507 + }, + { + "epoch": 18.178770949720672, + "grad_norm": 0.9141087532043457, + "learning_rate": 2.1801117318435755e-05, + "loss": 0.1725, + "step": 6508 + }, + { + "epoch": 18.181564245810055, + "grad_norm": 0.9323158860206604, + "learning_rate": 2.180446927374302e-05, + "loss": 0.1565, + "step": 6509 + }, + { + "epoch": 18.18435754189944, + "grad_norm": 1.4699829816818237, + "learning_rate": 2.180782122905028e-05, + "loss": 0.1593, + "step": 6510 + }, + { + "epoch": 18.187150837988828, + "grad_norm": 1.1214752197265625, + "learning_rate": 2.1811173184357543e-05, + "loss": 0.1605, + "step": 6511 + }, + { + "epoch": 18.189944134078214, + "grad_norm": 0.6363785862922668, + "learning_rate": 2.1814525139664807e-05, + "loss": 0.152, + "step": 6512 + }, + { + "epoch": 18.192737430167597, + "grad_norm": 0.8552382588386536, + "learning_rate": 2.181787709497207e-05, + "loss": 0.1589, + "step": 6513 + }, + { + "epoch": 18.195530726256983, + "grad_norm": 0.6033236384391785, + "learning_rate": 2.182122905027933e-05, + "loss": 0.1496, + "step": 6514 + }, + { + "epoch": 18.19832402234637, + "grad_norm": 1.2588881254196167, + "learning_rate": 2.1824581005586595e-05, + "loss": 0.1505, + "step": 6515 + }, + { + "epoch": 18.201117318435756, + "grad_norm": 1.0830405950546265, + "learning_rate": 2.1827932960893856e-05, + "loss": 0.1605, + "step": 6516 + }, + { + "epoch": 18.20391061452514, + "grad_norm": 11.930985450744629, + "learning_rate": 2.1831284916201117e-05, + "loss": 0.1897, + "step": 6517 + }, + { + "epoch": 18.206703910614525, + "grad_norm": 1.2917513847351074, + "learning_rate": 2.1834636871508382e-05, + "loss": 0.2265, + "step": 6518 + }, + { + "epoch": 18.20949720670391, + "grad_norm": 4.788974761962891, + "learning_rate": 2.1837988826815643e-05, + "loss": 0.2462, + "step": 6519 + }, + { + "epoch": 18.212290502793294, + "grad_norm": 0.7580523490905762, + "learning_rate": 2.1841340782122908e-05, + "loss": 0.1956, + "step": 6520 + }, + { + "epoch": 18.21508379888268, + "grad_norm": 0.7571823596954346, + "learning_rate": 2.184469273743017e-05, + "loss": 0.1751, + "step": 6521 + }, + { + "epoch": 18.217877094972067, + "grad_norm": 0.6986799836158752, + "learning_rate": 2.184804469273743e-05, + "loss": 0.1668, + "step": 6522 + }, + { + "epoch": 18.220670391061454, + "grad_norm": 0.6929610967636108, + "learning_rate": 2.1851396648044696e-05, + "loss": 0.187, + "step": 6523 + }, + { + "epoch": 18.223463687150836, + "grad_norm": 0.8552725911140442, + "learning_rate": 2.1854748603351957e-05, + "loss": 0.1756, + "step": 6524 + }, + { + "epoch": 18.226256983240223, + "grad_norm": 0.6944541931152344, + "learning_rate": 2.185810055865922e-05, + "loss": 0.1766, + "step": 6525 + }, + { + "epoch": 18.22905027932961, + "grad_norm": 1.0961240530014038, + "learning_rate": 2.1861452513966483e-05, + "loss": 0.1825, + "step": 6526 + }, + { + "epoch": 18.231843575418996, + "grad_norm": 2.385854721069336, + "learning_rate": 2.1864804469273744e-05, + "loss": 0.1631, + "step": 6527 + }, + { + "epoch": 18.23463687150838, + "grad_norm": 1.0430999994277954, + "learning_rate": 2.1868156424581006e-05, + "loss": 0.1455, + "step": 6528 + }, + { + "epoch": 18.237430167597765, + "grad_norm": 0.8715009689331055, + "learning_rate": 2.187150837988827e-05, + "loss": 0.1673, + "step": 6529 + }, + { + "epoch": 18.24022346368715, + "grad_norm": 1.0362776517868042, + "learning_rate": 2.1874860335195532e-05, + "loss": 0.1687, + "step": 6530 + }, + { + "epoch": 18.243016759776538, + "grad_norm": 0.7648055553436279, + "learning_rate": 2.1878212290502793e-05, + "loss": 0.1613, + "step": 6531 + }, + { + "epoch": 18.24581005586592, + "grad_norm": 0.7353103756904602, + "learning_rate": 2.1881564245810058e-05, + "loss": 0.1751, + "step": 6532 + }, + { + "epoch": 18.248603351955307, + "grad_norm": 1.1848242282867432, + "learning_rate": 2.188491620111732e-05, + "loss": 0.1621, + "step": 6533 + }, + { + "epoch": 18.251396648044693, + "grad_norm": 2.087217092514038, + "learning_rate": 2.188826815642458e-05, + "loss": 0.1654, + "step": 6534 + }, + { + "epoch": 18.25418994413408, + "grad_norm": 1.3140385150909424, + "learning_rate": 2.1891620111731845e-05, + "loss": 0.1785, + "step": 6535 + }, + { + "epoch": 18.256983240223462, + "grad_norm": 1.1559693813323975, + "learning_rate": 2.1894972067039107e-05, + "loss": 0.2022, + "step": 6536 + }, + { + "epoch": 18.25977653631285, + "grad_norm": 1.814704179763794, + "learning_rate": 2.1898324022346368e-05, + "loss": 0.1835, + "step": 6537 + }, + { + "epoch": 18.262569832402235, + "grad_norm": 0.7900218367576599, + "learning_rate": 2.1901675977653633e-05, + "loss": 0.1661, + "step": 6538 + }, + { + "epoch": 18.26536312849162, + "grad_norm": 1.6915947198867798, + "learning_rate": 2.1905027932960894e-05, + "loss": 0.1984, + "step": 6539 + }, + { + "epoch": 18.268156424581004, + "grad_norm": 1.78475022315979, + "learning_rate": 2.190837988826816e-05, + "loss": 0.1762, + "step": 6540 + }, + { + "epoch": 18.27094972067039, + "grad_norm": 1.982572317123413, + "learning_rate": 2.191173184357542e-05, + "loss": 0.1736, + "step": 6541 + }, + { + "epoch": 18.273743016759777, + "grad_norm": 1.1258094310760498, + "learning_rate": 2.191508379888268e-05, + "loss": 0.1777, + "step": 6542 + }, + { + "epoch": 18.276536312849164, + "grad_norm": 1.5007588863372803, + "learning_rate": 2.1918435754189946e-05, + "loss": 0.2051, + "step": 6543 + }, + { + "epoch": 18.279329608938546, + "grad_norm": 2.0084524154663086, + "learning_rate": 2.1921787709497208e-05, + "loss": 0.2464, + "step": 6544 + }, + { + "epoch": 18.282122905027933, + "grad_norm": 1.0344091653823853, + "learning_rate": 2.192513966480447e-05, + "loss": 0.2009, + "step": 6545 + }, + { + "epoch": 18.28491620111732, + "grad_norm": 0.7425658106803894, + "learning_rate": 2.1928491620111734e-05, + "loss": 0.1785, + "step": 6546 + }, + { + "epoch": 18.287709497206706, + "grad_norm": 0.6223273873329163, + "learning_rate": 2.1931843575418995e-05, + "loss": 0.164, + "step": 6547 + }, + { + "epoch": 18.29050279329609, + "grad_norm": 1.5762462615966797, + "learning_rate": 2.1935195530726256e-05, + "loss": 0.1847, + "step": 6548 + }, + { + "epoch": 18.293296089385475, + "grad_norm": 0.6750763654708862, + "learning_rate": 2.193854748603352e-05, + "loss": 0.1744, + "step": 6549 + }, + { + "epoch": 18.29608938547486, + "grad_norm": 0.5275968313217163, + "learning_rate": 2.1941899441340782e-05, + "loss": 0.1725, + "step": 6550 + }, + { + "epoch": 18.298882681564244, + "grad_norm": 0.7205382585525513, + "learning_rate": 2.1945251396648044e-05, + "loss": 0.1674, + "step": 6551 + }, + { + "epoch": 18.30167597765363, + "grad_norm": 1.6767871379852295, + "learning_rate": 2.194860335195531e-05, + "loss": 0.2002, + "step": 6552 + }, + { + "epoch": 18.304469273743017, + "grad_norm": 0.9924115538597107, + "learning_rate": 2.195195530726257e-05, + "loss": 0.161, + "step": 6553 + }, + { + "epoch": 18.307262569832403, + "grad_norm": 1.0417075157165527, + "learning_rate": 2.195530726256983e-05, + "loss": 0.2165, + "step": 6554 + }, + { + "epoch": 18.310055865921786, + "grad_norm": 0.5680922269821167, + "learning_rate": 2.1958659217877096e-05, + "loss": 0.1818, + "step": 6555 + }, + { + "epoch": 18.312849162011172, + "grad_norm": 0.71729975938797, + "learning_rate": 2.1962011173184357e-05, + "loss": 0.1924, + "step": 6556 + }, + { + "epoch": 18.31564245810056, + "grad_norm": 0.9954932928085327, + "learning_rate": 2.1965363128491622e-05, + "loss": 0.163, + "step": 6557 + }, + { + "epoch": 18.318435754189945, + "grad_norm": 1.4862611293792725, + "learning_rate": 2.1968715083798883e-05, + "loss": 0.2102, + "step": 6558 + }, + { + "epoch": 18.321229050279328, + "grad_norm": 0.9917794466018677, + "learning_rate": 2.1972067039106145e-05, + "loss": 0.1482, + "step": 6559 + }, + { + "epoch": 18.324022346368714, + "grad_norm": 1.1741236448287964, + "learning_rate": 2.197541899441341e-05, + "loss": 0.1708, + "step": 6560 + }, + { + "epoch": 18.3268156424581, + "grad_norm": 1.081438422203064, + "learning_rate": 2.197877094972067e-05, + "loss": 0.1896, + "step": 6561 + }, + { + "epoch": 18.329608938547487, + "grad_norm": 0.9710485339164734, + "learning_rate": 2.1982122905027932e-05, + "loss": 0.1625, + "step": 6562 + }, + { + "epoch": 18.33240223463687, + "grad_norm": 5.229148864746094, + "learning_rate": 2.1985474860335197e-05, + "loss": 0.1896, + "step": 6563 + }, + { + "epoch": 18.335195530726256, + "grad_norm": 0.6760509014129639, + "learning_rate": 2.1988826815642458e-05, + "loss": 0.1536, + "step": 6564 + }, + { + "epoch": 18.337988826815643, + "grad_norm": 2.499143600463867, + "learning_rate": 2.199217877094972e-05, + "loss": 0.153, + "step": 6565 + }, + { + "epoch": 18.34078212290503, + "grad_norm": 1.1474177837371826, + "learning_rate": 2.1995530726256984e-05, + "loss": 0.1845, + "step": 6566 + }, + { + "epoch": 18.343575418994412, + "grad_norm": 1.7606875896453857, + "learning_rate": 2.1998882681564246e-05, + "loss": 0.2114, + "step": 6567 + }, + { + "epoch": 18.3463687150838, + "grad_norm": 2.738887071609497, + "learning_rate": 2.2002234636871507e-05, + "loss": 0.2176, + "step": 6568 + }, + { + "epoch": 18.349162011173185, + "grad_norm": 1.7226967811584473, + "learning_rate": 2.2005586592178772e-05, + "loss": 0.3078, + "step": 6569 + }, + { + "epoch": 18.35195530726257, + "grad_norm": 0.5301074385643005, + "learning_rate": 2.2008938547486033e-05, + "loss": 0.1816, + "step": 6570 + }, + { + "epoch": 18.354748603351954, + "grad_norm": 0.627437174320221, + "learning_rate": 2.2012290502793294e-05, + "loss": 0.1725, + "step": 6571 + }, + { + "epoch": 18.35754189944134, + "grad_norm": 1.2895716428756714, + "learning_rate": 2.201564245810056e-05, + "loss": 0.1958, + "step": 6572 + }, + { + "epoch": 18.360335195530727, + "grad_norm": 0.6188374161720276, + "learning_rate": 2.201899441340782e-05, + "loss": 0.1845, + "step": 6573 + }, + { + "epoch": 18.363128491620113, + "grad_norm": 0.6965281367301941, + "learning_rate": 2.2022346368715082e-05, + "loss": 0.2195, + "step": 6574 + }, + { + "epoch": 18.365921787709496, + "grad_norm": 0.7012326121330261, + "learning_rate": 2.2025698324022347e-05, + "loss": 0.1698, + "step": 6575 + }, + { + "epoch": 18.368715083798882, + "grad_norm": 1.7476850748062134, + "learning_rate": 2.2029050279329608e-05, + "loss": 0.1695, + "step": 6576 + }, + { + "epoch": 18.37150837988827, + "grad_norm": 1.1827434301376343, + "learning_rate": 2.2032402234636873e-05, + "loss": 0.1949, + "step": 6577 + }, + { + "epoch": 18.374301675977655, + "grad_norm": 1.38129460811615, + "learning_rate": 2.2035754189944134e-05, + "loss": 0.1727, + "step": 6578 + }, + { + "epoch": 18.377094972067038, + "grad_norm": 0.8173670172691345, + "learning_rate": 2.2039106145251395e-05, + "loss": 0.1969, + "step": 6579 + }, + { + "epoch": 18.379888268156424, + "grad_norm": 0.8217169046401978, + "learning_rate": 2.204245810055866e-05, + "loss": 0.1762, + "step": 6580 + }, + { + "epoch": 18.38268156424581, + "grad_norm": 0.906395673751831, + "learning_rate": 2.204581005586592e-05, + "loss": 0.1895, + "step": 6581 + }, + { + "epoch": 18.385474860335197, + "grad_norm": 0.6973811388015747, + "learning_rate": 2.2049162011173183e-05, + "loss": 0.1618, + "step": 6582 + }, + { + "epoch": 18.38826815642458, + "grad_norm": 0.8256562948226929, + "learning_rate": 2.2052513966480447e-05, + "loss": 0.1757, + "step": 6583 + }, + { + "epoch": 18.391061452513966, + "grad_norm": 2.245246171951294, + "learning_rate": 2.205586592178771e-05, + "loss": 0.2116, + "step": 6584 + }, + { + "epoch": 18.393854748603353, + "grad_norm": 1.0112227201461792, + "learning_rate": 2.205921787709497e-05, + "loss": 0.1728, + "step": 6585 + }, + { + "epoch": 18.39664804469274, + "grad_norm": 1.404579997062683, + "learning_rate": 2.2062569832402235e-05, + "loss": 0.191, + "step": 6586 + }, + { + "epoch": 18.399441340782122, + "grad_norm": 0.9646676182746887, + "learning_rate": 2.2065921787709496e-05, + "loss": 0.1611, + "step": 6587 + }, + { + "epoch": 18.40223463687151, + "grad_norm": 1.1080317497253418, + "learning_rate": 2.2069273743016758e-05, + "loss": 0.1545, + "step": 6588 + }, + { + "epoch": 18.405027932960895, + "grad_norm": 1.0706168413162231, + "learning_rate": 2.2072625698324022e-05, + "loss": 0.1946, + "step": 6589 + }, + { + "epoch": 18.407821229050278, + "grad_norm": 1.859371542930603, + "learning_rate": 2.2075977653631284e-05, + "loss": 0.2014, + "step": 6590 + }, + { + "epoch": 18.410614525139664, + "grad_norm": 1.3562006950378418, + "learning_rate": 2.2079329608938545e-05, + "loss": 0.1847, + "step": 6591 + }, + { + "epoch": 18.41340782122905, + "grad_norm": 1.1503552198410034, + "learning_rate": 2.208268156424581e-05, + "loss": 0.1669, + "step": 6592 + }, + { + "epoch": 18.416201117318437, + "grad_norm": 1.4390994310379028, + "learning_rate": 2.208603351955307e-05, + "loss": 0.2096, + "step": 6593 + }, + { + "epoch": 18.41899441340782, + "grad_norm": 6.349584579467773, + "learning_rate": 2.2089385474860332e-05, + "loss": 0.286, + "step": 6594 + }, + { + "epoch": 18.421787709497206, + "grad_norm": 0.711837649345398, + "learning_rate": 2.2092737430167597e-05, + "loss": 0.1983, + "step": 6595 + }, + { + "epoch": 18.424581005586592, + "grad_norm": 0.7775872945785522, + "learning_rate": 2.2096089385474862e-05, + "loss": 0.1625, + "step": 6596 + }, + { + "epoch": 18.42737430167598, + "grad_norm": 0.7114835977554321, + "learning_rate": 2.2099441340782127e-05, + "loss": 0.1964, + "step": 6597 + }, + { + "epoch": 18.43016759776536, + "grad_norm": 0.7067684531211853, + "learning_rate": 2.2102793296089388e-05, + "loss": 0.1765, + "step": 6598 + }, + { + "epoch": 18.432960893854748, + "grad_norm": 0.7161216735839844, + "learning_rate": 2.210614525139665e-05, + "loss": 0.1633, + "step": 6599 + }, + { + "epoch": 18.435754189944134, + "grad_norm": 0.9323875904083252, + "learning_rate": 2.2109497206703914e-05, + "loss": 0.2034, + "step": 6600 + }, + { + "epoch": 18.43854748603352, + "grad_norm": 0.9747028350830078, + "learning_rate": 2.2112849162011175e-05, + "loss": 0.1724, + "step": 6601 + }, + { + "epoch": 18.441340782122904, + "grad_norm": 2.0002899169921875, + "learning_rate": 2.2116201117318437e-05, + "loss": 0.1685, + "step": 6602 + }, + { + "epoch": 18.44413407821229, + "grad_norm": 1.0667749643325806, + "learning_rate": 2.21195530726257e-05, + "loss": 0.1558, + "step": 6603 + }, + { + "epoch": 18.446927374301676, + "grad_norm": 1.0804308652877808, + "learning_rate": 2.2122905027932963e-05, + "loss": 0.1799, + "step": 6604 + }, + { + "epoch": 18.449720670391063, + "grad_norm": 1.1823842525482178, + "learning_rate": 2.2126256983240224e-05, + "loss": 0.2015, + "step": 6605 + }, + { + "epoch": 18.452513966480446, + "grad_norm": 2.0140442848205566, + "learning_rate": 2.212960893854749e-05, + "loss": 0.1923, + "step": 6606 + }, + { + "epoch": 18.455307262569832, + "grad_norm": 1.0263594388961792, + "learning_rate": 2.213296089385475e-05, + "loss": 0.1957, + "step": 6607 + }, + { + "epoch": 18.45810055865922, + "grad_norm": 1.2282031774520874, + "learning_rate": 2.213631284916201e-05, + "loss": 0.2029, + "step": 6608 + }, + { + "epoch": 18.460893854748605, + "grad_norm": 1.1898783445358276, + "learning_rate": 2.2139664804469276e-05, + "loss": 0.1647, + "step": 6609 + }, + { + "epoch": 18.463687150837988, + "grad_norm": 1.0133163928985596, + "learning_rate": 2.2143016759776538e-05, + "loss": 0.1799, + "step": 6610 + }, + { + "epoch": 18.466480446927374, + "grad_norm": 1.1314464807510376, + "learning_rate": 2.21463687150838e-05, + "loss": 0.1815, + "step": 6611 + }, + { + "epoch": 18.46927374301676, + "grad_norm": 1.6557033061981201, + "learning_rate": 2.2149720670391064e-05, + "loss": 0.2067, + "step": 6612 + }, + { + "epoch": 18.472067039106147, + "grad_norm": 4.219315528869629, + "learning_rate": 2.2153072625698325e-05, + "loss": 0.1821, + "step": 6613 + }, + { + "epoch": 18.47486033519553, + "grad_norm": 0.9419822692871094, + "learning_rate": 2.215642458100559e-05, + "loss": 0.2041, + "step": 6614 + }, + { + "epoch": 18.477653631284916, + "grad_norm": 0.9198123812675476, + "learning_rate": 2.215977653631285e-05, + "loss": 0.1851, + "step": 6615 + }, + { + "epoch": 18.480446927374302, + "grad_norm": 1.1655642986297607, + "learning_rate": 2.2163128491620113e-05, + "loss": 0.1678, + "step": 6616 + }, + { + "epoch": 18.48324022346369, + "grad_norm": 0.9556518793106079, + "learning_rate": 2.2166480446927377e-05, + "loss": 0.182, + "step": 6617 + }, + { + "epoch": 18.48603351955307, + "grad_norm": 1.6369916200637817, + "learning_rate": 2.216983240223464e-05, + "loss": 0.2241, + "step": 6618 + }, + { + "epoch": 18.488826815642458, + "grad_norm": 1.5373023748397827, + "learning_rate": 2.21731843575419e-05, + "loss": 0.2329, + "step": 6619 + }, + { + "epoch": 18.491620111731844, + "grad_norm": 0.5282483100891113, + "learning_rate": 2.2176536312849165e-05, + "loss": 0.182, + "step": 6620 + }, + { + "epoch": 18.49441340782123, + "grad_norm": 0.5234744548797607, + "learning_rate": 2.2179888268156426e-05, + "loss": 0.185, + "step": 6621 + }, + { + "epoch": 18.497206703910614, + "grad_norm": 0.7245776057243347, + "learning_rate": 2.2183240223463687e-05, + "loss": 0.2021, + "step": 6622 + }, + { + "epoch": 18.5, + "grad_norm": 0.6085922718048096, + "learning_rate": 2.2186592178770952e-05, + "loss": 0.1803, + "step": 6623 + }, + { + "epoch": 18.502793296089386, + "grad_norm": 1.6710795164108276, + "learning_rate": 2.2189944134078213e-05, + "loss": 0.2196, + "step": 6624 + }, + { + "epoch": 18.505586592178773, + "grad_norm": 0.7883060574531555, + "learning_rate": 2.2193296089385475e-05, + "loss": 0.1607, + "step": 6625 + }, + { + "epoch": 18.508379888268156, + "grad_norm": 0.8258941173553467, + "learning_rate": 2.219664804469274e-05, + "loss": 0.1838, + "step": 6626 + }, + { + "epoch": 18.511173184357542, + "grad_norm": 0.5414320230484009, + "learning_rate": 2.22e-05, + "loss": 0.1615, + "step": 6627 + }, + { + "epoch": 18.51396648044693, + "grad_norm": 0.5946111679077148, + "learning_rate": 2.2203351955307262e-05, + "loss": 0.1615, + "step": 6628 + }, + { + "epoch": 18.51675977653631, + "grad_norm": 0.6172095537185669, + "learning_rate": 2.2206703910614527e-05, + "loss": 0.1989, + "step": 6629 + }, + { + "epoch": 18.519553072625698, + "grad_norm": 0.6625424027442932, + "learning_rate": 2.221005586592179e-05, + "loss": 0.1734, + "step": 6630 + }, + { + "epoch": 18.522346368715084, + "grad_norm": 0.8833639621734619, + "learning_rate": 2.221340782122905e-05, + "loss": 0.1632, + "step": 6631 + }, + { + "epoch": 18.52513966480447, + "grad_norm": 1.242221713066101, + "learning_rate": 2.2216759776536314e-05, + "loss": 0.1501, + "step": 6632 + }, + { + "epoch": 18.527932960893853, + "grad_norm": 0.7606943249702454, + "learning_rate": 2.2220111731843576e-05, + "loss": 0.1522, + "step": 6633 + }, + { + "epoch": 18.53072625698324, + "grad_norm": 0.9464311599731445, + "learning_rate": 2.222346368715084e-05, + "loss": 0.1784, + "step": 6634 + }, + { + "epoch": 18.533519553072626, + "grad_norm": 0.8650925159454346, + "learning_rate": 2.2226815642458102e-05, + "loss": 0.1586, + "step": 6635 + }, + { + "epoch": 18.536312849162012, + "grad_norm": 1.2701150178909302, + "learning_rate": 2.2230167597765363e-05, + "loss": 0.1916, + "step": 6636 + }, + { + "epoch": 18.539106145251395, + "grad_norm": 1.3891226053237915, + "learning_rate": 2.2233519553072628e-05, + "loss": 0.1734, + "step": 6637 + }, + { + "epoch": 18.54189944134078, + "grad_norm": 5.073934555053711, + "learning_rate": 2.223687150837989e-05, + "loss": 0.1931, + "step": 6638 + }, + { + "epoch": 18.544692737430168, + "grad_norm": 0.7337828278541565, + "learning_rate": 2.224022346368715e-05, + "loss": 0.2043, + "step": 6639 + }, + { + "epoch": 18.547486033519554, + "grad_norm": 0.8581085801124573, + "learning_rate": 2.2243575418994415e-05, + "loss": 0.1915, + "step": 6640 + }, + { + "epoch": 18.550279329608937, + "grad_norm": 1.0387686491012573, + "learning_rate": 2.2246927374301677e-05, + "loss": 0.1934, + "step": 6641 + }, + { + "epoch": 18.553072625698324, + "grad_norm": 1.178272008895874, + "learning_rate": 2.2250279329608938e-05, + "loss": 0.2151, + "step": 6642 + }, + { + "epoch": 18.55586592178771, + "grad_norm": 1.2032743692398071, + "learning_rate": 2.2253631284916203e-05, + "loss": 0.1727, + "step": 6643 + }, + { + "epoch": 18.558659217877096, + "grad_norm": 2.5953540802001953, + "learning_rate": 2.2256983240223464e-05, + "loss": 0.2767, + "step": 6644 + }, + { + "epoch": 18.56145251396648, + "grad_norm": 0.5131906867027283, + "learning_rate": 2.2260335195530725e-05, + "loss": 0.1602, + "step": 6645 + }, + { + "epoch": 18.564245810055866, + "grad_norm": 0.708238422870636, + "learning_rate": 2.226368715083799e-05, + "loss": 0.1819, + "step": 6646 + }, + { + "epoch": 18.567039106145252, + "grad_norm": 0.6368159651756287, + "learning_rate": 2.226703910614525e-05, + "loss": 0.2054, + "step": 6647 + }, + { + "epoch": 18.56983240223464, + "grad_norm": 0.6882116794586182, + "learning_rate": 2.2270391061452513e-05, + "loss": 0.1464, + "step": 6648 + }, + { + "epoch": 18.57262569832402, + "grad_norm": 0.7810428142547607, + "learning_rate": 2.2273743016759778e-05, + "loss": 0.1755, + "step": 6649 + }, + { + "epoch": 18.575418994413408, + "grad_norm": 0.8049083352088928, + "learning_rate": 2.227709497206704e-05, + "loss": 0.1728, + "step": 6650 + }, + { + "epoch": 18.578212290502794, + "grad_norm": 0.596515417098999, + "learning_rate": 2.2280446927374304e-05, + "loss": 0.1872, + "step": 6651 + }, + { + "epoch": 18.58100558659218, + "grad_norm": 0.7255532145500183, + "learning_rate": 2.2283798882681565e-05, + "loss": 0.1757, + "step": 6652 + }, + { + "epoch": 18.583798882681563, + "grad_norm": 0.7157669067382812, + "learning_rate": 2.2287150837988826e-05, + "loss": 0.1496, + "step": 6653 + }, + { + "epoch": 18.58659217877095, + "grad_norm": 0.8716872930526733, + "learning_rate": 2.229050279329609e-05, + "loss": 0.1478, + "step": 6654 + }, + { + "epoch": 18.589385474860336, + "grad_norm": 0.9759521484375, + "learning_rate": 2.2293854748603352e-05, + "loss": 0.1829, + "step": 6655 + }, + { + "epoch": 18.592178770949722, + "grad_norm": 0.7654448747634888, + "learning_rate": 2.2297206703910614e-05, + "loss": 0.1591, + "step": 6656 + }, + { + "epoch": 18.594972067039105, + "grad_norm": 1.0717629194259644, + "learning_rate": 2.230055865921788e-05, + "loss": 0.2041, + "step": 6657 + }, + { + "epoch": 18.59776536312849, + "grad_norm": 0.774677574634552, + "learning_rate": 2.230391061452514e-05, + "loss": 0.1625, + "step": 6658 + }, + { + "epoch": 18.600558659217878, + "grad_norm": 0.9897159934043884, + "learning_rate": 2.23072625698324e-05, + "loss": 0.1499, + "step": 6659 + }, + { + "epoch": 18.60335195530726, + "grad_norm": 0.8901182413101196, + "learning_rate": 2.2310614525139666e-05, + "loss": 0.179, + "step": 6660 + }, + { + "epoch": 18.606145251396647, + "grad_norm": 1.083706021308899, + "learning_rate": 2.2313966480446927e-05, + "loss": 0.1536, + "step": 6661 + }, + { + "epoch": 18.608938547486034, + "grad_norm": 1.6295655965805054, + "learning_rate": 2.231731843575419e-05, + "loss": 0.1715, + "step": 6662 + }, + { + "epoch": 18.61173184357542, + "grad_norm": 0.6388189792633057, + "learning_rate": 2.2320670391061453e-05, + "loss": 0.1596, + "step": 6663 + }, + { + "epoch": 18.614525139664803, + "grad_norm": 1.0807539224624634, + "learning_rate": 2.2324022346368715e-05, + "loss": 0.1868, + "step": 6664 + }, + { + "epoch": 18.61731843575419, + "grad_norm": 1.1095974445343018, + "learning_rate": 2.2327374301675976e-05, + "loss": 0.2112, + "step": 6665 + }, + { + "epoch": 18.620111731843576, + "grad_norm": 1.0829147100448608, + "learning_rate": 2.233072625698324e-05, + "loss": 0.2008, + "step": 6666 + }, + { + "epoch": 18.622905027932962, + "grad_norm": 1.0406770706176758, + "learning_rate": 2.2334078212290502e-05, + "loss": 0.1809, + "step": 6667 + }, + { + "epoch": 18.625698324022345, + "grad_norm": 3.802800416946411, + "learning_rate": 2.2337430167597763e-05, + "loss": 0.2063, + "step": 6668 + }, + { + "epoch": 18.62849162011173, + "grad_norm": 7.222405433654785, + "learning_rate": 2.2340782122905028e-05, + "loss": 0.2541, + "step": 6669 + }, + { + "epoch": 18.631284916201118, + "grad_norm": 0.7345314025878906, + "learning_rate": 2.234413407821229e-05, + "loss": 0.1862, + "step": 6670 + }, + { + "epoch": 18.634078212290504, + "grad_norm": 0.9421355724334717, + "learning_rate": 2.2347486033519554e-05, + "loss": 0.1792, + "step": 6671 + }, + { + "epoch": 18.636871508379887, + "grad_norm": 0.6282586455345154, + "learning_rate": 2.2350837988826816e-05, + "loss": 0.1872, + "step": 6672 + }, + { + "epoch": 18.639664804469273, + "grad_norm": 0.6290165185928345, + "learning_rate": 2.2354189944134077e-05, + "loss": 0.1829, + "step": 6673 + }, + { + "epoch": 18.64245810055866, + "grad_norm": 1.1490585803985596, + "learning_rate": 2.2357541899441342e-05, + "loss": 0.21, + "step": 6674 + }, + { + "epoch": 18.645251396648046, + "grad_norm": 1.5831387042999268, + "learning_rate": 2.2360893854748603e-05, + "loss": 0.1623, + "step": 6675 + }, + { + "epoch": 18.64804469273743, + "grad_norm": 0.7485482692718506, + "learning_rate": 2.2364245810055864e-05, + "loss": 0.1652, + "step": 6676 + }, + { + "epoch": 18.650837988826815, + "grad_norm": 0.679096519947052, + "learning_rate": 2.236759776536313e-05, + "loss": 0.1786, + "step": 6677 + }, + { + "epoch": 18.6536312849162, + "grad_norm": 0.6718062162399292, + "learning_rate": 2.237094972067039e-05, + "loss": 0.1855, + "step": 6678 + }, + { + "epoch": 18.656424581005588, + "grad_norm": 1.357611060142517, + "learning_rate": 2.2374301675977652e-05, + "loss": 0.1783, + "step": 6679 + }, + { + "epoch": 18.65921787709497, + "grad_norm": 0.9233220815658569, + "learning_rate": 2.2377653631284917e-05, + "loss": 0.2066, + "step": 6680 + }, + { + "epoch": 18.662011173184357, + "grad_norm": 0.6260904669761658, + "learning_rate": 2.2381005586592178e-05, + "loss": 0.1432, + "step": 6681 + }, + { + "epoch": 18.664804469273744, + "grad_norm": 1.6762003898620605, + "learning_rate": 2.238435754189944e-05, + "loss": 0.1537, + "step": 6682 + }, + { + "epoch": 18.66759776536313, + "grad_norm": 1.1380743980407715, + "learning_rate": 2.2387709497206704e-05, + "loss": 0.1863, + "step": 6683 + }, + { + "epoch": 18.670391061452513, + "grad_norm": 1.6467101573944092, + "learning_rate": 2.2391061452513965e-05, + "loss": 0.1856, + "step": 6684 + }, + { + "epoch": 18.6731843575419, + "grad_norm": 2.3344578742980957, + "learning_rate": 2.2394413407821227e-05, + "loss": 0.1558, + "step": 6685 + }, + { + "epoch": 18.675977653631286, + "grad_norm": 1.2417271137237549, + "learning_rate": 2.239776536312849e-05, + "loss": 0.1746, + "step": 6686 + }, + { + "epoch": 18.678770949720672, + "grad_norm": 0.828464925289154, + "learning_rate": 2.2401117318435753e-05, + "loss": 0.1613, + "step": 6687 + }, + { + "epoch": 18.681564245810055, + "grad_norm": 1.478533387184143, + "learning_rate": 2.2404469273743014e-05, + "loss": 0.1471, + "step": 6688 + }, + { + "epoch": 18.68435754189944, + "grad_norm": 1.998786211013794, + "learning_rate": 2.240782122905028e-05, + "loss": 0.1818, + "step": 6689 + }, + { + "epoch": 18.687150837988828, + "grad_norm": 0.8573099970817566, + "learning_rate": 2.241117318435754e-05, + "loss": 0.1674, + "step": 6690 + }, + { + "epoch": 18.689944134078214, + "grad_norm": 1.3153414726257324, + "learning_rate": 2.2414525139664805e-05, + "loss": 0.1794, + "step": 6691 + }, + { + "epoch": 18.692737430167597, + "grad_norm": 2.831162929534912, + "learning_rate": 2.2417877094972066e-05, + "loss": 0.1863, + "step": 6692 + }, + { + "epoch": 18.695530726256983, + "grad_norm": 1.3129533529281616, + "learning_rate": 2.242122905027933e-05, + "loss": 0.1706, + "step": 6693 + }, + { + "epoch": 18.69832402234637, + "grad_norm": 2.2311806678771973, + "learning_rate": 2.2424581005586596e-05, + "loss": 0.247, + "step": 6694 + }, + { + "epoch": 18.701117318435756, + "grad_norm": 1.439873456954956, + "learning_rate": 2.2427932960893857e-05, + "loss": 0.2122, + "step": 6695 + }, + { + "epoch": 18.70391061452514, + "grad_norm": 0.8088756799697876, + "learning_rate": 2.243128491620112e-05, + "loss": 0.1796, + "step": 6696 + }, + { + "epoch": 18.706703910614525, + "grad_norm": 0.8094180226325989, + "learning_rate": 2.2434636871508383e-05, + "loss": 0.1626, + "step": 6697 + }, + { + "epoch": 18.70949720670391, + "grad_norm": 0.9938566088676453, + "learning_rate": 2.2437988826815645e-05, + "loss": 0.1799, + "step": 6698 + }, + { + "epoch": 18.712290502793294, + "grad_norm": 0.6476538777351379, + "learning_rate": 2.2441340782122906e-05, + "loss": 0.1852, + "step": 6699 + }, + { + "epoch": 18.71508379888268, + "grad_norm": 0.6114839911460876, + "learning_rate": 2.244469273743017e-05, + "loss": 0.1908, + "step": 6700 + }, + { + "epoch": 18.717877094972067, + "grad_norm": 0.5877167582511902, + "learning_rate": 2.2448044692737432e-05, + "loss": 0.1663, + "step": 6701 + }, + { + "epoch": 18.720670391061454, + "grad_norm": 0.7316798567771912, + "learning_rate": 2.2451396648044693e-05, + "loss": 0.2189, + "step": 6702 + }, + { + "epoch": 18.723463687150836, + "grad_norm": 2.1873464584350586, + "learning_rate": 2.2454748603351958e-05, + "loss": 0.1486, + "step": 6703 + }, + { + "epoch": 18.726256983240223, + "grad_norm": 0.8183286190032959, + "learning_rate": 2.245810055865922e-05, + "loss": 0.2344, + "step": 6704 + }, + { + "epoch": 18.72905027932961, + "grad_norm": 1.1384003162384033, + "learning_rate": 2.246145251396648e-05, + "loss": 0.1857, + "step": 6705 + }, + { + "epoch": 18.731843575418996, + "grad_norm": 0.638582706451416, + "learning_rate": 2.2464804469273745e-05, + "loss": 0.1637, + "step": 6706 + }, + { + "epoch": 18.73463687150838, + "grad_norm": 0.7699151039123535, + "learning_rate": 2.2468156424581007e-05, + "loss": 0.1482, + "step": 6707 + }, + { + "epoch": 18.737430167597765, + "grad_norm": 0.8830801248550415, + "learning_rate": 2.247150837988827e-05, + "loss": 0.1683, + "step": 6708 + }, + { + "epoch": 18.74022346368715, + "grad_norm": 0.9522262811660767, + "learning_rate": 2.2474860335195533e-05, + "loss": 0.1786, + "step": 6709 + }, + { + "epoch": 18.743016759776538, + "grad_norm": 0.6561604738235474, + "learning_rate": 2.2478212290502794e-05, + "loss": 0.1382, + "step": 6710 + }, + { + "epoch": 18.74581005586592, + "grad_norm": 0.8931381106376648, + "learning_rate": 2.248156424581006e-05, + "loss": 0.1812, + "step": 6711 + }, + { + "epoch": 18.748603351955307, + "grad_norm": 0.8736217617988586, + "learning_rate": 2.248491620111732e-05, + "loss": 0.1575, + "step": 6712 + }, + { + "epoch": 18.751396648044693, + "grad_norm": 1.4393537044525146, + "learning_rate": 2.248826815642458e-05, + "loss": 0.1678, + "step": 6713 + }, + { + "epoch": 18.75418994413408, + "grad_norm": 1.158219337463379, + "learning_rate": 2.2491620111731846e-05, + "loss": 0.1522, + "step": 6714 + }, + { + "epoch": 18.756983240223462, + "grad_norm": 1.0235567092895508, + "learning_rate": 2.2494972067039108e-05, + "loss": 0.1449, + "step": 6715 + }, + { + "epoch": 18.75977653631285, + "grad_norm": 1.5135390758514404, + "learning_rate": 2.249832402234637e-05, + "loss": 0.2033, + "step": 6716 + }, + { + "epoch": 18.762569832402235, + "grad_norm": 1.5724647045135498, + "learning_rate": 2.2501675977653634e-05, + "loss": 0.1854, + "step": 6717 + }, + { + "epoch": 18.76536312849162, + "grad_norm": 3.3288748264312744, + "learning_rate": 2.2505027932960895e-05, + "loss": 0.1749, + "step": 6718 + }, + { + "epoch": 18.768156424581004, + "grad_norm": 1.8485387563705444, + "learning_rate": 2.2508379888268156e-05, + "loss": 0.2533, + "step": 6719 + }, + { + "epoch": 18.77094972067039, + "grad_norm": 0.561972975730896, + "learning_rate": 2.251173184357542e-05, + "loss": 0.1785, + "step": 6720 + }, + { + "epoch": 18.773743016759777, + "grad_norm": 0.5668889880180359, + "learning_rate": 2.2515083798882683e-05, + "loss": 0.2079, + "step": 6721 + }, + { + "epoch": 18.776536312849164, + "grad_norm": 0.6091586351394653, + "learning_rate": 2.2518435754189944e-05, + "loss": 0.1851, + "step": 6722 + }, + { + "epoch": 18.779329608938546, + "grad_norm": 0.4532790780067444, + "learning_rate": 2.252178770949721e-05, + "loss": 0.1906, + "step": 6723 + }, + { + "epoch": 18.782122905027933, + "grad_norm": 0.9762935042381287, + "learning_rate": 2.252513966480447e-05, + "loss": 0.1883, + "step": 6724 + }, + { + "epoch": 18.78491620111732, + "grad_norm": 0.7392973303794861, + "learning_rate": 2.252849162011173e-05, + "loss": 0.1807, + "step": 6725 + }, + { + "epoch": 18.787709497206706, + "grad_norm": 0.6244463920593262, + "learning_rate": 2.2531843575418996e-05, + "loss": 0.2035, + "step": 6726 + }, + { + "epoch": 18.79050279329609, + "grad_norm": 1.0835851430892944, + "learning_rate": 2.2535195530726257e-05, + "loss": 0.1943, + "step": 6727 + }, + { + "epoch": 18.793296089385475, + "grad_norm": 0.8212046027183533, + "learning_rate": 2.2538547486033522e-05, + "loss": 0.1927, + "step": 6728 + }, + { + "epoch": 18.79608938547486, + "grad_norm": 0.705992579460144, + "learning_rate": 2.2541899441340783e-05, + "loss": 0.1823, + "step": 6729 + }, + { + "epoch": 18.798882681564244, + "grad_norm": 1.2828904390335083, + "learning_rate": 2.2545251396648045e-05, + "loss": 0.1885, + "step": 6730 + }, + { + "epoch": 18.80167597765363, + "grad_norm": 0.8512688279151917, + "learning_rate": 2.254860335195531e-05, + "loss": 0.1907, + "step": 6731 + }, + { + "epoch": 18.804469273743017, + "grad_norm": 1.1381951570510864, + "learning_rate": 2.255195530726257e-05, + "loss": 0.1445, + "step": 6732 + }, + { + "epoch": 18.807262569832403, + "grad_norm": 1.0353126525878906, + "learning_rate": 2.2555307262569832e-05, + "loss": 0.1786, + "step": 6733 + }, + { + "epoch": 18.810055865921786, + "grad_norm": 1.0090689659118652, + "learning_rate": 2.2558659217877097e-05, + "loss": 0.1857, + "step": 6734 + }, + { + "epoch": 18.812849162011172, + "grad_norm": 1.6306713819503784, + "learning_rate": 2.256201117318436e-05, + "loss": 0.1618, + "step": 6735 + }, + { + "epoch": 18.81564245810056, + "grad_norm": 0.9287089109420776, + "learning_rate": 2.256536312849162e-05, + "loss": 0.1437, + "step": 6736 + }, + { + "epoch": 18.818435754189945, + "grad_norm": 1.7660413980484009, + "learning_rate": 2.2568715083798884e-05, + "loss": 0.1818, + "step": 6737 + }, + { + "epoch": 18.821229050279328, + "grad_norm": 1.4161784648895264, + "learning_rate": 2.2572067039106146e-05, + "loss": 0.2016, + "step": 6738 + }, + { + "epoch": 18.824022346368714, + "grad_norm": 1.2386507987976074, + "learning_rate": 2.2575418994413407e-05, + "loss": 0.1833, + "step": 6739 + }, + { + "epoch": 18.8268156424581, + "grad_norm": 1.4579575061798096, + "learning_rate": 2.2578770949720672e-05, + "loss": 0.167, + "step": 6740 + }, + { + "epoch": 18.829608938547487, + "grad_norm": 5.236811637878418, + "learning_rate": 2.2582122905027933e-05, + "loss": 0.1993, + "step": 6741 + }, + { + "epoch": 18.83240223463687, + "grad_norm": 0.958784282207489, + "learning_rate": 2.2585474860335195e-05, + "loss": 0.1646, + "step": 6742 + }, + { + "epoch": 18.835195530726256, + "grad_norm": 1.617271065711975, + "learning_rate": 2.258882681564246e-05, + "loss": 0.2335, + "step": 6743 + }, + { + "epoch": 18.837988826815643, + "grad_norm": 1.9631959199905396, + "learning_rate": 2.259217877094972e-05, + "loss": 0.2619, + "step": 6744 + }, + { + "epoch": 18.84078212290503, + "grad_norm": 0.6915988922119141, + "learning_rate": 2.2595530726256985e-05, + "loss": 0.1824, + "step": 6745 + }, + { + "epoch": 18.843575418994412, + "grad_norm": 0.6696775555610657, + "learning_rate": 2.2598882681564247e-05, + "loss": 0.2107, + "step": 6746 + }, + { + "epoch": 18.8463687150838, + "grad_norm": 0.6977854371070862, + "learning_rate": 2.2602234636871508e-05, + "loss": 0.1821, + "step": 6747 + }, + { + "epoch": 18.849162011173185, + "grad_norm": 0.7072428464889526, + "learning_rate": 2.2605586592178773e-05, + "loss": 0.1753, + "step": 6748 + }, + { + "epoch": 18.85195530726257, + "grad_norm": 0.48961666226387024, + "learning_rate": 2.2608938547486034e-05, + "loss": 0.1534, + "step": 6749 + }, + { + "epoch": 18.854748603351954, + "grad_norm": 0.7856647968292236, + "learning_rate": 2.2612290502793295e-05, + "loss": 0.1871, + "step": 6750 + }, + { + "epoch": 18.85754189944134, + "grad_norm": 0.9878406524658203, + "learning_rate": 2.261564245810056e-05, + "loss": 0.1801, + "step": 6751 + }, + { + "epoch": 18.860335195530727, + "grad_norm": 0.6565583944320679, + "learning_rate": 2.261899441340782e-05, + "loss": 0.1721, + "step": 6752 + }, + { + "epoch": 18.863128491620113, + "grad_norm": 1.4271739721298218, + "learning_rate": 2.2622346368715083e-05, + "loss": 0.1761, + "step": 6753 + }, + { + "epoch": 18.865921787709496, + "grad_norm": 1.4816044569015503, + "learning_rate": 2.2625698324022348e-05, + "loss": 0.1869, + "step": 6754 + }, + { + "epoch": 18.868715083798882, + "grad_norm": 0.7116886973381042, + "learning_rate": 2.262905027932961e-05, + "loss": 0.1727, + "step": 6755 + }, + { + "epoch": 18.87150837988827, + "grad_norm": 0.6506154537200928, + "learning_rate": 2.263240223463687e-05, + "loss": 0.1883, + "step": 6756 + }, + { + "epoch": 18.874301675977655, + "grad_norm": 1.5036073923110962, + "learning_rate": 2.2635754189944135e-05, + "loss": 0.1547, + "step": 6757 + }, + { + "epoch": 18.877094972067038, + "grad_norm": 0.7585493922233582, + "learning_rate": 2.2639106145251396e-05, + "loss": 0.171, + "step": 6758 + }, + { + "epoch": 18.879888268156424, + "grad_norm": 1.8794846534729004, + "learning_rate": 2.2642458100558658e-05, + "loss": 0.1476, + "step": 6759 + }, + { + "epoch": 18.88268156424581, + "grad_norm": 0.7960659861564636, + "learning_rate": 2.2645810055865922e-05, + "loss": 0.1733, + "step": 6760 + }, + { + "epoch": 18.885474860335197, + "grad_norm": 0.7352858781814575, + "learning_rate": 2.2649162011173184e-05, + "loss": 0.1494, + "step": 6761 + }, + { + "epoch": 18.88826815642458, + "grad_norm": 1.2003321647644043, + "learning_rate": 2.2652513966480445e-05, + "loss": 0.1875, + "step": 6762 + }, + { + "epoch": 18.891061452513966, + "grad_norm": 1.2920717000961304, + "learning_rate": 2.265586592178771e-05, + "loss": 0.1564, + "step": 6763 + }, + { + "epoch": 18.893854748603353, + "grad_norm": 1.2444453239440918, + "learning_rate": 2.265921787709497e-05, + "loss": 0.205, + "step": 6764 + }, + { + "epoch": 18.89664804469274, + "grad_norm": 1.007473349571228, + "learning_rate": 2.2662569832402236e-05, + "loss": 0.2055, + "step": 6765 + }, + { + "epoch": 18.899441340782122, + "grad_norm": 3.480865478515625, + "learning_rate": 2.2665921787709497e-05, + "loss": 0.1506, + "step": 6766 + }, + { + "epoch": 18.90223463687151, + "grad_norm": 1.5143436193466187, + "learning_rate": 2.266927374301676e-05, + "loss": 0.1924, + "step": 6767 + }, + { + "epoch": 18.905027932960895, + "grad_norm": 4.152637004852295, + "learning_rate": 2.2672625698324023e-05, + "loss": 0.1772, + "step": 6768 + }, + { + "epoch": 18.907821229050278, + "grad_norm": 1.693211555480957, + "learning_rate": 2.2675977653631285e-05, + "loss": 0.2731, + "step": 6769 + }, + { + "epoch": 18.910614525139664, + "grad_norm": 0.9294683933258057, + "learning_rate": 2.2679329608938546e-05, + "loss": 0.1711, + "step": 6770 + }, + { + "epoch": 18.91340782122905, + "grad_norm": 1.491827368736267, + "learning_rate": 2.268268156424581e-05, + "loss": 0.1809, + "step": 6771 + }, + { + "epoch": 18.916201117318437, + "grad_norm": 0.5758061408996582, + "learning_rate": 2.2686033519553072e-05, + "loss": 0.1734, + "step": 6772 + }, + { + "epoch": 18.91899441340782, + "grad_norm": 0.9631050229072571, + "learning_rate": 2.2689385474860333e-05, + "loss": 0.1762, + "step": 6773 + }, + { + "epoch": 18.921787709497206, + "grad_norm": 1.0507643222808838, + "learning_rate": 2.2692737430167598e-05, + "loss": 0.1815, + "step": 6774 + }, + { + "epoch": 18.924581005586592, + "grad_norm": 0.7321004271507263, + "learning_rate": 2.269608938547486e-05, + "loss": 0.2123, + "step": 6775 + }, + { + "epoch": 18.92737430167598, + "grad_norm": 1.582513451576233, + "learning_rate": 2.269944134078212e-05, + "loss": 0.1893, + "step": 6776 + }, + { + "epoch": 18.93016759776536, + "grad_norm": 0.8518775105476379, + "learning_rate": 2.2702793296089386e-05, + "loss": 0.1502, + "step": 6777 + }, + { + "epoch": 18.932960893854748, + "grad_norm": 1.0387675762176514, + "learning_rate": 2.2706145251396647e-05, + "loss": 0.1518, + "step": 6778 + }, + { + "epoch": 18.935754189944134, + "grad_norm": 0.6001703143119812, + "learning_rate": 2.270949720670391e-05, + "loss": 0.1662, + "step": 6779 + }, + { + "epoch": 18.93854748603352, + "grad_norm": 1.0976885557174683, + "learning_rate": 2.2712849162011173e-05, + "loss": 0.1688, + "step": 6780 + }, + { + "epoch": 18.941340782122904, + "grad_norm": 0.8073830604553223, + "learning_rate": 2.2716201117318434e-05, + "loss": 0.1697, + "step": 6781 + }, + { + "epoch": 18.94413407821229, + "grad_norm": 0.5761579871177673, + "learning_rate": 2.2719553072625696e-05, + "loss": 0.1585, + "step": 6782 + }, + { + "epoch": 18.946927374301676, + "grad_norm": 0.5525195598602295, + "learning_rate": 2.272290502793296e-05, + "loss": 0.1661, + "step": 6783 + }, + { + "epoch": 18.949720670391063, + "grad_norm": 0.8434780240058899, + "learning_rate": 2.2726256983240222e-05, + "loss": 0.1439, + "step": 6784 + }, + { + "epoch": 18.952513966480446, + "grad_norm": 1.0768159627914429, + "learning_rate": 2.2729608938547487e-05, + "loss": 0.1783, + "step": 6785 + }, + { + "epoch": 18.955307262569832, + "grad_norm": 0.7646124362945557, + "learning_rate": 2.2732960893854748e-05, + "loss": 0.1732, + "step": 6786 + }, + { + "epoch": 18.95810055865922, + "grad_norm": 0.8955920934677124, + "learning_rate": 2.273631284916201e-05, + "loss": 0.1808, + "step": 6787 + }, + { + "epoch": 18.960893854748605, + "grad_norm": 0.8878370523452759, + "learning_rate": 2.2739664804469274e-05, + "loss": 0.1849, + "step": 6788 + }, + { + "epoch": 18.963687150837988, + "grad_norm": 1.2945703268051147, + "learning_rate": 2.274301675977654e-05, + "loss": 0.1654, + "step": 6789 + }, + { + "epoch": 18.966480446927374, + "grad_norm": 0.8006954193115234, + "learning_rate": 2.27463687150838e-05, + "loss": 0.1797, + "step": 6790 + }, + { + "epoch": 18.96927374301676, + "grad_norm": 0.9742876291275024, + "learning_rate": 2.2749720670391065e-05, + "loss": 0.1613, + "step": 6791 + }, + { + "epoch": 18.972067039106147, + "grad_norm": 0.8621468544006348, + "learning_rate": 2.2753072625698326e-05, + "loss": 0.1761, + "step": 6792 + }, + { + "epoch": 18.97486033519553, + "grad_norm": 1.4806816577911377, + "learning_rate": 2.2756424581005588e-05, + "loss": 0.232, + "step": 6793 + }, + { + "epoch": 18.977653631284916, + "grad_norm": 1.6684037446975708, + "learning_rate": 2.2759776536312852e-05, + "loss": 0.2096, + "step": 6794 + }, + { + "epoch": 18.980446927374302, + "grad_norm": 0.920448899269104, + "learning_rate": 2.2763128491620114e-05, + "loss": 0.2493, + "step": 6795 + }, + { + "epoch": 18.98324022346369, + "grad_norm": 0.7095656394958496, + "learning_rate": 2.2766480446927375e-05, + "loss": 0.2222, + "step": 6796 + }, + { + "epoch": 18.98603351955307, + "grad_norm": 0.7134034633636475, + "learning_rate": 2.276983240223464e-05, + "loss": 0.1493, + "step": 6797 + }, + { + "epoch": 18.988826815642458, + "grad_norm": 0.6896864771842957, + "learning_rate": 2.27731843575419e-05, + "loss": 0.2062, + "step": 6798 + }, + { + "epoch": 18.991620111731844, + "grad_norm": 0.7963417768478394, + "learning_rate": 2.2776536312849162e-05, + "loss": 0.1597, + "step": 6799 + }, + { + "epoch": 18.994413407821227, + "grad_norm": 1.095271110534668, + "learning_rate": 2.2779888268156427e-05, + "loss": 0.1836, + "step": 6800 + }, + { + "epoch": 18.997206703910614, + "grad_norm": 1.1248834133148193, + "learning_rate": 2.278324022346369e-05, + "loss": 0.1958, + "step": 6801 + }, + { + "epoch": 19.0, + "grad_norm": 1.1340689659118652, + "learning_rate": 2.2786592178770953e-05, + "loss": 0.1966, + "step": 6802 + }, + { + "epoch": 19.002793296089386, + "grad_norm": 0.7172737121582031, + "learning_rate": 2.2789944134078215e-05, + "loss": 0.1763, + "step": 6803 + }, + { + "epoch": 19.00558659217877, + "grad_norm": 0.5619004368782043, + "learning_rate": 2.2793296089385476e-05, + "loss": 0.2078, + "step": 6804 + }, + { + "epoch": 19.008379888268156, + "grad_norm": 0.702162504196167, + "learning_rate": 2.279664804469274e-05, + "loss": 0.1968, + "step": 6805 + }, + { + "epoch": 19.011173184357542, + "grad_norm": 1.3721144199371338, + "learning_rate": 2.2800000000000002e-05, + "loss": 0.189, + "step": 6806 + }, + { + "epoch": 19.01396648044693, + "grad_norm": 0.6362245082855225, + "learning_rate": 2.2803351955307263e-05, + "loss": 0.1783, + "step": 6807 + }, + { + "epoch": 19.01675977653631, + "grad_norm": 0.6704689860343933, + "learning_rate": 2.2806703910614528e-05, + "loss": 0.1618, + "step": 6808 + }, + { + "epoch": 19.019553072625698, + "grad_norm": 0.574946939945221, + "learning_rate": 2.281005586592179e-05, + "loss": 0.1952, + "step": 6809 + }, + { + "epoch": 19.022346368715084, + "grad_norm": 0.8615474104881287, + "learning_rate": 2.281340782122905e-05, + "loss": 0.1789, + "step": 6810 + }, + { + "epoch": 19.02513966480447, + "grad_norm": 0.595396101474762, + "learning_rate": 2.2816759776536315e-05, + "loss": 0.1592, + "step": 6811 + }, + { + "epoch": 19.027932960893853, + "grad_norm": 0.6304011344909668, + "learning_rate": 2.2820111731843577e-05, + "loss": 0.1526, + "step": 6812 + }, + { + "epoch": 19.03072625698324, + "grad_norm": 0.9633841514587402, + "learning_rate": 2.2823463687150838e-05, + "loss": 0.1892, + "step": 6813 + }, + { + "epoch": 19.033519553072626, + "grad_norm": 1.1449196338653564, + "learning_rate": 2.2826815642458103e-05, + "loss": 0.1817, + "step": 6814 + }, + { + "epoch": 19.036312849162012, + "grad_norm": 1.0722917318344116, + "learning_rate": 2.2830167597765364e-05, + "loss": 0.1917, + "step": 6815 + }, + { + "epoch": 19.039106145251395, + "grad_norm": 0.7206853628158569, + "learning_rate": 2.2833519553072626e-05, + "loss": 0.1514, + "step": 6816 + }, + { + "epoch": 19.04189944134078, + "grad_norm": 1.044886827468872, + "learning_rate": 2.283687150837989e-05, + "loss": 0.1577, + "step": 6817 + }, + { + "epoch": 19.044692737430168, + "grad_norm": 0.9015007019042969, + "learning_rate": 2.284022346368715e-05, + "loss": 0.1737, + "step": 6818 + }, + { + "epoch": 19.047486033519554, + "grad_norm": 1.0842074155807495, + "learning_rate": 2.2843575418994413e-05, + "loss": 0.1612, + "step": 6819 + }, + { + "epoch": 19.050279329608937, + "grad_norm": 0.8467515707015991, + "learning_rate": 2.2846927374301678e-05, + "loss": 0.1681, + "step": 6820 + }, + { + "epoch": 19.053072625698324, + "grad_norm": 0.8959486484527588, + "learning_rate": 2.285027932960894e-05, + "loss": 0.1663, + "step": 6821 + }, + { + "epoch": 19.05586592178771, + "grad_norm": 0.900285005569458, + "learning_rate": 2.2853631284916204e-05, + "loss": 0.1741, + "step": 6822 + }, + { + "epoch": 19.058659217877096, + "grad_norm": 1.1578705310821533, + "learning_rate": 2.2856983240223465e-05, + "loss": 0.2099, + "step": 6823 + }, + { + "epoch": 19.06145251396648, + "grad_norm": 0.8922029137611389, + "learning_rate": 2.2860335195530726e-05, + "loss": 0.1605, + "step": 6824 + }, + { + "epoch": 19.064245810055866, + "grad_norm": 0.948874294757843, + "learning_rate": 2.286368715083799e-05, + "loss": 0.1909, + "step": 6825 + }, + { + "epoch": 19.067039106145252, + "grad_norm": 12.880716323852539, + "learning_rate": 2.2867039106145253e-05, + "loss": 0.1979, + "step": 6826 + }, + { + "epoch": 19.06983240223464, + "grad_norm": 3.7902414798736572, + "learning_rate": 2.2870391061452514e-05, + "loss": 0.255, + "step": 6827 + }, + { + "epoch": 19.07262569832402, + "grad_norm": 0.44929203391075134, + "learning_rate": 2.287374301675978e-05, + "loss": 0.1693, + "step": 6828 + }, + { + "epoch": 19.075418994413408, + "grad_norm": 0.7469626069068909, + "learning_rate": 2.287709497206704e-05, + "loss": 0.1776, + "step": 6829 + }, + { + "epoch": 19.078212290502794, + "grad_norm": 0.8398376703262329, + "learning_rate": 2.28804469273743e-05, + "loss": 0.1856, + "step": 6830 + }, + { + "epoch": 19.08100558659218, + "grad_norm": 0.7010822892189026, + "learning_rate": 2.2883798882681566e-05, + "loss": 0.1659, + "step": 6831 + }, + { + "epoch": 19.083798882681563, + "grad_norm": 0.5399913191795349, + "learning_rate": 2.2887150837988827e-05, + "loss": 0.1403, + "step": 6832 + }, + { + "epoch": 19.08659217877095, + "grad_norm": 0.6643296480178833, + "learning_rate": 2.289050279329609e-05, + "loss": 0.1833, + "step": 6833 + }, + { + "epoch": 19.089385474860336, + "grad_norm": 1.088152527809143, + "learning_rate": 2.2893854748603353e-05, + "loss": 0.1684, + "step": 6834 + }, + { + "epoch": 19.092178770949722, + "grad_norm": 1.3576503992080688, + "learning_rate": 2.2897206703910615e-05, + "loss": 0.1509, + "step": 6835 + }, + { + "epoch": 19.094972067039105, + "grad_norm": 0.6726009845733643, + "learning_rate": 2.2900558659217876e-05, + "loss": 0.177, + "step": 6836 + }, + { + "epoch": 19.09776536312849, + "grad_norm": 0.5726436376571655, + "learning_rate": 2.290391061452514e-05, + "loss": 0.1715, + "step": 6837 + }, + { + "epoch": 19.100558659217878, + "grad_norm": 1.2028793096542358, + "learning_rate": 2.2907262569832402e-05, + "loss": 0.1965, + "step": 6838 + }, + { + "epoch": 19.10335195530726, + "grad_norm": 0.7775738835334778, + "learning_rate": 2.2910614525139667e-05, + "loss": 0.1749, + "step": 6839 + }, + { + "epoch": 19.106145251396647, + "grad_norm": 0.8018128275871277, + "learning_rate": 2.291396648044693e-05, + "loss": 0.1738, + "step": 6840 + }, + { + "epoch": 19.108938547486034, + "grad_norm": 0.8373948931694031, + "learning_rate": 2.291731843575419e-05, + "loss": 0.1711, + "step": 6841 + }, + { + "epoch": 19.11173184357542, + "grad_norm": 0.7913157343864441, + "learning_rate": 2.2920670391061454e-05, + "loss": 0.166, + "step": 6842 + }, + { + "epoch": 19.114525139664803, + "grad_norm": 1.0799381732940674, + "learning_rate": 2.2924022346368716e-05, + "loss": 0.185, + "step": 6843 + }, + { + "epoch": 19.11731843575419, + "grad_norm": 1.4225702285766602, + "learning_rate": 2.2927374301675977e-05, + "loss": 0.1701, + "step": 6844 + }, + { + "epoch": 19.120111731843576, + "grad_norm": 0.885611891746521, + "learning_rate": 2.2930726256983242e-05, + "loss": 0.1625, + "step": 6845 + }, + { + "epoch": 19.122905027932962, + "grad_norm": 0.8343682885169983, + "learning_rate": 2.2934078212290503e-05, + "loss": 0.1765, + "step": 6846 + }, + { + "epoch": 19.125698324022345, + "grad_norm": 0.9193997979164124, + "learning_rate": 2.2937430167597765e-05, + "loss": 0.1792, + "step": 6847 + }, + { + "epoch": 19.12849162011173, + "grad_norm": 1.1817981004714966, + "learning_rate": 2.294078212290503e-05, + "loss": 0.1603, + "step": 6848 + }, + { + "epoch": 19.131284916201118, + "grad_norm": 1.3022370338439941, + "learning_rate": 2.294413407821229e-05, + "loss": 0.1595, + "step": 6849 + }, + { + "epoch": 19.134078212290504, + "grad_norm": 0.8265297412872314, + "learning_rate": 2.2947486033519552e-05, + "loss": 0.1617, + "step": 6850 + }, + { + "epoch": 19.136871508379887, + "grad_norm": 2.0615694522857666, + "learning_rate": 2.2950837988826817e-05, + "loss": 0.1924, + "step": 6851 + }, + { + "epoch": 19.139664804469273, + "grad_norm": 1.2797751426696777, + "learning_rate": 2.2954189944134078e-05, + "loss": 0.2663, + "step": 6852 + }, + { + "epoch": 19.14245810055866, + "grad_norm": 0.9992491602897644, + "learning_rate": 2.295754189944134e-05, + "loss": 0.2319, + "step": 6853 + }, + { + "epoch": 19.145251396648046, + "grad_norm": 0.7760328054428101, + "learning_rate": 2.2960893854748604e-05, + "loss": 0.1823, + "step": 6854 + }, + { + "epoch": 19.14804469273743, + "grad_norm": 0.5963214039802551, + "learning_rate": 2.2964245810055865e-05, + "loss": 0.1598, + "step": 6855 + }, + { + "epoch": 19.150837988826815, + "grad_norm": 0.9472948908805847, + "learning_rate": 2.2967597765363127e-05, + "loss": 0.164, + "step": 6856 + }, + { + "epoch": 19.1536312849162, + "grad_norm": 0.772456705570221, + "learning_rate": 2.297094972067039e-05, + "loss": 0.1865, + "step": 6857 + }, + { + "epoch": 19.156424581005588, + "grad_norm": 1.7377331256866455, + "learning_rate": 2.2974301675977653e-05, + "loss": 0.1685, + "step": 6858 + }, + { + "epoch": 19.15921787709497, + "grad_norm": 0.6868610978126526, + "learning_rate": 2.2977653631284918e-05, + "loss": 0.1967, + "step": 6859 + }, + { + "epoch": 19.162011173184357, + "grad_norm": 0.8696832060813904, + "learning_rate": 2.298100558659218e-05, + "loss": 0.1808, + "step": 6860 + }, + { + "epoch": 19.164804469273744, + "grad_norm": 1.2787681818008423, + "learning_rate": 2.298435754189944e-05, + "loss": 0.1488, + "step": 6861 + }, + { + "epoch": 19.16759776536313, + "grad_norm": 0.9333193302154541, + "learning_rate": 2.2987709497206705e-05, + "loss": 0.2057, + "step": 6862 + }, + { + "epoch": 19.170391061452513, + "grad_norm": 0.938504695892334, + "learning_rate": 2.2991061452513966e-05, + "loss": 0.1639, + "step": 6863 + }, + { + "epoch": 19.1731843575419, + "grad_norm": 0.5578368902206421, + "learning_rate": 2.2994413407821228e-05, + "loss": 0.1791, + "step": 6864 + }, + { + "epoch": 19.175977653631286, + "grad_norm": 0.5020110011100769, + "learning_rate": 2.2997765363128492e-05, + "loss": 0.1638, + "step": 6865 + }, + { + "epoch": 19.178770949720672, + "grad_norm": 0.7087791562080383, + "learning_rate": 2.3001117318435754e-05, + "loss": 0.1662, + "step": 6866 + }, + { + "epoch": 19.181564245810055, + "grad_norm": 0.7539852261543274, + "learning_rate": 2.3004469273743015e-05, + "loss": 0.153, + "step": 6867 + }, + { + "epoch": 19.18435754189944, + "grad_norm": 0.8585681319236755, + "learning_rate": 2.300782122905028e-05, + "loss": 0.184, + "step": 6868 + }, + { + "epoch": 19.187150837988828, + "grad_norm": 0.8281576037406921, + "learning_rate": 2.301117318435754e-05, + "loss": 0.175, + "step": 6869 + }, + { + "epoch": 19.189944134078214, + "grad_norm": 0.892473578453064, + "learning_rate": 2.3014525139664803e-05, + "loss": 0.1623, + "step": 6870 + }, + { + "epoch": 19.192737430167597, + "grad_norm": 0.7686049342155457, + "learning_rate": 2.3017877094972067e-05, + "loss": 0.1993, + "step": 6871 + }, + { + "epoch": 19.195530726256983, + "grad_norm": 1.9562779664993286, + "learning_rate": 2.302122905027933e-05, + "loss": 0.1783, + "step": 6872 + }, + { + "epoch": 19.19832402234637, + "grad_norm": 3.9274542331695557, + "learning_rate": 2.302458100558659e-05, + "loss": 0.186, + "step": 6873 + }, + { + "epoch": 19.201117318435756, + "grad_norm": 0.9850323796272278, + "learning_rate": 2.3027932960893855e-05, + "loss": 0.1834, + "step": 6874 + }, + { + "epoch": 19.20391061452514, + "grad_norm": 2.3192272186279297, + "learning_rate": 2.3031284916201116e-05, + "loss": 0.1958, + "step": 6875 + }, + { + "epoch": 19.206703910614525, + "grad_norm": 3.2211289405822754, + "learning_rate": 2.3034636871508377e-05, + "loss": 0.1979, + "step": 6876 + }, + { + "epoch": 19.20949720670391, + "grad_norm": 2.389173984527588, + "learning_rate": 2.3037988826815642e-05, + "loss": 0.2355, + "step": 6877 + }, + { + "epoch": 19.212290502793294, + "grad_norm": 0.5844464898109436, + "learning_rate": 2.3041340782122903e-05, + "loss": 0.1692, + "step": 6878 + }, + { + "epoch": 19.21508379888268, + "grad_norm": 0.6774546504020691, + "learning_rate": 2.3044692737430168e-05, + "loss": 0.1791, + "step": 6879 + }, + { + "epoch": 19.217877094972067, + "grad_norm": 0.702168881893158, + "learning_rate": 2.304804469273743e-05, + "loss": 0.1626, + "step": 6880 + }, + { + "epoch": 19.220670391061454, + "grad_norm": 1.8706917762756348, + "learning_rate": 2.305139664804469e-05, + "loss": 0.191, + "step": 6881 + }, + { + "epoch": 19.223463687150836, + "grad_norm": 1.0569815635681152, + "learning_rate": 2.3054748603351956e-05, + "loss": 0.1986, + "step": 6882 + }, + { + "epoch": 19.226256983240223, + "grad_norm": 0.5932477712631226, + "learning_rate": 2.3058100558659217e-05, + "loss": 0.2002, + "step": 6883 + }, + { + "epoch": 19.22905027932961, + "grad_norm": 0.931361973285675, + "learning_rate": 2.306145251396648e-05, + "loss": 0.2113, + "step": 6884 + }, + { + "epoch": 19.231843575418996, + "grad_norm": 0.8533995151519775, + "learning_rate": 2.3064804469273743e-05, + "loss": 0.1743, + "step": 6885 + }, + { + "epoch": 19.23463687150838, + "grad_norm": 0.8313480019569397, + "learning_rate": 2.3068156424581008e-05, + "loss": 0.1851, + "step": 6886 + }, + { + "epoch": 19.237430167597765, + "grad_norm": 1.1980364322662354, + "learning_rate": 2.307150837988827e-05, + "loss": 0.1739, + "step": 6887 + }, + { + "epoch": 19.24022346368715, + "grad_norm": 1.6835918426513672, + "learning_rate": 2.3074860335195534e-05, + "loss": 0.1735, + "step": 6888 + }, + { + "epoch": 19.243016759776538, + "grad_norm": 3.5127651691436768, + "learning_rate": 2.3078212290502795e-05, + "loss": 0.1875, + "step": 6889 + }, + { + "epoch": 19.24581005586592, + "grad_norm": 0.8733412027359009, + "learning_rate": 2.3081564245810057e-05, + "loss": 0.1634, + "step": 6890 + }, + { + "epoch": 19.248603351955307, + "grad_norm": 0.8954377770423889, + "learning_rate": 2.308491620111732e-05, + "loss": 0.1787, + "step": 6891 + }, + { + "epoch": 19.251396648044693, + "grad_norm": 1.4943127632141113, + "learning_rate": 2.3088268156424583e-05, + "loss": 0.18, + "step": 6892 + }, + { + "epoch": 19.25418994413408, + "grad_norm": 0.9525713324546814, + "learning_rate": 2.3091620111731844e-05, + "loss": 0.1661, + "step": 6893 + }, + { + "epoch": 19.256983240223462, + "grad_norm": 0.9311386942863464, + "learning_rate": 2.309497206703911e-05, + "loss": 0.1657, + "step": 6894 + }, + { + "epoch": 19.25977653631285, + "grad_norm": 0.9445611238479614, + "learning_rate": 2.309832402234637e-05, + "loss": 0.161, + "step": 6895 + }, + { + "epoch": 19.262569832402235, + "grad_norm": 0.9806358814239502, + "learning_rate": 2.3101675977653635e-05, + "loss": 0.1493, + "step": 6896 + }, + { + "epoch": 19.26536312849162, + "grad_norm": 1.2243030071258545, + "learning_rate": 2.3105027932960896e-05, + "loss": 0.2197, + "step": 6897 + }, + { + "epoch": 19.268156424581004, + "grad_norm": 0.960353672504425, + "learning_rate": 2.3108379888268158e-05, + "loss": 0.2006, + "step": 6898 + }, + { + "epoch": 19.27094972067039, + "grad_norm": 1.2326257228851318, + "learning_rate": 2.3111731843575422e-05, + "loss": 0.1715, + "step": 6899 + }, + { + "epoch": 19.273743016759777, + "grad_norm": 1.3840599060058594, + "learning_rate": 2.3115083798882684e-05, + "loss": 0.1851, + "step": 6900 + }, + { + "epoch": 19.276536312849164, + "grad_norm": 3.4499683380126953, + "learning_rate": 2.3118435754189945e-05, + "loss": 0.1919, + "step": 6901 + }, + { + "epoch": 19.279329608938546, + "grad_norm": 2.3234612941741943, + "learning_rate": 2.312178770949721e-05, + "loss": 0.3144, + "step": 6902 + }, + { + "epoch": 19.282122905027933, + "grad_norm": 0.6623157262802124, + "learning_rate": 2.312513966480447e-05, + "loss": 0.1993, + "step": 6903 + }, + { + "epoch": 19.28491620111732, + "grad_norm": 0.5386640429496765, + "learning_rate": 2.3128491620111732e-05, + "loss": 0.1702, + "step": 6904 + }, + { + "epoch": 19.287709497206706, + "grad_norm": 0.7326470613479614, + "learning_rate": 2.3131843575418997e-05, + "loss": 0.1839, + "step": 6905 + }, + { + "epoch": 19.29050279329609, + "grad_norm": 0.7075898051261902, + "learning_rate": 2.313519553072626e-05, + "loss": 0.1799, + "step": 6906 + }, + { + "epoch": 19.293296089385475, + "grad_norm": 0.6175139546394348, + "learning_rate": 2.313854748603352e-05, + "loss": 0.1794, + "step": 6907 + }, + { + "epoch": 19.29608938547486, + "grad_norm": 1.0769612789154053, + "learning_rate": 2.3141899441340785e-05, + "loss": 0.1836, + "step": 6908 + }, + { + "epoch": 19.298882681564244, + "grad_norm": 0.6645368933677673, + "learning_rate": 2.3145251396648046e-05, + "loss": 0.1654, + "step": 6909 + }, + { + "epoch": 19.30167597765363, + "grad_norm": 0.6620663404464722, + "learning_rate": 2.3148603351955307e-05, + "loss": 0.2012, + "step": 6910 + }, + { + "epoch": 19.304469273743017, + "grad_norm": 0.8032721877098083, + "learning_rate": 2.3151955307262572e-05, + "loss": 0.1579, + "step": 6911 + }, + { + "epoch": 19.307262569832403, + "grad_norm": 0.9645483493804932, + "learning_rate": 2.3155307262569833e-05, + "loss": 0.2009, + "step": 6912 + }, + { + "epoch": 19.310055865921786, + "grad_norm": 0.7840367555618286, + "learning_rate": 2.3158659217877095e-05, + "loss": 0.1604, + "step": 6913 + }, + { + "epoch": 19.312849162011172, + "grad_norm": 0.9818944931030273, + "learning_rate": 2.316201117318436e-05, + "loss": 0.1706, + "step": 6914 + }, + { + "epoch": 19.31564245810056, + "grad_norm": 0.9691653251647949, + "learning_rate": 2.316536312849162e-05, + "loss": 0.1528, + "step": 6915 + }, + { + "epoch": 19.318435754189945, + "grad_norm": 1.0522100925445557, + "learning_rate": 2.3168715083798885e-05, + "loss": 0.1629, + "step": 6916 + }, + { + "epoch": 19.321229050279328, + "grad_norm": 1.466636300086975, + "learning_rate": 2.3172067039106147e-05, + "loss": 0.1756, + "step": 6917 + }, + { + "epoch": 19.324022346368714, + "grad_norm": 1.0476021766662598, + "learning_rate": 2.3175418994413408e-05, + "loss": 0.162, + "step": 6918 + }, + { + "epoch": 19.3268156424581, + "grad_norm": 2.947481393814087, + "learning_rate": 2.3178770949720673e-05, + "loss": 0.1668, + "step": 6919 + }, + { + "epoch": 19.329608938547487, + "grad_norm": 1.8607840538024902, + "learning_rate": 2.3182122905027934e-05, + "loss": 0.1897, + "step": 6920 + }, + { + "epoch": 19.33240223463687, + "grad_norm": 1.6308377981185913, + "learning_rate": 2.3185474860335196e-05, + "loss": 0.1798, + "step": 6921 + }, + { + "epoch": 19.335195530726256, + "grad_norm": 0.9641748070716858, + "learning_rate": 2.318882681564246e-05, + "loss": 0.1638, + "step": 6922 + }, + { + "epoch": 19.337988826815643, + "grad_norm": 2.8230268955230713, + "learning_rate": 2.319217877094972e-05, + "loss": 0.2091, + "step": 6923 + }, + { + "epoch": 19.34078212290503, + "grad_norm": 2.714028835296631, + "learning_rate": 2.3195530726256983e-05, + "loss": 0.1694, + "step": 6924 + }, + { + "epoch": 19.343575418994412, + "grad_norm": 1.010202169418335, + "learning_rate": 2.3198882681564248e-05, + "loss": 0.149, + "step": 6925 + }, + { + "epoch": 19.3463687150838, + "grad_norm": 1.1979694366455078, + "learning_rate": 2.320223463687151e-05, + "loss": 0.2023, + "step": 6926 + }, + { + "epoch": 19.349162011173185, + "grad_norm": 1.670149564743042, + "learning_rate": 2.320558659217877e-05, + "loss": 0.2195, + "step": 6927 + }, + { + "epoch": 19.35195530726257, + "grad_norm": 0.6427662968635559, + "learning_rate": 2.3208938547486035e-05, + "loss": 0.2155, + "step": 6928 + }, + { + "epoch": 19.354748603351954, + "grad_norm": 0.7266324758529663, + "learning_rate": 2.3212290502793296e-05, + "loss": 0.1687, + "step": 6929 + }, + { + "epoch": 19.35754189944134, + "grad_norm": 0.5822476148605347, + "learning_rate": 2.3215642458100558e-05, + "loss": 0.1825, + "step": 6930 + }, + { + "epoch": 19.360335195530727, + "grad_norm": 0.9457987546920776, + "learning_rate": 2.3218994413407823e-05, + "loss": 0.1902, + "step": 6931 + }, + { + "epoch": 19.363128491620113, + "grad_norm": 0.8645831942558289, + "learning_rate": 2.3222346368715084e-05, + "loss": 0.1609, + "step": 6932 + }, + { + "epoch": 19.365921787709496, + "grad_norm": 0.9243878126144409, + "learning_rate": 2.322569832402235e-05, + "loss": 0.1763, + "step": 6933 + }, + { + "epoch": 19.368715083798882, + "grad_norm": 0.6433614492416382, + "learning_rate": 2.322905027932961e-05, + "loss": 0.1876, + "step": 6934 + }, + { + "epoch": 19.37150837988827, + "grad_norm": 0.8251209855079651, + "learning_rate": 2.323240223463687e-05, + "loss": 0.2011, + "step": 6935 + }, + { + "epoch": 19.374301675977655, + "grad_norm": 0.776979923248291, + "learning_rate": 2.3235754189944136e-05, + "loss": 0.1453, + "step": 6936 + }, + { + "epoch": 19.377094972067038, + "grad_norm": 2.6410820484161377, + "learning_rate": 2.3239106145251397e-05, + "loss": 0.1805, + "step": 6937 + }, + { + "epoch": 19.379888268156424, + "grad_norm": 0.6784037947654724, + "learning_rate": 2.324245810055866e-05, + "loss": 0.1636, + "step": 6938 + }, + { + "epoch": 19.38268156424581, + "grad_norm": 1.0390244722366333, + "learning_rate": 2.3245810055865923e-05, + "loss": 0.1532, + "step": 6939 + }, + { + "epoch": 19.385474860335197, + "grad_norm": 0.7959606647491455, + "learning_rate": 2.3249162011173185e-05, + "loss": 0.1568, + "step": 6940 + }, + { + "epoch": 19.38826815642458, + "grad_norm": 0.8357700109481812, + "learning_rate": 2.3252513966480446e-05, + "loss": 0.1768, + "step": 6941 + }, + { + "epoch": 19.391061452513966, + "grad_norm": 0.8933786153793335, + "learning_rate": 2.325586592178771e-05, + "loss": 0.1626, + "step": 6942 + }, + { + "epoch": 19.393854748603353, + "grad_norm": 0.7958260178565979, + "learning_rate": 2.3259217877094972e-05, + "loss": 0.149, + "step": 6943 + }, + { + "epoch": 19.39664804469274, + "grad_norm": 1.871788501739502, + "learning_rate": 2.3262569832402234e-05, + "loss": 0.1953, + "step": 6944 + }, + { + "epoch": 19.399441340782122, + "grad_norm": 2.1579082012176514, + "learning_rate": 2.32659217877095e-05, + "loss": 0.1828, + "step": 6945 + }, + { + "epoch": 19.40223463687151, + "grad_norm": 0.8628382682800293, + "learning_rate": 2.326927374301676e-05, + "loss": 0.1606, + "step": 6946 + }, + { + "epoch": 19.405027932960895, + "grad_norm": 1.0874403715133667, + "learning_rate": 2.327262569832402e-05, + "loss": 0.1553, + "step": 6947 + }, + { + "epoch": 19.407821229050278, + "grad_norm": 0.882678210735321, + "learning_rate": 2.3275977653631286e-05, + "loss": 0.1551, + "step": 6948 + }, + { + "epoch": 19.410614525139664, + "grad_norm": 1.0417197942733765, + "learning_rate": 2.3279329608938547e-05, + "loss": 0.1605, + "step": 6949 + }, + { + "epoch": 19.41340782122905, + "grad_norm": 1.4940989017486572, + "learning_rate": 2.328268156424581e-05, + "loss": 0.1898, + "step": 6950 + }, + { + "epoch": 19.416201117318437, + "grad_norm": 2.6449427604675293, + "learning_rate": 2.3286033519553073e-05, + "loss": 0.2288, + "step": 6951 + }, + { + "epoch": 19.41899441340782, + "grad_norm": 2.6170573234558105, + "learning_rate": 2.3289385474860335e-05, + "loss": 0.2745, + "step": 6952 + }, + { + "epoch": 19.421787709497206, + "grad_norm": 1.006591796875, + "learning_rate": 2.32927374301676e-05, + "loss": 0.1688, + "step": 6953 + }, + { + "epoch": 19.424581005586592, + "grad_norm": 0.7509216070175171, + "learning_rate": 2.329608938547486e-05, + "loss": 0.1758, + "step": 6954 + }, + { + "epoch": 19.42737430167598, + "grad_norm": 0.7344080209732056, + "learning_rate": 2.3299441340782122e-05, + "loss": 0.1682, + "step": 6955 + }, + { + "epoch": 19.43016759776536, + "grad_norm": 1.2084076404571533, + "learning_rate": 2.3302793296089387e-05, + "loss": 0.1963, + "step": 6956 + }, + { + "epoch": 19.432960893854748, + "grad_norm": 0.6955239772796631, + "learning_rate": 2.3306145251396648e-05, + "loss": 0.1929, + "step": 6957 + }, + { + "epoch": 19.435754189944134, + "grad_norm": 0.8098305463790894, + "learning_rate": 2.330949720670391e-05, + "loss": 0.1783, + "step": 6958 + }, + { + "epoch": 19.43854748603352, + "grad_norm": 0.9135779142379761, + "learning_rate": 2.3312849162011174e-05, + "loss": 0.1956, + "step": 6959 + }, + { + "epoch": 19.441340782122904, + "grad_norm": 0.7495726346969604, + "learning_rate": 2.3316201117318435e-05, + "loss": 0.2051, + "step": 6960 + }, + { + "epoch": 19.44413407821229, + "grad_norm": 0.9847620725631714, + "learning_rate": 2.3319553072625697e-05, + "loss": 0.1671, + "step": 6961 + }, + { + "epoch": 19.446927374301676, + "grad_norm": 2.1936755180358887, + "learning_rate": 2.332290502793296e-05, + "loss": 0.1652, + "step": 6962 + }, + { + "epoch": 19.449720670391063, + "grad_norm": 2.818476915359497, + "learning_rate": 2.3326256983240223e-05, + "loss": 0.1476, + "step": 6963 + }, + { + "epoch": 19.452513966480446, + "grad_norm": 1.0220338106155396, + "learning_rate": 2.3329608938547484e-05, + "loss": 0.1643, + "step": 6964 + }, + { + "epoch": 19.455307262569832, + "grad_norm": 2.009242296218872, + "learning_rate": 2.333296089385475e-05, + "loss": 0.1731, + "step": 6965 + }, + { + "epoch": 19.45810055865922, + "grad_norm": 0.8311036825180054, + "learning_rate": 2.333631284916201e-05, + "loss": 0.1812, + "step": 6966 + }, + { + "epoch": 19.460893854748605, + "grad_norm": 0.9296649694442749, + "learning_rate": 2.333966480446927e-05, + "loss": 0.1571, + "step": 6967 + }, + { + "epoch": 19.463687150837988, + "grad_norm": 1.4275732040405273, + "learning_rate": 2.3343016759776536e-05, + "loss": 0.1737, + "step": 6968 + }, + { + "epoch": 19.466480446927374, + "grad_norm": 0.9006220102310181, + "learning_rate": 2.3346368715083798e-05, + "loss": 0.1859, + "step": 6969 + }, + { + "epoch": 19.46927374301676, + "grad_norm": 1.9320985078811646, + "learning_rate": 2.334972067039106e-05, + "loss": 0.1526, + "step": 6970 + }, + { + "epoch": 19.472067039106147, + "grad_norm": 1.8261640071868896, + "learning_rate": 2.3353072625698324e-05, + "loss": 0.181, + "step": 6971 + }, + { + "epoch": 19.47486033519553, + "grad_norm": 1.555362582206726, + "learning_rate": 2.3356424581005585e-05, + "loss": 0.1764, + "step": 6972 + }, + { + "epoch": 19.477653631284916, + "grad_norm": 1.3560712337493896, + "learning_rate": 2.335977653631285e-05, + "loss": 0.1741, + "step": 6973 + }, + { + "epoch": 19.480446927374302, + "grad_norm": 1.0661669969558716, + "learning_rate": 2.336312849162011e-05, + "loss": 0.185, + "step": 6974 + }, + { + "epoch": 19.48324022346369, + "grad_norm": 1.4649044275283813, + "learning_rate": 2.3366480446927373e-05, + "loss": 0.1951, + "step": 6975 + }, + { + "epoch": 19.48603351955307, + "grad_norm": 0.9307098388671875, + "learning_rate": 2.3369832402234637e-05, + "loss": 0.1668, + "step": 6976 + }, + { + "epoch": 19.488826815642458, + "grad_norm": 3.4808411598205566, + "learning_rate": 2.33731843575419e-05, + "loss": 0.2348, + "step": 6977 + }, + { + "epoch": 19.491620111731844, + "grad_norm": 0.8867921233177185, + "learning_rate": 2.337653631284916e-05, + "loss": 0.188, + "step": 6978 + }, + { + "epoch": 19.49441340782123, + "grad_norm": 0.7465406656265259, + "learning_rate": 2.3379888268156425e-05, + "loss": 0.1871, + "step": 6979 + }, + { + "epoch": 19.497206703910614, + "grad_norm": 0.6188945770263672, + "learning_rate": 2.3383240223463686e-05, + "loss": 0.1608, + "step": 6980 + }, + { + "epoch": 19.5, + "grad_norm": 0.476329505443573, + "learning_rate": 2.338659217877095e-05, + "loss": 0.181, + "step": 6981 + }, + { + "epoch": 19.502793296089386, + "grad_norm": 0.9983512163162231, + "learning_rate": 2.3389944134078216e-05, + "loss": 0.178, + "step": 6982 + }, + { + "epoch": 19.505586592178773, + "grad_norm": 0.587064802646637, + "learning_rate": 2.3393296089385477e-05, + "loss": 0.172, + "step": 6983 + }, + { + "epoch": 19.508379888268156, + "grad_norm": 1.143992304801941, + "learning_rate": 2.3396648044692738e-05, + "loss": 0.1842, + "step": 6984 + }, + { + "epoch": 19.511173184357542, + "grad_norm": 0.7603588700294495, + "learning_rate": 2.3400000000000003e-05, + "loss": 0.1871, + "step": 6985 + }, + { + "epoch": 19.51396648044693, + "grad_norm": 0.6280167698860168, + "learning_rate": 2.3403351955307264e-05, + "loss": 0.1816, + "step": 6986 + }, + { + "epoch": 19.51675977653631, + "grad_norm": 0.7175262570381165, + "learning_rate": 2.3406703910614526e-05, + "loss": 0.1681, + "step": 6987 + }, + { + "epoch": 19.519553072625698, + "grad_norm": 1.1042481660842896, + "learning_rate": 2.341005586592179e-05, + "loss": 0.183, + "step": 6988 + }, + { + "epoch": 19.522346368715084, + "grad_norm": 0.8931612968444824, + "learning_rate": 2.3413407821229052e-05, + "loss": 0.1979, + "step": 6989 + }, + { + "epoch": 19.52513966480447, + "grad_norm": 1.3061977624893188, + "learning_rate": 2.3416759776536316e-05, + "loss": 0.1706, + "step": 6990 + }, + { + "epoch": 19.527932960893853, + "grad_norm": 0.6848676800727844, + "learning_rate": 2.3420111731843578e-05, + "loss": 0.1638, + "step": 6991 + }, + { + "epoch": 19.53072625698324, + "grad_norm": 0.969746470451355, + "learning_rate": 2.342346368715084e-05, + "loss": 0.1591, + "step": 6992 + }, + { + "epoch": 19.533519553072626, + "grad_norm": 0.8271951675415039, + "learning_rate": 2.3426815642458104e-05, + "loss": 0.1794, + "step": 6993 + }, + { + "epoch": 19.536312849162012, + "grad_norm": 0.9795354008674622, + "learning_rate": 2.3430167597765365e-05, + "loss": 0.1724, + "step": 6994 + }, + { + "epoch": 19.539106145251395, + "grad_norm": 1.192864179611206, + "learning_rate": 2.3433519553072627e-05, + "loss": 0.1781, + "step": 6995 + }, + { + "epoch": 19.54189944134078, + "grad_norm": 0.9506837725639343, + "learning_rate": 2.343687150837989e-05, + "loss": 0.1723, + "step": 6996 + }, + { + "epoch": 19.544692737430168, + "grad_norm": 1.5753856897354126, + "learning_rate": 2.3440223463687153e-05, + "loss": 0.1783, + "step": 6997 + }, + { + "epoch": 19.547486033519554, + "grad_norm": 1.0980275869369507, + "learning_rate": 2.3443575418994414e-05, + "loss": 0.2054, + "step": 6998 + }, + { + "epoch": 19.550279329608937, + "grad_norm": 1.1298408508300781, + "learning_rate": 2.344692737430168e-05, + "loss": 0.1732, + "step": 6999 + }, + { + "epoch": 19.553072625698324, + "grad_norm": 1.299847960472107, + "learning_rate": 2.345027932960894e-05, + "loss": 0.1732, + "step": 7000 + }, + { + "epoch": 19.553072625698324, + "eval_cer": 0.19346338908710026, + "eval_loss": 0.27098867297172546, + "eval_runtime": 23.6742, + "eval_samples_per_second": 191.686, + "eval_steps_per_second": 0.634, + "eval_wer": 0.3436775509063993, + "step": 7000 + }, + { + "epoch": 19.55586592178771, + "grad_norm": 1.6343648433685303, + "learning_rate": 2.34536312849162e-05, + "loss": 0.2086, + "step": 7001 + }, + { + "epoch": 19.558659217877096, + "grad_norm": 6.274414539337158, + "learning_rate": 2.3456983240223466e-05, + "loss": 0.2355, + "step": 7002 + }, + { + "epoch": 19.56145251396648, + "grad_norm": 0.7837162613868713, + "learning_rate": 2.3460335195530728e-05, + "loss": 0.2011, + "step": 7003 + }, + { + "epoch": 19.564245810055866, + "grad_norm": 2.4675590991973877, + "learning_rate": 2.346368715083799e-05, + "loss": 0.1754, + "step": 7004 + }, + { + "epoch": 19.567039106145252, + "grad_norm": 0.967023491859436, + "learning_rate": 2.3467039106145254e-05, + "loss": 0.2045, + "step": 7005 + }, + { + "epoch": 19.56983240223464, + "grad_norm": 0.5713855624198914, + "learning_rate": 2.3470391061452515e-05, + "loss": 0.1987, + "step": 7006 + }, + { + "epoch": 19.57262569832402, + "grad_norm": 1.7139480113983154, + "learning_rate": 2.3473743016759776e-05, + "loss": 0.2021, + "step": 7007 + }, + { + "epoch": 19.575418994413408, + "grad_norm": 0.8103799223899841, + "learning_rate": 2.347709497206704e-05, + "loss": 0.2119, + "step": 7008 + }, + { + "epoch": 19.578212290502794, + "grad_norm": 0.6392827033996582, + "learning_rate": 2.3480446927374302e-05, + "loss": 0.1776, + "step": 7009 + }, + { + "epoch": 19.58100558659218, + "grad_norm": 0.789204478263855, + "learning_rate": 2.3483798882681567e-05, + "loss": 0.1473, + "step": 7010 + }, + { + "epoch": 19.583798882681563, + "grad_norm": 1.4158953428268433, + "learning_rate": 2.348715083798883e-05, + "loss": 0.1685, + "step": 7011 + }, + { + "epoch": 19.58659217877095, + "grad_norm": 0.6202448606491089, + "learning_rate": 2.349050279329609e-05, + "loss": 0.1804, + "step": 7012 + }, + { + "epoch": 19.589385474860336, + "grad_norm": 1.4943188428878784, + "learning_rate": 2.3493854748603355e-05, + "loss": 0.1774, + "step": 7013 + }, + { + "epoch": 19.592178770949722, + "grad_norm": 3.222291946411133, + "learning_rate": 2.3497206703910616e-05, + "loss": 0.1616, + "step": 7014 + }, + { + "epoch": 19.594972067039105, + "grad_norm": 0.6577155590057373, + "learning_rate": 2.3500558659217877e-05, + "loss": 0.1553, + "step": 7015 + }, + { + "epoch": 19.59776536312849, + "grad_norm": 0.717879593372345, + "learning_rate": 2.3503910614525142e-05, + "loss": 0.1746, + "step": 7016 + }, + { + "epoch": 19.600558659217878, + "grad_norm": 0.9056254029273987, + "learning_rate": 2.3507262569832403e-05, + "loss": 0.1729, + "step": 7017 + }, + { + "epoch": 19.60335195530726, + "grad_norm": 0.6958282589912415, + "learning_rate": 2.3510614525139665e-05, + "loss": 0.1396, + "step": 7018 + }, + { + "epoch": 19.606145251396647, + "grad_norm": 1.0400491952896118, + "learning_rate": 2.351396648044693e-05, + "loss": 0.1629, + "step": 7019 + }, + { + "epoch": 19.608938547486034, + "grad_norm": 2.7446024417877197, + "learning_rate": 2.351731843575419e-05, + "loss": 0.1679, + "step": 7020 + }, + { + "epoch": 19.61173184357542, + "grad_norm": 1.0088529586791992, + "learning_rate": 2.3520670391061452e-05, + "loss": 0.15, + "step": 7021 + }, + { + "epoch": 19.614525139664803, + "grad_norm": 0.7930108308792114, + "learning_rate": 2.3524022346368717e-05, + "loss": 0.1991, + "step": 7022 + }, + { + "epoch": 19.61731843575419, + "grad_norm": 0.9185044169425964, + "learning_rate": 2.3527374301675978e-05, + "loss": 0.2152, + "step": 7023 + }, + { + "epoch": 19.620111731843576, + "grad_norm": 1.6111767292022705, + "learning_rate": 2.353072625698324e-05, + "loss": 0.1781, + "step": 7024 + }, + { + "epoch": 19.622905027932962, + "grad_norm": 1.7347513437271118, + "learning_rate": 2.3534078212290504e-05, + "loss": 0.2271, + "step": 7025 + }, + { + "epoch": 19.625698324022345, + "grad_norm": 1.1357284784317017, + "learning_rate": 2.3537430167597766e-05, + "loss": 0.2035, + "step": 7026 + }, + { + "epoch": 19.62849162011173, + "grad_norm": 1.379331111907959, + "learning_rate": 2.354078212290503e-05, + "loss": 0.248, + "step": 7027 + }, + { + "epoch": 19.631284916201118, + "grad_norm": 0.7445060014724731, + "learning_rate": 2.354413407821229e-05, + "loss": 0.1714, + "step": 7028 + }, + { + "epoch": 19.634078212290504, + "grad_norm": 0.5250019431114197, + "learning_rate": 2.3547486033519553e-05, + "loss": 0.1613, + "step": 7029 + }, + { + "epoch": 19.636871508379887, + "grad_norm": 0.7122596502304077, + "learning_rate": 2.3550837988826818e-05, + "loss": 0.2028, + "step": 7030 + }, + { + "epoch": 19.639664804469273, + "grad_norm": 0.5755768418312073, + "learning_rate": 2.355418994413408e-05, + "loss": 0.1751, + "step": 7031 + }, + { + "epoch": 19.64245810055866, + "grad_norm": 0.7145944237709045, + "learning_rate": 2.355754189944134e-05, + "loss": 0.2035, + "step": 7032 + }, + { + "epoch": 19.645251396648046, + "grad_norm": 0.7312837243080139, + "learning_rate": 2.3560893854748605e-05, + "loss": 0.1755, + "step": 7033 + }, + { + "epoch": 19.64804469273743, + "grad_norm": 0.671195924282074, + "learning_rate": 2.3564245810055866e-05, + "loss": 0.1744, + "step": 7034 + }, + { + "epoch": 19.650837988826815, + "grad_norm": 0.8609539270401001, + "learning_rate": 2.3567597765363128e-05, + "loss": 0.1775, + "step": 7035 + }, + { + "epoch": 19.6536312849162, + "grad_norm": 0.6857690215110779, + "learning_rate": 2.3570949720670393e-05, + "loss": 0.1332, + "step": 7036 + }, + { + "epoch": 19.656424581005588, + "grad_norm": 0.6729062795639038, + "learning_rate": 2.3574301675977654e-05, + "loss": 0.1716, + "step": 7037 + }, + { + "epoch": 19.65921787709497, + "grad_norm": 0.7228321433067322, + "learning_rate": 2.3577653631284915e-05, + "loss": 0.1869, + "step": 7038 + }, + { + "epoch": 19.662011173184357, + "grad_norm": 0.7164409160614014, + "learning_rate": 2.358100558659218e-05, + "loss": 0.1774, + "step": 7039 + }, + { + "epoch": 19.664804469273744, + "grad_norm": 1.2172420024871826, + "learning_rate": 2.358435754189944e-05, + "loss": 0.199, + "step": 7040 + }, + { + "epoch": 19.66759776536313, + "grad_norm": 1.070784330368042, + "learning_rate": 2.3587709497206703e-05, + "loss": 0.1715, + "step": 7041 + }, + { + "epoch": 19.670391061452513, + "grad_norm": 1.2195768356323242, + "learning_rate": 2.3591061452513967e-05, + "loss": 0.1946, + "step": 7042 + }, + { + "epoch": 19.6731843575419, + "grad_norm": 1.0301662683486938, + "learning_rate": 2.359441340782123e-05, + "loss": 0.1645, + "step": 7043 + }, + { + "epoch": 19.675977653631286, + "grad_norm": 0.8098085522651672, + "learning_rate": 2.359776536312849e-05, + "loss": 0.2091, + "step": 7044 + }, + { + "epoch": 19.678770949720672, + "grad_norm": 0.7983869314193726, + "learning_rate": 2.3601117318435755e-05, + "loss": 0.1609, + "step": 7045 + }, + { + "epoch": 19.681564245810055, + "grad_norm": 3.7078638076782227, + "learning_rate": 2.3604469273743016e-05, + "loss": 0.1603, + "step": 7046 + }, + { + "epoch": 19.68435754189944, + "grad_norm": 1.267618179321289, + "learning_rate": 2.360782122905028e-05, + "loss": 0.1513, + "step": 7047 + }, + { + "epoch": 19.687150837988828, + "grad_norm": 1.5402214527130127, + "learning_rate": 2.3611173184357542e-05, + "loss": 0.1604, + "step": 7048 + }, + { + "epoch": 19.689944134078214, + "grad_norm": 1.341744065284729, + "learning_rate": 2.3614525139664804e-05, + "loss": 0.185, + "step": 7049 + }, + { + "epoch": 19.692737430167597, + "grad_norm": 1.1810765266418457, + "learning_rate": 2.361787709497207e-05, + "loss": 0.162, + "step": 7050 + }, + { + "epoch": 19.695530726256983, + "grad_norm": 2.28888201713562, + "learning_rate": 2.362122905027933e-05, + "loss": 0.1685, + "step": 7051 + }, + { + "epoch": 19.69832402234637, + "grad_norm": 1.8017157316207886, + "learning_rate": 2.362458100558659e-05, + "loss": 0.2463, + "step": 7052 + }, + { + "epoch": 19.701117318435756, + "grad_norm": 0.60808926820755, + "learning_rate": 2.3627932960893856e-05, + "loss": 0.1718, + "step": 7053 + }, + { + "epoch": 19.70391061452514, + "grad_norm": 0.8289648294448853, + "learning_rate": 2.3631284916201117e-05, + "loss": 0.1763, + "step": 7054 + }, + { + "epoch": 19.706703910614525, + "grad_norm": 0.5317052006721497, + "learning_rate": 2.363463687150838e-05, + "loss": 0.1792, + "step": 7055 + }, + { + "epoch": 19.70949720670391, + "grad_norm": 1.4214812517166138, + "learning_rate": 2.3637988826815643e-05, + "loss": 0.1651, + "step": 7056 + }, + { + "epoch": 19.712290502793294, + "grad_norm": 0.8538981676101685, + "learning_rate": 2.3641340782122905e-05, + "loss": 0.1935, + "step": 7057 + }, + { + "epoch": 19.71508379888268, + "grad_norm": 0.510391891002655, + "learning_rate": 2.3644692737430166e-05, + "loss": 0.1535, + "step": 7058 + }, + { + "epoch": 19.717877094972067, + "grad_norm": 1.0381782054901123, + "learning_rate": 2.364804469273743e-05, + "loss": 0.1859, + "step": 7059 + }, + { + "epoch": 19.720670391061454, + "grad_norm": 2.628239631652832, + "learning_rate": 2.3651396648044692e-05, + "loss": 0.1719, + "step": 7060 + }, + { + "epoch": 19.723463687150836, + "grad_norm": 0.7371566295623779, + "learning_rate": 2.3654748603351953e-05, + "loss": 0.1685, + "step": 7061 + }, + { + "epoch": 19.726256983240223, + "grad_norm": 0.7755251526832581, + "learning_rate": 2.3658100558659218e-05, + "loss": 0.2163, + "step": 7062 + }, + { + "epoch": 19.72905027932961, + "grad_norm": 0.7458909749984741, + "learning_rate": 2.366145251396648e-05, + "loss": 0.1664, + "step": 7063 + }, + { + "epoch": 19.731843575418996, + "grad_norm": 1.1905498504638672, + "learning_rate": 2.366480446927374e-05, + "loss": 0.1889, + "step": 7064 + }, + { + "epoch": 19.73463687150838, + "grad_norm": 0.7312001585960388, + "learning_rate": 2.3668156424581005e-05, + "loss": 0.1601, + "step": 7065 + }, + { + "epoch": 19.737430167597765, + "grad_norm": 0.7321117520332336, + "learning_rate": 2.3671508379888267e-05, + "loss": 0.1687, + "step": 7066 + }, + { + "epoch": 19.74022346368715, + "grad_norm": 0.660643458366394, + "learning_rate": 2.367486033519553e-05, + "loss": 0.1525, + "step": 7067 + }, + { + "epoch": 19.743016759776538, + "grad_norm": 0.7293267250061035, + "learning_rate": 2.3678212290502793e-05, + "loss": 0.1627, + "step": 7068 + }, + { + "epoch": 19.74581005586592, + "grad_norm": 0.9076694250106812, + "learning_rate": 2.3681564245810054e-05, + "loss": 0.1901, + "step": 7069 + }, + { + "epoch": 19.748603351955307, + "grad_norm": 1.0267351865768433, + "learning_rate": 2.368491620111732e-05, + "loss": 0.1707, + "step": 7070 + }, + { + "epoch": 19.751396648044693, + "grad_norm": 0.8089665174484253, + "learning_rate": 2.368826815642458e-05, + "loss": 0.1588, + "step": 7071 + }, + { + "epoch": 19.75418994413408, + "grad_norm": 1.303009033203125, + "learning_rate": 2.369162011173184e-05, + "loss": 0.1725, + "step": 7072 + }, + { + "epoch": 19.756983240223462, + "grad_norm": 0.8969444036483765, + "learning_rate": 2.3694972067039106e-05, + "loss": 0.1661, + "step": 7073 + }, + { + "epoch": 19.75977653631285, + "grad_norm": 1.1569768190383911, + "learning_rate": 2.3698324022346368e-05, + "loss": 0.2124, + "step": 7074 + }, + { + "epoch": 19.762569832402235, + "grad_norm": 1.3742624521255493, + "learning_rate": 2.370167597765363e-05, + "loss": 0.197, + "step": 7075 + }, + { + "epoch": 19.76536312849162, + "grad_norm": 1.4541922807693481, + "learning_rate": 2.3705027932960894e-05, + "loss": 0.194, + "step": 7076 + }, + { + "epoch": 19.768156424581004, + "grad_norm": 2.0282142162323, + "learning_rate": 2.3708379888268155e-05, + "loss": 0.2197, + "step": 7077 + }, + { + "epoch": 19.77094972067039, + "grad_norm": 0.6258147358894348, + "learning_rate": 2.371173184357542e-05, + "loss": 0.1901, + "step": 7078 + }, + { + "epoch": 19.773743016759777, + "grad_norm": 0.5484924912452698, + "learning_rate": 2.3715083798882685e-05, + "loss": 0.1842, + "step": 7079 + }, + { + "epoch": 19.776536312849164, + "grad_norm": 0.5556193590164185, + "learning_rate": 2.3718435754189946e-05, + "loss": 0.1657, + "step": 7080 + }, + { + "epoch": 19.779329608938546, + "grad_norm": 0.4046329855918884, + "learning_rate": 2.3721787709497207e-05, + "loss": 0.1502, + "step": 7081 + }, + { + "epoch": 19.782122905027933, + "grad_norm": 0.8677686452865601, + "learning_rate": 2.3725139664804472e-05, + "loss": 0.1889, + "step": 7082 + }, + { + "epoch": 19.78491620111732, + "grad_norm": 1.6597260236740112, + "learning_rate": 2.3728491620111733e-05, + "loss": 0.1663, + "step": 7083 + }, + { + "epoch": 19.787709497206706, + "grad_norm": 0.6772783994674683, + "learning_rate": 2.3731843575418998e-05, + "loss": 0.1784, + "step": 7084 + }, + { + "epoch": 19.79050279329609, + "grad_norm": 1.0501360893249512, + "learning_rate": 2.373519553072626e-05, + "loss": 0.1542, + "step": 7085 + }, + { + "epoch": 19.793296089385475, + "grad_norm": 0.590410053730011, + "learning_rate": 2.373854748603352e-05, + "loss": 0.1597, + "step": 7086 + }, + { + "epoch": 19.79608938547486, + "grad_norm": 0.6214460134506226, + "learning_rate": 2.3741899441340786e-05, + "loss": 0.221, + "step": 7087 + }, + { + "epoch": 19.798882681564244, + "grad_norm": 0.6070120334625244, + "learning_rate": 2.3745251396648047e-05, + "loss": 0.1488, + "step": 7088 + }, + { + "epoch": 19.80167597765363, + "grad_norm": 1.052262783050537, + "learning_rate": 2.3748603351955308e-05, + "loss": 0.1662, + "step": 7089 + }, + { + "epoch": 19.804469273743017, + "grad_norm": 0.7628524303436279, + "learning_rate": 2.3751955307262573e-05, + "loss": 0.1731, + "step": 7090 + }, + { + "epoch": 19.807262569832403, + "grad_norm": 0.8773073554039001, + "learning_rate": 2.3755307262569834e-05, + "loss": 0.192, + "step": 7091 + }, + { + "epoch": 19.810055865921786, + "grad_norm": 0.6047481298446655, + "learning_rate": 2.3758659217877096e-05, + "loss": 0.1664, + "step": 7092 + }, + { + "epoch": 19.812849162011172, + "grad_norm": 0.7896928191184998, + "learning_rate": 2.376201117318436e-05, + "loss": 0.1923, + "step": 7093 + }, + { + "epoch": 19.81564245810056, + "grad_norm": 0.5907187461853027, + "learning_rate": 2.3765363128491622e-05, + "loss": 0.1687, + "step": 7094 + }, + { + "epoch": 19.818435754189945, + "grad_norm": 0.9537122845649719, + "learning_rate": 2.3768715083798883e-05, + "loss": 0.1776, + "step": 7095 + }, + { + "epoch": 19.821229050279328, + "grad_norm": 1.3170573711395264, + "learning_rate": 2.3772067039106148e-05, + "loss": 0.167, + "step": 7096 + }, + { + "epoch": 19.824022346368714, + "grad_norm": 0.9531773924827576, + "learning_rate": 2.377541899441341e-05, + "loss": 0.1563, + "step": 7097 + }, + { + "epoch": 19.8268156424581, + "grad_norm": 1.1783701181411743, + "learning_rate": 2.377877094972067e-05, + "loss": 0.1797, + "step": 7098 + }, + { + "epoch": 19.829608938547487, + "grad_norm": 1.1214468479156494, + "learning_rate": 2.3782122905027935e-05, + "loss": 0.1821, + "step": 7099 + }, + { + "epoch": 19.83240223463687, + "grad_norm": 1.0067596435546875, + "learning_rate": 2.3785474860335197e-05, + "loss": 0.1832, + "step": 7100 + }, + { + "epoch": 19.835195530726256, + "grad_norm": 1.3714131116867065, + "learning_rate": 2.3788826815642458e-05, + "loss": 0.2243, + "step": 7101 + }, + { + "epoch": 19.837988826815643, + "grad_norm": 10.641332626342773, + "learning_rate": 2.3792178770949723e-05, + "loss": 0.2547, + "step": 7102 + }, + { + "epoch": 19.84078212290503, + "grad_norm": 1.0041911602020264, + "learning_rate": 2.3795530726256984e-05, + "loss": 0.1753, + "step": 7103 + }, + { + "epoch": 19.843575418994412, + "grad_norm": 0.4645034968852997, + "learning_rate": 2.379888268156425e-05, + "loss": 0.1605, + "step": 7104 + }, + { + "epoch": 19.8463687150838, + "grad_norm": 0.46517136693000793, + "learning_rate": 2.380223463687151e-05, + "loss": 0.2028, + "step": 7105 + }, + { + "epoch": 19.849162011173185, + "grad_norm": 0.6951001882553101, + "learning_rate": 2.380558659217877e-05, + "loss": 0.1456, + "step": 7106 + }, + { + "epoch": 19.85195530726257, + "grad_norm": 0.6209039688110352, + "learning_rate": 2.3808938547486036e-05, + "loss": 0.1859, + "step": 7107 + }, + { + "epoch": 19.854748603351954, + "grad_norm": 0.46214672923088074, + "learning_rate": 2.3812290502793298e-05, + "loss": 0.1763, + "step": 7108 + }, + { + "epoch": 19.85754189944134, + "grad_norm": 1.070048451423645, + "learning_rate": 2.381564245810056e-05, + "loss": 0.154, + "step": 7109 + }, + { + "epoch": 19.860335195530727, + "grad_norm": 0.6142211556434631, + "learning_rate": 2.3818994413407824e-05, + "loss": 0.1487, + "step": 7110 + }, + { + "epoch": 19.863128491620113, + "grad_norm": 0.8521487712860107, + "learning_rate": 2.3822346368715085e-05, + "loss": 0.1702, + "step": 7111 + }, + { + "epoch": 19.865921787709496, + "grad_norm": 0.7007567286491394, + "learning_rate": 2.3825698324022346e-05, + "loss": 0.1682, + "step": 7112 + }, + { + "epoch": 19.868715083798882, + "grad_norm": 0.7637000679969788, + "learning_rate": 2.382905027932961e-05, + "loss": 0.2041, + "step": 7113 + }, + { + "epoch": 19.87150837988827, + "grad_norm": 1.4018632173538208, + "learning_rate": 2.3832402234636872e-05, + "loss": 0.1712, + "step": 7114 + }, + { + "epoch": 19.874301675977655, + "grad_norm": 0.7694666981697083, + "learning_rate": 2.3835754189944134e-05, + "loss": 0.1614, + "step": 7115 + }, + { + "epoch": 19.877094972067038, + "grad_norm": 0.8510320782661438, + "learning_rate": 2.38391061452514e-05, + "loss": 0.1901, + "step": 7116 + }, + { + "epoch": 19.879888268156424, + "grad_norm": 0.7745225429534912, + "learning_rate": 2.384245810055866e-05, + "loss": 0.1418, + "step": 7117 + }, + { + "epoch": 19.88268156424581, + "grad_norm": 0.7740775942802429, + "learning_rate": 2.384581005586592e-05, + "loss": 0.1727, + "step": 7118 + }, + { + "epoch": 19.885474860335197, + "grad_norm": 0.7883382439613342, + "learning_rate": 2.3849162011173186e-05, + "loss": 0.17, + "step": 7119 + }, + { + "epoch": 19.88826815642458, + "grad_norm": 0.9805196523666382, + "learning_rate": 2.3852513966480447e-05, + "loss": 0.166, + "step": 7120 + }, + { + "epoch": 19.891061452513966, + "grad_norm": 1.2484411001205444, + "learning_rate": 2.3855865921787712e-05, + "loss": 0.1619, + "step": 7121 + }, + { + "epoch": 19.893854748603353, + "grad_norm": 0.7947695851325989, + "learning_rate": 2.3859217877094973e-05, + "loss": 0.1783, + "step": 7122 + }, + { + "epoch": 19.89664804469274, + "grad_norm": 1.0842812061309814, + "learning_rate": 2.3862569832402235e-05, + "loss": 0.1506, + "step": 7123 + }, + { + "epoch": 19.899441340782122, + "grad_norm": 0.9172155857086182, + "learning_rate": 2.38659217877095e-05, + "loss": 0.1545, + "step": 7124 + }, + { + "epoch": 19.90223463687151, + "grad_norm": 0.9505378603935242, + "learning_rate": 2.386927374301676e-05, + "loss": 0.1937, + "step": 7125 + }, + { + "epoch": 19.905027932960895, + "grad_norm": 1.6201999187469482, + "learning_rate": 2.3872625698324022e-05, + "loss": 0.2155, + "step": 7126 + }, + { + "epoch": 19.907821229050278, + "grad_norm": 2.083340883255005, + "learning_rate": 2.3875977653631287e-05, + "loss": 0.2805, + "step": 7127 + }, + { + "epoch": 19.910614525139664, + "grad_norm": 0.42051762342453003, + "learning_rate": 2.3879329608938548e-05, + "loss": 0.1816, + "step": 7128 + }, + { + "epoch": 19.91340782122905, + "grad_norm": 0.4197305738925934, + "learning_rate": 2.388268156424581e-05, + "loss": 0.1777, + "step": 7129 + }, + { + "epoch": 19.916201117318437, + "grad_norm": 0.6050782203674316, + "learning_rate": 2.3886033519553074e-05, + "loss": 0.1734, + "step": 7130 + }, + { + "epoch": 19.91899441340782, + "grad_norm": 0.6505265831947327, + "learning_rate": 2.3889385474860336e-05, + "loss": 0.1702, + "step": 7131 + }, + { + "epoch": 19.921787709497206, + "grad_norm": 0.6604123711585999, + "learning_rate": 2.3892737430167597e-05, + "loss": 0.2076, + "step": 7132 + }, + { + "epoch": 19.924581005586592, + "grad_norm": 0.6025567650794983, + "learning_rate": 2.389608938547486e-05, + "loss": 0.1683, + "step": 7133 + }, + { + "epoch": 19.92737430167598, + "grad_norm": 0.692002534866333, + "learning_rate": 2.3899441340782123e-05, + "loss": 0.1687, + "step": 7134 + }, + { + "epoch": 19.93016759776536, + "grad_norm": 0.7479541301727295, + "learning_rate": 2.3902793296089384e-05, + "loss": 0.1527, + "step": 7135 + }, + { + "epoch": 19.932960893854748, + "grad_norm": 0.5042833089828491, + "learning_rate": 2.390614525139665e-05, + "loss": 0.1663, + "step": 7136 + }, + { + "epoch": 19.935754189944134, + "grad_norm": 0.4989471733570099, + "learning_rate": 2.390949720670391e-05, + "loss": 0.1654, + "step": 7137 + }, + { + "epoch": 19.93854748603352, + "grad_norm": 0.9113882780075073, + "learning_rate": 2.3912849162011172e-05, + "loss": 0.191, + "step": 7138 + }, + { + "epoch": 19.941340782122904, + "grad_norm": 0.9373453259468079, + "learning_rate": 2.3916201117318436e-05, + "loss": 0.1669, + "step": 7139 + }, + { + "epoch": 19.94413407821229, + "grad_norm": 0.7807818055152893, + "learning_rate": 2.3919553072625698e-05, + "loss": 0.1608, + "step": 7140 + }, + { + "epoch": 19.946927374301676, + "grad_norm": 0.5629586577415466, + "learning_rate": 2.3922905027932963e-05, + "loss": 0.16, + "step": 7141 + }, + { + "epoch": 19.949720670391063, + "grad_norm": 0.7299330830574036, + "learning_rate": 2.3926256983240224e-05, + "loss": 0.1928, + "step": 7142 + }, + { + "epoch": 19.952513966480446, + "grad_norm": 1.1426470279693604, + "learning_rate": 2.3929608938547485e-05, + "loss": 0.1517, + "step": 7143 + }, + { + "epoch": 19.955307262569832, + "grad_norm": 0.727867841720581, + "learning_rate": 2.393296089385475e-05, + "loss": 0.1394, + "step": 7144 + }, + { + "epoch": 19.95810055865922, + "grad_norm": 0.6477013230323792, + "learning_rate": 2.393631284916201e-05, + "loss": 0.1481, + "step": 7145 + }, + { + "epoch": 19.960893854748605, + "grad_norm": 1.323077917098999, + "learning_rate": 2.3939664804469273e-05, + "loss": 0.1834, + "step": 7146 + }, + { + "epoch": 19.963687150837988, + "grad_norm": 1.0703915357589722, + "learning_rate": 2.3943016759776537e-05, + "loss": 0.1517, + "step": 7147 + }, + { + "epoch": 19.966480446927374, + "grad_norm": 1.2907564640045166, + "learning_rate": 2.39463687150838e-05, + "loss": 0.1869, + "step": 7148 + }, + { + "epoch": 19.96927374301676, + "grad_norm": 1.1569823026657104, + "learning_rate": 2.394972067039106e-05, + "loss": 0.1863, + "step": 7149 + }, + { + "epoch": 19.972067039106147, + "grad_norm": 1.8919814825057983, + "learning_rate": 2.3953072625698325e-05, + "loss": 0.1878, + "step": 7150 + }, + { + "epoch": 19.97486033519553, + "grad_norm": 1.9297492504119873, + "learning_rate": 2.3956424581005586e-05, + "loss": 0.1887, + "step": 7151 + }, + { + "epoch": 19.977653631284916, + "grad_norm": 3.081409454345703, + "learning_rate": 2.3959776536312848e-05, + "loss": 0.226, + "step": 7152 + }, + { + "epoch": 19.980446927374302, + "grad_norm": 0.5800550580024719, + "learning_rate": 2.3963128491620112e-05, + "loss": 0.2058, + "step": 7153 + }, + { + "epoch": 19.98324022346369, + "grad_norm": 0.9814385771751404, + "learning_rate": 2.3966480446927374e-05, + "loss": 0.1792, + "step": 7154 + }, + { + "epoch": 19.98603351955307, + "grad_norm": 0.6091160178184509, + "learning_rate": 2.3969832402234635e-05, + "loss": 0.1744, + "step": 7155 + }, + { + "epoch": 19.988826815642458, + "grad_norm": 1.0140187740325928, + "learning_rate": 2.39731843575419e-05, + "loss": 0.2088, + "step": 7156 + }, + { + "epoch": 19.991620111731844, + "grad_norm": 0.7624630928039551, + "learning_rate": 2.397653631284916e-05, + "loss": 0.1689, + "step": 7157 + }, + { + "epoch": 19.994413407821227, + "grad_norm": 3.8300278186798096, + "learning_rate": 2.3979888268156422e-05, + "loss": 0.1717, + "step": 7158 + }, + { + "epoch": 19.997206703910614, + "grad_norm": 0.7639839053153992, + "learning_rate": 2.3983240223463687e-05, + "loss": 0.1762, + "step": 7159 + }, + { + "epoch": 20.0, + "grad_norm": 2.3868401050567627, + "learning_rate": 2.398659217877095e-05, + "loss": 0.1955, + "step": 7160 + }, + { + "epoch": 20.002793296089386, + "grad_norm": 0.61519855260849, + "learning_rate": 2.3989944134078213e-05, + "loss": 0.2048, + "step": 7161 + }, + { + "epoch": 20.00558659217877, + "grad_norm": 0.41054442524909973, + "learning_rate": 2.3993296089385475e-05, + "loss": 0.1615, + "step": 7162 + }, + { + "epoch": 20.008379888268156, + "grad_norm": 0.6490452885627747, + "learning_rate": 2.3996648044692736e-05, + "loss": 0.1577, + "step": 7163 + }, + { + "epoch": 20.011173184357542, + "grad_norm": 0.5626227855682373, + "learning_rate": 2.4e-05, + "loss": 0.1775, + "step": 7164 + }, + { + "epoch": 20.01396648044693, + "grad_norm": 0.7038726806640625, + "learning_rate": 2.4003351955307262e-05, + "loss": 0.1927, + "step": 7165 + }, + { + "epoch": 20.01675977653631, + "grad_norm": 0.6124482750892639, + "learning_rate": 2.4006703910614523e-05, + "loss": 0.1456, + "step": 7166 + }, + { + "epoch": 20.019553072625698, + "grad_norm": 0.5438107252120972, + "learning_rate": 2.4010055865921788e-05, + "loss": 0.1583, + "step": 7167 + }, + { + "epoch": 20.022346368715084, + "grad_norm": 0.4599384069442749, + "learning_rate": 2.401340782122905e-05, + "loss": 0.161, + "step": 7168 + }, + { + "epoch": 20.02513966480447, + "grad_norm": 0.8765257596969604, + "learning_rate": 2.401675977653631e-05, + "loss": 0.1911, + "step": 7169 + }, + { + "epoch": 20.027932960893853, + "grad_norm": 1.032969355583191, + "learning_rate": 2.4020111731843575e-05, + "loss": 0.1813, + "step": 7170 + }, + { + "epoch": 20.03072625698324, + "grad_norm": 0.47000205516815186, + "learning_rate": 2.4023463687150837e-05, + "loss": 0.1658, + "step": 7171 + }, + { + "epoch": 20.033519553072626, + "grad_norm": 0.6781328916549683, + "learning_rate": 2.4026815642458098e-05, + "loss": 0.1745, + "step": 7172 + }, + { + "epoch": 20.036312849162012, + "grad_norm": 0.9579741954803467, + "learning_rate": 2.4030167597765363e-05, + "loss": 0.141, + "step": 7173 + }, + { + "epoch": 20.039106145251395, + "grad_norm": 0.5862113833427429, + "learning_rate": 2.4033519553072628e-05, + "loss": 0.1547, + "step": 7174 + }, + { + "epoch": 20.04189944134078, + "grad_norm": 1.011606216430664, + "learning_rate": 2.403687150837989e-05, + "loss": 0.1595, + "step": 7175 + }, + { + "epoch": 20.044692737430168, + "grad_norm": 1.0426089763641357, + "learning_rate": 2.4040223463687154e-05, + "loss": 0.2004, + "step": 7176 + }, + { + "epoch": 20.047486033519554, + "grad_norm": 0.8240382075309753, + "learning_rate": 2.4043575418994415e-05, + "loss": 0.1765, + "step": 7177 + }, + { + "epoch": 20.050279329608937, + "grad_norm": 0.6782767176628113, + "learning_rate": 2.404692737430168e-05, + "loss": 0.1615, + "step": 7178 + }, + { + "epoch": 20.053072625698324, + "grad_norm": 1.006145715713501, + "learning_rate": 2.405027932960894e-05, + "loss": 0.1545, + "step": 7179 + }, + { + "epoch": 20.05586592178771, + "grad_norm": 0.6595909595489502, + "learning_rate": 2.4053631284916202e-05, + "loss": 0.1537, + "step": 7180 + }, + { + "epoch": 20.058659217877096, + "grad_norm": 0.8688907623291016, + "learning_rate": 2.4056983240223467e-05, + "loss": 0.1441, + "step": 7181 + }, + { + "epoch": 20.06145251396648, + "grad_norm": 1.0907950401306152, + "learning_rate": 2.406033519553073e-05, + "loss": 0.1501, + "step": 7182 + }, + { + "epoch": 20.064245810055866, + "grad_norm": 0.9255315661430359, + "learning_rate": 2.406368715083799e-05, + "loss": 0.146, + "step": 7183 + }, + { + "epoch": 20.067039106145252, + "grad_norm": 1.1489713191986084, + "learning_rate": 2.4067039106145255e-05, + "loss": 0.1598, + "step": 7184 + }, + { + "epoch": 20.06983240223464, + "grad_norm": 1.314004898071289, + "learning_rate": 2.4070391061452516e-05, + "loss": 0.232, + "step": 7185 + }, + { + "epoch": 20.07262569832402, + "grad_norm": 0.6744851469993591, + "learning_rate": 2.4073743016759777e-05, + "loss": 0.1907, + "step": 7186 + }, + { + "epoch": 20.075418994413408, + "grad_norm": 0.6830453872680664, + "learning_rate": 2.4077094972067042e-05, + "loss": 0.1942, + "step": 7187 + }, + { + "epoch": 20.078212290502794, + "grad_norm": 0.6430268287658691, + "learning_rate": 2.4080446927374303e-05, + "loss": 0.2023, + "step": 7188 + }, + { + "epoch": 20.08100558659218, + "grad_norm": 0.5391658544540405, + "learning_rate": 2.4083798882681565e-05, + "loss": 0.173, + "step": 7189 + }, + { + "epoch": 20.083798882681563, + "grad_norm": 0.6770532131195068, + "learning_rate": 2.408715083798883e-05, + "loss": 0.1948, + "step": 7190 + }, + { + "epoch": 20.08659217877095, + "grad_norm": 0.5507566332817078, + "learning_rate": 2.409050279329609e-05, + "loss": 0.1696, + "step": 7191 + }, + { + "epoch": 20.089385474860336, + "grad_norm": 0.6334704160690308, + "learning_rate": 2.4093854748603352e-05, + "loss": 0.1551, + "step": 7192 + }, + { + "epoch": 20.092178770949722, + "grad_norm": 1.4038410186767578, + "learning_rate": 2.4097206703910617e-05, + "loss": 0.1794, + "step": 7193 + }, + { + "epoch": 20.094972067039105, + "grad_norm": 0.6805074214935303, + "learning_rate": 2.4100558659217878e-05, + "loss": 0.1552, + "step": 7194 + }, + { + "epoch": 20.09776536312849, + "grad_norm": 0.8782157301902771, + "learning_rate": 2.410391061452514e-05, + "loss": 0.1752, + "step": 7195 + }, + { + "epoch": 20.100558659217878, + "grad_norm": 0.7009027600288391, + "learning_rate": 2.4107262569832404e-05, + "loss": 0.1687, + "step": 7196 + }, + { + "epoch": 20.10335195530726, + "grad_norm": 0.5764031410217285, + "learning_rate": 2.4110614525139666e-05, + "loss": 0.1593, + "step": 7197 + }, + { + "epoch": 20.106145251396647, + "grad_norm": 0.6134417653083801, + "learning_rate": 2.411396648044693e-05, + "loss": 0.1703, + "step": 7198 + }, + { + "epoch": 20.108938547486034, + "grad_norm": 1.2562706470489502, + "learning_rate": 2.4117318435754192e-05, + "loss": 0.1626, + "step": 7199 + }, + { + "epoch": 20.11173184357542, + "grad_norm": 0.6783420443534851, + "learning_rate": 2.4120670391061453e-05, + "loss": 0.1476, + "step": 7200 + }, + { + "epoch": 20.114525139664803, + "grad_norm": 0.6893381476402283, + "learning_rate": 2.4124022346368718e-05, + "loss": 0.1657, + "step": 7201 + }, + { + "epoch": 20.11731843575419, + "grad_norm": 0.9269212484359741, + "learning_rate": 2.412737430167598e-05, + "loss": 0.1816, + "step": 7202 + }, + { + "epoch": 20.120111731843576, + "grad_norm": 0.77527916431427, + "learning_rate": 2.413072625698324e-05, + "loss": 0.1721, + "step": 7203 + }, + { + "epoch": 20.122905027932962, + "grad_norm": 0.6343909502029419, + "learning_rate": 2.4134078212290505e-05, + "loss": 0.1761, + "step": 7204 + }, + { + "epoch": 20.125698324022345, + "grad_norm": 1.0309758186340332, + "learning_rate": 2.4137430167597767e-05, + "loss": 0.1585, + "step": 7205 + }, + { + "epoch": 20.12849162011173, + "grad_norm": 1.135887622833252, + "learning_rate": 2.4140782122905028e-05, + "loss": 0.1959, + "step": 7206 + }, + { + "epoch": 20.131284916201118, + "grad_norm": 1.106117844581604, + "learning_rate": 2.4144134078212293e-05, + "loss": 0.1626, + "step": 7207 + }, + { + "epoch": 20.134078212290504, + "grad_norm": 1.8772125244140625, + "learning_rate": 2.4147486033519554e-05, + "loss": 0.1797, + "step": 7208 + }, + { + "epoch": 20.136871508379887, + "grad_norm": 1.4730199575424194, + "learning_rate": 2.4150837988826815e-05, + "loss": 0.2197, + "step": 7209 + }, + { + "epoch": 20.139664804469273, + "grad_norm": 2.008441925048828, + "learning_rate": 2.415418994413408e-05, + "loss": 0.2379, + "step": 7210 + }, + { + "epoch": 20.14245810055866, + "grad_norm": 0.4349430203437805, + "learning_rate": 2.415754189944134e-05, + "loss": 0.146, + "step": 7211 + }, + { + "epoch": 20.145251396648046, + "grad_norm": 0.5725405216217041, + "learning_rate": 2.4160893854748603e-05, + "loss": 0.156, + "step": 7212 + }, + { + "epoch": 20.14804469273743, + "grad_norm": 0.4250563383102417, + "learning_rate": 2.4164245810055868e-05, + "loss": 0.1646, + "step": 7213 + }, + { + "epoch": 20.150837988826815, + "grad_norm": 0.5250257849693298, + "learning_rate": 2.416759776536313e-05, + "loss": 0.1606, + "step": 7214 + }, + { + "epoch": 20.1536312849162, + "grad_norm": 1.518868327140808, + "learning_rate": 2.4170949720670394e-05, + "loss": 0.1569, + "step": 7215 + }, + { + "epoch": 20.156424581005588, + "grad_norm": 0.6838036179542542, + "learning_rate": 2.4174301675977655e-05, + "loss": 0.1517, + "step": 7216 + }, + { + "epoch": 20.15921787709497, + "grad_norm": 0.5672019124031067, + "learning_rate": 2.4177653631284916e-05, + "loss": 0.1606, + "step": 7217 + }, + { + "epoch": 20.162011173184357, + "grad_norm": 1.1969047784805298, + "learning_rate": 2.418100558659218e-05, + "loss": 0.1783, + "step": 7218 + }, + { + "epoch": 20.164804469273744, + "grad_norm": 0.5713597536087036, + "learning_rate": 2.4184357541899442e-05, + "loss": 0.1432, + "step": 7219 + }, + { + "epoch": 20.16759776536313, + "grad_norm": 0.6705187559127808, + "learning_rate": 2.4187709497206704e-05, + "loss": 0.1799, + "step": 7220 + }, + { + "epoch": 20.170391061452513, + "grad_norm": 0.9665130376815796, + "learning_rate": 2.419106145251397e-05, + "loss": 0.1803, + "step": 7221 + }, + { + "epoch": 20.1731843575419, + "grad_norm": 0.5831063985824585, + "learning_rate": 2.419441340782123e-05, + "loss": 0.1735, + "step": 7222 + }, + { + "epoch": 20.175977653631286, + "grad_norm": 0.7931868433952332, + "learning_rate": 2.419776536312849e-05, + "loss": 0.1604, + "step": 7223 + }, + { + "epoch": 20.178770949720672, + "grad_norm": 0.7248826622962952, + "learning_rate": 2.4201117318435756e-05, + "loss": 0.1577, + "step": 7224 + }, + { + "epoch": 20.181564245810055, + "grad_norm": 1.3757652044296265, + "learning_rate": 2.4204469273743017e-05, + "loss": 0.1699, + "step": 7225 + }, + { + "epoch": 20.18435754189944, + "grad_norm": 0.9463686347007751, + "learning_rate": 2.420782122905028e-05, + "loss": 0.1594, + "step": 7226 + }, + { + "epoch": 20.187150837988828, + "grad_norm": 0.9438828229904175, + "learning_rate": 2.4211173184357543e-05, + "loss": 0.1831, + "step": 7227 + }, + { + "epoch": 20.189944134078214, + "grad_norm": 0.8699836730957031, + "learning_rate": 2.4214525139664805e-05, + "loss": 0.1967, + "step": 7228 + }, + { + "epoch": 20.192737430167597, + "grad_norm": 0.7467378377914429, + "learning_rate": 2.4217877094972066e-05, + "loss": 0.1603, + "step": 7229 + }, + { + "epoch": 20.195530726256983, + "grad_norm": 0.919251561164856, + "learning_rate": 2.422122905027933e-05, + "loss": 0.1951, + "step": 7230 + }, + { + "epoch": 20.19832402234637, + "grad_norm": 0.6998412609100342, + "learning_rate": 2.4224581005586592e-05, + "loss": 0.1876, + "step": 7231 + }, + { + "epoch": 20.201117318435756, + "grad_norm": 0.8500011563301086, + "learning_rate": 2.4227932960893853e-05, + "loss": 0.1635, + "step": 7232 + }, + { + "epoch": 20.20391061452514, + "grad_norm": 1.0353529453277588, + "learning_rate": 2.4231284916201118e-05, + "loss": 0.1902, + "step": 7233 + }, + { + "epoch": 20.206703910614525, + "grad_norm": 2.048818349838257, + "learning_rate": 2.423463687150838e-05, + "loss": 0.2079, + "step": 7234 + }, + { + "epoch": 20.20949720670391, + "grad_norm": 1.187036156654358, + "learning_rate": 2.4237988826815644e-05, + "loss": 0.2567, + "step": 7235 + }, + { + "epoch": 20.212290502793294, + "grad_norm": 0.5229787826538086, + "learning_rate": 2.4241340782122906e-05, + "loss": 0.1764, + "step": 7236 + }, + { + "epoch": 20.21508379888268, + "grad_norm": 0.4859652519226074, + "learning_rate": 2.4244692737430167e-05, + "loss": 0.1517, + "step": 7237 + }, + { + "epoch": 20.217877094972067, + "grad_norm": 0.48867589235305786, + "learning_rate": 2.424804469273743e-05, + "loss": 0.1829, + "step": 7238 + }, + { + "epoch": 20.220670391061454, + "grad_norm": 0.4462956488132477, + "learning_rate": 2.4251396648044693e-05, + "loss": 0.1461, + "step": 7239 + }, + { + "epoch": 20.223463687150836, + "grad_norm": 0.4621028006076813, + "learning_rate": 2.4254748603351954e-05, + "loss": 0.1789, + "step": 7240 + }, + { + "epoch": 20.226256983240223, + "grad_norm": 1.6981010437011719, + "learning_rate": 2.425810055865922e-05, + "loss": 0.186, + "step": 7241 + }, + { + "epoch": 20.22905027932961, + "grad_norm": 0.694505512714386, + "learning_rate": 2.426145251396648e-05, + "loss": 0.1793, + "step": 7242 + }, + { + "epoch": 20.231843575418996, + "grad_norm": 0.5685461759567261, + "learning_rate": 2.4264804469273742e-05, + "loss": 0.1815, + "step": 7243 + }, + { + "epoch": 20.23463687150838, + "grad_norm": 0.6137434840202332, + "learning_rate": 2.4268156424581006e-05, + "loss": 0.1626, + "step": 7244 + }, + { + "epoch": 20.237430167597765, + "grad_norm": 0.9935303330421448, + "learning_rate": 2.4271508379888268e-05, + "loss": 0.195, + "step": 7245 + }, + { + "epoch": 20.24022346368715, + "grad_norm": 0.5671684145927429, + "learning_rate": 2.427486033519553e-05, + "loss": 0.1605, + "step": 7246 + }, + { + "epoch": 20.243016759776538, + "grad_norm": 1.0131093263626099, + "learning_rate": 2.4278212290502794e-05, + "loss": 0.1457, + "step": 7247 + }, + { + "epoch": 20.24581005586592, + "grad_norm": 0.5498825907707214, + "learning_rate": 2.4281564245810055e-05, + "loss": 0.1689, + "step": 7248 + }, + { + "epoch": 20.248603351955307, + "grad_norm": 0.9523348212242126, + "learning_rate": 2.4284916201117317e-05, + "loss": 0.1424, + "step": 7249 + }, + { + "epoch": 20.251396648044693, + "grad_norm": 0.6776517629623413, + "learning_rate": 2.428826815642458e-05, + "loss": 0.1841, + "step": 7250 + }, + { + "epoch": 20.25418994413408, + "grad_norm": 0.965962827205658, + "learning_rate": 2.4291620111731843e-05, + "loss": 0.1638, + "step": 7251 + }, + { + "epoch": 20.256983240223462, + "grad_norm": 0.812049388885498, + "learning_rate": 2.4294972067039104e-05, + "loss": 0.1856, + "step": 7252 + }, + { + "epoch": 20.25977653631285, + "grad_norm": 0.6389148831367493, + "learning_rate": 2.429832402234637e-05, + "loss": 0.1488, + "step": 7253 + }, + { + "epoch": 20.262569832402235, + "grad_norm": 0.9957918524742126, + "learning_rate": 2.430167597765363e-05, + "loss": 0.1816, + "step": 7254 + }, + { + "epoch": 20.26536312849162, + "grad_norm": 0.6952267289161682, + "learning_rate": 2.4305027932960895e-05, + "loss": 0.1639, + "step": 7255 + }, + { + "epoch": 20.268156424581004, + "grad_norm": 0.7661445140838623, + "learning_rate": 2.4308379888268156e-05, + "loss": 0.1746, + "step": 7256 + }, + { + "epoch": 20.27094972067039, + "grad_norm": 0.8507143259048462, + "learning_rate": 2.4311731843575418e-05, + "loss": 0.1851, + "step": 7257 + }, + { + "epoch": 20.273743016759777, + "grad_norm": 1.0866618156433105, + "learning_rate": 2.4315083798882682e-05, + "loss": 0.1704, + "step": 7258 + }, + { + "epoch": 20.276536312849164, + "grad_norm": 1.0667026042938232, + "learning_rate": 2.4318435754189944e-05, + "loss": 0.1887, + "step": 7259 + }, + { + "epoch": 20.279329608938546, + "grad_norm": 1.4642583131790161, + "learning_rate": 2.4321787709497205e-05, + "loss": 0.2417, + "step": 7260 + }, + { + "epoch": 20.282122905027933, + "grad_norm": 0.8003265857696533, + "learning_rate": 2.432513966480447e-05, + "loss": 0.1798, + "step": 7261 + }, + { + "epoch": 20.28491620111732, + "grad_norm": 0.6904001235961914, + "learning_rate": 2.432849162011173e-05, + "loss": 0.1808, + "step": 7262 + }, + { + "epoch": 20.287709497206706, + "grad_norm": 0.4117068350315094, + "learning_rate": 2.4331843575418992e-05, + "loss": 0.1534, + "step": 7263 + }, + { + "epoch": 20.29050279329609, + "grad_norm": 0.6426831483840942, + "learning_rate": 2.4335195530726257e-05, + "loss": 0.1763, + "step": 7264 + }, + { + "epoch": 20.293296089385475, + "grad_norm": 0.5149931907653809, + "learning_rate": 2.433854748603352e-05, + "loss": 0.1664, + "step": 7265 + }, + { + "epoch": 20.29608938547486, + "grad_norm": 0.6579505205154419, + "learning_rate": 2.434189944134078e-05, + "loss": 0.1844, + "step": 7266 + }, + { + "epoch": 20.298882681564244, + "grad_norm": 0.7798484563827515, + "learning_rate": 2.4345251396648045e-05, + "loss": 0.1691, + "step": 7267 + }, + { + "epoch": 20.30167597765363, + "grad_norm": 0.6524451375007629, + "learning_rate": 2.4348603351955306e-05, + "loss": 0.1596, + "step": 7268 + }, + { + "epoch": 20.304469273743017, + "grad_norm": 0.45775893330574036, + "learning_rate": 2.4351955307262567e-05, + "loss": 0.1375, + "step": 7269 + }, + { + "epoch": 20.307262569832403, + "grad_norm": 0.7475041151046753, + "learning_rate": 2.4355307262569832e-05, + "loss": 0.1604, + "step": 7270 + }, + { + "epoch": 20.310055865921786, + "grad_norm": 0.5455397367477417, + "learning_rate": 2.4358659217877097e-05, + "loss": 0.1646, + "step": 7271 + }, + { + "epoch": 20.312849162011172, + "grad_norm": 2.5762248039245605, + "learning_rate": 2.436201117318436e-05, + "loss": 0.1828, + "step": 7272 + }, + { + "epoch": 20.31564245810056, + "grad_norm": 0.6095021963119507, + "learning_rate": 2.4365363128491623e-05, + "loss": 0.1685, + "step": 7273 + }, + { + "epoch": 20.318435754189945, + "grad_norm": 0.6538267731666565, + "learning_rate": 2.4368715083798884e-05, + "loss": 0.1656, + "step": 7274 + }, + { + "epoch": 20.321229050279328, + "grad_norm": 0.619361937046051, + "learning_rate": 2.437206703910615e-05, + "loss": 0.1442, + "step": 7275 + }, + { + "epoch": 20.324022346368714, + "grad_norm": 0.8890283107757568, + "learning_rate": 2.437541899441341e-05, + "loss": 0.1633, + "step": 7276 + }, + { + "epoch": 20.3268156424581, + "grad_norm": 1.542181372642517, + "learning_rate": 2.437877094972067e-05, + "loss": 0.1607, + "step": 7277 + }, + { + "epoch": 20.329608938547487, + "grad_norm": 0.8082472085952759, + "learning_rate": 2.4382122905027936e-05, + "loss": 0.1602, + "step": 7278 + }, + { + "epoch": 20.33240223463687, + "grad_norm": 0.8343743681907654, + "learning_rate": 2.4385474860335198e-05, + "loss": 0.1615, + "step": 7279 + }, + { + "epoch": 20.335195530726256, + "grad_norm": 1.9286526441574097, + "learning_rate": 2.438882681564246e-05, + "loss": 0.1929, + "step": 7280 + }, + { + "epoch": 20.337988826815643, + "grad_norm": 1.458625316619873, + "learning_rate": 2.4392178770949724e-05, + "loss": 0.1635, + "step": 7281 + }, + { + "epoch": 20.34078212290503, + "grad_norm": 1.075823426246643, + "learning_rate": 2.4395530726256985e-05, + "loss": 0.1927, + "step": 7282 + }, + { + "epoch": 20.343575418994412, + "grad_norm": 0.826828122138977, + "learning_rate": 2.4398882681564246e-05, + "loss": 0.1746, + "step": 7283 + }, + { + "epoch": 20.3463687150838, + "grad_norm": 1.024996280670166, + "learning_rate": 2.440223463687151e-05, + "loss": 0.1999, + "step": 7284 + }, + { + "epoch": 20.349162011173185, + "grad_norm": 1.8599257469177246, + "learning_rate": 2.4405586592178772e-05, + "loss": 0.2516, + "step": 7285 + }, + { + "epoch": 20.35195530726257, + "grad_norm": 0.5857642889022827, + "learning_rate": 2.4408938547486034e-05, + "loss": 0.1874, + "step": 7286 + }, + { + "epoch": 20.354748603351954, + "grad_norm": 0.7376865744590759, + "learning_rate": 2.44122905027933e-05, + "loss": 0.1834, + "step": 7287 + }, + { + "epoch": 20.35754189944134, + "grad_norm": 0.6216188669204712, + "learning_rate": 2.441564245810056e-05, + "loss": 0.18, + "step": 7288 + }, + { + "epoch": 20.360335195530727, + "grad_norm": 0.5133812427520752, + "learning_rate": 2.441899441340782e-05, + "loss": 0.1667, + "step": 7289 + }, + { + "epoch": 20.363128491620113, + "grad_norm": 0.6422765254974365, + "learning_rate": 2.4422346368715086e-05, + "loss": 0.1643, + "step": 7290 + }, + { + "epoch": 20.365921787709496, + "grad_norm": 0.5210520625114441, + "learning_rate": 2.4425698324022347e-05, + "loss": 0.1629, + "step": 7291 + }, + { + "epoch": 20.368715083798882, + "grad_norm": 0.6934716105461121, + "learning_rate": 2.4429050279329612e-05, + "loss": 0.192, + "step": 7292 + }, + { + "epoch": 20.37150837988827, + "grad_norm": 0.5782346129417419, + "learning_rate": 2.4432402234636873e-05, + "loss": 0.1527, + "step": 7293 + }, + { + "epoch": 20.374301675977655, + "grad_norm": 0.42932674288749695, + "learning_rate": 2.4435754189944135e-05, + "loss": 0.1476, + "step": 7294 + }, + { + "epoch": 20.377094972067038, + "grad_norm": 0.7579542994499207, + "learning_rate": 2.44391061452514e-05, + "loss": 0.1913, + "step": 7295 + }, + { + "epoch": 20.379888268156424, + "grad_norm": 0.5900919437408447, + "learning_rate": 2.444245810055866e-05, + "loss": 0.1717, + "step": 7296 + }, + { + "epoch": 20.38268156424581, + "grad_norm": 0.8653086423873901, + "learning_rate": 2.4445810055865922e-05, + "loss": 0.1794, + "step": 7297 + }, + { + "epoch": 20.385474860335197, + "grad_norm": 1.122644066810608, + "learning_rate": 2.4449162011173187e-05, + "loss": 0.1865, + "step": 7298 + }, + { + "epoch": 20.38826815642458, + "grad_norm": 0.7683152556419373, + "learning_rate": 2.4452513966480448e-05, + "loss": 0.1556, + "step": 7299 + }, + { + "epoch": 20.391061452513966, + "grad_norm": 0.8476088643074036, + "learning_rate": 2.445586592178771e-05, + "loss": 0.1564, + "step": 7300 + }, + { + "epoch": 20.393854748603353, + "grad_norm": 0.9940696358680725, + "learning_rate": 2.4459217877094974e-05, + "loss": 0.1511, + "step": 7301 + }, + { + "epoch": 20.39664804469274, + "grad_norm": 1.1216533184051514, + "learning_rate": 2.4462569832402236e-05, + "loss": 0.1535, + "step": 7302 + }, + { + "epoch": 20.399441340782122, + "grad_norm": 1.4158662557601929, + "learning_rate": 2.4465921787709497e-05, + "loss": 0.1605, + "step": 7303 + }, + { + "epoch": 20.40223463687151, + "grad_norm": 0.8883798718452454, + "learning_rate": 2.4469273743016762e-05, + "loss": 0.1648, + "step": 7304 + }, + { + "epoch": 20.405027932960895, + "grad_norm": 0.7102013826370239, + "learning_rate": 2.4472625698324023e-05, + "loss": 0.1571, + "step": 7305 + }, + { + "epoch": 20.407821229050278, + "grad_norm": 0.9862778186798096, + "learning_rate": 2.4475977653631284e-05, + "loss": 0.1618, + "step": 7306 + }, + { + "epoch": 20.410614525139664, + "grad_norm": 1.651241660118103, + "learning_rate": 2.447932960893855e-05, + "loss": 0.1839, + "step": 7307 + }, + { + "epoch": 20.41340782122905, + "grad_norm": 0.6489852666854858, + "learning_rate": 2.448268156424581e-05, + "loss": 0.1506, + "step": 7308 + }, + { + "epoch": 20.416201117318437, + "grad_norm": 1.4753797054290771, + "learning_rate": 2.4486033519553075e-05, + "loss": 0.2154, + "step": 7309 + }, + { + "epoch": 20.41899441340782, + "grad_norm": 1.363762378692627, + "learning_rate": 2.4489385474860337e-05, + "loss": 0.2587, + "step": 7310 + }, + { + "epoch": 20.421787709497206, + "grad_norm": 1.0870394706726074, + "learning_rate": 2.4492737430167598e-05, + "loss": 0.1751, + "step": 7311 + }, + { + "epoch": 20.424581005586592, + "grad_norm": 0.4929102659225464, + "learning_rate": 2.4496089385474863e-05, + "loss": 0.1806, + "step": 7312 + }, + { + "epoch": 20.42737430167598, + "grad_norm": 1.0264811515808105, + "learning_rate": 2.4499441340782124e-05, + "loss": 0.1832, + "step": 7313 + }, + { + "epoch": 20.43016759776536, + "grad_norm": 0.7583464980125427, + "learning_rate": 2.4502793296089385e-05, + "loss": 0.1736, + "step": 7314 + }, + { + "epoch": 20.432960893854748, + "grad_norm": 0.6233735680580139, + "learning_rate": 2.450614525139665e-05, + "loss": 0.1874, + "step": 7315 + }, + { + "epoch": 20.435754189944134, + "grad_norm": 0.7129375338554382, + "learning_rate": 2.450949720670391e-05, + "loss": 0.206, + "step": 7316 + }, + { + "epoch": 20.43854748603352, + "grad_norm": 0.9507898092269897, + "learning_rate": 2.4512849162011173e-05, + "loss": 0.2015, + "step": 7317 + }, + { + "epoch": 20.441340782122904, + "grad_norm": 0.7894779443740845, + "learning_rate": 2.4516201117318438e-05, + "loss": 0.1817, + "step": 7318 + }, + { + "epoch": 20.44413407821229, + "grad_norm": 0.7469631433486938, + "learning_rate": 2.45195530726257e-05, + "loss": 0.1597, + "step": 7319 + }, + { + "epoch": 20.446927374301676, + "grad_norm": 0.8570680618286133, + "learning_rate": 2.452290502793296e-05, + "loss": 0.1596, + "step": 7320 + }, + { + "epoch": 20.449720670391063, + "grad_norm": 0.6290203928947449, + "learning_rate": 2.4526256983240225e-05, + "loss": 0.1542, + "step": 7321 + }, + { + "epoch": 20.452513966480446, + "grad_norm": 1.0106810331344604, + "learning_rate": 2.4529608938547486e-05, + "loss": 0.1785, + "step": 7322 + }, + { + "epoch": 20.455307262569832, + "grad_norm": 0.8443183898925781, + "learning_rate": 2.4532960893854748e-05, + "loss": 0.1536, + "step": 7323 + }, + { + "epoch": 20.45810055865922, + "grad_norm": 2.199228286743164, + "learning_rate": 2.4536312849162012e-05, + "loss": 0.1973, + "step": 7324 + }, + { + "epoch": 20.460893854748605, + "grad_norm": 0.5094379782676697, + "learning_rate": 2.4539664804469274e-05, + "loss": 0.1398, + "step": 7325 + }, + { + "epoch": 20.463687150837988, + "grad_norm": 0.7860713005065918, + "learning_rate": 2.4543016759776535e-05, + "loss": 0.1454, + "step": 7326 + }, + { + "epoch": 20.466480446927374, + "grad_norm": 0.6831288933753967, + "learning_rate": 2.45463687150838e-05, + "loss": 0.1545, + "step": 7327 + }, + { + "epoch": 20.46927374301676, + "grad_norm": 1.1245063543319702, + "learning_rate": 2.454972067039106e-05, + "loss": 0.1928, + "step": 7328 + }, + { + "epoch": 20.472067039106147, + "grad_norm": 1.4414846897125244, + "learning_rate": 2.4553072625698326e-05, + "loss": 0.1824, + "step": 7329 + }, + { + "epoch": 20.47486033519553, + "grad_norm": 1.5563138723373413, + "learning_rate": 2.4556424581005587e-05, + "loss": 0.1651, + "step": 7330 + }, + { + "epoch": 20.477653631284916, + "grad_norm": 1.1461706161499023, + "learning_rate": 2.455977653631285e-05, + "loss": 0.1629, + "step": 7331 + }, + { + "epoch": 20.480446927374302, + "grad_norm": 1.0165942907333374, + "learning_rate": 2.4563128491620113e-05, + "loss": 0.1803, + "step": 7332 + }, + { + "epoch": 20.48324022346369, + "grad_norm": 1.7796685695648193, + "learning_rate": 2.4566480446927375e-05, + "loss": 0.1491, + "step": 7333 + }, + { + "epoch": 20.48603351955307, + "grad_norm": 2.50408673286438, + "learning_rate": 2.4569832402234636e-05, + "loss": 0.1924, + "step": 7334 + }, + { + "epoch": 20.488826815642458, + "grad_norm": 1.673673152923584, + "learning_rate": 2.45731843575419e-05, + "loss": 0.2056, + "step": 7335 + }, + { + "epoch": 20.491620111731844, + "grad_norm": 0.6392183899879456, + "learning_rate": 2.4576536312849162e-05, + "loss": 0.1702, + "step": 7336 + }, + { + "epoch": 20.49441340782123, + "grad_norm": 0.6467755436897278, + "learning_rate": 2.4579888268156423e-05, + "loss": 0.1775, + "step": 7337 + }, + { + "epoch": 20.497206703910614, + "grad_norm": 0.6671670079231262, + "learning_rate": 2.4583240223463688e-05, + "loss": 0.2126, + "step": 7338 + }, + { + "epoch": 20.5, + "grad_norm": 0.5246102213859558, + "learning_rate": 2.458659217877095e-05, + "loss": 0.1901, + "step": 7339 + }, + { + "epoch": 20.502793296089386, + "grad_norm": 0.6172130107879639, + "learning_rate": 2.458994413407821e-05, + "loss": 0.1872, + "step": 7340 + }, + { + "epoch": 20.505586592178773, + "grad_norm": 0.6036800742149353, + "learning_rate": 2.4593296089385476e-05, + "loss": 0.1784, + "step": 7341 + }, + { + "epoch": 20.508379888268156, + "grad_norm": 0.9313324093818665, + "learning_rate": 2.4596648044692737e-05, + "loss": 0.1932, + "step": 7342 + }, + { + "epoch": 20.511173184357542, + "grad_norm": 0.4709869921207428, + "learning_rate": 2.4599999999999998e-05, + "loss": 0.1569, + "step": 7343 + }, + { + "epoch": 20.51396648044693, + "grad_norm": 0.9404955506324768, + "learning_rate": 2.4603351955307263e-05, + "loss": 0.1739, + "step": 7344 + }, + { + "epoch": 20.51675977653631, + "grad_norm": 0.8395499587059021, + "learning_rate": 2.4606703910614524e-05, + "loss": 0.1667, + "step": 7345 + }, + { + "epoch": 20.519553072625698, + "grad_norm": 1.503735899925232, + "learning_rate": 2.4610055865921786e-05, + "loss": 0.1647, + "step": 7346 + }, + { + "epoch": 20.522346368715084, + "grad_norm": 0.728862464427948, + "learning_rate": 2.461340782122905e-05, + "loss": 0.1749, + "step": 7347 + }, + { + "epoch": 20.52513966480447, + "grad_norm": 0.5939711332321167, + "learning_rate": 2.4616759776536312e-05, + "loss": 0.1532, + "step": 7348 + }, + { + "epoch": 20.527932960893853, + "grad_norm": 1.5798299312591553, + "learning_rate": 2.4620111731843576e-05, + "loss": 0.1716, + "step": 7349 + }, + { + "epoch": 20.53072625698324, + "grad_norm": 2.298173189163208, + "learning_rate": 2.4623463687150838e-05, + "loss": 0.1677, + "step": 7350 + }, + { + "epoch": 20.533519553072626, + "grad_norm": 1.0109812021255493, + "learning_rate": 2.46268156424581e-05, + "loss": 0.1631, + "step": 7351 + }, + { + "epoch": 20.536312849162012, + "grad_norm": 0.8819444179534912, + "learning_rate": 2.4630167597765364e-05, + "loss": 0.1782, + "step": 7352 + }, + { + "epoch": 20.539106145251395, + "grad_norm": 2.270177125930786, + "learning_rate": 2.4633519553072625e-05, + "loss": 0.1597, + "step": 7353 + }, + { + "epoch": 20.54189944134078, + "grad_norm": 1.4417927265167236, + "learning_rate": 2.4636871508379887e-05, + "loss": 0.1726, + "step": 7354 + }, + { + "epoch": 20.544692737430168, + "grad_norm": 1.0863862037658691, + "learning_rate": 2.464022346368715e-05, + "loss": 0.1397, + "step": 7355 + }, + { + "epoch": 20.547486033519554, + "grad_norm": 1.1923902034759521, + "learning_rate": 2.4643575418994413e-05, + "loss": 0.174, + "step": 7356 + }, + { + "epoch": 20.550279329608937, + "grad_norm": 1.1960972547531128, + "learning_rate": 2.4646927374301674e-05, + "loss": 0.1962, + "step": 7357 + }, + { + "epoch": 20.553072625698324, + "grad_norm": 1.9651217460632324, + "learning_rate": 2.465027932960894e-05, + "loss": 0.1811, + "step": 7358 + }, + { + "epoch": 20.55586592178771, + "grad_norm": 1.1703851222991943, + "learning_rate": 2.46536312849162e-05, + "loss": 0.1859, + "step": 7359 + }, + { + "epoch": 20.558659217877096, + "grad_norm": 1.6236518621444702, + "learning_rate": 2.465698324022346e-05, + "loss": 0.2134, + "step": 7360 + }, + { + "epoch": 20.56145251396648, + "grad_norm": 0.800959050655365, + "learning_rate": 2.4660335195530726e-05, + "loss": 0.1737, + "step": 7361 + }, + { + "epoch": 20.564245810055866, + "grad_norm": 1.5326471328735352, + "learning_rate": 2.4663687150837988e-05, + "loss": 0.196, + "step": 7362 + }, + { + "epoch": 20.567039106145252, + "grad_norm": 0.9877214431762695, + "learning_rate": 2.466703910614525e-05, + "loss": 0.1781, + "step": 7363 + }, + { + "epoch": 20.56983240223464, + "grad_norm": 0.4529157876968384, + "learning_rate": 2.4670391061452514e-05, + "loss": 0.1435, + "step": 7364 + }, + { + "epoch": 20.57262569832402, + "grad_norm": 1.7770507335662842, + "learning_rate": 2.4673743016759775e-05, + "loss": 0.1916, + "step": 7365 + }, + { + "epoch": 20.575418994413408, + "grad_norm": 0.7784431576728821, + "learning_rate": 2.467709497206704e-05, + "loss": 0.182, + "step": 7366 + }, + { + "epoch": 20.578212290502794, + "grad_norm": 1.326188564300537, + "learning_rate": 2.4680446927374304e-05, + "loss": 0.1829, + "step": 7367 + }, + { + "epoch": 20.58100558659218, + "grad_norm": 0.7248008847236633, + "learning_rate": 2.4683798882681566e-05, + "loss": 0.1767, + "step": 7368 + }, + { + "epoch": 20.583798882681563, + "grad_norm": 0.854132890701294, + "learning_rate": 2.468715083798883e-05, + "loss": 0.1896, + "step": 7369 + }, + { + "epoch": 20.58659217877095, + "grad_norm": 1.1993522644042969, + "learning_rate": 2.4690502793296092e-05, + "loss": 0.1716, + "step": 7370 + }, + { + "epoch": 20.589385474860336, + "grad_norm": 0.7786386609077454, + "learning_rate": 2.4693854748603353e-05, + "loss": 0.179, + "step": 7371 + }, + { + "epoch": 20.592178770949722, + "grad_norm": 0.7147282361984253, + "learning_rate": 2.4697206703910618e-05, + "loss": 0.1698, + "step": 7372 + }, + { + "epoch": 20.594972067039105, + "grad_norm": 0.7605031728744507, + "learning_rate": 2.470055865921788e-05, + "loss": 0.1518, + "step": 7373 + }, + { + "epoch": 20.59776536312849, + "grad_norm": 0.8008034229278564, + "learning_rate": 2.470391061452514e-05, + "loss": 0.1651, + "step": 7374 + }, + { + "epoch": 20.600558659217878, + "grad_norm": 1.158530592918396, + "learning_rate": 2.4707262569832405e-05, + "loss": 0.1805, + "step": 7375 + }, + { + "epoch": 20.60335195530726, + "grad_norm": 1.9630532264709473, + "learning_rate": 2.4710614525139667e-05, + "loss": 0.1626, + "step": 7376 + }, + { + "epoch": 20.606145251396647, + "grad_norm": 0.7807005643844604, + "learning_rate": 2.4713966480446928e-05, + "loss": 0.1706, + "step": 7377 + }, + { + "epoch": 20.608938547486034, + "grad_norm": 1.063786506652832, + "learning_rate": 2.4717318435754193e-05, + "loss": 0.1687, + "step": 7378 + }, + { + "epoch": 20.61173184357542, + "grad_norm": 1.1144468784332275, + "learning_rate": 2.4720670391061454e-05, + "loss": 0.1834, + "step": 7379 + }, + { + "epoch": 20.614525139664803, + "grad_norm": 0.8509934544563293, + "learning_rate": 2.4724022346368715e-05, + "loss": 0.1483, + "step": 7380 + }, + { + "epoch": 20.61731843575419, + "grad_norm": 1.24480402469635, + "learning_rate": 2.472737430167598e-05, + "loss": 0.1546, + "step": 7381 + }, + { + "epoch": 20.620111731843576, + "grad_norm": 0.8450750708580017, + "learning_rate": 2.473072625698324e-05, + "loss": 0.1748, + "step": 7382 + }, + { + "epoch": 20.622905027932962, + "grad_norm": 1.2707113027572632, + "learning_rate": 2.4734078212290503e-05, + "loss": 0.1879, + "step": 7383 + }, + { + "epoch": 20.625698324022345, + "grad_norm": 2.0281901359558105, + "learning_rate": 2.4737430167597768e-05, + "loss": 0.1831, + "step": 7384 + }, + { + "epoch": 20.62849162011173, + "grad_norm": 2.5433123111724854, + "learning_rate": 2.474078212290503e-05, + "loss": 0.2737, + "step": 7385 + }, + { + "epoch": 20.631284916201118, + "grad_norm": 1.4208136796951294, + "learning_rate": 2.4744134078212294e-05, + "loss": 0.1632, + "step": 7386 + }, + { + "epoch": 20.634078212290504, + "grad_norm": 0.6162134408950806, + "learning_rate": 2.4747486033519555e-05, + "loss": 0.1685, + "step": 7387 + }, + { + "epoch": 20.636871508379887, + "grad_norm": 0.6478732824325562, + "learning_rate": 2.4750837988826816e-05, + "loss": 0.1745, + "step": 7388 + }, + { + "epoch": 20.639664804469273, + "grad_norm": 0.8224806785583496, + "learning_rate": 2.475418994413408e-05, + "loss": 0.1999, + "step": 7389 + }, + { + "epoch": 20.64245810055866, + "grad_norm": 0.8901958465576172, + "learning_rate": 2.4757541899441342e-05, + "loss": 0.1782, + "step": 7390 + }, + { + "epoch": 20.645251396648046, + "grad_norm": 1.378720998764038, + "learning_rate": 2.4760893854748604e-05, + "loss": 0.2046, + "step": 7391 + }, + { + "epoch": 20.64804469273743, + "grad_norm": 0.7326593399047852, + "learning_rate": 2.476424581005587e-05, + "loss": 0.1456, + "step": 7392 + }, + { + "epoch": 20.650837988826815, + "grad_norm": 0.6679059267044067, + "learning_rate": 2.476759776536313e-05, + "loss": 0.1831, + "step": 7393 + }, + { + "epoch": 20.6536312849162, + "grad_norm": 0.621074914932251, + "learning_rate": 2.477094972067039e-05, + "loss": 0.1629, + "step": 7394 + }, + { + "epoch": 20.656424581005588, + "grad_norm": 0.8466845750808716, + "learning_rate": 2.4774301675977656e-05, + "loss": 0.1639, + "step": 7395 + }, + { + "epoch": 20.65921787709497, + "grad_norm": 0.7260837554931641, + "learning_rate": 2.4777653631284917e-05, + "loss": 0.1627, + "step": 7396 + }, + { + "epoch": 20.662011173184357, + "grad_norm": 0.6293861269950867, + "learning_rate": 2.478100558659218e-05, + "loss": 0.1828, + "step": 7397 + }, + { + "epoch": 20.664804469273744, + "grad_norm": 0.5230247974395752, + "learning_rate": 2.4784357541899443e-05, + "loss": 0.1705, + "step": 7398 + }, + { + "epoch": 20.66759776536313, + "grad_norm": 0.896108090877533, + "learning_rate": 2.4787709497206705e-05, + "loss": 0.1977, + "step": 7399 + }, + { + "epoch": 20.670391061452513, + "grad_norm": 1.297526478767395, + "learning_rate": 2.4791061452513966e-05, + "loss": 0.1335, + "step": 7400 + }, + { + "epoch": 20.6731843575419, + "grad_norm": 0.6386539340019226, + "learning_rate": 2.479441340782123e-05, + "loss": 0.1752, + "step": 7401 + }, + { + "epoch": 20.675977653631286, + "grad_norm": 0.8792681694030762, + "learning_rate": 2.4797765363128492e-05, + "loss": 0.1462, + "step": 7402 + }, + { + "epoch": 20.678770949720672, + "grad_norm": 0.8164882659912109, + "learning_rate": 2.4801117318435757e-05, + "loss": 0.1736, + "step": 7403 + }, + { + "epoch": 20.681564245810055, + "grad_norm": 0.7149460911750793, + "learning_rate": 2.4804469273743018e-05, + "loss": 0.1325, + "step": 7404 + }, + { + "epoch": 20.68435754189944, + "grad_norm": 0.7507341504096985, + "learning_rate": 2.480782122905028e-05, + "loss": 0.1863, + "step": 7405 + }, + { + "epoch": 20.687150837988828, + "grad_norm": 1.0219967365264893, + "learning_rate": 2.4811173184357544e-05, + "loss": 0.1887, + "step": 7406 + }, + { + "epoch": 20.689944134078214, + "grad_norm": 1.6698451042175293, + "learning_rate": 2.4814525139664806e-05, + "loss": 0.1821, + "step": 7407 + }, + { + "epoch": 20.692737430167597, + "grad_norm": 2.147526741027832, + "learning_rate": 2.4817877094972067e-05, + "loss": 0.1981, + "step": 7408 + }, + { + "epoch": 20.695530726256983, + "grad_norm": 0.8978058099746704, + "learning_rate": 2.4821229050279332e-05, + "loss": 0.1648, + "step": 7409 + }, + { + "epoch": 20.69832402234637, + "grad_norm": 3.0040910243988037, + "learning_rate": 2.4824581005586593e-05, + "loss": 0.2283, + "step": 7410 + }, + { + "epoch": 20.701117318435756, + "grad_norm": 0.759616494178772, + "learning_rate": 2.4827932960893854e-05, + "loss": 0.1874, + "step": 7411 + }, + { + "epoch": 20.70391061452514, + "grad_norm": 0.4501548409461975, + "learning_rate": 2.483128491620112e-05, + "loss": 0.1567, + "step": 7412 + }, + { + "epoch": 20.706703910614525, + "grad_norm": 0.5413023829460144, + "learning_rate": 2.483463687150838e-05, + "loss": 0.1713, + "step": 7413 + }, + { + "epoch": 20.70949720670391, + "grad_norm": 0.6946665048599243, + "learning_rate": 2.4837988826815642e-05, + "loss": 0.1878, + "step": 7414 + }, + { + "epoch": 20.712290502793294, + "grad_norm": 0.5650476217269897, + "learning_rate": 2.4841340782122907e-05, + "loss": 0.1551, + "step": 7415 + }, + { + "epoch": 20.71508379888268, + "grad_norm": 1.0425302982330322, + "learning_rate": 2.4844692737430168e-05, + "loss": 0.1928, + "step": 7416 + }, + { + "epoch": 20.717877094972067, + "grad_norm": 0.454826682806015, + "learning_rate": 2.484804469273743e-05, + "loss": 0.1674, + "step": 7417 + }, + { + "epoch": 20.720670391061454, + "grad_norm": 0.8656522631645203, + "learning_rate": 2.4851396648044694e-05, + "loss": 0.1925, + "step": 7418 + }, + { + "epoch": 20.723463687150836, + "grad_norm": 1.643303394317627, + "learning_rate": 2.4854748603351955e-05, + "loss": 0.1551, + "step": 7419 + }, + { + "epoch": 20.726256983240223, + "grad_norm": 0.9217292666435242, + "learning_rate": 2.4858100558659217e-05, + "loss": 0.1854, + "step": 7420 + }, + { + "epoch": 20.72905027932961, + "grad_norm": 0.7229206562042236, + "learning_rate": 2.486145251396648e-05, + "loss": 0.1574, + "step": 7421 + }, + { + "epoch": 20.731843575418996, + "grad_norm": 0.6950011253356934, + "learning_rate": 2.4864804469273743e-05, + "loss": 0.1614, + "step": 7422 + }, + { + "epoch": 20.73463687150838, + "grad_norm": 1.003848671913147, + "learning_rate": 2.4868156424581008e-05, + "loss": 0.1889, + "step": 7423 + }, + { + "epoch": 20.737430167597765, + "grad_norm": 1.2001631259918213, + "learning_rate": 2.487150837988827e-05, + "loss": 0.1638, + "step": 7424 + }, + { + "epoch": 20.74022346368715, + "grad_norm": 1.0017588138580322, + "learning_rate": 2.487486033519553e-05, + "loss": 0.1377, + "step": 7425 + }, + { + "epoch": 20.743016759776538, + "grad_norm": 0.6768364906311035, + "learning_rate": 2.4878212290502795e-05, + "loss": 0.1408, + "step": 7426 + }, + { + "epoch": 20.74581005586592, + "grad_norm": 1.1269067525863647, + "learning_rate": 2.4881564245810056e-05, + "loss": 0.1613, + "step": 7427 + }, + { + "epoch": 20.748603351955307, + "grad_norm": 2.0170702934265137, + "learning_rate": 2.4884916201117318e-05, + "loss": 0.1463, + "step": 7428 + }, + { + "epoch": 20.751396648044693, + "grad_norm": 0.9876300692558289, + "learning_rate": 2.4888268156424582e-05, + "loss": 0.1502, + "step": 7429 + }, + { + "epoch": 20.75418994413408, + "grad_norm": 0.7776789665222168, + "learning_rate": 2.4891620111731844e-05, + "loss": 0.1879, + "step": 7430 + }, + { + "epoch": 20.756983240223462, + "grad_norm": 0.9100198149681091, + "learning_rate": 2.4894972067039105e-05, + "loss": 0.1449, + "step": 7431 + }, + { + "epoch": 20.75977653631285, + "grad_norm": 1.6666736602783203, + "learning_rate": 2.489832402234637e-05, + "loss": 0.1563, + "step": 7432 + }, + { + "epoch": 20.762569832402235, + "grad_norm": 0.9323330521583557, + "learning_rate": 2.490167597765363e-05, + "loss": 0.1567, + "step": 7433 + }, + { + "epoch": 20.76536312849162, + "grad_norm": 1.1774739027023315, + "learning_rate": 2.4905027932960892e-05, + "loss": 0.1784, + "step": 7434 + }, + { + "epoch": 20.768156424581004, + "grad_norm": 4.261353969573975, + "learning_rate": 2.4908379888268157e-05, + "loss": 0.246, + "step": 7435 + }, + { + "epoch": 20.77094972067039, + "grad_norm": 0.8061210513114929, + "learning_rate": 2.491173184357542e-05, + "loss": 0.1797, + "step": 7436 + }, + { + "epoch": 20.773743016759777, + "grad_norm": 0.793525755405426, + "learning_rate": 2.491508379888268e-05, + "loss": 0.1709, + "step": 7437 + }, + { + "epoch": 20.776536312849164, + "grad_norm": 0.5429149270057678, + "learning_rate": 2.4918435754189945e-05, + "loss": 0.1934, + "step": 7438 + }, + { + "epoch": 20.779329608938546, + "grad_norm": 0.5176782011985779, + "learning_rate": 2.4921787709497206e-05, + "loss": 0.1668, + "step": 7439 + }, + { + "epoch": 20.782122905027933, + "grad_norm": 0.8837960362434387, + "learning_rate": 2.4925139664804467e-05, + "loss": 0.1759, + "step": 7440 + }, + { + "epoch": 20.78491620111732, + "grad_norm": 0.7653173804283142, + "learning_rate": 2.4928491620111732e-05, + "loss": 0.1922, + "step": 7441 + }, + { + "epoch": 20.787709497206706, + "grad_norm": 3.625697135925293, + "learning_rate": 2.4931843575418993e-05, + "loss": 0.1543, + "step": 7442 + }, + { + "epoch": 20.79050279329609, + "grad_norm": 0.7548379898071289, + "learning_rate": 2.4935195530726258e-05, + "loss": 0.175, + "step": 7443 + }, + { + "epoch": 20.793296089385475, + "grad_norm": 1.4809361696243286, + "learning_rate": 2.493854748603352e-05, + "loss": 0.1737, + "step": 7444 + }, + { + "epoch": 20.79608938547486, + "grad_norm": 0.6250826716423035, + "learning_rate": 2.494189944134078e-05, + "loss": 0.1589, + "step": 7445 + }, + { + "epoch": 20.798882681564244, + "grad_norm": 0.7351654767990112, + "learning_rate": 2.4945251396648046e-05, + "loss": 0.2302, + "step": 7446 + }, + { + "epoch": 20.80167597765363, + "grad_norm": 0.9426538944244385, + "learning_rate": 2.4948603351955307e-05, + "loss": 0.1642, + "step": 7447 + }, + { + "epoch": 20.804469273743017, + "grad_norm": 0.6565791368484497, + "learning_rate": 2.4951955307262568e-05, + "loss": 0.1415, + "step": 7448 + }, + { + "epoch": 20.807262569832403, + "grad_norm": 0.7588261365890503, + "learning_rate": 2.4955307262569833e-05, + "loss": 0.1608, + "step": 7449 + }, + { + "epoch": 20.810055865921786, + "grad_norm": 1.3683723211288452, + "learning_rate": 2.4958659217877094e-05, + "loss": 0.1892, + "step": 7450 + }, + { + "epoch": 20.812849162011172, + "grad_norm": 0.7353014945983887, + "learning_rate": 2.4962011173184356e-05, + "loss": 0.1531, + "step": 7451 + }, + { + "epoch": 20.81564245810056, + "grad_norm": 0.8497700095176697, + "learning_rate": 2.496536312849162e-05, + "loss": 0.1891, + "step": 7452 + }, + { + "epoch": 20.818435754189945, + "grad_norm": 1.7789686918258667, + "learning_rate": 2.4968715083798882e-05, + "loss": 0.1848, + "step": 7453 + }, + { + "epoch": 20.821229050279328, + "grad_norm": 1.0376248359680176, + "learning_rate": 2.4972067039106143e-05, + "loss": 0.1791, + "step": 7454 + }, + { + "epoch": 20.824022346368714, + "grad_norm": 1.203192114830017, + "learning_rate": 2.4975418994413408e-05, + "loss": 0.1827, + "step": 7455 + }, + { + "epoch": 20.8268156424581, + "grad_norm": 0.6582546830177307, + "learning_rate": 2.497877094972067e-05, + "loss": 0.176, + "step": 7456 + }, + { + "epoch": 20.829608938547487, + "grad_norm": 2.151252508163452, + "learning_rate": 2.498212290502793e-05, + "loss": 0.1643, + "step": 7457 + }, + { + "epoch": 20.83240223463687, + "grad_norm": 1.1536120176315308, + "learning_rate": 2.4985474860335195e-05, + "loss": 0.1749, + "step": 7458 + }, + { + "epoch": 20.835195530726256, + "grad_norm": 1.7963380813598633, + "learning_rate": 2.4988826815642457e-05, + "loss": 0.2049, + "step": 7459 + }, + { + "epoch": 20.837988826815643, + "grad_norm": 2.9868667125701904, + "learning_rate": 2.499217877094972e-05, + "loss": 0.2745, + "step": 7460 + }, + { + "epoch": 20.84078212290503, + "grad_norm": 0.6971890926361084, + "learning_rate": 2.4995530726256983e-05, + "loss": 0.2348, + "step": 7461 + }, + { + "epoch": 20.843575418994412, + "grad_norm": 0.8118026852607727, + "learning_rate": 2.4998882681564247e-05, + "loss": 0.1976, + "step": 7462 + }, + { + "epoch": 20.8463687150838, + "grad_norm": 0.6770991683006287, + "learning_rate": 2.500223463687151e-05, + "loss": 0.1675, + "step": 7463 + }, + { + "epoch": 20.849162011173185, + "grad_norm": 0.5991693139076233, + "learning_rate": 2.5005586592178773e-05, + "loss": 0.1802, + "step": 7464 + }, + { + "epoch": 20.85195530726257, + "grad_norm": 0.5252550840377808, + "learning_rate": 2.5008938547486035e-05, + "loss": 0.1868, + "step": 7465 + }, + { + "epoch": 20.854748603351954, + "grad_norm": 0.4586496949195862, + "learning_rate": 2.50122905027933e-05, + "loss": 0.1547, + "step": 7466 + }, + { + "epoch": 20.85754189944134, + "grad_norm": 0.9771348834037781, + "learning_rate": 2.501564245810056e-05, + "loss": 0.1777, + "step": 7467 + }, + { + "epoch": 20.860335195530727, + "grad_norm": 0.5789086818695068, + "learning_rate": 2.5018994413407822e-05, + "loss": 0.1491, + "step": 7468 + }, + { + "epoch": 20.863128491620113, + "grad_norm": 0.7095538973808289, + "learning_rate": 2.5022346368715087e-05, + "loss": 0.18, + "step": 7469 + }, + { + "epoch": 20.865921787709496, + "grad_norm": 0.7986768484115601, + "learning_rate": 2.502569832402235e-05, + "loss": 0.1581, + "step": 7470 + }, + { + "epoch": 20.868715083798882, + "grad_norm": 0.8633471727371216, + "learning_rate": 2.502905027932961e-05, + "loss": 0.1666, + "step": 7471 + }, + { + "epoch": 20.87150837988827, + "grad_norm": 0.7789058089256287, + "learning_rate": 2.5032402234636874e-05, + "loss": 0.2, + "step": 7472 + }, + { + "epoch": 20.874301675977655, + "grad_norm": 1.0025207996368408, + "learning_rate": 2.5035754189944136e-05, + "loss": 0.1523, + "step": 7473 + }, + { + "epoch": 20.877094972067038, + "grad_norm": 0.8804805278778076, + "learning_rate": 2.5039106145251397e-05, + "loss": 0.1703, + "step": 7474 + }, + { + "epoch": 20.879888268156424, + "grad_norm": 0.8477663397789001, + "learning_rate": 2.5042458100558662e-05, + "loss": 0.1975, + "step": 7475 + }, + { + "epoch": 20.88268156424581, + "grad_norm": 1.6500921249389648, + "learning_rate": 2.5045810055865923e-05, + "loss": 0.1592, + "step": 7476 + }, + { + "epoch": 20.885474860335197, + "grad_norm": 0.9993670582771301, + "learning_rate": 2.5049162011173185e-05, + "loss": 0.1622, + "step": 7477 + }, + { + "epoch": 20.88826815642458, + "grad_norm": 0.8457367420196533, + "learning_rate": 2.505251396648045e-05, + "loss": 0.1199, + "step": 7478 + }, + { + "epoch": 20.891061452513966, + "grad_norm": 0.8248480558395386, + "learning_rate": 2.505586592178771e-05, + "loss": 0.1501, + "step": 7479 + }, + { + "epoch": 20.893854748603353, + "grad_norm": 1.011444330215454, + "learning_rate": 2.5059217877094975e-05, + "loss": 0.1702, + "step": 7480 + }, + { + "epoch": 20.89664804469274, + "grad_norm": 2.2607202529907227, + "learning_rate": 2.5062569832402237e-05, + "loss": 0.1947, + "step": 7481 + }, + { + "epoch": 20.899441340782122, + "grad_norm": 0.9409871697425842, + "learning_rate": 2.5065921787709498e-05, + "loss": 0.1473, + "step": 7482 + }, + { + "epoch": 20.90223463687151, + "grad_norm": 2.1449713706970215, + "learning_rate": 2.5069273743016763e-05, + "loss": 0.2072, + "step": 7483 + }, + { + "epoch": 20.905027932960895, + "grad_norm": 1.6189316511154175, + "learning_rate": 2.5072625698324024e-05, + "loss": 0.1999, + "step": 7484 + }, + { + "epoch": 20.907821229050278, + "grad_norm": 1.4467216730117798, + "learning_rate": 2.5075977653631285e-05, + "loss": 0.2649, + "step": 7485 + }, + { + "epoch": 20.910614525139664, + "grad_norm": 0.7218950390815735, + "learning_rate": 2.507932960893855e-05, + "loss": 0.2148, + "step": 7486 + }, + { + "epoch": 20.91340782122905, + "grad_norm": 0.54013991355896, + "learning_rate": 2.508268156424581e-05, + "loss": 0.1567, + "step": 7487 + }, + { + "epoch": 20.916201117318437, + "grad_norm": 0.5888855457305908, + "learning_rate": 2.5086033519553073e-05, + "loss": 0.1898, + "step": 7488 + }, + { + "epoch": 20.91899441340782, + "grad_norm": 0.6822313070297241, + "learning_rate": 2.5089385474860338e-05, + "loss": 0.1881, + "step": 7489 + }, + { + "epoch": 20.921787709497206, + "grad_norm": 1.0876966714859009, + "learning_rate": 2.50927374301676e-05, + "loss": 0.1714, + "step": 7490 + }, + { + "epoch": 20.924581005586592, + "grad_norm": 0.6288774609565735, + "learning_rate": 2.509608938547486e-05, + "loss": 0.1816, + "step": 7491 + }, + { + "epoch": 20.92737430167598, + "grad_norm": 0.5862246155738831, + "learning_rate": 2.5099441340782125e-05, + "loss": 0.1596, + "step": 7492 + }, + { + "epoch": 20.93016759776536, + "grad_norm": 0.9456610679626465, + "learning_rate": 2.5102793296089386e-05, + "loss": 0.1644, + "step": 7493 + }, + { + "epoch": 20.932960893854748, + "grad_norm": 0.6046572327613831, + "learning_rate": 2.5106145251396648e-05, + "loss": 0.1516, + "step": 7494 + }, + { + "epoch": 20.935754189944134, + "grad_norm": 0.7294389605522156, + "learning_rate": 2.5109497206703912e-05, + "loss": 0.18, + "step": 7495 + }, + { + "epoch": 20.93854748603352, + "grad_norm": 0.46669211983680725, + "learning_rate": 2.5112849162011174e-05, + "loss": 0.1693, + "step": 7496 + }, + { + "epoch": 20.941340782122904, + "grad_norm": 0.4991873800754547, + "learning_rate": 2.511620111731844e-05, + "loss": 0.1723, + "step": 7497 + }, + { + "epoch": 20.94413407821229, + "grad_norm": 0.5228559970855713, + "learning_rate": 2.51195530726257e-05, + "loss": 0.1405, + "step": 7498 + }, + { + "epoch": 20.946927374301676, + "grad_norm": 0.5702192783355713, + "learning_rate": 2.512290502793296e-05, + "loss": 0.1644, + "step": 7499 + }, + { + "epoch": 20.949720670391063, + "grad_norm": 0.7964890599250793, + "learning_rate": 2.5126256983240226e-05, + "loss": 0.1857, + "step": 7500 + }, + { + "epoch": 20.952513966480446, + "grad_norm": 0.7737938165664673, + "learning_rate": 2.5129608938547487e-05, + "loss": 0.1712, + "step": 7501 + }, + { + "epoch": 20.955307262569832, + "grad_norm": 1.1989119052886963, + "learning_rate": 2.513296089385475e-05, + "loss": 0.1479, + "step": 7502 + }, + { + "epoch": 20.95810055865922, + "grad_norm": 0.7314482927322388, + "learning_rate": 2.5136312849162013e-05, + "loss": 0.193, + "step": 7503 + }, + { + "epoch": 20.960893854748605, + "grad_norm": 1.086031198501587, + "learning_rate": 2.5139664804469275e-05, + "loss": 0.1696, + "step": 7504 + }, + { + "epoch": 20.963687150837988, + "grad_norm": 0.6956580877304077, + "learning_rate": 2.5143016759776536e-05, + "loss": 0.1386, + "step": 7505 + }, + { + "epoch": 20.966480446927374, + "grad_norm": 0.845199465751648, + "learning_rate": 2.51463687150838e-05, + "loss": 0.1974, + "step": 7506 + }, + { + "epoch": 20.96927374301676, + "grad_norm": 1.2132642269134521, + "learning_rate": 2.5149720670391062e-05, + "loss": 0.1541, + "step": 7507 + }, + { + "epoch": 20.972067039106147, + "grad_norm": 2.1576855182647705, + "learning_rate": 2.5153072625698324e-05, + "loss": 0.2105, + "step": 7508 + }, + { + "epoch": 20.97486033519553, + "grad_norm": 1.3216408491134644, + "learning_rate": 2.5156424581005588e-05, + "loss": 0.1856, + "step": 7509 + }, + { + "epoch": 20.977653631284916, + "grad_norm": 2.3486642837524414, + "learning_rate": 2.515977653631285e-05, + "loss": 0.2236, + "step": 7510 + }, + { + "epoch": 20.980446927374302, + "grad_norm": 0.6733744144439697, + "learning_rate": 2.516312849162011e-05, + "loss": 0.1838, + "step": 7511 + }, + { + "epoch": 20.98324022346369, + "grad_norm": 0.575633704662323, + "learning_rate": 2.5166480446927376e-05, + "loss": 0.1705, + "step": 7512 + }, + { + "epoch": 20.98603351955307, + "grad_norm": 0.7688231468200684, + "learning_rate": 2.5169832402234637e-05, + "loss": 0.1912, + "step": 7513 + }, + { + "epoch": 20.988826815642458, + "grad_norm": 1.2432513236999512, + "learning_rate": 2.51731843575419e-05, + "loss": 0.176, + "step": 7514 + }, + { + "epoch": 20.991620111731844, + "grad_norm": 0.9133241772651672, + "learning_rate": 2.5176536312849163e-05, + "loss": 0.1481, + "step": 7515 + }, + { + "epoch": 20.994413407821227, + "grad_norm": 1.2173962593078613, + "learning_rate": 2.5179888268156424e-05, + "loss": 0.1551, + "step": 7516 + }, + { + "epoch": 20.997206703910614, + "grad_norm": 0.883695662021637, + "learning_rate": 2.518324022346369e-05, + "loss": 0.1914, + "step": 7517 + }, + { + "epoch": 21.0, + "grad_norm": 1.1730304956436157, + "learning_rate": 2.518659217877095e-05, + "loss": 0.1835, + "step": 7518 + }, + { + "epoch": 21.002793296089386, + "grad_norm": 0.5812371969223022, + "learning_rate": 2.5189944134078212e-05, + "loss": 0.1766, + "step": 7519 + }, + { + "epoch": 21.00558659217877, + "grad_norm": 0.5017976760864258, + "learning_rate": 2.5193296089385477e-05, + "loss": 0.1737, + "step": 7520 + }, + { + "epoch": 21.008379888268156, + "grad_norm": 0.44714978337287903, + "learning_rate": 2.5196648044692738e-05, + "loss": 0.1728, + "step": 7521 + }, + { + "epoch": 21.011173184357542, + "grad_norm": 0.5826650857925415, + "learning_rate": 2.52e-05, + "loss": 0.1555, + "step": 7522 + }, + { + "epoch": 21.01396648044693, + "grad_norm": 1.047879934310913, + "learning_rate": 2.5203351955307264e-05, + "loss": 0.192, + "step": 7523 + }, + { + "epoch": 21.01675977653631, + "grad_norm": 0.5309848785400391, + "learning_rate": 2.5206703910614525e-05, + "loss": 0.1466, + "step": 7524 + }, + { + "epoch": 21.019553072625698, + "grad_norm": 0.540006697177887, + "learning_rate": 2.5210055865921787e-05, + "loss": 0.1293, + "step": 7525 + }, + { + "epoch": 21.022346368715084, + "grad_norm": 0.6683850884437561, + "learning_rate": 2.521340782122905e-05, + "loss": 0.1594, + "step": 7526 + }, + { + "epoch": 21.02513966480447, + "grad_norm": 0.4131088852882385, + "learning_rate": 2.5216759776536313e-05, + "loss": 0.1665, + "step": 7527 + }, + { + "epoch": 21.027932960893853, + "grad_norm": 0.6647536754608154, + "learning_rate": 2.5220111731843574e-05, + "loss": 0.1941, + "step": 7528 + }, + { + "epoch": 21.03072625698324, + "grad_norm": 0.606874942779541, + "learning_rate": 2.522346368715084e-05, + "loss": 0.169, + "step": 7529 + }, + { + "epoch": 21.033519553072626, + "grad_norm": 0.5157326459884644, + "learning_rate": 2.52268156424581e-05, + "loss": 0.1464, + "step": 7530 + }, + { + "epoch": 21.036312849162012, + "grad_norm": 0.5149962306022644, + "learning_rate": 2.523016759776536e-05, + "loss": 0.1448, + "step": 7531 + }, + { + "epoch": 21.039106145251395, + "grad_norm": 0.8205741047859192, + "learning_rate": 2.5233519553072626e-05, + "loss": 0.1514, + "step": 7532 + }, + { + "epoch": 21.04189944134078, + "grad_norm": 0.599885880947113, + "learning_rate": 2.5236871508379888e-05, + "loss": 0.151, + "step": 7533 + }, + { + "epoch": 21.044692737430168, + "grad_norm": 1.1086379289627075, + "learning_rate": 2.524022346368715e-05, + "loss": 0.1643, + "step": 7534 + }, + { + "epoch": 21.047486033519554, + "grad_norm": 0.8281517624855042, + "learning_rate": 2.5243575418994414e-05, + "loss": 0.1341, + "step": 7535 + }, + { + "epoch": 21.050279329608937, + "grad_norm": 0.730055034160614, + "learning_rate": 2.5246927374301675e-05, + "loss": 0.1756, + "step": 7536 + }, + { + "epoch": 21.053072625698324, + "grad_norm": 2.2964885234832764, + "learning_rate": 2.525027932960894e-05, + "loss": 0.1461, + "step": 7537 + }, + { + "epoch": 21.05586592178771, + "grad_norm": 0.7904701828956604, + "learning_rate": 2.52536312849162e-05, + "loss": 0.1527, + "step": 7538 + }, + { + "epoch": 21.058659217877096, + "grad_norm": 1.034803867340088, + "learning_rate": 2.5256983240223462e-05, + "loss": 0.1652, + "step": 7539 + }, + { + "epoch": 21.06145251396648, + "grad_norm": 1.0651918649673462, + "learning_rate": 2.5260335195530727e-05, + "loss": 0.1713, + "step": 7540 + }, + { + "epoch": 21.064245810055866, + "grad_norm": 1.3776992559432983, + "learning_rate": 2.526368715083799e-05, + "loss": 0.1937, + "step": 7541 + }, + { + "epoch": 21.067039106145252, + "grad_norm": 0.8773573637008667, + "learning_rate": 2.526703910614525e-05, + "loss": 0.1625, + "step": 7542 + }, + { + "epoch": 21.06983240223464, + "grad_norm": 1.2533490657806396, + "learning_rate": 2.5270391061452515e-05, + "loss": 0.2167, + "step": 7543 + }, + { + "epoch": 21.07262569832402, + "grad_norm": 0.424712210893631, + "learning_rate": 2.5273743016759776e-05, + "loss": 0.178, + "step": 7544 + }, + { + "epoch": 21.075418994413408, + "grad_norm": 0.5351057648658752, + "learning_rate": 2.5277094972067037e-05, + "loss": 0.156, + "step": 7545 + }, + { + "epoch": 21.078212290502794, + "grad_norm": 0.604992687702179, + "learning_rate": 2.5280446927374302e-05, + "loss": 0.166, + "step": 7546 + }, + { + "epoch": 21.08100558659218, + "grad_norm": 0.43343833088874817, + "learning_rate": 2.5283798882681563e-05, + "loss": 0.1731, + "step": 7547 + }, + { + "epoch": 21.083798882681563, + "grad_norm": 0.6464236378669739, + "learning_rate": 2.5287150837988825e-05, + "loss": 0.1711, + "step": 7548 + }, + { + "epoch": 21.08659217877095, + "grad_norm": 0.5030801296234131, + "learning_rate": 2.529050279329609e-05, + "loss": 0.1442, + "step": 7549 + }, + { + "epoch": 21.089385474860336, + "grad_norm": 1.392884373664856, + "learning_rate": 2.529385474860335e-05, + "loss": 0.2053, + "step": 7550 + }, + { + "epoch": 21.092178770949722, + "grad_norm": 0.45898640155792236, + "learning_rate": 2.5297206703910612e-05, + "loss": 0.1523, + "step": 7551 + }, + { + "epoch": 21.094972067039105, + "grad_norm": 0.8584231734275818, + "learning_rate": 2.5300558659217877e-05, + "loss": 0.1454, + "step": 7552 + }, + { + "epoch": 21.09776536312849, + "grad_norm": 0.8958854675292969, + "learning_rate": 2.5303910614525138e-05, + "loss": 0.1516, + "step": 7553 + }, + { + "epoch": 21.100558659217878, + "grad_norm": 0.9231472015380859, + "learning_rate": 2.5307262569832403e-05, + "loss": 0.1694, + "step": 7554 + }, + { + "epoch": 21.10335195530726, + "grad_norm": 0.8347711563110352, + "learning_rate": 2.5310614525139664e-05, + "loss": 0.1596, + "step": 7555 + }, + { + "epoch": 21.106145251396647, + "grad_norm": 0.5873692035675049, + "learning_rate": 2.5313966480446926e-05, + "loss": 0.1441, + "step": 7556 + }, + { + "epoch": 21.108938547486034, + "grad_norm": 0.5462577939033508, + "learning_rate": 2.531731843575419e-05, + "loss": 0.1585, + "step": 7557 + }, + { + "epoch": 21.11173184357542, + "grad_norm": 1.265139102935791, + "learning_rate": 2.5320670391061452e-05, + "loss": 0.154, + "step": 7558 + }, + { + "epoch": 21.114525139664803, + "grad_norm": 1.123019814491272, + "learning_rate": 2.5324022346368716e-05, + "loss": 0.1552, + "step": 7559 + }, + { + "epoch": 21.11731843575419, + "grad_norm": 1.0132043361663818, + "learning_rate": 2.532737430167598e-05, + "loss": 0.1628, + "step": 7560 + }, + { + "epoch": 21.120111731843576, + "grad_norm": 1.3188437223434448, + "learning_rate": 2.5330726256983243e-05, + "loss": 0.157, + "step": 7561 + }, + { + "epoch": 21.122905027932962, + "grad_norm": 0.9396587610244751, + "learning_rate": 2.5334078212290504e-05, + "loss": 0.1434, + "step": 7562 + }, + { + "epoch": 21.125698324022345, + "grad_norm": 0.8374066352844238, + "learning_rate": 2.533743016759777e-05, + "loss": 0.1772, + "step": 7563 + }, + { + "epoch": 21.12849162011173, + "grad_norm": 1.4041168689727783, + "learning_rate": 2.534078212290503e-05, + "loss": 0.1992, + "step": 7564 + }, + { + "epoch": 21.131284916201118, + "grad_norm": 1.518479585647583, + "learning_rate": 2.534413407821229e-05, + "loss": 0.1568, + "step": 7565 + }, + { + "epoch": 21.134078212290504, + "grad_norm": 0.8266905546188354, + "learning_rate": 2.5347486033519556e-05, + "loss": 0.161, + "step": 7566 + }, + { + "epoch": 21.136871508379887, + "grad_norm": 1.708020806312561, + "learning_rate": 2.5350837988826817e-05, + "loss": 0.1604, + "step": 7567 + }, + { + "epoch": 21.139664804469273, + "grad_norm": 1.7107677459716797, + "learning_rate": 2.535418994413408e-05, + "loss": 0.2315, + "step": 7568 + }, + { + "epoch": 21.14245810055866, + "grad_norm": 0.9428108334541321, + "learning_rate": 2.5357541899441343e-05, + "loss": 0.1902, + "step": 7569 + }, + { + "epoch": 21.145251396648046, + "grad_norm": 0.9141278862953186, + "learning_rate": 2.5360893854748605e-05, + "loss": 0.1733, + "step": 7570 + }, + { + "epoch": 21.14804469273743, + "grad_norm": 0.7228692770004272, + "learning_rate": 2.5364245810055866e-05, + "loss": 0.1647, + "step": 7571 + }, + { + "epoch": 21.150837988826815, + "grad_norm": 0.7447243332862854, + "learning_rate": 2.536759776536313e-05, + "loss": 0.19, + "step": 7572 + }, + { + "epoch": 21.1536312849162, + "grad_norm": 0.7316471338272095, + "learning_rate": 2.5370949720670392e-05, + "loss": 0.1618, + "step": 7573 + }, + { + "epoch": 21.156424581005588, + "grad_norm": 0.7060441374778748, + "learning_rate": 2.5374301675977657e-05, + "loss": 0.1736, + "step": 7574 + }, + { + "epoch": 21.15921787709497, + "grad_norm": 0.848702073097229, + "learning_rate": 2.537765363128492e-05, + "loss": 0.1439, + "step": 7575 + }, + { + "epoch": 21.162011173184357, + "grad_norm": 0.8861293196678162, + "learning_rate": 2.538100558659218e-05, + "loss": 0.1599, + "step": 7576 + }, + { + "epoch": 21.164804469273744, + "grad_norm": 0.9120028614997864, + "learning_rate": 2.5384357541899444e-05, + "loss": 0.1582, + "step": 7577 + }, + { + "epoch": 21.16759776536313, + "grad_norm": 2.583442449569702, + "learning_rate": 2.5387709497206706e-05, + "loss": 0.1481, + "step": 7578 + }, + { + "epoch": 21.170391061452513, + "grad_norm": 0.5948621034622192, + "learning_rate": 2.5391061452513967e-05, + "loss": 0.1646, + "step": 7579 + }, + { + "epoch": 21.1731843575419, + "grad_norm": 0.7683195471763611, + "learning_rate": 2.5394413407821232e-05, + "loss": 0.1458, + "step": 7580 + }, + { + "epoch": 21.175977653631286, + "grad_norm": 1.1228301525115967, + "learning_rate": 2.5397765363128493e-05, + "loss": 0.1821, + "step": 7581 + }, + { + "epoch": 21.178770949720672, + "grad_norm": 0.8567558526992798, + "learning_rate": 2.5401117318435755e-05, + "loss": 0.1772, + "step": 7582 + }, + { + "epoch": 21.181564245810055, + "grad_norm": 0.57715904712677, + "learning_rate": 2.540446927374302e-05, + "loss": 0.1537, + "step": 7583 + }, + { + "epoch": 21.18435754189944, + "grad_norm": 1.8277618885040283, + "learning_rate": 2.540782122905028e-05, + "loss": 0.1728, + "step": 7584 + }, + { + "epoch": 21.187150837988828, + "grad_norm": 1.1730068922042847, + "learning_rate": 2.5411173184357542e-05, + "loss": 0.1489, + "step": 7585 + }, + { + "epoch": 21.189944134078214, + "grad_norm": 1.4657429456710815, + "learning_rate": 2.5414525139664807e-05, + "loss": 0.179, + "step": 7586 + }, + { + "epoch": 21.192737430167597, + "grad_norm": 0.6303529739379883, + "learning_rate": 2.5417877094972068e-05, + "loss": 0.1538, + "step": 7587 + }, + { + "epoch": 21.195530726256983, + "grad_norm": 2.531046152114868, + "learning_rate": 2.542122905027933e-05, + "loss": 0.1376, + "step": 7588 + }, + { + "epoch": 21.19832402234637, + "grad_norm": 1.1073977947235107, + "learning_rate": 2.5424581005586594e-05, + "loss": 0.155, + "step": 7589 + }, + { + "epoch": 21.201117318435756, + "grad_norm": 0.8335393667221069, + "learning_rate": 2.5427932960893855e-05, + "loss": 0.157, + "step": 7590 + }, + { + "epoch": 21.20391061452514, + "grad_norm": 1.383588194847107, + "learning_rate": 2.543128491620112e-05, + "loss": 0.1962, + "step": 7591 + }, + { + "epoch": 21.206703910614525, + "grad_norm": 2.658134937286377, + "learning_rate": 2.543463687150838e-05, + "loss": 0.1538, + "step": 7592 + }, + { + "epoch": 21.20949720670391, + "grad_norm": 2.661034107208252, + "learning_rate": 2.5437988826815643e-05, + "loss": 0.2151, + "step": 7593 + }, + { + "epoch": 21.212290502793294, + "grad_norm": 1.54359769821167, + "learning_rate": 2.5441340782122908e-05, + "loss": 0.1709, + "step": 7594 + }, + { + "epoch": 21.21508379888268, + "grad_norm": 0.7841529250144958, + "learning_rate": 2.544469273743017e-05, + "loss": 0.1674, + "step": 7595 + }, + { + "epoch": 21.217877094972067, + "grad_norm": 0.5435138940811157, + "learning_rate": 2.544804469273743e-05, + "loss": 0.1588, + "step": 7596 + }, + { + "epoch": 21.220670391061454, + "grad_norm": 1.2725781202316284, + "learning_rate": 2.5451396648044695e-05, + "loss": 0.1696, + "step": 7597 + }, + { + "epoch": 21.223463687150836, + "grad_norm": 0.7395046353340149, + "learning_rate": 2.5454748603351956e-05, + "loss": 0.1896, + "step": 7598 + }, + { + "epoch": 21.226256983240223, + "grad_norm": 0.5810819268226624, + "learning_rate": 2.5458100558659218e-05, + "loss": 0.1597, + "step": 7599 + }, + { + "epoch": 21.22905027932961, + "grad_norm": 2.623837947845459, + "learning_rate": 2.5461452513966482e-05, + "loss": 0.1631, + "step": 7600 + }, + { + "epoch": 21.231843575418996, + "grad_norm": 0.6425523161888123, + "learning_rate": 2.5464804469273744e-05, + "loss": 0.1536, + "step": 7601 + }, + { + "epoch": 21.23463687150838, + "grad_norm": 0.7084157466888428, + "learning_rate": 2.5468156424581005e-05, + "loss": 0.1406, + "step": 7602 + }, + { + "epoch": 21.237430167597765, + "grad_norm": 0.8144797086715698, + "learning_rate": 2.547150837988827e-05, + "loss": 0.1638, + "step": 7603 + }, + { + "epoch": 21.24022346368715, + "grad_norm": 0.7236500978469849, + "learning_rate": 2.547486033519553e-05, + "loss": 0.1503, + "step": 7604 + }, + { + "epoch": 21.243016759776538, + "grad_norm": 1.1716355085372925, + "learning_rate": 2.5478212290502793e-05, + "loss": 0.1858, + "step": 7605 + }, + { + "epoch": 21.24581005586592, + "grad_norm": 0.6915284395217896, + "learning_rate": 2.5481564245810057e-05, + "loss": 0.1429, + "step": 7606 + }, + { + "epoch": 21.248603351955307, + "grad_norm": 1.0403928756713867, + "learning_rate": 2.548491620111732e-05, + "loss": 0.1671, + "step": 7607 + }, + { + "epoch": 21.251396648044693, + "grad_norm": 1.0166468620300293, + "learning_rate": 2.548826815642458e-05, + "loss": 0.1576, + "step": 7608 + }, + { + "epoch": 21.25418994413408, + "grad_norm": 0.7484316825866699, + "learning_rate": 2.5491620111731845e-05, + "loss": 0.1516, + "step": 7609 + }, + { + "epoch": 21.256983240223462, + "grad_norm": 0.9118483662605286, + "learning_rate": 2.5494972067039106e-05, + "loss": 0.1877, + "step": 7610 + }, + { + "epoch": 21.25977653631285, + "grad_norm": 1.7341879606246948, + "learning_rate": 2.549832402234637e-05, + "loss": 0.1738, + "step": 7611 + }, + { + "epoch": 21.262569832402235, + "grad_norm": 1.8442147970199585, + "learning_rate": 2.5501675977653632e-05, + "loss": 0.1546, + "step": 7612 + }, + { + "epoch": 21.26536312849162, + "grad_norm": 1.1944420337677002, + "learning_rate": 2.5505027932960894e-05, + "loss": 0.1583, + "step": 7613 + }, + { + "epoch": 21.268156424581004, + "grad_norm": 1.1833322048187256, + "learning_rate": 2.5508379888268158e-05, + "loss": 0.1645, + "step": 7614 + }, + { + "epoch": 21.27094972067039, + "grad_norm": 3.150165319442749, + "learning_rate": 2.551173184357542e-05, + "loss": 0.1881, + "step": 7615 + }, + { + "epoch": 21.273743016759777, + "grad_norm": 0.8169300556182861, + "learning_rate": 2.551508379888268e-05, + "loss": 0.1607, + "step": 7616 + }, + { + "epoch": 21.276536312849164, + "grad_norm": 0.9336345791816711, + "learning_rate": 2.5518435754189946e-05, + "loss": 0.1885, + "step": 7617 + }, + { + "epoch": 21.279329608938546, + "grad_norm": 1.579567790031433, + "learning_rate": 2.5521787709497207e-05, + "loss": 0.2386, + "step": 7618 + }, + { + "epoch": 21.282122905027933, + "grad_norm": 0.4437618553638458, + "learning_rate": 2.552513966480447e-05, + "loss": 0.1643, + "step": 7619 + }, + { + "epoch": 21.28491620111732, + "grad_norm": 0.7359052896499634, + "learning_rate": 2.5528491620111733e-05, + "loss": 0.1518, + "step": 7620 + }, + { + "epoch": 21.287709497206706, + "grad_norm": 0.8630828857421875, + "learning_rate": 2.5531843575418994e-05, + "loss": 0.2023, + "step": 7621 + }, + { + "epoch": 21.29050279329609, + "grad_norm": 0.5002668499946594, + "learning_rate": 2.5535195530726256e-05, + "loss": 0.1935, + "step": 7622 + }, + { + "epoch": 21.293296089385475, + "grad_norm": 0.4236806333065033, + "learning_rate": 2.553854748603352e-05, + "loss": 0.1677, + "step": 7623 + }, + { + "epoch": 21.29608938547486, + "grad_norm": 0.5983018279075623, + "learning_rate": 2.5541899441340782e-05, + "loss": 0.1671, + "step": 7624 + }, + { + "epoch": 21.298882681564244, + "grad_norm": 2.48429536819458, + "learning_rate": 2.5545251396648043e-05, + "loss": 0.1499, + "step": 7625 + }, + { + "epoch": 21.30167597765363, + "grad_norm": 0.5584551095962524, + "learning_rate": 2.5548603351955308e-05, + "loss": 0.1736, + "step": 7626 + }, + { + "epoch": 21.304469273743017, + "grad_norm": 0.6951791048049927, + "learning_rate": 2.555195530726257e-05, + "loss": 0.1709, + "step": 7627 + }, + { + "epoch": 21.307262569832403, + "grad_norm": 0.5816057920455933, + "learning_rate": 2.555530726256983e-05, + "loss": 0.1677, + "step": 7628 + }, + { + "epoch": 21.310055865921786, + "grad_norm": 0.6737315654754639, + "learning_rate": 2.5558659217877095e-05, + "loss": 0.1899, + "step": 7629 + }, + { + "epoch": 21.312849162011172, + "grad_norm": 0.7342494130134583, + "learning_rate": 2.5562011173184357e-05, + "loss": 0.1505, + "step": 7630 + }, + { + "epoch": 21.31564245810056, + "grad_norm": 0.8137749433517456, + "learning_rate": 2.556536312849162e-05, + "loss": 0.1407, + "step": 7631 + }, + { + "epoch": 21.318435754189945, + "grad_norm": 1.2587740421295166, + "learning_rate": 2.5568715083798883e-05, + "loss": 0.1674, + "step": 7632 + }, + { + "epoch": 21.321229050279328, + "grad_norm": 0.5620272755622864, + "learning_rate": 2.5572067039106144e-05, + "loss": 0.1474, + "step": 7633 + }, + { + "epoch": 21.324022346368714, + "grad_norm": 0.7482175230979919, + "learning_rate": 2.557541899441341e-05, + "loss": 0.1429, + "step": 7634 + }, + { + "epoch": 21.3268156424581, + "grad_norm": 0.7335797548294067, + "learning_rate": 2.557877094972067e-05, + "loss": 0.1806, + "step": 7635 + }, + { + "epoch": 21.329608938547487, + "grad_norm": 0.9186086058616638, + "learning_rate": 2.558212290502793e-05, + "loss": 0.1863, + "step": 7636 + }, + { + "epoch": 21.33240223463687, + "grad_norm": 1.3152698278427124, + "learning_rate": 2.5585474860335196e-05, + "loss": 0.1578, + "step": 7637 + }, + { + "epoch": 21.335195530726256, + "grad_norm": 0.769132673740387, + "learning_rate": 2.5588826815642458e-05, + "loss": 0.1764, + "step": 7638 + }, + { + "epoch": 21.337988826815643, + "grad_norm": 0.686496376991272, + "learning_rate": 2.559217877094972e-05, + "loss": 0.1395, + "step": 7639 + }, + { + "epoch": 21.34078212290503, + "grad_norm": 1.0953322649002075, + "learning_rate": 2.5595530726256984e-05, + "loss": 0.1428, + "step": 7640 + }, + { + "epoch": 21.343575418994412, + "grad_norm": 0.967597246170044, + "learning_rate": 2.5598882681564245e-05, + "loss": 0.1525, + "step": 7641 + }, + { + "epoch": 21.3463687150838, + "grad_norm": 3.8811261653900146, + "learning_rate": 2.5602234636871506e-05, + "loss": 0.1822, + "step": 7642 + }, + { + "epoch": 21.349162011173185, + "grad_norm": 1.9658013582229614, + "learning_rate": 2.560558659217877e-05, + "loss": 0.2425, + "step": 7643 + }, + { + "epoch": 21.35195530726257, + "grad_norm": 0.42978742718696594, + "learning_rate": 2.5608938547486032e-05, + "loss": 0.1799, + "step": 7644 + }, + { + "epoch": 21.354748603351954, + "grad_norm": 0.530407190322876, + "learning_rate": 2.5612290502793294e-05, + "loss": 0.1652, + "step": 7645 + }, + { + "epoch": 21.35754189944134, + "grad_norm": 0.5481172204017639, + "learning_rate": 2.561564245810056e-05, + "loss": 0.177, + "step": 7646 + }, + { + "epoch": 21.360335195530727, + "grad_norm": 0.6757632493972778, + "learning_rate": 2.561899441340782e-05, + "loss": 0.1559, + "step": 7647 + }, + { + "epoch": 21.363128491620113, + "grad_norm": 1.097920298576355, + "learning_rate": 2.5622346368715085e-05, + "loss": 0.2034, + "step": 7648 + }, + { + "epoch": 21.365921787709496, + "grad_norm": 0.4903882145881653, + "learning_rate": 2.5625698324022346e-05, + "loss": 0.1835, + "step": 7649 + }, + { + "epoch": 21.368715083798882, + "grad_norm": 0.9423296451568604, + "learning_rate": 2.5629050279329607e-05, + "loss": 0.1864, + "step": 7650 + }, + { + "epoch": 21.37150837988827, + "grad_norm": 0.5609569549560547, + "learning_rate": 2.5632402234636872e-05, + "loss": 0.1613, + "step": 7651 + }, + { + "epoch": 21.374301675977655, + "grad_norm": 0.7150386571884155, + "learning_rate": 2.5635754189944133e-05, + "loss": 0.1666, + "step": 7652 + }, + { + "epoch": 21.377094972067038, + "grad_norm": 0.5193126797676086, + "learning_rate": 2.5639106145251395e-05, + "loss": 0.1616, + "step": 7653 + }, + { + "epoch": 21.379888268156424, + "grad_norm": 0.6491525173187256, + "learning_rate": 2.564245810055866e-05, + "loss": 0.1598, + "step": 7654 + }, + { + "epoch": 21.38268156424581, + "grad_norm": 1.1652121543884277, + "learning_rate": 2.5645810055865924e-05, + "loss": 0.1774, + "step": 7655 + }, + { + "epoch": 21.385474860335197, + "grad_norm": 0.4637206494808197, + "learning_rate": 2.5649162011173186e-05, + "loss": 0.1498, + "step": 7656 + }, + { + "epoch": 21.38826815642458, + "grad_norm": 1.2919756174087524, + "learning_rate": 2.565251396648045e-05, + "loss": 0.1751, + "step": 7657 + }, + { + "epoch": 21.391061452513966, + "grad_norm": 0.9410260319709778, + "learning_rate": 2.565586592178771e-05, + "loss": 0.1676, + "step": 7658 + }, + { + "epoch": 21.393854748603353, + "grad_norm": 0.8268300294876099, + "learning_rate": 2.5659217877094973e-05, + "loss": 0.1552, + "step": 7659 + }, + { + "epoch": 21.39664804469274, + "grad_norm": 0.9050667881965637, + "learning_rate": 2.5662569832402238e-05, + "loss": 0.1726, + "step": 7660 + }, + { + "epoch": 21.399441340782122, + "grad_norm": 1.1003172397613525, + "learning_rate": 2.56659217877095e-05, + "loss": 0.1737, + "step": 7661 + }, + { + "epoch": 21.40223463687151, + "grad_norm": 0.847176194190979, + "learning_rate": 2.566927374301676e-05, + "loss": 0.16, + "step": 7662 + }, + { + "epoch": 21.405027932960895, + "grad_norm": 2.092766284942627, + "learning_rate": 2.5672625698324025e-05, + "loss": 0.1625, + "step": 7663 + }, + { + "epoch": 21.407821229050278, + "grad_norm": 1.5360548496246338, + "learning_rate": 2.5675977653631286e-05, + "loss": 0.2152, + "step": 7664 + }, + { + "epoch": 21.410614525139664, + "grad_norm": 0.7297159433364868, + "learning_rate": 2.5679329608938548e-05, + "loss": 0.187, + "step": 7665 + }, + { + "epoch": 21.41340782122905, + "grad_norm": 1.015587568283081, + "learning_rate": 2.5682681564245813e-05, + "loss": 0.1697, + "step": 7666 + }, + { + "epoch": 21.416201117318437, + "grad_norm": 1.177230954170227, + "learning_rate": 2.5686033519553074e-05, + "loss": 0.1665, + "step": 7667 + }, + { + "epoch": 21.41899441340782, + "grad_norm": 1.6979113817214966, + "learning_rate": 2.568938547486034e-05, + "loss": 0.2335, + "step": 7668 + }, + { + "epoch": 21.421787709497206, + "grad_norm": 0.6041009426116943, + "learning_rate": 2.56927374301676e-05, + "loss": 0.1678, + "step": 7669 + }, + { + "epoch": 21.424581005586592, + "grad_norm": 0.6326160430908203, + "learning_rate": 2.569608938547486e-05, + "loss": 0.1872, + "step": 7670 + }, + { + "epoch": 21.42737430167598, + "grad_norm": 1.0541030168533325, + "learning_rate": 2.5699441340782126e-05, + "loss": 0.1639, + "step": 7671 + }, + { + "epoch": 21.43016759776536, + "grad_norm": 0.5231131315231323, + "learning_rate": 2.5702793296089387e-05, + "loss": 0.1482, + "step": 7672 + }, + { + "epoch": 21.432960893854748, + "grad_norm": 0.8797820806503296, + "learning_rate": 2.570614525139665e-05, + "loss": 0.1922, + "step": 7673 + }, + { + "epoch": 21.435754189944134, + "grad_norm": 0.8835678100585938, + "learning_rate": 2.5709497206703913e-05, + "loss": 0.1928, + "step": 7674 + }, + { + "epoch": 21.43854748603352, + "grad_norm": 1.3719569444656372, + "learning_rate": 2.5712849162011175e-05, + "loss": 0.1659, + "step": 7675 + }, + { + "epoch": 21.441340782122904, + "grad_norm": 0.8267677426338196, + "learning_rate": 2.5716201117318436e-05, + "loss": 0.1756, + "step": 7676 + }, + { + "epoch": 21.44413407821229, + "grad_norm": 0.6564570665359497, + "learning_rate": 2.57195530726257e-05, + "loss": 0.1641, + "step": 7677 + }, + { + "epoch": 21.446927374301676, + "grad_norm": 0.5404424667358398, + "learning_rate": 2.5722905027932962e-05, + "loss": 0.1833, + "step": 7678 + }, + { + "epoch": 21.449720670391063, + "grad_norm": 0.7168551087379456, + "learning_rate": 2.5726256983240224e-05, + "loss": 0.1769, + "step": 7679 + }, + { + "epoch": 21.452513966480446, + "grad_norm": 0.5268176794052124, + "learning_rate": 2.572960893854749e-05, + "loss": 0.1672, + "step": 7680 + }, + { + "epoch": 21.455307262569832, + "grad_norm": 0.4257758855819702, + "learning_rate": 2.573296089385475e-05, + "loss": 0.1484, + "step": 7681 + }, + { + "epoch": 21.45810055865922, + "grad_norm": 1.432535171508789, + "learning_rate": 2.573631284916201e-05, + "loss": 0.1746, + "step": 7682 + }, + { + "epoch": 21.460893854748605, + "grad_norm": 0.7110638618469238, + "learning_rate": 2.5739664804469276e-05, + "loss": 0.1705, + "step": 7683 + }, + { + "epoch": 21.463687150837988, + "grad_norm": 0.7348331809043884, + "learning_rate": 2.5743016759776537e-05, + "loss": 0.1697, + "step": 7684 + }, + { + "epoch": 21.466480446927374, + "grad_norm": 1.6159015893936157, + "learning_rate": 2.5746368715083802e-05, + "loss": 0.1802, + "step": 7685 + }, + { + "epoch": 21.46927374301676, + "grad_norm": 1.111865758895874, + "learning_rate": 2.5749720670391063e-05, + "loss": 0.1296, + "step": 7686 + }, + { + "epoch": 21.472067039106147, + "grad_norm": 1.2675189971923828, + "learning_rate": 2.5753072625698325e-05, + "loss": 0.1852, + "step": 7687 + }, + { + "epoch": 21.47486033519553, + "grad_norm": 1.3448681831359863, + "learning_rate": 2.575642458100559e-05, + "loss": 0.1675, + "step": 7688 + }, + { + "epoch": 21.477653631284916, + "grad_norm": 0.8250183463096619, + "learning_rate": 2.575977653631285e-05, + "loss": 0.1754, + "step": 7689 + }, + { + "epoch": 21.480446927374302, + "grad_norm": 2.0905961990356445, + "learning_rate": 2.5763128491620112e-05, + "loss": 0.1636, + "step": 7690 + }, + { + "epoch": 21.48324022346369, + "grad_norm": 1.2468020915985107, + "learning_rate": 2.5766480446927377e-05, + "loss": 0.1596, + "step": 7691 + }, + { + "epoch": 21.48603351955307, + "grad_norm": 1.324833631515503, + "learning_rate": 2.5769832402234638e-05, + "loss": 0.1861, + "step": 7692 + }, + { + "epoch": 21.488826815642458, + "grad_norm": 1.7186894416809082, + "learning_rate": 2.57731843575419e-05, + "loss": 0.2032, + "step": 7693 + }, + { + "epoch": 21.491620111731844, + "grad_norm": 0.9654924869537354, + "learning_rate": 2.5776536312849164e-05, + "loss": 0.185, + "step": 7694 + }, + { + "epoch": 21.49441340782123, + "grad_norm": 1.1346319913864136, + "learning_rate": 2.5779888268156425e-05, + "loss": 0.1887, + "step": 7695 + }, + { + "epoch": 21.497206703910614, + "grad_norm": 1.6205062866210938, + "learning_rate": 2.5783240223463687e-05, + "loss": 0.1708, + "step": 7696 + }, + { + "epoch": 21.5, + "grad_norm": 0.6169498562812805, + "learning_rate": 2.578659217877095e-05, + "loss": 0.1982, + "step": 7697 + }, + { + "epoch": 21.502793296089386, + "grad_norm": 0.9497862458229065, + "learning_rate": 2.5789944134078213e-05, + "loss": 0.1765, + "step": 7698 + }, + { + "epoch": 21.505586592178773, + "grad_norm": 1.62716543674469, + "learning_rate": 2.5793296089385474e-05, + "loss": 0.2122, + "step": 7699 + }, + { + "epoch": 21.508379888268156, + "grad_norm": 0.7986272573471069, + "learning_rate": 2.579664804469274e-05, + "loss": 0.1957, + "step": 7700 + }, + { + "epoch": 21.511173184357542, + "grad_norm": 0.85270094871521, + "learning_rate": 2.58e-05, + "loss": 0.1657, + "step": 7701 + }, + { + "epoch": 21.51396648044693, + "grad_norm": 0.493850439786911, + "learning_rate": 2.580335195530726e-05, + "loss": 0.1595, + "step": 7702 + }, + { + "epoch": 21.51675977653631, + "grad_norm": 0.7370725274085999, + "learning_rate": 2.5806703910614526e-05, + "loss": 0.1951, + "step": 7703 + }, + { + "epoch": 21.519553072625698, + "grad_norm": 1.0247735977172852, + "learning_rate": 2.5810055865921788e-05, + "loss": 0.1606, + "step": 7704 + }, + { + "epoch": 21.522346368715084, + "grad_norm": 0.5359322428703308, + "learning_rate": 2.5813407821229052e-05, + "loss": 0.1872, + "step": 7705 + }, + { + "epoch": 21.52513966480447, + "grad_norm": 0.7644962072372437, + "learning_rate": 2.5816759776536314e-05, + "loss": 0.1709, + "step": 7706 + }, + { + "epoch": 21.527932960893853, + "grad_norm": 0.6194400191307068, + "learning_rate": 2.5820111731843575e-05, + "loss": 0.1733, + "step": 7707 + }, + { + "epoch": 21.53072625698324, + "grad_norm": 0.43529239296913147, + "learning_rate": 2.582346368715084e-05, + "loss": 0.1278, + "step": 7708 + }, + { + "epoch": 21.533519553072626, + "grad_norm": 0.7558255195617676, + "learning_rate": 2.58268156424581e-05, + "loss": 0.155, + "step": 7709 + }, + { + "epoch": 21.536312849162012, + "grad_norm": 0.8187474012374878, + "learning_rate": 2.5830167597765363e-05, + "loss": 0.1632, + "step": 7710 + }, + { + "epoch": 21.539106145251395, + "grad_norm": 0.8247483968734741, + "learning_rate": 2.5833519553072627e-05, + "loss": 0.175, + "step": 7711 + }, + { + "epoch": 21.54189944134078, + "grad_norm": 0.8103248476982117, + "learning_rate": 2.583687150837989e-05, + "loss": 0.1467, + "step": 7712 + }, + { + "epoch": 21.544692737430168, + "grad_norm": 0.6739637851715088, + "learning_rate": 2.584022346368715e-05, + "loss": 0.1772, + "step": 7713 + }, + { + "epoch": 21.547486033519554, + "grad_norm": 1.9398597478866577, + "learning_rate": 2.5843575418994415e-05, + "loss": 0.1721, + "step": 7714 + }, + { + "epoch": 21.550279329608937, + "grad_norm": 0.9569180607795715, + "learning_rate": 2.5846927374301676e-05, + "loss": 0.1777, + "step": 7715 + }, + { + "epoch": 21.553072625698324, + "grad_norm": 4.2460174560546875, + "learning_rate": 2.5850279329608937e-05, + "loss": 0.1946, + "step": 7716 + }, + { + "epoch": 21.55586592178771, + "grad_norm": 2.2417914867401123, + "learning_rate": 2.5853631284916202e-05, + "loss": 0.18, + "step": 7717 + }, + { + "epoch": 21.558659217877096, + "grad_norm": 1.8113455772399902, + "learning_rate": 2.5856983240223464e-05, + "loss": 0.239, + "step": 7718 + }, + { + "epoch": 21.56145251396648, + "grad_norm": 0.6757938861846924, + "learning_rate": 2.5860335195530725e-05, + "loss": 0.2285, + "step": 7719 + }, + { + "epoch": 21.564245810055866, + "grad_norm": 0.47917285561561584, + "learning_rate": 2.586368715083799e-05, + "loss": 0.1829, + "step": 7720 + }, + { + "epoch": 21.567039106145252, + "grad_norm": 0.5308159589767456, + "learning_rate": 2.586703910614525e-05, + "loss": 0.1662, + "step": 7721 + }, + { + "epoch": 21.56983240223464, + "grad_norm": 0.6150697469711304, + "learning_rate": 2.5870391061452512e-05, + "loss": 0.1867, + "step": 7722 + }, + { + "epoch": 21.57262569832402, + "grad_norm": 0.5404456257820129, + "learning_rate": 2.5873743016759777e-05, + "loss": 0.167, + "step": 7723 + }, + { + "epoch": 21.575418994413408, + "grad_norm": 0.6356098055839539, + "learning_rate": 2.587709497206704e-05, + "loss": 0.1652, + "step": 7724 + }, + { + "epoch": 21.578212290502794, + "grad_norm": 0.5820078253746033, + "learning_rate": 2.5880446927374303e-05, + "loss": 0.177, + "step": 7725 + }, + { + "epoch": 21.58100558659218, + "grad_norm": 0.8829872012138367, + "learning_rate": 2.5883798882681564e-05, + "loss": 0.1893, + "step": 7726 + }, + { + "epoch": 21.583798882681563, + "grad_norm": 0.8503148555755615, + "learning_rate": 2.5887150837988826e-05, + "loss": 0.176, + "step": 7727 + }, + { + "epoch": 21.58659217877095, + "grad_norm": 0.609172523021698, + "learning_rate": 2.589050279329609e-05, + "loss": 0.2086, + "step": 7728 + }, + { + "epoch": 21.589385474860336, + "grad_norm": 0.7411264777183533, + "learning_rate": 2.5893854748603352e-05, + "loss": 0.1815, + "step": 7729 + }, + { + "epoch": 21.592178770949722, + "grad_norm": 0.6599768996238708, + "learning_rate": 2.5897206703910613e-05, + "loss": 0.1756, + "step": 7730 + }, + { + "epoch": 21.594972067039105, + "grad_norm": 0.5626360177993774, + "learning_rate": 2.5900558659217878e-05, + "loss": 0.1672, + "step": 7731 + }, + { + "epoch": 21.59776536312849, + "grad_norm": 0.9015595316886902, + "learning_rate": 2.590391061452514e-05, + "loss": 0.1539, + "step": 7732 + }, + { + "epoch": 21.600558659217878, + "grad_norm": 0.5690361261367798, + "learning_rate": 2.59072625698324e-05, + "loss": 0.1542, + "step": 7733 + }, + { + "epoch": 21.60335195530726, + "grad_norm": 0.6163727045059204, + "learning_rate": 2.5910614525139665e-05, + "loss": 0.1365, + "step": 7734 + }, + { + "epoch": 21.606145251396647, + "grad_norm": 0.6907418370246887, + "learning_rate": 2.5913966480446927e-05, + "loss": 0.1465, + "step": 7735 + }, + { + "epoch": 21.608938547486034, + "grad_norm": 0.9201969504356384, + "learning_rate": 2.5917318435754188e-05, + "loss": 0.1334, + "step": 7736 + }, + { + "epoch": 21.61173184357542, + "grad_norm": 1.1259077787399292, + "learning_rate": 2.5920670391061453e-05, + "loss": 0.1671, + "step": 7737 + }, + { + "epoch": 21.614525139664803, + "grad_norm": 0.9638356566429138, + "learning_rate": 2.5924022346368714e-05, + "loss": 0.1894, + "step": 7738 + }, + { + "epoch": 21.61731843575419, + "grad_norm": 1.1225947141647339, + "learning_rate": 2.5927374301675975e-05, + "loss": 0.1759, + "step": 7739 + }, + { + "epoch": 21.620111731843576, + "grad_norm": 0.8139482140541077, + "learning_rate": 2.593072625698324e-05, + "loss": 0.1578, + "step": 7740 + }, + { + "epoch": 21.622905027932962, + "grad_norm": 0.786419153213501, + "learning_rate": 2.59340782122905e-05, + "loss": 0.1624, + "step": 7741 + }, + { + "epoch": 21.625698324022345, + "grad_norm": 1.069106101989746, + "learning_rate": 2.5937430167597766e-05, + "loss": 0.2035, + "step": 7742 + }, + { + "epoch": 21.62849162011173, + "grad_norm": 1.8549046516418457, + "learning_rate": 2.5940782122905028e-05, + "loss": 0.2262, + "step": 7743 + }, + { + "epoch": 21.631284916201118, + "grad_norm": 0.5082066059112549, + "learning_rate": 2.594413407821229e-05, + "loss": 0.1892, + "step": 7744 + }, + { + "epoch": 21.634078212290504, + "grad_norm": 0.8624891638755798, + "learning_rate": 2.5947486033519554e-05, + "loss": 0.2155, + "step": 7745 + }, + { + "epoch": 21.636871508379887, + "grad_norm": 0.6818066835403442, + "learning_rate": 2.5950837988826815e-05, + "loss": 0.178, + "step": 7746 + }, + { + "epoch": 21.639664804469273, + "grad_norm": 0.3969781696796417, + "learning_rate": 2.5954189944134076e-05, + "loss": 0.1699, + "step": 7747 + }, + { + "epoch": 21.64245810055866, + "grad_norm": 1.2158668041229248, + "learning_rate": 2.595754189944134e-05, + "loss": 0.1644, + "step": 7748 + }, + { + "epoch": 21.645251396648046, + "grad_norm": 0.3955950140953064, + "learning_rate": 2.5960893854748602e-05, + "loss": 0.1608, + "step": 7749 + }, + { + "epoch": 21.64804469273743, + "grad_norm": 0.5899773836135864, + "learning_rate": 2.5964245810055864e-05, + "loss": 0.2006, + "step": 7750 + }, + { + "epoch": 21.650837988826815, + "grad_norm": 0.7492181062698364, + "learning_rate": 2.596759776536313e-05, + "loss": 0.1861, + "step": 7751 + }, + { + "epoch": 21.6536312849162, + "grad_norm": 0.5311159491539001, + "learning_rate": 2.5970949720670393e-05, + "loss": 0.165, + "step": 7752 + }, + { + "epoch": 21.656424581005588, + "grad_norm": 0.4636228084564209, + "learning_rate": 2.5974301675977655e-05, + "loss": 0.171, + "step": 7753 + }, + { + "epoch": 21.65921787709497, + "grad_norm": 0.9943622946739197, + "learning_rate": 2.597765363128492e-05, + "loss": 0.205, + "step": 7754 + }, + { + "epoch": 21.662011173184357, + "grad_norm": 1.7560997009277344, + "learning_rate": 2.598100558659218e-05, + "loss": 0.1463, + "step": 7755 + }, + { + "epoch": 21.664804469273744, + "grad_norm": 0.45058557391166687, + "learning_rate": 2.5984357541899442e-05, + "loss": 0.1445, + "step": 7756 + }, + { + "epoch": 21.66759776536313, + "grad_norm": 0.5629660487174988, + "learning_rate": 2.5987709497206707e-05, + "loss": 0.1642, + "step": 7757 + }, + { + "epoch": 21.670391061452513, + "grad_norm": 0.7865996956825256, + "learning_rate": 2.5991061452513968e-05, + "loss": 0.1535, + "step": 7758 + }, + { + "epoch": 21.6731843575419, + "grad_norm": 1.0691745281219482, + "learning_rate": 2.599441340782123e-05, + "loss": 0.154, + "step": 7759 + }, + { + "epoch": 21.675977653631286, + "grad_norm": 0.7595177888870239, + "learning_rate": 2.5997765363128494e-05, + "loss": 0.1751, + "step": 7760 + }, + { + "epoch": 21.678770949720672, + "grad_norm": 0.9132778644561768, + "learning_rate": 2.6001117318435756e-05, + "loss": 0.1694, + "step": 7761 + }, + { + "epoch": 21.681564245810055, + "grad_norm": 0.9855597019195557, + "learning_rate": 2.600446927374302e-05, + "loss": 0.1969, + "step": 7762 + }, + { + "epoch": 21.68435754189944, + "grad_norm": 0.8296615481376648, + "learning_rate": 2.600782122905028e-05, + "loss": 0.1661, + "step": 7763 + }, + { + "epoch": 21.687150837988828, + "grad_norm": 1.2108677625656128, + "learning_rate": 2.6011173184357543e-05, + "loss": 0.1977, + "step": 7764 + }, + { + "epoch": 21.689944134078214, + "grad_norm": 1.1626962423324585, + "learning_rate": 2.6014525139664808e-05, + "loss": 0.1404, + "step": 7765 + }, + { + "epoch": 21.692737430167597, + "grad_norm": 0.7857217788696289, + "learning_rate": 2.601787709497207e-05, + "loss": 0.1713, + "step": 7766 + }, + { + "epoch": 21.695530726256983, + "grad_norm": 1.138536810874939, + "learning_rate": 2.602122905027933e-05, + "loss": 0.1775, + "step": 7767 + }, + { + "epoch": 21.69832402234637, + "grad_norm": 1.6907974481582642, + "learning_rate": 2.6024581005586595e-05, + "loss": 0.2505, + "step": 7768 + }, + { + "epoch": 21.701117318435756, + "grad_norm": 1.544473648071289, + "learning_rate": 2.6027932960893856e-05, + "loss": 0.1728, + "step": 7769 + }, + { + "epoch": 21.70391061452514, + "grad_norm": 0.9045861959457397, + "learning_rate": 2.6031284916201118e-05, + "loss": 0.1843, + "step": 7770 + }, + { + "epoch": 21.706703910614525, + "grad_norm": 1.3645986318588257, + "learning_rate": 2.6034636871508383e-05, + "loss": 0.1481, + "step": 7771 + }, + { + "epoch": 21.70949720670391, + "grad_norm": 0.6640289425849915, + "learning_rate": 2.6037988826815644e-05, + "loss": 0.1598, + "step": 7772 + }, + { + "epoch": 21.712290502793294, + "grad_norm": 0.6439160108566284, + "learning_rate": 2.6041340782122905e-05, + "loss": 0.184, + "step": 7773 + }, + { + "epoch": 21.71508379888268, + "grad_norm": 0.604657769203186, + "learning_rate": 2.604469273743017e-05, + "loss": 0.146, + "step": 7774 + }, + { + "epoch": 21.717877094972067, + "grad_norm": 0.5711581110954285, + "learning_rate": 2.604804469273743e-05, + "loss": 0.1801, + "step": 7775 + }, + { + "epoch": 21.720670391061454, + "grad_norm": 0.768743634223938, + "learning_rate": 2.6051396648044693e-05, + "loss": 0.1592, + "step": 7776 + }, + { + "epoch": 21.723463687150836, + "grad_norm": 0.7518887519836426, + "learning_rate": 2.6054748603351957e-05, + "loss": 0.1514, + "step": 7777 + }, + { + "epoch": 21.726256983240223, + "grad_norm": 0.6405645608901978, + "learning_rate": 2.605810055865922e-05, + "loss": 0.1418, + "step": 7778 + }, + { + "epoch": 21.72905027932961, + "grad_norm": 0.7745922207832336, + "learning_rate": 2.6061452513966484e-05, + "loss": 0.1919, + "step": 7779 + }, + { + "epoch": 21.731843575418996, + "grad_norm": 0.6966911554336548, + "learning_rate": 2.6064804469273745e-05, + "loss": 0.2105, + "step": 7780 + }, + { + "epoch": 21.73463687150838, + "grad_norm": 0.6444169878959656, + "learning_rate": 2.6068156424581006e-05, + "loss": 0.1453, + "step": 7781 + }, + { + "epoch": 21.737430167597765, + "grad_norm": 0.6693825721740723, + "learning_rate": 2.607150837988827e-05, + "loss": 0.1595, + "step": 7782 + }, + { + "epoch": 21.74022346368715, + "grad_norm": 0.8023471832275391, + "learning_rate": 2.6074860335195532e-05, + "loss": 0.1502, + "step": 7783 + }, + { + "epoch": 21.743016759776538, + "grad_norm": 1.1007559299468994, + "learning_rate": 2.6078212290502794e-05, + "loss": 0.158, + "step": 7784 + }, + { + "epoch": 21.74581005586592, + "grad_norm": 0.7764101028442383, + "learning_rate": 2.608156424581006e-05, + "loss": 0.1457, + "step": 7785 + }, + { + "epoch": 21.748603351955307, + "grad_norm": 1.0429787635803223, + "learning_rate": 2.608491620111732e-05, + "loss": 0.1477, + "step": 7786 + }, + { + "epoch": 21.751396648044693, + "grad_norm": 1.300158977508545, + "learning_rate": 2.608826815642458e-05, + "loss": 0.1673, + "step": 7787 + }, + { + "epoch": 21.75418994413408, + "grad_norm": 1.0175738334655762, + "learning_rate": 2.6091620111731846e-05, + "loss": 0.1528, + "step": 7788 + }, + { + "epoch": 21.756983240223462, + "grad_norm": 1.0015062093734741, + "learning_rate": 2.6094972067039107e-05, + "loss": 0.1399, + "step": 7789 + }, + { + "epoch": 21.75977653631285, + "grad_norm": 0.7634696960449219, + "learning_rate": 2.609832402234637e-05, + "loss": 0.1514, + "step": 7790 + }, + { + "epoch": 21.762569832402235, + "grad_norm": 1.3089159727096558, + "learning_rate": 2.6101675977653633e-05, + "loss": 0.1798, + "step": 7791 + }, + { + "epoch": 21.76536312849162, + "grad_norm": 1.272267460823059, + "learning_rate": 2.6105027932960895e-05, + "loss": 0.1971, + "step": 7792 + }, + { + "epoch": 21.768156424581004, + "grad_norm": 1.7579312324523926, + "learning_rate": 2.6108379888268156e-05, + "loss": 0.2339, + "step": 7793 + }, + { + "epoch": 21.77094972067039, + "grad_norm": 0.9138770699501038, + "learning_rate": 2.611173184357542e-05, + "loss": 0.1645, + "step": 7794 + }, + { + "epoch": 21.773743016759777, + "grad_norm": 0.4265211224555969, + "learning_rate": 2.6115083798882682e-05, + "loss": 0.1547, + "step": 7795 + }, + { + "epoch": 21.776536312849164, + "grad_norm": 0.5170164704322815, + "learning_rate": 2.6118435754189943e-05, + "loss": 0.1704, + "step": 7796 + }, + { + "epoch": 21.779329608938546, + "grad_norm": 0.7257465720176697, + "learning_rate": 2.6121787709497208e-05, + "loss": 0.1639, + "step": 7797 + }, + { + "epoch": 21.782122905027933, + "grad_norm": 0.5306664705276489, + "learning_rate": 2.612513966480447e-05, + "loss": 0.1965, + "step": 7798 + }, + { + "epoch": 21.78491620111732, + "grad_norm": 0.7129229307174683, + "learning_rate": 2.6128491620111734e-05, + "loss": 0.1793, + "step": 7799 + }, + { + "epoch": 21.787709497206706, + "grad_norm": 0.7835840582847595, + "learning_rate": 2.6131843575418995e-05, + "loss": 0.1964, + "step": 7800 + }, + { + "epoch": 21.79050279329609, + "grad_norm": 0.6245186924934387, + "learning_rate": 2.6135195530726257e-05, + "loss": 0.2006, + "step": 7801 + }, + { + "epoch": 21.793296089385475, + "grad_norm": 1.0638221502304077, + "learning_rate": 2.613854748603352e-05, + "loss": 0.1465, + "step": 7802 + }, + { + "epoch": 21.79608938547486, + "grad_norm": 0.9328100681304932, + "learning_rate": 2.6141899441340783e-05, + "loss": 0.1907, + "step": 7803 + }, + { + "epoch": 21.798882681564244, + "grad_norm": 0.6893723011016846, + "learning_rate": 2.6145251396648044e-05, + "loss": 0.1335, + "step": 7804 + }, + { + "epoch": 21.80167597765363, + "grad_norm": 1.7624566555023193, + "learning_rate": 2.614860335195531e-05, + "loss": 0.1585, + "step": 7805 + }, + { + "epoch": 21.804469273743017, + "grad_norm": 0.6934065222740173, + "learning_rate": 2.615195530726257e-05, + "loss": 0.1646, + "step": 7806 + }, + { + "epoch": 21.807262569832403, + "grad_norm": 1.3225336074829102, + "learning_rate": 2.615530726256983e-05, + "loss": 0.1553, + "step": 7807 + }, + { + "epoch": 21.810055865921786, + "grad_norm": 0.9309976696968079, + "learning_rate": 2.6158659217877096e-05, + "loss": 0.1575, + "step": 7808 + }, + { + "epoch": 21.812849162011172, + "grad_norm": 0.6996243596076965, + "learning_rate": 2.6162011173184358e-05, + "loss": 0.1599, + "step": 7809 + }, + { + "epoch": 21.81564245810056, + "grad_norm": 0.7372048497200012, + "learning_rate": 2.616536312849162e-05, + "loss": 0.2025, + "step": 7810 + }, + { + "epoch": 21.818435754189945, + "grad_norm": 1.1844816207885742, + "learning_rate": 2.6168715083798884e-05, + "loss": 0.1379, + "step": 7811 + }, + { + "epoch": 21.821229050279328, + "grad_norm": 0.9261242151260376, + "learning_rate": 2.6172067039106145e-05, + "loss": 0.1689, + "step": 7812 + }, + { + "epoch": 21.824022346368714, + "grad_norm": 0.8296582698822021, + "learning_rate": 2.6175418994413407e-05, + "loss": 0.1681, + "step": 7813 + }, + { + "epoch": 21.8268156424581, + "grad_norm": 1.3249186277389526, + "learning_rate": 2.617877094972067e-05, + "loss": 0.183, + "step": 7814 + }, + { + "epoch": 21.829608938547487, + "grad_norm": 0.8132324814796448, + "learning_rate": 2.6182122905027933e-05, + "loss": 0.1574, + "step": 7815 + }, + { + "epoch": 21.83240223463687, + "grad_norm": 1.2290916442871094, + "learning_rate": 2.6185474860335194e-05, + "loss": 0.1726, + "step": 7816 + }, + { + "epoch": 21.835195530726256, + "grad_norm": 0.9991129636764526, + "learning_rate": 2.618882681564246e-05, + "loss": 0.2328, + "step": 7817 + }, + { + "epoch": 21.837988826815643, + "grad_norm": 2.4099619388580322, + "learning_rate": 2.619217877094972e-05, + "loss": 0.2178, + "step": 7818 + }, + { + "epoch": 21.84078212290503, + "grad_norm": 0.55148845911026, + "learning_rate": 2.6195530726256985e-05, + "loss": 0.1884, + "step": 7819 + }, + { + "epoch": 21.843575418994412, + "grad_norm": 0.48479920625686646, + "learning_rate": 2.6198882681564246e-05, + "loss": 0.1595, + "step": 7820 + }, + { + "epoch": 21.8463687150838, + "grad_norm": 0.9879711866378784, + "learning_rate": 2.6202234636871507e-05, + "loss": 0.144, + "step": 7821 + }, + { + "epoch": 21.849162011173185, + "grad_norm": 0.6489811539649963, + "learning_rate": 2.6205586592178772e-05, + "loss": 0.1762, + "step": 7822 + }, + { + "epoch": 21.85195530726257, + "grad_norm": 0.5258278846740723, + "learning_rate": 2.6208938547486034e-05, + "loss": 0.1413, + "step": 7823 + }, + { + "epoch": 21.854748603351954, + "grad_norm": 0.4998035132884979, + "learning_rate": 2.6212290502793295e-05, + "loss": 0.1665, + "step": 7824 + }, + { + "epoch": 21.85754189944134, + "grad_norm": 0.5584720373153687, + "learning_rate": 2.621564245810056e-05, + "loss": 0.1676, + "step": 7825 + }, + { + "epoch": 21.860335195530727, + "grad_norm": 0.8814937472343445, + "learning_rate": 2.621899441340782e-05, + "loss": 0.1598, + "step": 7826 + }, + { + "epoch": 21.863128491620113, + "grad_norm": 0.7379825711250305, + "learning_rate": 2.6222346368715082e-05, + "loss": 0.1633, + "step": 7827 + }, + { + "epoch": 21.865921787709496, + "grad_norm": 0.457662433385849, + "learning_rate": 2.6225698324022347e-05, + "loss": 0.1514, + "step": 7828 + }, + { + "epoch": 21.868715083798882, + "grad_norm": 0.5338233709335327, + "learning_rate": 2.622905027932961e-05, + "loss": 0.1305, + "step": 7829 + }, + { + "epoch": 21.87150837988827, + "grad_norm": 1.298097848892212, + "learning_rate": 2.623240223463687e-05, + "loss": 0.1577, + "step": 7830 + }, + { + "epoch": 21.874301675977655, + "grad_norm": 0.6746712327003479, + "learning_rate": 2.6235754189944134e-05, + "loss": 0.1642, + "step": 7831 + }, + { + "epoch": 21.877094972067038, + "grad_norm": 0.6356415152549744, + "learning_rate": 2.6239106145251396e-05, + "loss": 0.1827, + "step": 7832 + }, + { + "epoch": 21.879888268156424, + "grad_norm": 0.8733199834823608, + "learning_rate": 2.6242458100558657e-05, + "loss": 0.1407, + "step": 7833 + }, + { + "epoch": 21.88268156424581, + "grad_norm": 0.7100284099578857, + "learning_rate": 2.6245810055865922e-05, + "loss": 0.1576, + "step": 7834 + }, + { + "epoch": 21.885474860335197, + "grad_norm": 0.7292144298553467, + "learning_rate": 2.6249162011173183e-05, + "loss": 0.1847, + "step": 7835 + }, + { + "epoch": 21.88826815642458, + "grad_norm": 2.0663704872131348, + "learning_rate": 2.6252513966480448e-05, + "loss": 0.1557, + "step": 7836 + }, + { + "epoch": 21.891061452513966, + "grad_norm": 0.6443936228752136, + "learning_rate": 2.625586592178771e-05, + "loss": 0.1653, + "step": 7837 + }, + { + "epoch": 21.893854748603353, + "grad_norm": 1.139275074005127, + "learning_rate": 2.625921787709497e-05, + "loss": 0.1586, + "step": 7838 + }, + { + "epoch": 21.89664804469274, + "grad_norm": 0.7205286622047424, + "learning_rate": 2.6262569832402235e-05, + "loss": 0.1629, + "step": 7839 + }, + { + "epoch": 21.899441340782122, + "grad_norm": 0.8915383815765381, + "learning_rate": 2.6265921787709497e-05, + "loss": 0.156, + "step": 7840 + }, + { + "epoch": 21.90223463687151, + "grad_norm": 1.571757197380066, + "learning_rate": 2.6269273743016758e-05, + "loss": 0.1842, + "step": 7841 + }, + { + "epoch": 21.905027932960895, + "grad_norm": 6.937084197998047, + "learning_rate": 2.6272625698324023e-05, + "loss": 0.201, + "step": 7842 + }, + { + "epoch": 21.907821229050278, + "grad_norm": 1.8235082626342773, + "learning_rate": 2.6275977653631284e-05, + "loss": 0.2445, + "step": 7843 + }, + { + "epoch": 21.910614525139664, + "grad_norm": 0.4081736207008362, + "learning_rate": 2.6279329608938545e-05, + "loss": 0.1831, + "step": 7844 + }, + { + "epoch": 21.91340782122905, + "grad_norm": 0.7917836904525757, + "learning_rate": 2.628268156424581e-05, + "loss": 0.1788, + "step": 7845 + }, + { + "epoch": 21.916201117318437, + "grad_norm": 0.9535239934921265, + "learning_rate": 2.628603351955307e-05, + "loss": 0.2137, + "step": 7846 + }, + { + "epoch": 21.91899441340782, + "grad_norm": 0.6565654873847961, + "learning_rate": 2.6289385474860336e-05, + "loss": 0.1815, + "step": 7847 + }, + { + "epoch": 21.921787709497206, + "grad_norm": 0.6542697548866272, + "learning_rate": 2.62927374301676e-05, + "loss": 0.1908, + "step": 7848 + }, + { + "epoch": 21.924581005586592, + "grad_norm": 0.8679887056350708, + "learning_rate": 2.6296089385474862e-05, + "loss": 0.1706, + "step": 7849 + }, + { + "epoch": 21.92737430167598, + "grad_norm": 0.6847873330116272, + "learning_rate": 2.6299441340782124e-05, + "loss": 0.179, + "step": 7850 + }, + { + "epoch": 21.93016759776536, + "grad_norm": 1.0022799968719482, + "learning_rate": 2.630279329608939e-05, + "loss": 0.1617, + "step": 7851 + }, + { + "epoch": 21.932960893854748, + "grad_norm": 0.8524408340454102, + "learning_rate": 2.630614525139665e-05, + "loss": 0.1629, + "step": 7852 + }, + { + "epoch": 21.935754189944134, + "grad_norm": 0.9325610399246216, + "learning_rate": 2.630949720670391e-05, + "loss": 0.177, + "step": 7853 + }, + { + "epoch": 21.93854748603352, + "grad_norm": 0.6609992384910583, + "learning_rate": 2.6312849162011176e-05, + "loss": 0.1668, + "step": 7854 + }, + { + "epoch": 21.941340782122904, + "grad_norm": 0.7186777591705322, + "learning_rate": 2.6316201117318437e-05, + "loss": 0.159, + "step": 7855 + }, + { + "epoch": 21.94413407821229, + "grad_norm": 0.7891885042190552, + "learning_rate": 2.6319553072625702e-05, + "loss": 0.1992, + "step": 7856 + }, + { + "epoch": 21.946927374301676, + "grad_norm": 0.5868457555770874, + "learning_rate": 2.6322905027932963e-05, + "loss": 0.1586, + "step": 7857 + }, + { + "epoch": 21.949720670391063, + "grad_norm": 1.1813751459121704, + "learning_rate": 2.6326256983240225e-05, + "loss": 0.1777, + "step": 7858 + }, + { + "epoch": 21.952513966480446, + "grad_norm": 0.7648119926452637, + "learning_rate": 2.632960893854749e-05, + "loss": 0.1511, + "step": 7859 + }, + { + "epoch": 21.955307262569832, + "grad_norm": 0.9312558174133301, + "learning_rate": 2.633296089385475e-05, + "loss": 0.1721, + "step": 7860 + }, + { + "epoch": 21.95810055865922, + "grad_norm": 0.7215638160705566, + "learning_rate": 2.6336312849162012e-05, + "loss": 0.1575, + "step": 7861 + }, + { + "epoch": 21.960893854748605, + "grad_norm": 3.375340700149536, + "learning_rate": 2.6339664804469277e-05, + "loss": 0.1574, + "step": 7862 + }, + { + "epoch": 21.963687150837988, + "grad_norm": 1.0222119092941284, + "learning_rate": 2.6343016759776538e-05, + "loss": 0.1516, + "step": 7863 + }, + { + "epoch": 21.966480446927374, + "grad_norm": 0.935189962387085, + "learning_rate": 2.63463687150838e-05, + "loss": 0.1804, + "step": 7864 + }, + { + "epoch": 21.96927374301676, + "grad_norm": 2.7159199714660645, + "learning_rate": 2.6349720670391064e-05, + "loss": 0.1502, + "step": 7865 + }, + { + "epoch": 21.972067039106147, + "grad_norm": 1.4849170446395874, + "learning_rate": 2.6353072625698326e-05, + "loss": 0.183, + "step": 7866 + }, + { + "epoch": 21.97486033519553, + "grad_norm": 1.7754586935043335, + "learning_rate": 2.6356424581005587e-05, + "loss": 0.185, + "step": 7867 + }, + { + "epoch": 21.977653631284916, + "grad_norm": 4.503955364227295, + "learning_rate": 2.635977653631285e-05, + "loss": 0.2612, + "step": 7868 + }, + { + "epoch": 21.980446927374302, + "grad_norm": 0.5966487526893616, + "learning_rate": 2.6363128491620113e-05, + "loss": 0.1928, + "step": 7869 + }, + { + "epoch": 21.98324022346369, + "grad_norm": 0.9294701218605042, + "learning_rate": 2.6366480446927374e-05, + "loss": 0.1728, + "step": 7870 + }, + { + "epoch": 21.98603351955307, + "grad_norm": 0.6523683667182922, + "learning_rate": 2.636983240223464e-05, + "loss": 0.1672, + "step": 7871 + }, + { + "epoch": 21.988826815642458, + "grad_norm": 0.5796045660972595, + "learning_rate": 2.63731843575419e-05, + "loss": 0.156, + "step": 7872 + }, + { + "epoch": 21.991620111731844, + "grad_norm": 0.6035365462303162, + "learning_rate": 2.6376536312849165e-05, + "loss": 0.1754, + "step": 7873 + }, + { + "epoch": 21.994413407821227, + "grad_norm": 1.8075237274169922, + "learning_rate": 2.6379888268156426e-05, + "loss": 0.1572, + "step": 7874 + }, + { + "epoch": 21.997206703910614, + "grad_norm": 1.9696401357650757, + "learning_rate": 2.6383240223463688e-05, + "loss": 0.1642, + "step": 7875 + }, + { + "epoch": 22.0, + "grad_norm": 0.9906492233276367, + "learning_rate": 2.6386592178770953e-05, + "loss": 0.2, + "step": 7876 + }, + { + "epoch": 22.002793296089386, + "grad_norm": 0.4237697124481201, + "learning_rate": 2.6389944134078214e-05, + "loss": 0.1811, + "step": 7877 + }, + { + "epoch": 22.00558659217877, + "grad_norm": 0.5136610269546509, + "learning_rate": 2.6393296089385475e-05, + "loss": 0.1693, + "step": 7878 + }, + { + "epoch": 22.008379888268156, + "grad_norm": 0.6668696403503418, + "learning_rate": 2.639664804469274e-05, + "loss": 0.1659, + "step": 7879 + }, + { + "epoch": 22.011173184357542, + "grad_norm": 0.4744291603565216, + "learning_rate": 2.64e-05, + "loss": 0.1804, + "step": 7880 + }, + { + "epoch": 22.01396648044693, + "grad_norm": 0.478912889957428, + "learning_rate": 2.6403351955307263e-05, + "loss": 0.1611, + "step": 7881 + }, + { + "epoch": 22.01675977653631, + "grad_norm": 0.6100389361381531, + "learning_rate": 2.6406703910614527e-05, + "loss": 0.1593, + "step": 7882 + }, + { + "epoch": 22.019553072625698, + "grad_norm": 0.9962566494941711, + "learning_rate": 2.641005586592179e-05, + "loss": 0.2066, + "step": 7883 + }, + { + "epoch": 22.022346368715084, + "grad_norm": 0.6838847994804382, + "learning_rate": 2.641340782122905e-05, + "loss": 0.1529, + "step": 7884 + }, + { + "epoch": 22.02513966480447, + "grad_norm": 0.9703115224838257, + "learning_rate": 2.6416759776536315e-05, + "loss": 0.1694, + "step": 7885 + }, + { + "epoch": 22.027932960893853, + "grad_norm": 0.56429123878479, + "learning_rate": 2.6420111731843576e-05, + "loss": 0.1641, + "step": 7886 + }, + { + "epoch": 22.03072625698324, + "grad_norm": 0.5601845383644104, + "learning_rate": 2.6423463687150838e-05, + "loss": 0.1881, + "step": 7887 + }, + { + "epoch": 22.033519553072626, + "grad_norm": 0.7585433125495911, + "learning_rate": 2.6426815642458102e-05, + "loss": 0.1642, + "step": 7888 + }, + { + "epoch": 22.036312849162012, + "grad_norm": 0.6029527187347412, + "learning_rate": 2.6430167597765364e-05, + "loss": 0.15, + "step": 7889 + }, + { + "epoch": 22.039106145251395, + "grad_norm": 0.5234736800193787, + "learning_rate": 2.6433519553072625e-05, + "loss": 0.1843, + "step": 7890 + }, + { + "epoch": 22.04189944134078, + "grad_norm": 0.726692259311676, + "learning_rate": 2.643687150837989e-05, + "loss": 0.1553, + "step": 7891 + }, + { + "epoch": 22.044692737430168, + "grad_norm": 0.6134770512580872, + "learning_rate": 2.644022346368715e-05, + "loss": 0.1433, + "step": 7892 + }, + { + "epoch": 22.047486033519554, + "grad_norm": 0.6395728588104248, + "learning_rate": 2.6443575418994416e-05, + "loss": 0.1468, + "step": 7893 + }, + { + "epoch": 22.050279329608937, + "grad_norm": 5.285223007202148, + "learning_rate": 2.6446927374301677e-05, + "loss": 0.1446, + "step": 7894 + }, + { + "epoch": 22.053072625698324, + "grad_norm": 0.960630476474762, + "learning_rate": 2.645027932960894e-05, + "loss": 0.1422, + "step": 7895 + }, + { + "epoch": 22.05586592178771, + "grad_norm": 0.8763542771339417, + "learning_rate": 2.6453631284916203e-05, + "loss": 0.151, + "step": 7896 + }, + { + "epoch": 22.058659217877096, + "grad_norm": 1.3528679609298706, + "learning_rate": 2.6456983240223465e-05, + "loss": 0.1565, + "step": 7897 + }, + { + "epoch": 22.06145251396648, + "grad_norm": 1.2949496507644653, + "learning_rate": 2.6460335195530726e-05, + "loss": 0.1469, + "step": 7898 + }, + { + "epoch": 22.064245810055866, + "grad_norm": 2.3557982444763184, + "learning_rate": 2.646368715083799e-05, + "loss": 0.1549, + "step": 7899 + }, + { + "epoch": 22.067039106145252, + "grad_norm": 1.4667967557907104, + "learning_rate": 2.6467039106145252e-05, + "loss": 0.1749, + "step": 7900 + }, + { + "epoch": 22.06983240223464, + "grad_norm": 1.5758122205734253, + "learning_rate": 2.6470391061452513e-05, + "loss": 0.2096, + "step": 7901 + }, + { + "epoch": 22.07262569832402, + "grad_norm": 0.839803159236908, + "learning_rate": 2.6473743016759778e-05, + "loss": 0.1684, + "step": 7902 + }, + { + "epoch": 22.075418994413408, + "grad_norm": 0.5568279027938843, + "learning_rate": 2.647709497206704e-05, + "loss": 0.1524, + "step": 7903 + }, + { + "epoch": 22.078212290502794, + "grad_norm": 0.8372209668159485, + "learning_rate": 2.64804469273743e-05, + "loss": 0.1439, + "step": 7904 + }, + { + "epoch": 22.08100558659218, + "grad_norm": 0.7860749363899231, + "learning_rate": 2.6483798882681565e-05, + "loss": 0.1505, + "step": 7905 + }, + { + "epoch": 22.083798882681563, + "grad_norm": 1.086247205734253, + "learning_rate": 2.6487150837988827e-05, + "loss": 0.2082, + "step": 7906 + }, + { + "epoch": 22.08659217877095, + "grad_norm": 0.7068296670913696, + "learning_rate": 2.6490502793296088e-05, + "loss": 0.159, + "step": 7907 + }, + { + "epoch": 22.089385474860336, + "grad_norm": 3.6472740173339844, + "learning_rate": 2.6493854748603353e-05, + "loss": 0.1804, + "step": 7908 + }, + { + "epoch": 22.092178770949722, + "grad_norm": 0.8783243894577026, + "learning_rate": 2.6497206703910614e-05, + "loss": 0.1568, + "step": 7909 + }, + { + "epoch": 22.094972067039105, + "grad_norm": 0.4723097085952759, + "learning_rate": 2.6500558659217876e-05, + "loss": 0.1724, + "step": 7910 + }, + { + "epoch": 22.09776536312849, + "grad_norm": 0.9125997424125671, + "learning_rate": 2.650391061452514e-05, + "loss": 0.1754, + "step": 7911 + }, + { + "epoch": 22.100558659217878, + "grad_norm": 0.9117453694343567, + "learning_rate": 2.65072625698324e-05, + "loss": 0.1424, + "step": 7912 + }, + { + "epoch": 22.10335195530726, + "grad_norm": 0.4290664494037628, + "learning_rate": 2.6510614525139666e-05, + "loss": 0.1496, + "step": 7913 + }, + { + "epoch": 22.106145251396647, + "grad_norm": 0.5028756260871887, + "learning_rate": 2.6513966480446928e-05, + "loss": 0.1399, + "step": 7914 + }, + { + "epoch": 22.108938547486034, + "grad_norm": 0.7490841746330261, + "learning_rate": 2.651731843575419e-05, + "loss": 0.1567, + "step": 7915 + }, + { + "epoch": 22.11173184357542, + "grad_norm": 0.6924324035644531, + "learning_rate": 2.6520670391061454e-05, + "loss": 0.1349, + "step": 7916 + }, + { + "epoch": 22.114525139664803, + "grad_norm": 0.6080396175384521, + "learning_rate": 2.6524022346368715e-05, + "loss": 0.151, + "step": 7917 + }, + { + "epoch": 22.11731843575419, + "grad_norm": 0.9845266938209534, + "learning_rate": 2.6527374301675977e-05, + "loss": 0.1646, + "step": 7918 + }, + { + "epoch": 22.120111731843576, + "grad_norm": 1.5881942510604858, + "learning_rate": 2.653072625698324e-05, + "loss": 0.1706, + "step": 7919 + }, + { + "epoch": 22.122905027932962, + "grad_norm": 1.295616865158081, + "learning_rate": 2.6534078212290503e-05, + "loss": 0.187, + "step": 7920 + }, + { + "epoch": 22.125698324022345, + "grad_norm": 0.7685617208480835, + "learning_rate": 2.6537430167597764e-05, + "loss": 0.1873, + "step": 7921 + }, + { + "epoch": 22.12849162011173, + "grad_norm": 0.8132511377334595, + "learning_rate": 2.654078212290503e-05, + "loss": 0.1382, + "step": 7922 + }, + { + "epoch": 22.131284916201118, + "grad_norm": 0.8639641404151917, + "learning_rate": 2.654413407821229e-05, + "loss": 0.1647, + "step": 7923 + }, + { + "epoch": 22.134078212290504, + "grad_norm": 1.0057343244552612, + "learning_rate": 2.654748603351955e-05, + "loss": 0.1747, + "step": 7924 + }, + { + "epoch": 22.136871508379887, + "grad_norm": 1.1537292003631592, + "learning_rate": 2.6550837988826816e-05, + "loss": 0.1556, + "step": 7925 + }, + { + "epoch": 22.139664804469273, + "grad_norm": 2.6073951721191406, + "learning_rate": 2.6554189944134077e-05, + "loss": 0.2115, + "step": 7926 + }, + { + "epoch": 22.14245810055866, + "grad_norm": 0.5901563763618469, + "learning_rate": 2.655754189944134e-05, + "loss": 0.2098, + "step": 7927 + }, + { + "epoch": 22.145251396648046, + "grad_norm": 0.6029103398323059, + "learning_rate": 2.6560893854748604e-05, + "loss": 0.1983, + "step": 7928 + }, + { + "epoch": 22.14804469273743, + "grad_norm": 1.2558470964431763, + "learning_rate": 2.6564245810055865e-05, + "loss": 0.1742, + "step": 7929 + }, + { + "epoch": 22.150837988826815, + "grad_norm": 0.45540371537208557, + "learning_rate": 2.656759776536313e-05, + "loss": 0.185, + "step": 7930 + }, + { + "epoch": 22.1536312849162, + "grad_norm": 1.0885956287384033, + "learning_rate": 2.657094972067039e-05, + "loss": 0.1849, + "step": 7931 + }, + { + "epoch": 22.156424581005588, + "grad_norm": 0.5614975094795227, + "learning_rate": 2.6574301675977652e-05, + "loss": 0.1754, + "step": 7932 + }, + { + "epoch": 22.15921787709497, + "grad_norm": 0.6331824064254761, + "learning_rate": 2.6577653631284917e-05, + "loss": 0.1721, + "step": 7933 + }, + { + "epoch": 22.162011173184357, + "grad_norm": 0.5809019207954407, + "learning_rate": 2.658100558659218e-05, + "loss": 0.1728, + "step": 7934 + }, + { + "epoch": 22.164804469273744, + "grad_norm": 0.5154022574424744, + "learning_rate": 2.658435754189944e-05, + "loss": 0.139, + "step": 7935 + }, + { + "epoch": 22.16759776536313, + "grad_norm": 0.5877636671066284, + "learning_rate": 2.6587709497206704e-05, + "loss": 0.181, + "step": 7936 + }, + { + "epoch": 22.170391061452513, + "grad_norm": 0.8054257035255432, + "learning_rate": 2.6591061452513966e-05, + "loss": 0.1668, + "step": 7937 + }, + { + "epoch": 22.1731843575419, + "grad_norm": 0.7663404941558838, + "learning_rate": 2.6594413407821227e-05, + "loss": 0.1562, + "step": 7938 + }, + { + "epoch": 22.175977653631286, + "grad_norm": 0.965268611907959, + "learning_rate": 2.6597765363128492e-05, + "loss": 0.1647, + "step": 7939 + }, + { + "epoch": 22.178770949720672, + "grad_norm": 1.2410606145858765, + "learning_rate": 2.6601117318435753e-05, + "loss": 0.1446, + "step": 7940 + }, + { + "epoch": 22.181564245810055, + "grad_norm": 0.7466762065887451, + "learning_rate": 2.6604469273743015e-05, + "loss": 0.1468, + "step": 7941 + }, + { + "epoch": 22.18435754189944, + "grad_norm": 1.2268697023391724, + "learning_rate": 2.660782122905028e-05, + "loss": 0.1603, + "step": 7942 + }, + { + "epoch": 22.187150837988828, + "grad_norm": 0.6330245733261108, + "learning_rate": 2.661117318435754e-05, + "loss": 0.1369, + "step": 7943 + }, + { + "epoch": 22.189944134078214, + "grad_norm": 0.6648063063621521, + "learning_rate": 2.6614525139664805e-05, + "loss": 0.1599, + "step": 7944 + }, + { + "epoch": 22.192737430167597, + "grad_norm": 1.107960820198059, + "learning_rate": 2.661787709497207e-05, + "loss": 0.1721, + "step": 7945 + }, + { + "epoch": 22.195530726256983, + "grad_norm": 1.3340113162994385, + "learning_rate": 2.662122905027933e-05, + "loss": 0.1667, + "step": 7946 + }, + { + "epoch": 22.19832402234637, + "grad_norm": 0.9522841572761536, + "learning_rate": 2.6624581005586593e-05, + "loss": 0.157, + "step": 7947 + }, + { + "epoch": 22.201117318435756, + "grad_norm": 2.8796303272247314, + "learning_rate": 2.6627932960893858e-05, + "loss": 0.1593, + "step": 7948 + }, + { + "epoch": 22.20391061452514, + "grad_norm": 0.8559346795082092, + "learning_rate": 2.663128491620112e-05, + "loss": 0.1567, + "step": 7949 + }, + { + "epoch": 22.206703910614525, + "grad_norm": 0.8659203052520752, + "learning_rate": 2.6634636871508384e-05, + "loss": 0.1587, + "step": 7950 + }, + { + "epoch": 22.20949720670391, + "grad_norm": 2.879755735397339, + "learning_rate": 2.6637988826815645e-05, + "loss": 0.2683, + "step": 7951 + }, + { + "epoch": 22.212290502793294, + "grad_norm": 0.6027741432189941, + "learning_rate": 2.6641340782122906e-05, + "loss": 0.1853, + "step": 7952 + }, + { + "epoch": 22.21508379888268, + "grad_norm": 0.5981945395469666, + "learning_rate": 2.664469273743017e-05, + "loss": 0.1707, + "step": 7953 + }, + { + "epoch": 22.217877094972067, + "grad_norm": 0.5736843943595886, + "learning_rate": 2.6648044692737432e-05, + "loss": 0.1721, + "step": 7954 + }, + { + "epoch": 22.220670391061454, + "grad_norm": 1.5127372741699219, + "learning_rate": 2.6651396648044694e-05, + "loss": 0.1577, + "step": 7955 + }, + { + "epoch": 22.223463687150836, + "grad_norm": 0.5974329113960266, + "learning_rate": 2.665474860335196e-05, + "loss": 0.1478, + "step": 7956 + }, + { + "epoch": 22.226256983240223, + "grad_norm": 0.4945756196975708, + "learning_rate": 2.665810055865922e-05, + "loss": 0.186, + "step": 7957 + }, + { + "epoch": 22.22905027932961, + "grad_norm": 1.1989701986312866, + "learning_rate": 2.666145251396648e-05, + "loss": 0.1937, + "step": 7958 + }, + { + "epoch": 22.231843575418996, + "grad_norm": 0.6308298707008362, + "learning_rate": 2.6664804469273746e-05, + "loss": 0.1357, + "step": 7959 + }, + { + "epoch": 22.23463687150838, + "grad_norm": 1.0688064098358154, + "learning_rate": 2.6668156424581007e-05, + "loss": 0.1448, + "step": 7960 + }, + { + "epoch": 22.237430167597765, + "grad_norm": 0.7172343134880066, + "learning_rate": 2.667150837988827e-05, + "loss": 0.1758, + "step": 7961 + }, + { + "epoch": 22.24022346368715, + "grad_norm": 2.657715320587158, + "learning_rate": 2.6674860335195533e-05, + "loss": 0.1628, + "step": 7962 + }, + { + "epoch": 22.243016759776538, + "grad_norm": 0.602561891078949, + "learning_rate": 2.6678212290502795e-05, + "loss": 0.1489, + "step": 7963 + }, + { + "epoch": 22.24581005586592, + "grad_norm": 1.2476184368133545, + "learning_rate": 2.6681564245810056e-05, + "loss": 0.1604, + "step": 7964 + }, + { + "epoch": 22.248603351955307, + "grad_norm": 0.8561045527458191, + "learning_rate": 2.668491620111732e-05, + "loss": 0.1884, + "step": 7965 + }, + { + "epoch": 22.251396648044693, + "grad_norm": 0.7554227709770203, + "learning_rate": 2.6688268156424582e-05, + "loss": 0.1347, + "step": 7966 + }, + { + "epoch": 22.25418994413408, + "grad_norm": 0.9117192029953003, + "learning_rate": 2.6691620111731847e-05, + "loss": 0.1424, + "step": 7967 + }, + { + "epoch": 22.256983240223462, + "grad_norm": 0.8556345701217651, + "learning_rate": 2.6694972067039108e-05, + "loss": 0.1668, + "step": 7968 + }, + { + "epoch": 22.25977653631285, + "grad_norm": 1.033816933631897, + "learning_rate": 2.669832402234637e-05, + "loss": 0.156, + "step": 7969 + }, + { + "epoch": 22.262569832402235, + "grad_norm": 4.444972515106201, + "learning_rate": 2.6701675977653634e-05, + "loss": 0.1649, + "step": 7970 + }, + { + "epoch": 22.26536312849162, + "grad_norm": 0.5616012811660767, + "learning_rate": 2.6705027932960896e-05, + "loss": 0.1483, + "step": 7971 + }, + { + "epoch": 22.268156424581004, + "grad_norm": 0.9038326740264893, + "learning_rate": 2.6708379888268157e-05, + "loss": 0.1908, + "step": 7972 + }, + { + "epoch": 22.27094972067039, + "grad_norm": 1.3689532279968262, + "learning_rate": 2.671173184357542e-05, + "loss": 0.1836, + "step": 7973 + }, + { + "epoch": 22.273743016759777, + "grad_norm": 1.2178508043289185, + "learning_rate": 2.6715083798882683e-05, + "loss": 0.1824, + "step": 7974 + }, + { + "epoch": 22.276536312849164, + "grad_norm": 1.0856226682662964, + "learning_rate": 2.6718435754189944e-05, + "loss": 0.2029, + "step": 7975 + }, + { + "epoch": 22.279329608938546, + "grad_norm": 2.8208653926849365, + "learning_rate": 2.672178770949721e-05, + "loss": 0.2194, + "step": 7976 + }, + { + "epoch": 22.282122905027933, + "grad_norm": 0.7358446717262268, + "learning_rate": 2.672513966480447e-05, + "loss": 0.1965, + "step": 7977 + }, + { + "epoch": 22.28491620111732, + "grad_norm": 0.5570719838142395, + "learning_rate": 2.6728491620111732e-05, + "loss": 0.1598, + "step": 7978 + }, + { + "epoch": 22.287709497206706, + "grad_norm": 0.6084880232810974, + "learning_rate": 2.6731843575418997e-05, + "loss": 0.1984, + "step": 7979 + }, + { + "epoch": 22.29050279329609, + "grad_norm": 0.4775005877017975, + "learning_rate": 2.6735195530726258e-05, + "loss": 0.1657, + "step": 7980 + }, + { + "epoch": 22.293296089385475, + "grad_norm": 0.6780799627304077, + "learning_rate": 2.673854748603352e-05, + "loss": 0.1971, + "step": 7981 + }, + { + "epoch": 22.29608938547486, + "grad_norm": 0.6231633424758911, + "learning_rate": 2.6741899441340784e-05, + "loss": 0.1474, + "step": 7982 + }, + { + "epoch": 22.298882681564244, + "grad_norm": 0.7835116982460022, + "learning_rate": 2.6745251396648045e-05, + "loss": 0.1649, + "step": 7983 + }, + { + "epoch": 22.30167597765363, + "grad_norm": 1.1081002950668335, + "learning_rate": 2.6748603351955307e-05, + "loss": 0.1613, + "step": 7984 + }, + { + "epoch": 22.304469273743017, + "grad_norm": 0.707419753074646, + "learning_rate": 2.675195530726257e-05, + "loss": 0.1687, + "step": 7985 + }, + { + "epoch": 22.307262569832403, + "grad_norm": 0.6702539920806885, + "learning_rate": 2.6755307262569833e-05, + "loss": 0.1519, + "step": 7986 + }, + { + "epoch": 22.310055865921786, + "grad_norm": 0.8961595296859741, + "learning_rate": 2.6758659217877097e-05, + "loss": 0.1812, + "step": 7987 + }, + { + "epoch": 22.312849162011172, + "grad_norm": 0.8046408891677856, + "learning_rate": 2.676201117318436e-05, + "loss": 0.1597, + "step": 7988 + }, + { + "epoch": 22.31564245810056, + "grad_norm": 0.4920606017112732, + "learning_rate": 2.676536312849162e-05, + "loss": 0.1543, + "step": 7989 + }, + { + "epoch": 22.318435754189945, + "grad_norm": 1.0036081075668335, + "learning_rate": 2.6768715083798885e-05, + "loss": 0.1553, + "step": 7990 + }, + { + "epoch": 22.321229050279328, + "grad_norm": 0.5673649907112122, + "learning_rate": 2.6772067039106146e-05, + "loss": 0.1271, + "step": 7991 + }, + { + "epoch": 22.324022346368714, + "grad_norm": 0.9360150098800659, + "learning_rate": 2.6775418994413408e-05, + "loss": 0.1759, + "step": 7992 + }, + { + "epoch": 22.3268156424581, + "grad_norm": 1.667060136795044, + "learning_rate": 2.6778770949720672e-05, + "loss": 0.1276, + "step": 7993 + }, + { + "epoch": 22.329608938547487, + "grad_norm": 0.8462363481521606, + "learning_rate": 2.6782122905027934e-05, + "loss": 0.1605, + "step": 7994 + }, + { + "epoch": 22.33240223463687, + "grad_norm": 0.792304515838623, + "learning_rate": 2.6785474860335195e-05, + "loss": 0.1755, + "step": 7995 + }, + { + "epoch": 22.335195530726256, + "grad_norm": 3.8465874195098877, + "learning_rate": 2.678882681564246e-05, + "loss": 0.1625, + "step": 7996 + }, + { + "epoch": 22.337988826815643, + "grad_norm": 0.7650177478790283, + "learning_rate": 2.679217877094972e-05, + "loss": 0.156, + "step": 7997 + }, + { + "epoch": 22.34078212290503, + "grad_norm": 1.3190665245056152, + "learning_rate": 2.6795530726256982e-05, + "loss": 0.1697, + "step": 7998 + }, + { + "epoch": 22.343575418994412, + "grad_norm": 1.5974950790405273, + "learning_rate": 2.6798882681564247e-05, + "loss": 0.2146, + "step": 7999 + }, + { + "epoch": 22.3463687150838, + "grad_norm": 0.9608680009841919, + "learning_rate": 2.680223463687151e-05, + "loss": 0.1638, + "step": 8000 + }, + { + "epoch": 22.3463687150838, + "eval_cer": 0.19267781742005172, + "eval_loss": 0.26574966311454773, + "eval_runtime": 23.5257, + "eval_samples_per_second": 192.896, + "eval_steps_per_second": 0.638, + "eval_wer": 0.3426440602217815, + "step": 8000 + }, + { + "epoch": 22.349162011173185, + "grad_norm": 5.252028942108154, + "learning_rate": 2.680558659217877e-05, + "loss": 0.217, + "step": 8001 + }, + { + "epoch": 22.35195530726257, + "grad_norm": 0.5274025797843933, + "learning_rate": 2.6808938547486035e-05, + "loss": 0.1643, + "step": 8002 + }, + { + "epoch": 22.354748603351954, + "grad_norm": 0.6012386679649353, + "learning_rate": 2.6812290502793296e-05, + "loss": 0.1903, + "step": 8003 + }, + { + "epoch": 22.35754189944134, + "grad_norm": 0.7253398895263672, + "learning_rate": 2.6815642458100557e-05, + "loss": 0.179, + "step": 8004 + }, + { + "epoch": 22.360335195530727, + "grad_norm": 5.509706974029541, + "learning_rate": 2.6818994413407822e-05, + "loss": 0.1588, + "step": 8005 + }, + { + "epoch": 22.363128491620113, + "grad_norm": 0.543000340461731, + "learning_rate": 2.6822346368715083e-05, + "loss": 0.1872, + "step": 8006 + }, + { + "epoch": 22.365921787709496, + "grad_norm": 1.1368498802185059, + "learning_rate": 2.6825698324022348e-05, + "loss": 0.164, + "step": 8007 + }, + { + "epoch": 22.368715083798882, + "grad_norm": 0.5629104375839233, + "learning_rate": 2.682905027932961e-05, + "loss": 0.1836, + "step": 8008 + }, + { + "epoch": 22.37150837988827, + "grad_norm": 0.6294392943382263, + "learning_rate": 2.683240223463687e-05, + "loss": 0.1706, + "step": 8009 + }, + { + "epoch": 22.374301675977655, + "grad_norm": 0.44267261028289795, + "learning_rate": 2.6835754189944135e-05, + "loss": 0.1346, + "step": 8010 + }, + { + "epoch": 22.377094972067038, + "grad_norm": 0.5635685324668884, + "learning_rate": 2.6839106145251397e-05, + "loss": 0.1679, + "step": 8011 + }, + { + "epoch": 22.379888268156424, + "grad_norm": 0.7355695366859436, + "learning_rate": 2.6842458100558658e-05, + "loss": 0.1461, + "step": 8012 + }, + { + "epoch": 22.38268156424581, + "grad_norm": 0.6693527698516846, + "learning_rate": 2.6845810055865923e-05, + "loss": 0.1776, + "step": 8013 + }, + { + "epoch": 22.385474860335197, + "grad_norm": 0.576533317565918, + "learning_rate": 2.6849162011173184e-05, + "loss": 0.148, + "step": 8014 + }, + { + "epoch": 22.38826815642458, + "grad_norm": 0.6956793665885925, + "learning_rate": 2.6852513966480446e-05, + "loss": 0.1691, + "step": 8015 + }, + { + "epoch": 22.391061452513966, + "grad_norm": 0.562690258026123, + "learning_rate": 2.685586592178771e-05, + "loss": 0.1442, + "step": 8016 + }, + { + "epoch": 22.393854748603353, + "grad_norm": 1.3190083503723145, + "learning_rate": 2.685921787709497e-05, + "loss": 0.167, + "step": 8017 + }, + { + "epoch": 22.39664804469274, + "grad_norm": 0.6435993909835815, + "learning_rate": 2.6862569832402233e-05, + "loss": 0.1685, + "step": 8018 + }, + { + "epoch": 22.399441340782122, + "grad_norm": 0.8988540172576904, + "learning_rate": 2.6865921787709498e-05, + "loss": 0.1434, + "step": 8019 + }, + { + "epoch": 22.40223463687151, + "grad_norm": 0.7222833633422852, + "learning_rate": 2.686927374301676e-05, + "loss": 0.1457, + "step": 8020 + }, + { + "epoch": 22.405027932960895, + "grad_norm": 0.8080371618270874, + "learning_rate": 2.687262569832402e-05, + "loss": 0.1512, + "step": 8021 + }, + { + "epoch": 22.407821229050278, + "grad_norm": 0.5990322232246399, + "learning_rate": 2.6875977653631285e-05, + "loss": 0.164, + "step": 8022 + }, + { + "epoch": 22.410614525139664, + "grad_norm": 0.7144290208816528, + "learning_rate": 2.6879329608938547e-05, + "loss": 0.1545, + "step": 8023 + }, + { + "epoch": 22.41340782122905, + "grad_norm": 0.9640109539031982, + "learning_rate": 2.688268156424581e-05, + "loss": 0.1862, + "step": 8024 + }, + { + "epoch": 22.416201117318437, + "grad_norm": 1.2787353992462158, + "learning_rate": 2.6886033519553073e-05, + "loss": 0.1691, + "step": 8025 + }, + { + "epoch": 22.41899441340782, + "grad_norm": 2.226045846939087, + "learning_rate": 2.6889385474860334e-05, + "loss": 0.2164, + "step": 8026 + }, + { + "epoch": 22.421787709497206, + "grad_norm": 0.46513861417770386, + "learning_rate": 2.68927374301676e-05, + "loss": 0.184, + "step": 8027 + }, + { + "epoch": 22.424581005586592, + "grad_norm": 0.6207253932952881, + "learning_rate": 2.689608938547486e-05, + "loss": 0.1791, + "step": 8028 + }, + { + "epoch": 22.42737430167598, + "grad_norm": 0.5405367612838745, + "learning_rate": 2.689944134078212e-05, + "loss": 0.1619, + "step": 8029 + }, + { + "epoch": 22.43016759776536, + "grad_norm": 0.6521279215812683, + "learning_rate": 2.6902793296089386e-05, + "loss": 0.1518, + "step": 8030 + }, + { + "epoch": 22.432960893854748, + "grad_norm": 0.6723282337188721, + "learning_rate": 2.6906145251396647e-05, + "loss": 0.1945, + "step": 8031 + }, + { + "epoch": 22.435754189944134, + "grad_norm": 0.8180859684944153, + "learning_rate": 2.690949720670391e-05, + "loss": 0.1476, + "step": 8032 + }, + { + "epoch": 22.43854748603352, + "grad_norm": 0.9533007144927979, + "learning_rate": 2.6912849162011174e-05, + "loss": 0.1723, + "step": 8033 + }, + { + "epoch": 22.441340782122904, + "grad_norm": 0.8895072937011719, + "learning_rate": 2.6916201117318435e-05, + "loss": 0.1663, + "step": 8034 + }, + { + "epoch": 22.44413407821229, + "grad_norm": 0.5895541906356812, + "learning_rate": 2.6919553072625696e-05, + "loss": 0.1555, + "step": 8035 + }, + { + "epoch": 22.446927374301676, + "grad_norm": 0.4986507296562195, + "learning_rate": 2.692290502793296e-05, + "loss": 0.1561, + "step": 8036 + }, + { + "epoch": 22.449720670391063, + "grad_norm": 0.6531625986099243, + "learning_rate": 2.6926256983240222e-05, + "loss": 0.2094, + "step": 8037 + }, + { + "epoch": 22.452513966480446, + "grad_norm": 1.4065672159194946, + "learning_rate": 2.6929608938547484e-05, + "loss": 0.1486, + "step": 8038 + }, + { + "epoch": 22.455307262569832, + "grad_norm": 0.5743274092674255, + "learning_rate": 2.693296089385475e-05, + "loss": 0.1391, + "step": 8039 + }, + { + "epoch": 22.45810055865922, + "grad_norm": 0.6903788447380066, + "learning_rate": 2.6936312849162013e-05, + "loss": 0.1479, + "step": 8040 + }, + { + "epoch": 22.460893854748605, + "grad_norm": 0.7494829893112183, + "learning_rate": 2.6939664804469274e-05, + "loss": 0.1502, + "step": 8041 + }, + { + "epoch": 22.463687150837988, + "grad_norm": 0.5256777405738831, + "learning_rate": 2.694301675977654e-05, + "loss": 0.1678, + "step": 8042 + }, + { + "epoch": 22.466480446927374, + "grad_norm": 1.0644842386245728, + "learning_rate": 2.69463687150838e-05, + "loss": 0.1604, + "step": 8043 + }, + { + "epoch": 22.46927374301676, + "grad_norm": 1.1446343660354614, + "learning_rate": 2.6949720670391065e-05, + "loss": 0.1404, + "step": 8044 + }, + { + "epoch": 22.472067039106147, + "grad_norm": 1.6263923645019531, + "learning_rate": 2.6953072625698327e-05, + "loss": 0.1492, + "step": 8045 + }, + { + "epoch": 22.47486033519553, + "grad_norm": 1.051964521408081, + "learning_rate": 2.6956424581005588e-05, + "loss": 0.1944, + "step": 8046 + }, + { + "epoch": 22.477653631284916, + "grad_norm": 1.0073049068450928, + "learning_rate": 2.6959776536312853e-05, + "loss": 0.1377, + "step": 8047 + }, + { + "epoch": 22.480446927374302, + "grad_norm": 1.325864553451538, + "learning_rate": 2.6963128491620114e-05, + "loss": 0.1479, + "step": 8048 + }, + { + "epoch": 22.48324022346369, + "grad_norm": 2.3536927700042725, + "learning_rate": 2.6966480446927375e-05, + "loss": 0.1615, + "step": 8049 + }, + { + "epoch": 22.48603351955307, + "grad_norm": 1.3725669384002686, + "learning_rate": 2.696983240223464e-05, + "loss": 0.2256, + "step": 8050 + }, + { + "epoch": 22.488826815642458, + "grad_norm": 1.6911951303482056, + "learning_rate": 2.69731843575419e-05, + "loss": 0.2093, + "step": 8051 + }, + { + "epoch": 22.491620111731844, + "grad_norm": 0.5153712630271912, + "learning_rate": 2.6976536312849163e-05, + "loss": 0.1762, + "step": 8052 + }, + { + "epoch": 22.49441340782123, + "grad_norm": 0.46846574544906616, + "learning_rate": 2.6979888268156428e-05, + "loss": 0.158, + "step": 8053 + }, + { + "epoch": 22.497206703910614, + "grad_norm": 0.5914360284805298, + "learning_rate": 2.698324022346369e-05, + "loss": 0.1546, + "step": 8054 + }, + { + "epoch": 22.5, + "grad_norm": 0.5383375883102417, + "learning_rate": 2.698659217877095e-05, + "loss": 0.1638, + "step": 8055 + }, + { + "epoch": 22.502793296089386, + "grad_norm": 0.578696072101593, + "learning_rate": 2.6989944134078215e-05, + "loss": 0.1621, + "step": 8056 + }, + { + "epoch": 22.505586592178773, + "grad_norm": 0.5124533772468567, + "learning_rate": 2.6993296089385476e-05, + "loss": 0.1796, + "step": 8057 + }, + { + "epoch": 22.508379888268156, + "grad_norm": 0.3518007695674896, + "learning_rate": 2.6996648044692738e-05, + "loss": 0.1465, + "step": 8058 + }, + { + "epoch": 22.511173184357542, + "grad_norm": 0.4778892993927002, + "learning_rate": 2.7000000000000002e-05, + "loss": 0.165, + "step": 8059 + }, + { + "epoch": 22.51396648044693, + "grad_norm": 0.5690687298774719, + "learning_rate": 2.7003351955307264e-05, + "loss": 0.1549, + "step": 8060 + }, + { + "epoch": 22.51675977653631, + "grad_norm": 0.5605872273445129, + "learning_rate": 2.700670391061453e-05, + "loss": 0.1751, + "step": 8061 + }, + { + "epoch": 22.519553072625698, + "grad_norm": 0.586986243724823, + "learning_rate": 2.701005586592179e-05, + "loss": 0.1746, + "step": 8062 + }, + { + "epoch": 22.522346368715084, + "grad_norm": 0.48871952295303345, + "learning_rate": 2.701340782122905e-05, + "loss": 0.1665, + "step": 8063 + }, + { + "epoch": 22.52513966480447, + "grad_norm": 1.9742398262023926, + "learning_rate": 2.7016759776536316e-05, + "loss": 0.1773, + "step": 8064 + }, + { + "epoch": 22.527932960893853, + "grad_norm": 0.5121692419052124, + "learning_rate": 2.7020111731843577e-05, + "loss": 0.1599, + "step": 8065 + }, + { + "epoch": 22.53072625698324, + "grad_norm": 0.8228868246078491, + "learning_rate": 2.702346368715084e-05, + "loss": 0.1935, + "step": 8066 + }, + { + "epoch": 22.533519553072626, + "grad_norm": 1.050992488861084, + "learning_rate": 2.7026815642458103e-05, + "loss": 0.1618, + "step": 8067 + }, + { + "epoch": 22.536312849162012, + "grad_norm": 0.5415428280830383, + "learning_rate": 2.7030167597765365e-05, + "loss": 0.1354, + "step": 8068 + }, + { + "epoch": 22.539106145251395, + "grad_norm": 1.0306841135025024, + "learning_rate": 2.7033519553072626e-05, + "loss": 0.1392, + "step": 8069 + }, + { + "epoch": 22.54189944134078, + "grad_norm": 1.0900558233261108, + "learning_rate": 2.703687150837989e-05, + "loss": 0.158, + "step": 8070 + }, + { + "epoch": 22.544692737430168, + "grad_norm": 0.6733890175819397, + "learning_rate": 2.7040223463687152e-05, + "loss": 0.1655, + "step": 8071 + }, + { + "epoch": 22.547486033519554, + "grad_norm": 1.1381843090057373, + "learning_rate": 2.7043575418994413e-05, + "loss": 0.1966, + "step": 8072 + }, + { + "epoch": 22.550279329608937, + "grad_norm": 4.6218461990356445, + "learning_rate": 2.7046927374301678e-05, + "loss": 0.1528, + "step": 8073 + }, + { + "epoch": 22.553072625698324, + "grad_norm": 1.1191354990005493, + "learning_rate": 2.705027932960894e-05, + "loss": 0.1501, + "step": 8074 + }, + { + "epoch": 22.55586592178771, + "grad_norm": 0.8762264251708984, + "learning_rate": 2.70536312849162e-05, + "loss": 0.2093, + "step": 8075 + }, + { + "epoch": 22.558659217877096, + "grad_norm": 1.5136194229125977, + "learning_rate": 2.7056983240223466e-05, + "loss": 0.2144, + "step": 8076 + }, + { + "epoch": 22.56145251396648, + "grad_norm": 0.4687478840351105, + "learning_rate": 2.7060335195530727e-05, + "loss": 0.1623, + "step": 8077 + }, + { + "epoch": 22.564245810055866, + "grad_norm": 0.40935781598091125, + "learning_rate": 2.7063687150837988e-05, + "loss": 0.1657, + "step": 8078 + }, + { + "epoch": 22.567039106145252, + "grad_norm": 0.4381384253501892, + "learning_rate": 2.7067039106145253e-05, + "loss": 0.176, + "step": 8079 + }, + { + "epoch": 22.56983240223464, + "grad_norm": 0.5606831312179565, + "learning_rate": 2.7070391061452514e-05, + "loss": 0.2109, + "step": 8080 + }, + { + "epoch": 22.57262569832402, + "grad_norm": 0.6534865498542786, + "learning_rate": 2.707374301675978e-05, + "loss": 0.1525, + "step": 8081 + }, + { + "epoch": 22.575418994413408, + "grad_norm": 0.5222630500793457, + "learning_rate": 2.707709497206704e-05, + "loss": 0.1529, + "step": 8082 + }, + { + "epoch": 22.578212290502794, + "grad_norm": 0.5138198733329773, + "learning_rate": 2.7080446927374302e-05, + "loss": 0.1411, + "step": 8083 + }, + { + "epoch": 22.58100558659218, + "grad_norm": 0.6705204844474792, + "learning_rate": 2.7083798882681567e-05, + "loss": 0.2014, + "step": 8084 + }, + { + "epoch": 22.583798882681563, + "grad_norm": 0.802949070930481, + "learning_rate": 2.7087150837988828e-05, + "loss": 0.1276, + "step": 8085 + }, + { + "epoch": 22.58659217877095, + "grad_norm": 1.5442326068878174, + "learning_rate": 2.709050279329609e-05, + "loss": 0.1929, + "step": 8086 + }, + { + "epoch": 22.589385474860336, + "grad_norm": 0.5781466960906982, + "learning_rate": 2.7093854748603354e-05, + "loss": 0.1423, + "step": 8087 + }, + { + "epoch": 22.592178770949722, + "grad_norm": 0.4301846921443939, + "learning_rate": 2.7097206703910615e-05, + "loss": 0.1227, + "step": 8088 + }, + { + "epoch": 22.594972067039105, + "grad_norm": 0.73747718334198, + "learning_rate": 2.7100558659217877e-05, + "loss": 0.1543, + "step": 8089 + }, + { + "epoch": 22.59776536312849, + "grad_norm": 0.8354029655456543, + "learning_rate": 2.710391061452514e-05, + "loss": 0.1666, + "step": 8090 + }, + { + "epoch": 22.600558659217878, + "grad_norm": 0.649728000164032, + "learning_rate": 2.7107262569832403e-05, + "loss": 0.1411, + "step": 8091 + }, + { + "epoch": 22.60335195530726, + "grad_norm": 0.6639249324798584, + "learning_rate": 2.7110614525139664e-05, + "loss": 0.1554, + "step": 8092 + }, + { + "epoch": 22.606145251396647, + "grad_norm": 0.6148180365562439, + "learning_rate": 2.711396648044693e-05, + "loss": 0.1751, + "step": 8093 + }, + { + "epoch": 22.608938547486034, + "grad_norm": 0.8091557621955872, + "learning_rate": 2.711731843575419e-05, + "loss": 0.2284, + "step": 8094 + }, + { + "epoch": 22.61173184357542, + "grad_norm": 0.9655852913856506, + "learning_rate": 2.712067039106145e-05, + "loss": 0.1722, + "step": 8095 + }, + { + "epoch": 22.614525139664803, + "grad_norm": 1.5337556600570679, + "learning_rate": 2.7124022346368716e-05, + "loss": 0.1586, + "step": 8096 + }, + { + "epoch": 22.61731843575419, + "grad_norm": 0.8638274669647217, + "learning_rate": 2.7127374301675978e-05, + "loss": 0.165, + "step": 8097 + }, + { + "epoch": 22.620111731843576, + "grad_norm": 1.44559907913208, + "learning_rate": 2.713072625698324e-05, + "loss": 0.1575, + "step": 8098 + }, + { + "epoch": 22.622905027932962, + "grad_norm": 0.8365350365638733, + "learning_rate": 2.7134078212290504e-05, + "loss": 0.1639, + "step": 8099 + }, + { + "epoch": 22.625698324022345, + "grad_norm": 7.047539234161377, + "learning_rate": 2.7137430167597765e-05, + "loss": 0.1958, + "step": 8100 + }, + { + "epoch": 22.62849162011173, + "grad_norm": 2.0491721630096436, + "learning_rate": 2.714078212290503e-05, + "loss": 0.1876, + "step": 8101 + }, + { + "epoch": 22.631284916201118, + "grad_norm": 0.44019272923469543, + "learning_rate": 2.714413407821229e-05, + "loss": 0.1781, + "step": 8102 + }, + { + "epoch": 22.634078212290504, + "grad_norm": 0.5284161567687988, + "learning_rate": 2.7147486033519552e-05, + "loss": 0.18, + "step": 8103 + }, + { + "epoch": 22.636871508379887, + "grad_norm": 0.5631770491600037, + "learning_rate": 2.7150837988826817e-05, + "loss": 0.1389, + "step": 8104 + }, + { + "epoch": 22.639664804469273, + "grad_norm": 0.7956151366233826, + "learning_rate": 2.715418994413408e-05, + "loss": 0.1734, + "step": 8105 + }, + { + "epoch": 22.64245810055866, + "grad_norm": 0.5730973482131958, + "learning_rate": 2.715754189944134e-05, + "loss": 0.1873, + "step": 8106 + }, + { + "epoch": 22.645251396648046, + "grad_norm": 0.8273904919624329, + "learning_rate": 2.7160893854748605e-05, + "loss": 0.1757, + "step": 8107 + }, + { + "epoch": 22.64804469273743, + "grad_norm": 0.5929660201072693, + "learning_rate": 2.7164245810055866e-05, + "loss": 0.1582, + "step": 8108 + }, + { + "epoch": 22.650837988826815, + "grad_norm": 0.6267151236534119, + "learning_rate": 2.7167597765363127e-05, + "loss": 0.1694, + "step": 8109 + }, + { + "epoch": 22.6536312849162, + "grad_norm": 0.9056423902511597, + "learning_rate": 2.7170949720670392e-05, + "loss": 0.2023, + "step": 8110 + }, + { + "epoch": 22.656424581005588, + "grad_norm": 0.8291394114494324, + "learning_rate": 2.7174301675977653e-05, + "loss": 0.1725, + "step": 8111 + }, + { + "epoch": 22.65921787709497, + "grad_norm": 0.6988617777824402, + "learning_rate": 2.7177653631284915e-05, + "loss": 0.1395, + "step": 8112 + }, + { + "epoch": 22.662011173184357, + "grad_norm": 0.6284992694854736, + "learning_rate": 2.718100558659218e-05, + "loss": 0.2018, + "step": 8113 + }, + { + "epoch": 22.664804469273744, + "grad_norm": 0.7747178077697754, + "learning_rate": 2.718435754189944e-05, + "loss": 0.15, + "step": 8114 + }, + { + "epoch": 22.66759776536313, + "grad_norm": 0.8998814225196838, + "learning_rate": 2.7187709497206702e-05, + "loss": 0.1656, + "step": 8115 + }, + { + "epoch": 22.670391061452513, + "grad_norm": 0.6636459231376648, + "learning_rate": 2.7191061452513967e-05, + "loss": 0.1397, + "step": 8116 + }, + { + "epoch": 22.6731843575419, + "grad_norm": 1.5182644128799438, + "learning_rate": 2.7194413407821228e-05, + "loss": 0.1759, + "step": 8117 + }, + { + "epoch": 22.675977653631286, + "grad_norm": 0.8766854405403137, + "learning_rate": 2.7197765363128493e-05, + "loss": 0.1741, + "step": 8118 + }, + { + "epoch": 22.678770949720672, + "grad_norm": 1.2920781373977661, + "learning_rate": 2.7201117318435754e-05, + "loss": 0.171, + "step": 8119 + }, + { + "epoch": 22.681564245810055, + "grad_norm": 0.730093240737915, + "learning_rate": 2.7204469273743016e-05, + "loss": 0.1352, + "step": 8120 + }, + { + "epoch": 22.68435754189944, + "grad_norm": 2.0650126934051514, + "learning_rate": 2.720782122905028e-05, + "loss": 0.1649, + "step": 8121 + }, + { + "epoch": 22.687150837988828, + "grad_norm": 1.1069514751434326, + "learning_rate": 2.721117318435754e-05, + "loss": 0.1533, + "step": 8122 + }, + { + "epoch": 22.689944134078214, + "grad_norm": 1.1128751039505005, + "learning_rate": 2.7214525139664803e-05, + "loss": 0.1812, + "step": 8123 + }, + { + "epoch": 22.692737430167597, + "grad_norm": 0.84516841173172, + "learning_rate": 2.7217877094972068e-05, + "loss": 0.1746, + "step": 8124 + }, + { + "epoch": 22.695530726256983, + "grad_norm": 1.157692790031433, + "learning_rate": 2.722122905027933e-05, + "loss": 0.1721, + "step": 8125 + }, + { + "epoch": 22.69832402234637, + "grad_norm": 2.1414313316345215, + "learning_rate": 2.722458100558659e-05, + "loss": 0.2424, + "step": 8126 + }, + { + "epoch": 22.701117318435756, + "grad_norm": 0.5983080863952637, + "learning_rate": 2.7227932960893855e-05, + "loss": 0.1613, + "step": 8127 + }, + { + "epoch": 22.70391061452514, + "grad_norm": 0.45787423849105835, + "learning_rate": 2.7231284916201117e-05, + "loss": 0.1781, + "step": 8128 + }, + { + "epoch": 22.706703910614525, + "grad_norm": 1.0753670930862427, + "learning_rate": 2.7234636871508378e-05, + "loss": 0.1912, + "step": 8129 + }, + { + "epoch": 22.70949720670391, + "grad_norm": 0.5587481260299683, + "learning_rate": 2.7237988826815643e-05, + "loss": 0.1812, + "step": 8130 + }, + { + "epoch": 22.712290502793294, + "grad_norm": 1.203225016593933, + "learning_rate": 2.7241340782122904e-05, + "loss": 0.1559, + "step": 8131 + }, + { + "epoch": 22.71508379888268, + "grad_norm": 0.5577345490455627, + "learning_rate": 2.7244692737430165e-05, + "loss": 0.1611, + "step": 8132 + }, + { + "epoch": 22.717877094972067, + "grad_norm": 0.7509884834289551, + "learning_rate": 2.724804469273743e-05, + "loss": 0.1585, + "step": 8133 + }, + { + "epoch": 22.720670391061454, + "grad_norm": 0.4699751138687134, + "learning_rate": 2.725139664804469e-05, + "loss": 0.154, + "step": 8134 + }, + { + "epoch": 22.723463687150836, + "grad_norm": 0.6746963262557983, + "learning_rate": 2.7254748603351953e-05, + "loss": 0.1565, + "step": 8135 + }, + { + "epoch": 22.726256983240223, + "grad_norm": 1.8994220495224, + "learning_rate": 2.7258100558659217e-05, + "loss": 0.1674, + "step": 8136 + }, + { + "epoch": 22.72905027932961, + "grad_norm": 0.7517855167388916, + "learning_rate": 2.7261452513966482e-05, + "loss": 0.1688, + "step": 8137 + }, + { + "epoch": 22.731843575418996, + "grad_norm": 1.0036289691925049, + "learning_rate": 2.7264804469273747e-05, + "loss": 0.1494, + "step": 8138 + }, + { + "epoch": 22.73463687150838, + "grad_norm": 0.7565425038337708, + "learning_rate": 2.7268156424581008e-05, + "loss": 0.1689, + "step": 8139 + }, + { + "epoch": 22.737430167597765, + "grad_norm": 1.193384051322937, + "learning_rate": 2.727150837988827e-05, + "loss": 0.1587, + "step": 8140 + }, + { + "epoch": 22.74022346368715, + "grad_norm": 2.1872520446777344, + "learning_rate": 2.7274860335195534e-05, + "loss": 0.163, + "step": 8141 + }, + { + "epoch": 22.743016759776538, + "grad_norm": 0.6246809363365173, + "learning_rate": 2.7278212290502796e-05, + "loss": 0.136, + "step": 8142 + }, + { + "epoch": 22.74581005586592, + "grad_norm": 0.7556729912757874, + "learning_rate": 2.7281564245810057e-05, + "loss": 0.1759, + "step": 8143 + }, + { + "epoch": 22.748603351955307, + "grad_norm": 0.5653927326202393, + "learning_rate": 2.7284916201117322e-05, + "loss": 0.1385, + "step": 8144 + }, + { + "epoch": 22.751396648044693, + "grad_norm": 0.8717997074127197, + "learning_rate": 2.7288268156424583e-05, + "loss": 0.1681, + "step": 8145 + }, + { + "epoch": 22.75418994413408, + "grad_norm": 0.7813356518745422, + "learning_rate": 2.7291620111731844e-05, + "loss": 0.137, + "step": 8146 + }, + { + "epoch": 22.756983240223462, + "grad_norm": 1.172320008277893, + "learning_rate": 2.729497206703911e-05, + "loss": 0.204, + "step": 8147 + }, + { + "epoch": 22.75977653631285, + "grad_norm": 1.0103176832199097, + "learning_rate": 2.729832402234637e-05, + "loss": 0.1618, + "step": 8148 + }, + { + "epoch": 22.762569832402235, + "grad_norm": 1.2112774848937988, + "learning_rate": 2.7301675977653632e-05, + "loss": 0.2057, + "step": 8149 + }, + { + "epoch": 22.76536312849162, + "grad_norm": 1.3975750207901, + "learning_rate": 2.7305027932960897e-05, + "loss": 0.1816, + "step": 8150 + }, + { + "epoch": 22.768156424581004, + "grad_norm": 1.8716868162155151, + "learning_rate": 2.7308379888268158e-05, + "loss": 0.228, + "step": 8151 + }, + { + "epoch": 22.77094972067039, + "grad_norm": 0.4762079417705536, + "learning_rate": 2.731173184357542e-05, + "loss": 0.1601, + "step": 8152 + }, + { + "epoch": 22.773743016759777, + "grad_norm": 1.3883670568466187, + "learning_rate": 2.7315083798882684e-05, + "loss": 0.198, + "step": 8153 + }, + { + "epoch": 22.776536312849164, + "grad_norm": 0.6603648662567139, + "learning_rate": 2.7318435754189945e-05, + "loss": 0.1735, + "step": 8154 + }, + { + "epoch": 22.779329608938546, + "grad_norm": 0.553817093372345, + "learning_rate": 2.732178770949721e-05, + "loss": 0.1928, + "step": 8155 + }, + { + "epoch": 22.782122905027933, + "grad_norm": 0.5719030499458313, + "learning_rate": 2.732513966480447e-05, + "loss": 0.1647, + "step": 8156 + }, + { + "epoch": 22.78491620111732, + "grad_norm": 0.4838067889213562, + "learning_rate": 2.7328491620111733e-05, + "loss": 0.1631, + "step": 8157 + }, + { + "epoch": 22.787709497206706, + "grad_norm": 0.446083128452301, + "learning_rate": 2.7331843575418998e-05, + "loss": 0.1668, + "step": 8158 + }, + { + "epoch": 22.79050279329609, + "grad_norm": 0.8296679854393005, + "learning_rate": 2.733519553072626e-05, + "loss": 0.154, + "step": 8159 + }, + { + "epoch": 22.793296089385475, + "grad_norm": 0.6725515723228455, + "learning_rate": 2.733854748603352e-05, + "loss": 0.155, + "step": 8160 + }, + { + "epoch": 22.79608938547486, + "grad_norm": 0.4534919559955597, + "learning_rate": 2.7341899441340785e-05, + "loss": 0.1808, + "step": 8161 + }, + { + "epoch": 22.798882681564244, + "grad_norm": 0.6563641428947449, + "learning_rate": 2.7345251396648046e-05, + "loss": 0.1634, + "step": 8162 + }, + { + "epoch": 22.80167597765363, + "grad_norm": 0.8542342782020569, + "learning_rate": 2.7348603351955308e-05, + "loss": 0.1535, + "step": 8163 + }, + { + "epoch": 22.804469273743017, + "grad_norm": 0.6780477166175842, + "learning_rate": 2.7351955307262572e-05, + "loss": 0.1814, + "step": 8164 + }, + { + "epoch": 22.807262569832403, + "grad_norm": 0.6139163374900818, + "learning_rate": 2.7355307262569834e-05, + "loss": 0.1476, + "step": 8165 + }, + { + "epoch": 22.810055865921786, + "grad_norm": 0.6608590483665466, + "learning_rate": 2.7358659217877095e-05, + "loss": 0.1698, + "step": 8166 + }, + { + "epoch": 22.812849162011172, + "grad_norm": 0.6666707992553711, + "learning_rate": 2.736201117318436e-05, + "loss": 0.1364, + "step": 8167 + }, + { + "epoch": 22.81564245810056, + "grad_norm": 0.5666492581367493, + "learning_rate": 2.736536312849162e-05, + "loss": 0.1659, + "step": 8168 + }, + { + "epoch": 22.818435754189945, + "grad_norm": 0.8156938552856445, + "learning_rate": 2.7368715083798882e-05, + "loss": 0.1796, + "step": 8169 + }, + { + "epoch": 22.821229050279328, + "grad_norm": 1.480959177017212, + "learning_rate": 2.7372067039106147e-05, + "loss": 0.1293, + "step": 8170 + }, + { + "epoch": 22.824022346368714, + "grad_norm": 1.046487808227539, + "learning_rate": 2.737541899441341e-05, + "loss": 0.1634, + "step": 8171 + }, + { + "epoch": 22.8268156424581, + "grad_norm": 0.8561245799064636, + "learning_rate": 2.737877094972067e-05, + "loss": 0.1747, + "step": 8172 + }, + { + "epoch": 22.829608938547487, + "grad_norm": 0.7338825464248657, + "learning_rate": 2.7382122905027935e-05, + "loss": 0.1777, + "step": 8173 + }, + { + "epoch": 22.83240223463687, + "grad_norm": 0.8703963160514832, + "learning_rate": 2.7385474860335196e-05, + "loss": 0.1503, + "step": 8174 + }, + { + "epoch": 22.835195530726256, + "grad_norm": 1.411960482597351, + "learning_rate": 2.738882681564246e-05, + "loss": 0.1845, + "step": 8175 + }, + { + "epoch": 22.837988826815643, + "grad_norm": 3.5941972732543945, + "learning_rate": 2.7392178770949722e-05, + "loss": 0.2665, + "step": 8176 + }, + { + "epoch": 22.84078212290503, + "grad_norm": 0.6007026433944702, + "learning_rate": 2.7395530726256983e-05, + "loss": 0.1954, + "step": 8177 + }, + { + "epoch": 22.843575418994412, + "grad_norm": 0.42631542682647705, + "learning_rate": 2.7398882681564248e-05, + "loss": 0.1569, + "step": 8178 + }, + { + "epoch": 22.8463687150838, + "grad_norm": 1.408673882484436, + "learning_rate": 2.740223463687151e-05, + "loss": 0.17, + "step": 8179 + }, + { + "epoch": 22.849162011173185, + "grad_norm": 0.707449197769165, + "learning_rate": 2.740558659217877e-05, + "loss": 0.1613, + "step": 8180 + }, + { + "epoch": 22.85195530726257, + "grad_norm": 0.5053158402442932, + "learning_rate": 2.7408938547486036e-05, + "loss": 0.1681, + "step": 8181 + }, + { + "epoch": 22.854748603351954, + "grad_norm": 0.5763833522796631, + "learning_rate": 2.7412290502793297e-05, + "loss": 0.1671, + "step": 8182 + }, + { + "epoch": 22.85754189944134, + "grad_norm": 0.5124368667602539, + "learning_rate": 2.7415642458100558e-05, + "loss": 0.1851, + "step": 8183 + }, + { + "epoch": 22.860335195530727, + "grad_norm": 1.314307689666748, + "learning_rate": 2.7418994413407823e-05, + "loss": 0.178, + "step": 8184 + }, + { + "epoch": 22.863128491620113, + "grad_norm": 0.8020538091659546, + "learning_rate": 2.7422346368715084e-05, + "loss": 0.1866, + "step": 8185 + }, + { + "epoch": 22.865921787709496, + "grad_norm": 0.6609870791435242, + "learning_rate": 2.7425698324022346e-05, + "loss": 0.1576, + "step": 8186 + }, + { + "epoch": 22.868715083798882, + "grad_norm": 0.5649736523628235, + "learning_rate": 2.742905027932961e-05, + "loss": 0.1807, + "step": 8187 + }, + { + "epoch": 22.87150837988827, + "grad_norm": 0.6708289980888367, + "learning_rate": 2.7432402234636872e-05, + "loss": 0.1678, + "step": 8188 + }, + { + "epoch": 22.874301675977655, + "grad_norm": 0.7368927001953125, + "learning_rate": 2.7435754189944133e-05, + "loss": 0.1662, + "step": 8189 + }, + { + "epoch": 22.877094972067038, + "grad_norm": 0.9275481104850769, + "learning_rate": 2.7439106145251398e-05, + "loss": 0.1628, + "step": 8190 + }, + { + "epoch": 22.879888268156424, + "grad_norm": 0.6255871653556824, + "learning_rate": 2.744245810055866e-05, + "loss": 0.1398, + "step": 8191 + }, + { + "epoch": 22.88268156424581, + "grad_norm": 1.0364327430725098, + "learning_rate": 2.744581005586592e-05, + "loss": 0.1659, + "step": 8192 + }, + { + "epoch": 22.885474860335197, + "grad_norm": 0.5886762738227844, + "learning_rate": 2.7449162011173185e-05, + "loss": 0.1523, + "step": 8193 + }, + { + "epoch": 22.88826815642458, + "grad_norm": 1.2509371042251587, + "learning_rate": 2.7452513966480447e-05, + "loss": 0.1655, + "step": 8194 + }, + { + "epoch": 22.891061452513966, + "grad_norm": 1.6123626232147217, + "learning_rate": 2.745586592178771e-05, + "loss": 0.1765, + "step": 8195 + }, + { + "epoch": 22.893854748603353, + "grad_norm": 1.37398362159729, + "learning_rate": 2.7459217877094973e-05, + "loss": 0.1678, + "step": 8196 + }, + { + "epoch": 22.89664804469274, + "grad_norm": 0.735478401184082, + "learning_rate": 2.7462569832402234e-05, + "loss": 0.1375, + "step": 8197 + }, + { + "epoch": 22.899441340782122, + "grad_norm": 0.9359827041625977, + "learning_rate": 2.74659217877095e-05, + "loss": 0.1546, + "step": 8198 + }, + { + "epoch": 22.90223463687151, + "grad_norm": 0.9659619331359863, + "learning_rate": 2.746927374301676e-05, + "loss": 0.1621, + "step": 8199 + }, + { + "epoch": 22.905027932960895, + "grad_norm": 1.2498114109039307, + "learning_rate": 2.747262569832402e-05, + "loss": 0.1909, + "step": 8200 + }, + { + "epoch": 22.907821229050278, + "grad_norm": 1.6757324934005737, + "learning_rate": 2.7475977653631286e-05, + "loss": 0.2198, + "step": 8201 + }, + { + "epoch": 22.910614525139664, + "grad_norm": 0.48202449083328247, + "learning_rate": 2.7479329608938548e-05, + "loss": 0.1815, + "step": 8202 + }, + { + "epoch": 22.91340782122905, + "grad_norm": 0.7887973785400391, + "learning_rate": 2.748268156424581e-05, + "loss": 0.1869, + "step": 8203 + }, + { + "epoch": 22.916201117318437, + "grad_norm": 0.7327297329902649, + "learning_rate": 2.7486033519553074e-05, + "loss": 0.1765, + "step": 8204 + }, + { + "epoch": 22.91899441340782, + "grad_norm": 1.5500210523605347, + "learning_rate": 2.7489385474860335e-05, + "loss": 0.1526, + "step": 8205 + }, + { + "epoch": 22.921787709497206, + "grad_norm": 0.46983322501182556, + "learning_rate": 2.7492737430167596e-05, + "loss": 0.1808, + "step": 8206 + }, + { + "epoch": 22.924581005586592, + "grad_norm": 0.9832980036735535, + "learning_rate": 2.749608938547486e-05, + "loss": 0.1792, + "step": 8207 + }, + { + "epoch": 22.92737430167598, + "grad_norm": 0.6934425234794617, + "learning_rate": 2.7499441340782122e-05, + "loss": 0.1767, + "step": 8208 + }, + { + "epoch": 22.93016759776536, + "grad_norm": 1.5572909116744995, + "learning_rate": 2.7502793296089384e-05, + "loss": 0.1787, + "step": 8209 + }, + { + "epoch": 22.932960893854748, + "grad_norm": 1.0113829374313354, + "learning_rate": 2.750614525139665e-05, + "loss": 0.1453, + "step": 8210 + }, + { + "epoch": 22.935754189944134, + "grad_norm": 0.7610708475112915, + "learning_rate": 2.750949720670391e-05, + "loss": 0.1698, + "step": 8211 + }, + { + "epoch": 22.93854748603352, + "grad_norm": 0.9768586158752441, + "learning_rate": 2.7512849162011175e-05, + "loss": 0.1784, + "step": 8212 + }, + { + "epoch": 22.941340782122904, + "grad_norm": 0.9546639323234558, + "learning_rate": 2.7516201117318436e-05, + "loss": 0.1482, + "step": 8213 + }, + { + "epoch": 22.94413407821229, + "grad_norm": 0.8512697219848633, + "learning_rate": 2.7519553072625697e-05, + "loss": 0.1566, + "step": 8214 + }, + { + "epoch": 22.946927374301676, + "grad_norm": 0.9509080648422241, + "learning_rate": 2.7522905027932962e-05, + "loss": 0.1544, + "step": 8215 + }, + { + "epoch": 22.949720670391063, + "grad_norm": Infinity, + "learning_rate": 2.7522905027932962e-05, + "loss": 0.1721, + "step": 8216 + }, + { + "epoch": 22.952513966480446, + "grad_norm": 0.7834169268608093, + "learning_rate": 2.7526256983240223e-05, + "loss": 0.1612, + "step": 8217 + }, + { + "epoch": 22.955307262569832, + "grad_norm": 1.1801906824111938, + "learning_rate": 2.7529608938547485e-05, + "loss": 0.16, + "step": 8218 + }, + { + "epoch": 22.95810055865922, + "grad_norm": 1.0501708984375, + "learning_rate": 2.753296089385475e-05, + "loss": 0.1402, + "step": 8219 + }, + { + "epoch": 22.960893854748605, + "grad_norm": 0.7384483814239502, + "learning_rate": 2.753631284916201e-05, + "loss": 0.1499, + "step": 8220 + }, + { + "epoch": 22.963687150837988, + "grad_norm": 2.2157630920410156, + "learning_rate": 2.7539664804469272e-05, + "loss": 0.1486, + "step": 8221 + }, + { + "epoch": 22.966480446927374, + "grad_norm": 0.7371834516525269, + "learning_rate": 2.7543016759776537e-05, + "loss": 0.1649, + "step": 8222 + }, + { + "epoch": 22.96927374301676, + "grad_norm": 0.9268589615821838, + "learning_rate": 2.7546368715083798e-05, + "loss": 0.1632, + "step": 8223 + }, + { + "epoch": 22.972067039106147, + "grad_norm": 1.2294124364852905, + "learning_rate": 2.754972067039106e-05, + "loss": 0.1509, + "step": 8224 + }, + { + "epoch": 22.97486033519553, + "grad_norm": 1.162361741065979, + "learning_rate": 2.7553072625698324e-05, + "loss": 0.1836, + "step": 8225 + }, + { + "epoch": 22.977653631284916, + "grad_norm": 1.7212977409362793, + "learning_rate": 2.7556424581005586e-05, + "loss": 0.2658, + "step": 8226 + }, + { + "epoch": 22.980446927374302, + "grad_norm": 0.47551682591438293, + "learning_rate": 2.7559776536312847e-05, + "loss": 0.1721, + "step": 8227 + }, + { + "epoch": 22.98324022346369, + "grad_norm": 0.6439698338508606, + "learning_rate": 2.756312849162011e-05, + "loss": 0.172, + "step": 8228 + }, + { + "epoch": 22.98603351955307, + "grad_norm": 0.8736734986305237, + "learning_rate": 2.7566480446927373e-05, + "loss": 0.1655, + "step": 8229 + }, + { + "epoch": 22.988826815642458, + "grad_norm": 9.790510177612305, + "learning_rate": 2.7569832402234634e-05, + "loss": 0.1874, + "step": 8230 + }, + { + "epoch": 22.991620111731844, + "grad_norm": 0.8083372116088867, + "learning_rate": 2.75731843575419e-05, + "loss": 0.1604, + "step": 8231 + }, + { + "epoch": 22.994413407821227, + "grad_norm": 0.984838604927063, + "learning_rate": 2.757653631284916e-05, + "loss": 0.2101, + "step": 8232 + }, + { + "epoch": 22.997206703910614, + "grad_norm": 1.3086717128753662, + "learning_rate": 2.7579888268156425e-05, + "loss": 0.17, + "step": 8233 + }, + { + "epoch": 23.0, + "grad_norm": 1.5257940292358398, + "learning_rate": 2.758324022346369e-05, + "loss": 0.2064, + "step": 8234 + }, + { + "epoch": 23.002793296089386, + "grad_norm": 0.4993387758731842, + "learning_rate": 2.758659217877095e-05, + "loss": 0.1475, + "step": 8235 + }, + { + "epoch": 23.00558659217877, + "grad_norm": 0.5970611572265625, + "learning_rate": 2.7589944134078216e-05, + "loss": 0.1755, + "step": 8236 + }, + { + "epoch": 23.008379888268156, + "grad_norm": 0.5083764791488647, + "learning_rate": 2.7593296089385477e-05, + "loss": 0.1452, + "step": 8237 + }, + { + "epoch": 23.011173184357542, + "grad_norm": 0.6344144940376282, + "learning_rate": 2.759664804469274e-05, + "loss": 0.1632, + "step": 8238 + }, + { + "epoch": 23.01396648044693, + "grad_norm": 0.6044331192970276, + "learning_rate": 2.7600000000000003e-05, + "loss": 0.1697, + "step": 8239 + }, + { + "epoch": 23.01675977653631, + "grad_norm": 0.562511682510376, + "learning_rate": 2.7603351955307265e-05, + "loss": 0.1527, + "step": 8240 + }, + { + "epoch": 23.019553072625698, + "grad_norm": 1.9142109155654907, + "learning_rate": 2.7606703910614526e-05, + "loss": 0.164, + "step": 8241 + }, + { + "epoch": 23.022346368715084, + "grad_norm": 0.45197275280952454, + "learning_rate": 2.761005586592179e-05, + "loss": 0.1666, + "step": 8242 + }, + { + "epoch": 23.02513966480447, + "grad_norm": 0.6826145648956299, + "learning_rate": 2.7613407821229052e-05, + "loss": 0.162, + "step": 8243 + }, + { + "epoch": 23.027932960893853, + "grad_norm": 0.6830765604972839, + "learning_rate": 2.7616759776536314e-05, + "loss": 0.177, + "step": 8244 + }, + { + "epoch": 23.03072625698324, + "grad_norm": 0.6487274169921875, + "learning_rate": 2.7620111731843578e-05, + "loss": 0.1355, + "step": 8245 + }, + { + "epoch": 23.033519553072626, + "grad_norm": 0.8027136325836182, + "learning_rate": 2.762346368715084e-05, + "loss": 0.1539, + "step": 8246 + }, + { + "epoch": 23.036312849162012, + "grad_norm": 0.9419196844100952, + "learning_rate": 2.76268156424581e-05, + "loss": 0.1557, + "step": 8247 + }, + { + "epoch": 23.039106145251395, + "grad_norm": 0.43646037578582764, + "learning_rate": 2.7630167597765366e-05, + "loss": 0.1618, + "step": 8248 + }, + { + "epoch": 23.04189944134078, + "grad_norm": 0.5927084684371948, + "learning_rate": 2.7633519553072627e-05, + "loss": 0.1305, + "step": 8249 + }, + { + "epoch": 23.044692737430168, + "grad_norm": 0.8411561250686646, + "learning_rate": 2.7636871508379892e-05, + "loss": 0.1526, + "step": 8250 + }, + { + "epoch": 23.047486033519554, + "grad_norm": 1.0116937160491943, + "learning_rate": 2.7640223463687153e-05, + "loss": 0.1638, + "step": 8251 + }, + { + "epoch": 23.050279329608937, + "grad_norm": 0.6732376217842102, + "learning_rate": 2.7643575418994414e-05, + "loss": 0.1375, + "step": 8252 + }, + { + "epoch": 23.053072625698324, + "grad_norm": 1.025994062423706, + "learning_rate": 2.764692737430168e-05, + "loss": 0.1645, + "step": 8253 + }, + { + "epoch": 23.05586592178771, + "grad_norm": 1.3189808130264282, + "learning_rate": 2.765027932960894e-05, + "loss": 0.1757, + "step": 8254 + }, + { + "epoch": 23.058659217877096, + "grad_norm": 0.846630871295929, + "learning_rate": 2.7653631284916202e-05, + "loss": 0.1919, + "step": 8255 + }, + { + "epoch": 23.06145251396648, + "grad_norm": 0.8973835110664368, + "learning_rate": 2.7656983240223467e-05, + "loss": 0.1431, + "step": 8256 + }, + { + "epoch": 23.064245810055866, + "grad_norm": 0.7983786463737488, + "learning_rate": 2.7660335195530728e-05, + "loss": 0.1834, + "step": 8257 + }, + { + "epoch": 23.067039106145252, + "grad_norm": 1.4022127389907837, + "learning_rate": 2.766368715083799e-05, + "loss": 0.184, + "step": 8258 + }, + { + "epoch": 23.06983240223464, + "grad_norm": 1.3512821197509766, + "learning_rate": 2.7667039106145254e-05, + "loss": 0.22, + "step": 8259 + }, + { + "epoch": 23.07262569832402, + "grad_norm": 0.606860876083374, + "learning_rate": 2.7670391061452515e-05, + "loss": 0.1876, + "step": 8260 + }, + { + "epoch": 23.075418994413408, + "grad_norm": 0.5478688478469849, + "learning_rate": 2.7673743016759777e-05, + "loss": 0.1775, + "step": 8261 + }, + { + "epoch": 23.078212290502794, + "grad_norm": 0.45565545558929443, + "learning_rate": 2.767709497206704e-05, + "loss": 0.1421, + "step": 8262 + }, + { + "epoch": 23.08100558659218, + "grad_norm": 0.6802995204925537, + "learning_rate": 2.7680446927374303e-05, + "loss": 0.1679, + "step": 8263 + }, + { + "epoch": 23.083798882681563, + "grad_norm": 0.7113327383995056, + "learning_rate": 2.7683798882681564e-05, + "loss": 0.16, + "step": 8264 + }, + { + "epoch": 23.08659217877095, + "grad_norm": 0.614973247051239, + "learning_rate": 2.768715083798883e-05, + "loss": 0.1648, + "step": 8265 + }, + { + "epoch": 23.089385474860336, + "grad_norm": 1.3611236810684204, + "learning_rate": 2.769050279329609e-05, + "loss": 0.1786, + "step": 8266 + }, + { + "epoch": 23.092178770949722, + "grad_norm": 0.9463368058204651, + "learning_rate": 2.769385474860335e-05, + "loss": 0.1379, + "step": 8267 + }, + { + "epoch": 23.094972067039105, + "grad_norm": 0.4757532477378845, + "learning_rate": 2.7697206703910616e-05, + "loss": 0.1457, + "step": 8268 + }, + { + "epoch": 23.09776536312849, + "grad_norm": 0.6496172547340393, + "learning_rate": 2.7700558659217878e-05, + "loss": 0.1669, + "step": 8269 + }, + { + "epoch": 23.100558659217878, + "grad_norm": 1.8657981157302856, + "learning_rate": 2.7703910614525142e-05, + "loss": 0.1689, + "step": 8270 + }, + { + "epoch": 23.10335195530726, + "grad_norm": 0.7579984068870544, + "learning_rate": 2.7707262569832404e-05, + "loss": 0.1394, + "step": 8271 + }, + { + "epoch": 23.106145251396647, + "grad_norm": 1.6667025089263916, + "learning_rate": 2.7710614525139665e-05, + "loss": 0.1447, + "step": 8272 + }, + { + "epoch": 23.108938547486034, + "grad_norm": 0.687347412109375, + "learning_rate": 2.771396648044693e-05, + "loss": 0.1918, + "step": 8273 + }, + { + "epoch": 23.11173184357542, + "grad_norm": 0.6662672758102417, + "learning_rate": 2.771731843575419e-05, + "loss": 0.13, + "step": 8274 + }, + { + "epoch": 23.114525139664803, + "grad_norm": 0.7999485731124878, + "learning_rate": 2.7720670391061452e-05, + "loss": 0.1369, + "step": 8275 + }, + { + "epoch": 23.11731843575419, + "grad_norm": 0.9198694229125977, + "learning_rate": 2.7724022346368717e-05, + "loss": 0.1437, + "step": 8276 + }, + { + "epoch": 23.120111731843576, + "grad_norm": 1.4203752279281616, + "learning_rate": 2.772737430167598e-05, + "loss": 0.1442, + "step": 8277 + }, + { + "epoch": 23.122905027932962, + "grad_norm": 1.6847628355026245, + "learning_rate": 2.773072625698324e-05, + "loss": 0.1848, + "step": 8278 + }, + { + "epoch": 23.125698324022345, + "grad_norm": 1.100326657295227, + "learning_rate": 2.7734078212290505e-05, + "loss": 0.172, + "step": 8279 + }, + { + "epoch": 23.12849162011173, + "grad_norm": 1.3424084186553955, + "learning_rate": 2.7737430167597766e-05, + "loss": 0.1744, + "step": 8280 + }, + { + "epoch": 23.131284916201118, + "grad_norm": 0.6818543672561646, + "learning_rate": 2.7740782122905027e-05, + "loss": 0.1378, + "step": 8281 + }, + { + "epoch": 23.134078212290504, + "grad_norm": 1.1649537086486816, + "learning_rate": 2.7744134078212292e-05, + "loss": 0.154, + "step": 8282 + }, + { + "epoch": 23.136871508379887, + "grad_norm": 0.979681134223938, + "learning_rate": 2.7747486033519553e-05, + "loss": 0.1705, + "step": 8283 + }, + { + "epoch": 23.139664804469273, + "grad_norm": 1.4674488306045532, + "learning_rate": 2.7750837988826815e-05, + "loss": 0.2118, + "step": 8284 + }, + { + "epoch": 23.14245810055866, + "grad_norm": 0.4259682595729828, + "learning_rate": 2.775418994413408e-05, + "loss": 0.1667, + "step": 8285 + }, + { + "epoch": 23.145251396648046, + "grad_norm": 0.3658018112182617, + "learning_rate": 2.775754189944134e-05, + "loss": 0.1683, + "step": 8286 + }, + { + "epoch": 23.14804469273743, + "grad_norm": 0.4307640492916107, + "learning_rate": 2.7760893854748602e-05, + "loss": 0.182, + "step": 8287 + }, + { + "epoch": 23.150837988826815, + "grad_norm": 0.6479714512825012, + "learning_rate": 2.7764245810055867e-05, + "loss": 0.1651, + "step": 8288 + }, + { + "epoch": 23.1536312849162, + "grad_norm": 0.5905746817588806, + "learning_rate": 2.7767597765363128e-05, + "loss": 0.1556, + "step": 8289 + }, + { + "epoch": 23.156424581005588, + "grad_norm": 0.4545576572418213, + "learning_rate": 2.7770949720670393e-05, + "loss": 0.1408, + "step": 8290 + }, + { + "epoch": 23.15921787709497, + "grad_norm": 0.5117279887199402, + "learning_rate": 2.7774301675977654e-05, + "loss": 0.1681, + "step": 8291 + }, + { + "epoch": 23.162011173184357, + "grad_norm": 0.5683522820472717, + "learning_rate": 2.7777653631284916e-05, + "loss": 0.1494, + "step": 8292 + }, + { + "epoch": 23.164804469273744, + "grad_norm": 0.5834077000617981, + "learning_rate": 2.778100558659218e-05, + "loss": 0.1462, + "step": 8293 + }, + { + "epoch": 23.16759776536313, + "grad_norm": 1.2753914594650269, + "learning_rate": 2.7784357541899442e-05, + "loss": 0.1704, + "step": 8294 + }, + { + "epoch": 23.170391061452513, + "grad_norm": 0.5014309883117676, + "learning_rate": 2.7787709497206703e-05, + "loss": 0.1573, + "step": 8295 + }, + { + "epoch": 23.1731843575419, + "grad_norm": 4.6094746589660645, + "learning_rate": 2.7791061452513968e-05, + "loss": 0.1318, + "step": 8296 + }, + { + "epoch": 23.175977653631286, + "grad_norm": 0.5683315992355347, + "learning_rate": 2.779441340782123e-05, + "loss": 0.1304, + "step": 8297 + }, + { + "epoch": 23.178770949720672, + "grad_norm": 0.5506522059440613, + "learning_rate": 2.779776536312849e-05, + "loss": 0.1454, + "step": 8298 + }, + { + "epoch": 23.181564245810055, + "grad_norm": 0.758265495300293, + "learning_rate": 2.7801117318435755e-05, + "loss": 0.1241, + "step": 8299 + }, + { + "epoch": 23.18435754189944, + "grad_norm": 0.6459625363349915, + "learning_rate": 2.7804469273743017e-05, + "loss": 0.1222, + "step": 8300 + }, + { + "epoch": 23.187150837988828, + "grad_norm": 0.6446770429611206, + "learning_rate": 2.7807821229050278e-05, + "loss": 0.1449, + "step": 8301 + }, + { + "epoch": 23.189944134078214, + "grad_norm": 0.727302610874176, + "learning_rate": 2.7811173184357543e-05, + "loss": 0.1763, + "step": 8302 + }, + { + "epoch": 23.192737430167597, + "grad_norm": 1.7462918758392334, + "learning_rate": 2.7814525139664804e-05, + "loss": 0.1473, + "step": 8303 + }, + { + "epoch": 23.195530726256983, + "grad_norm": 0.7358245253562927, + "learning_rate": 2.7817877094972065e-05, + "loss": 0.1651, + "step": 8304 + }, + { + "epoch": 23.19832402234637, + "grad_norm": 2.550828456878662, + "learning_rate": 2.782122905027933e-05, + "loss": 0.1362, + "step": 8305 + }, + { + "epoch": 23.201117318435756, + "grad_norm": 0.6770806312561035, + "learning_rate": 2.782458100558659e-05, + "loss": 0.1509, + "step": 8306 + }, + { + "epoch": 23.20391061452514, + "grad_norm": 0.9030964374542236, + "learning_rate": 2.7827932960893856e-05, + "loss": 0.1524, + "step": 8307 + }, + { + "epoch": 23.206703910614525, + "grad_norm": 1.9979761838912964, + "learning_rate": 2.7831284916201118e-05, + "loss": 0.1861, + "step": 8308 + }, + { + "epoch": 23.20949720670391, + "grad_norm": 1.5619280338287354, + "learning_rate": 2.783463687150838e-05, + "loss": 0.2205, + "step": 8309 + }, + { + "epoch": 23.212290502793294, + "grad_norm": 0.6094651222229004, + "learning_rate": 2.7837988826815644e-05, + "loss": 0.1657, + "step": 8310 + }, + { + "epoch": 23.21508379888268, + "grad_norm": 0.6509326100349426, + "learning_rate": 2.7841340782122905e-05, + "loss": 0.171, + "step": 8311 + }, + { + "epoch": 23.217877094972067, + "grad_norm": 0.6108382940292358, + "learning_rate": 2.7844692737430166e-05, + "loss": 0.1607, + "step": 8312 + }, + { + "epoch": 23.220670391061454, + "grad_norm": 0.6103011965751648, + "learning_rate": 2.784804469273743e-05, + "loss": 0.1793, + "step": 8313 + }, + { + "epoch": 23.223463687150836, + "grad_norm": 0.5065670013427734, + "learning_rate": 2.7851396648044692e-05, + "loss": 0.1557, + "step": 8314 + }, + { + "epoch": 23.226256983240223, + "grad_norm": 0.7821817398071289, + "learning_rate": 2.7854748603351954e-05, + "loss": 0.1512, + "step": 8315 + }, + { + "epoch": 23.22905027932961, + "grad_norm": 0.5630698204040527, + "learning_rate": 2.785810055865922e-05, + "loss": 0.1611, + "step": 8316 + }, + { + "epoch": 23.231843575418996, + "grad_norm": 0.5731140971183777, + "learning_rate": 2.786145251396648e-05, + "loss": 0.1344, + "step": 8317 + }, + { + "epoch": 23.23463687150838, + "grad_norm": 0.47886550426483154, + "learning_rate": 2.786480446927374e-05, + "loss": 0.1445, + "step": 8318 + }, + { + "epoch": 23.237430167597765, + "grad_norm": 0.5447503924369812, + "learning_rate": 2.7868156424581006e-05, + "loss": 0.1326, + "step": 8319 + }, + { + "epoch": 23.24022346368715, + "grad_norm": 0.5589126348495483, + "learning_rate": 2.7871508379888267e-05, + "loss": 0.1644, + "step": 8320 + }, + { + "epoch": 23.243016759776538, + "grad_norm": 0.6207829117774963, + "learning_rate": 2.787486033519553e-05, + "loss": 0.1449, + "step": 8321 + }, + { + "epoch": 23.24581005586592, + "grad_norm": 0.650687038898468, + "learning_rate": 2.7878212290502793e-05, + "loss": 0.136, + "step": 8322 + }, + { + "epoch": 23.248603351955307, + "grad_norm": 0.7519351840019226, + "learning_rate": 2.7881564245810055e-05, + "loss": 0.1264, + "step": 8323 + }, + { + "epoch": 23.251396648044693, + "grad_norm": 0.8076945543289185, + "learning_rate": 2.7884916201117316e-05, + "loss": 0.148, + "step": 8324 + }, + { + "epoch": 23.25418994413408, + "grad_norm": 0.532965362071991, + "learning_rate": 2.788826815642458e-05, + "loss": 0.1393, + "step": 8325 + }, + { + "epoch": 23.256983240223462, + "grad_norm": 1.0560709238052368, + "learning_rate": 2.7891620111731842e-05, + "loss": 0.1272, + "step": 8326 + }, + { + "epoch": 23.25977653631285, + "grad_norm": 0.7042596936225891, + "learning_rate": 2.7894972067039107e-05, + "loss": 0.1546, + "step": 8327 + }, + { + "epoch": 23.262569832402235, + "grad_norm": 0.6225483417510986, + "learning_rate": 2.7898324022346368e-05, + "loss": 0.1462, + "step": 8328 + }, + { + "epoch": 23.26536312849162, + "grad_norm": 1.0720832347869873, + "learning_rate": 2.790167597765363e-05, + "loss": 0.1294, + "step": 8329 + }, + { + "epoch": 23.268156424581004, + "grad_norm": 0.7438483238220215, + "learning_rate": 2.7905027932960894e-05, + "loss": 0.1623, + "step": 8330 + }, + { + "epoch": 23.27094972067039, + "grad_norm": 1.0244909524917603, + "learning_rate": 2.790837988826816e-05, + "loss": 0.1714, + "step": 8331 + }, + { + "epoch": 23.273743016759777, + "grad_norm": 1.0812582969665527, + "learning_rate": 2.791173184357542e-05, + "loss": 0.1524, + "step": 8332 + }, + { + "epoch": 23.276536312849164, + "grad_norm": 0.6911754012107849, + "learning_rate": 2.7915083798882685e-05, + "loss": 0.1384, + "step": 8333 + }, + { + "epoch": 23.279329608938546, + "grad_norm": 2.1513311862945557, + "learning_rate": 2.7918435754189946e-05, + "loss": 0.1984, + "step": 8334 + }, + { + "epoch": 23.282122905027933, + "grad_norm": 0.9908499121665955, + "learning_rate": 2.7921787709497208e-05, + "loss": 0.1488, + "step": 8335 + }, + { + "epoch": 23.28491620111732, + "grad_norm": 0.5384692549705505, + "learning_rate": 2.7925139664804472e-05, + "loss": 0.1586, + "step": 8336 + }, + { + "epoch": 23.287709497206706, + "grad_norm": 1.6557960510253906, + "learning_rate": 2.7928491620111734e-05, + "loss": 0.1683, + "step": 8337 + }, + { + "epoch": 23.29050279329609, + "grad_norm": 0.6216103434562683, + "learning_rate": 2.7931843575418995e-05, + "loss": 0.1722, + "step": 8338 + }, + { + "epoch": 23.293296089385475, + "grad_norm": 0.7971453666687012, + "learning_rate": 2.793519553072626e-05, + "loss": 0.1918, + "step": 8339 + }, + { + "epoch": 23.29608938547486, + "grad_norm": 0.5888506174087524, + "learning_rate": 2.793854748603352e-05, + "loss": 0.1663, + "step": 8340 + }, + { + "epoch": 23.298882681564244, + "grad_norm": 0.773284375667572, + "learning_rate": 2.7941899441340783e-05, + "loss": 0.1577, + "step": 8341 + }, + { + "epoch": 23.30167597765363, + "grad_norm": 3.0588483810424805, + "learning_rate": 2.7945251396648047e-05, + "loss": 0.155, + "step": 8342 + }, + { + "epoch": 23.304469273743017, + "grad_norm": 0.7112618684768677, + "learning_rate": 2.794860335195531e-05, + "loss": 0.148, + "step": 8343 + }, + { + "epoch": 23.307262569832403, + "grad_norm": 0.7883642315864563, + "learning_rate": 2.7951955307262573e-05, + "loss": 0.1681, + "step": 8344 + }, + { + "epoch": 23.310055865921786, + "grad_norm": 0.8427278399467468, + "learning_rate": 2.7955307262569835e-05, + "loss": 0.1613, + "step": 8345 + }, + { + "epoch": 23.312849162011172, + "grad_norm": 0.5973252058029175, + "learning_rate": 2.7958659217877096e-05, + "loss": 0.1633, + "step": 8346 + }, + { + "epoch": 23.31564245810056, + "grad_norm": 0.6556832194328308, + "learning_rate": 2.796201117318436e-05, + "loss": 0.1761, + "step": 8347 + }, + { + "epoch": 23.318435754189945, + "grad_norm": 0.7691040635108948, + "learning_rate": 2.7965363128491622e-05, + "loss": 0.1375, + "step": 8348 + }, + { + "epoch": 23.321229050279328, + "grad_norm": 1.2183581590652466, + "learning_rate": 2.7968715083798884e-05, + "loss": 0.1558, + "step": 8349 + }, + { + "epoch": 23.324022346368714, + "grad_norm": 0.660189688205719, + "learning_rate": 2.7972067039106148e-05, + "loss": 0.1683, + "step": 8350 + }, + { + "epoch": 23.3268156424581, + "grad_norm": 0.9618346095085144, + "learning_rate": 2.797541899441341e-05, + "loss": 0.1416, + "step": 8351 + }, + { + "epoch": 23.329608938547487, + "grad_norm": 1.2816177606582642, + "learning_rate": 2.797877094972067e-05, + "loss": 0.1512, + "step": 8352 + }, + { + "epoch": 23.33240223463687, + "grad_norm": 0.9719724059104919, + "learning_rate": 2.7982122905027936e-05, + "loss": 0.1282, + "step": 8353 + }, + { + "epoch": 23.335195530726256, + "grad_norm": 0.8878934383392334, + "learning_rate": 2.7985474860335197e-05, + "loss": 0.1588, + "step": 8354 + }, + { + "epoch": 23.337988826815643, + "grad_norm": 0.8023217916488647, + "learning_rate": 2.798882681564246e-05, + "loss": 0.1291, + "step": 8355 + }, + { + "epoch": 23.34078212290503, + "grad_norm": 0.8013315200805664, + "learning_rate": 2.7992178770949723e-05, + "loss": 0.1436, + "step": 8356 + }, + { + "epoch": 23.343575418994412, + "grad_norm": 0.7932252287864685, + "learning_rate": 2.7995530726256984e-05, + "loss": 0.1382, + "step": 8357 + }, + { + "epoch": 23.3463687150838, + "grad_norm": 9.181938171386719, + "learning_rate": 2.7998882681564246e-05, + "loss": 0.1828, + "step": 8358 + }, + { + "epoch": 23.349162011173185, + "grad_norm": 1.6422390937805176, + "learning_rate": 2.800223463687151e-05, + "loss": 0.2069, + "step": 8359 + }, + { + "epoch": 23.35195530726257, + "grad_norm": 0.5259152054786682, + "learning_rate": 2.8005586592178772e-05, + "loss": 0.1529, + "step": 8360 + }, + { + "epoch": 23.354748603351954, + "grad_norm": 0.5223890542984009, + "learning_rate": 2.8008938547486033e-05, + "loss": 0.1683, + "step": 8361 + }, + { + "epoch": 23.35754189944134, + "grad_norm": 0.6987924575805664, + "learning_rate": 2.8012290502793298e-05, + "loss": 0.1875, + "step": 8362 + }, + { + "epoch": 23.360335195530727, + "grad_norm": 0.5029965043067932, + "learning_rate": 2.801564245810056e-05, + "loss": 0.1564, + "step": 8363 + }, + { + "epoch": 23.363128491620113, + "grad_norm": 0.974287211894989, + "learning_rate": 2.8018994413407824e-05, + "loss": 0.1392, + "step": 8364 + }, + { + "epoch": 23.365921787709496, + "grad_norm": 0.7684655785560608, + "learning_rate": 2.8022346368715085e-05, + "loss": 0.1833, + "step": 8365 + }, + { + "epoch": 23.368715083798882, + "grad_norm": 0.5615746378898621, + "learning_rate": 2.8025698324022347e-05, + "loss": 0.1634, + "step": 8366 + }, + { + "epoch": 23.37150837988827, + "grad_norm": 1.0188912153244019, + "learning_rate": 2.802905027932961e-05, + "loss": 0.1563, + "step": 8367 + }, + { + "epoch": 23.374301675977655, + "grad_norm": 1.1772767305374146, + "learning_rate": 2.8032402234636873e-05, + "loss": 0.1321, + "step": 8368 + }, + { + "epoch": 23.377094972067038, + "grad_norm": 0.8661177158355713, + "learning_rate": 2.8035754189944134e-05, + "loss": 0.148, + "step": 8369 + }, + { + "epoch": 23.379888268156424, + "grad_norm": 0.7041946649551392, + "learning_rate": 2.80391061452514e-05, + "loss": 0.1475, + "step": 8370 + }, + { + "epoch": 23.38268156424581, + "grad_norm": 0.7026709318161011, + "learning_rate": 2.804245810055866e-05, + "loss": 0.1929, + "step": 8371 + }, + { + "epoch": 23.385474860335197, + "grad_norm": 1.4014673233032227, + "learning_rate": 2.804581005586592e-05, + "loss": 0.1438, + "step": 8372 + }, + { + "epoch": 23.38826815642458, + "grad_norm": 0.8070633411407471, + "learning_rate": 2.8049162011173186e-05, + "loss": 0.1619, + "step": 8373 + }, + { + "epoch": 23.391061452513966, + "grad_norm": 0.7178633809089661, + "learning_rate": 2.8052513966480448e-05, + "loss": 0.1356, + "step": 8374 + }, + { + "epoch": 23.393854748603353, + "grad_norm": 0.6724171042442322, + "learning_rate": 2.805586592178771e-05, + "loss": 0.1584, + "step": 8375 + }, + { + "epoch": 23.39664804469274, + "grad_norm": 0.7470574378967285, + "learning_rate": 2.8059217877094974e-05, + "loss": 0.1778, + "step": 8376 + }, + { + "epoch": 23.399441340782122, + "grad_norm": 1.2169073820114136, + "learning_rate": 2.8062569832402235e-05, + "loss": 0.1542, + "step": 8377 + }, + { + "epoch": 23.40223463687151, + "grad_norm": 1.0145456790924072, + "learning_rate": 2.8065921787709496e-05, + "loss": 0.134, + "step": 8378 + }, + { + "epoch": 23.405027932960895, + "grad_norm": 0.6263037919998169, + "learning_rate": 2.806927374301676e-05, + "loss": 0.1621, + "step": 8379 + }, + { + "epoch": 23.407821229050278, + "grad_norm": 2.807863235473633, + "learning_rate": 2.8072625698324022e-05, + "loss": 0.168, + "step": 8380 + }, + { + "epoch": 23.410614525139664, + "grad_norm": 0.9047666788101196, + "learning_rate": 2.8075977653631284e-05, + "loss": 0.1501, + "step": 8381 + }, + { + "epoch": 23.41340782122905, + "grad_norm": 1.0654468536376953, + "learning_rate": 2.807932960893855e-05, + "loss": 0.1857, + "step": 8382 + }, + { + "epoch": 23.416201117318437, + "grad_norm": 1.2783920764923096, + "learning_rate": 2.808268156424581e-05, + "loss": 0.1756, + "step": 8383 + }, + { + "epoch": 23.41899441340782, + "grad_norm": 1.366176962852478, + "learning_rate": 2.8086033519553075e-05, + "loss": 0.1992, + "step": 8384 + }, + { + "epoch": 23.421787709497206, + "grad_norm": 0.4462856948375702, + "learning_rate": 2.8089385474860336e-05, + "loss": 0.1682, + "step": 8385 + }, + { + "epoch": 23.424581005586592, + "grad_norm": 0.5486913919448853, + "learning_rate": 2.8092737430167597e-05, + "loss": 0.1516, + "step": 8386 + }, + { + "epoch": 23.42737430167598, + "grad_norm": 0.632377028465271, + "learning_rate": 2.8096089385474862e-05, + "loss": 0.2018, + "step": 8387 + }, + { + "epoch": 23.43016759776536, + "grad_norm": 0.3838822841644287, + "learning_rate": 2.8099441340782123e-05, + "loss": 0.1538, + "step": 8388 + }, + { + "epoch": 23.432960893854748, + "grad_norm": 0.5435609817504883, + "learning_rate": 2.8102793296089385e-05, + "loss": 0.2024, + "step": 8389 + }, + { + "epoch": 23.435754189944134, + "grad_norm": 0.5977042317390442, + "learning_rate": 2.810614525139665e-05, + "loss": 0.1765, + "step": 8390 + }, + { + "epoch": 23.43854748603352, + "grad_norm": 0.4949083626270294, + "learning_rate": 2.810949720670391e-05, + "loss": 0.1566, + "step": 8391 + }, + { + "epoch": 23.441340782122904, + "grad_norm": 0.84336918592453, + "learning_rate": 2.8112849162011172e-05, + "loss": 0.1876, + "step": 8392 + }, + { + "epoch": 23.44413407821229, + "grad_norm": 0.6138424873352051, + "learning_rate": 2.8116201117318437e-05, + "loss": 0.1596, + "step": 8393 + }, + { + "epoch": 23.446927374301676, + "grad_norm": 0.7786412835121155, + "learning_rate": 2.8119553072625698e-05, + "loss": 0.1724, + "step": 8394 + }, + { + "epoch": 23.449720670391063, + "grad_norm": 0.667448103427887, + "learning_rate": 2.812290502793296e-05, + "loss": 0.1347, + "step": 8395 + }, + { + "epoch": 23.452513966480446, + "grad_norm": 0.7007943391799927, + "learning_rate": 2.8126256983240224e-05, + "loss": 0.1798, + "step": 8396 + }, + { + "epoch": 23.455307262569832, + "grad_norm": 0.8806066513061523, + "learning_rate": 2.8129608938547486e-05, + "loss": 0.129, + "step": 8397 + }, + { + "epoch": 23.45810055865922, + "grad_norm": 0.6882654428482056, + "learning_rate": 2.8132960893854747e-05, + "loss": 0.1507, + "step": 8398 + }, + { + "epoch": 23.460893854748605, + "grad_norm": 0.7211576104164124, + "learning_rate": 2.8136312849162012e-05, + "loss": 0.1881, + "step": 8399 + }, + { + "epoch": 23.463687150837988, + "grad_norm": 0.7117528915405273, + "learning_rate": 2.8139664804469273e-05, + "loss": 0.1439, + "step": 8400 + }, + { + "epoch": 23.466480446927374, + "grad_norm": 0.8596707582473755, + "learning_rate": 2.8143016759776538e-05, + "loss": 0.1406, + "step": 8401 + }, + { + "epoch": 23.46927374301676, + "grad_norm": 0.6506999135017395, + "learning_rate": 2.81463687150838e-05, + "loss": 0.1519, + "step": 8402 + }, + { + "epoch": 23.472067039106147, + "grad_norm": 4.49125862121582, + "learning_rate": 2.814972067039106e-05, + "loss": 0.1617, + "step": 8403 + }, + { + "epoch": 23.47486033519553, + "grad_norm": 0.7575588822364807, + "learning_rate": 2.8153072625698325e-05, + "loss": 0.1438, + "step": 8404 + }, + { + "epoch": 23.477653631284916, + "grad_norm": 1.2977269887924194, + "learning_rate": 2.8156424581005587e-05, + "loss": 0.1705, + "step": 8405 + }, + { + "epoch": 23.480446927374302, + "grad_norm": 1.4212923049926758, + "learning_rate": 2.8159776536312848e-05, + "loss": 0.1524, + "step": 8406 + }, + { + "epoch": 23.48324022346369, + "grad_norm": 0.722599983215332, + "learning_rate": 2.8163128491620113e-05, + "loss": 0.1419, + "step": 8407 + }, + { + "epoch": 23.48603351955307, + "grad_norm": 0.8221191167831421, + "learning_rate": 2.8166480446927374e-05, + "loss": 0.1738, + "step": 8408 + }, + { + "epoch": 23.488826815642458, + "grad_norm": 2.3358354568481445, + "learning_rate": 2.8169832402234635e-05, + "loss": 0.2253, + "step": 8409 + }, + { + "epoch": 23.491620111731844, + "grad_norm": 0.9933751225471497, + "learning_rate": 2.81731843575419e-05, + "loss": 0.1682, + "step": 8410 + }, + { + "epoch": 23.49441340782123, + "grad_norm": 0.44258108735084534, + "learning_rate": 2.817653631284916e-05, + "loss": 0.146, + "step": 8411 + }, + { + "epoch": 23.497206703910614, + "grad_norm": 0.4029732346534729, + "learning_rate": 2.8179888268156423e-05, + "loss": 0.162, + "step": 8412 + }, + { + "epoch": 23.5, + "grad_norm": 0.7378058433532715, + "learning_rate": 2.8183240223463688e-05, + "loss": 0.1379, + "step": 8413 + }, + { + "epoch": 23.502793296089386, + "grad_norm": 0.7702720165252686, + "learning_rate": 2.818659217877095e-05, + "loss": 0.1545, + "step": 8414 + }, + { + "epoch": 23.505586592178773, + "grad_norm": 0.5525854229927063, + "learning_rate": 2.818994413407821e-05, + "loss": 0.1812, + "step": 8415 + }, + { + "epoch": 23.508379888268156, + "grad_norm": 0.6488432884216309, + "learning_rate": 2.8193296089385475e-05, + "loss": 0.1669, + "step": 8416 + }, + { + "epoch": 23.511173184357542, + "grad_norm": 0.4792442321777344, + "learning_rate": 2.8196648044692736e-05, + "loss": 0.171, + "step": 8417 + }, + { + "epoch": 23.51396648044693, + "grad_norm": 0.4549412429332733, + "learning_rate": 2.8199999999999998e-05, + "loss": 0.1693, + "step": 8418 + }, + { + "epoch": 23.51675977653631, + "grad_norm": 0.5082424879074097, + "learning_rate": 2.8203351955307262e-05, + "loss": 0.1607, + "step": 8419 + }, + { + "epoch": 23.519553072625698, + "grad_norm": 0.6946405172348022, + "learning_rate": 2.8206703910614524e-05, + "loss": 0.1524, + "step": 8420 + }, + { + "epoch": 23.522346368715084, + "grad_norm": 0.5879225134849548, + "learning_rate": 2.821005586592179e-05, + "loss": 0.1499, + "step": 8421 + }, + { + "epoch": 23.52513966480447, + "grad_norm": 0.7558637261390686, + "learning_rate": 2.821340782122905e-05, + "loss": 0.1637, + "step": 8422 + }, + { + "epoch": 23.527932960893853, + "grad_norm": 0.9675734639167786, + "learning_rate": 2.821675977653631e-05, + "loss": 0.1812, + "step": 8423 + }, + { + "epoch": 23.53072625698324, + "grad_norm": 1.387252688407898, + "learning_rate": 2.8220111731843576e-05, + "loss": 0.1339, + "step": 8424 + }, + { + "epoch": 23.533519553072626, + "grad_norm": 0.6842560172080994, + "learning_rate": 2.8223463687150837e-05, + "loss": 0.1589, + "step": 8425 + }, + { + "epoch": 23.536312849162012, + "grad_norm": 0.5776386260986328, + "learning_rate": 2.8226815642458102e-05, + "loss": 0.1604, + "step": 8426 + }, + { + "epoch": 23.539106145251395, + "grad_norm": 0.8097773790359497, + "learning_rate": 2.8230167597765367e-05, + "loss": 0.1571, + "step": 8427 + }, + { + "epoch": 23.54189944134078, + "grad_norm": 0.7223955988883972, + "learning_rate": 2.8233519553072628e-05, + "loss": 0.1499, + "step": 8428 + }, + { + "epoch": 23.544692737430168, + "grad_norm": 0.9186801314353943, + "learning_rate": 2.823687150837989e-05, + "loss": 0.1653, + "step": 8429 + }, + { + "epoch": 23.547486033519554, + "grad_norm": 3.70755672454834, + "learning_rate": 2.8240223463687154e-05, + "loss": 0.133, + "step": 8430 + }, + { + "epoch": 23.550279329608937, + "grad_norm": 4.378608703613281, + "learning_rate": 2.8243575418994415e-05, + "loss": 0.1522, + "step": 8431 + }, + { + "epoch": 23.553072625698324, + "grad_norm": 1.4965442419052124, + "learning_rate": 2.8246927374301677e-05, + "loss": 0.1558, + "step": 8432 + }, + { + "epoch": 23.55586592178771, + "grad_norm": 1.789857268333435, + "learning_rate": 2.825027932960894e-05, + "loss": 0.2026, + "step": 8433 + }, + { + "epoch": 23.558659217877096, + "grad_norm": 1.7814948558807373, + "learning_rate": 2.8253631284916203e-05, + "loss": 0.2639, + "step": 8434 + }, + { + "epoch": 23.56145251396648, + "grad_norm": 0.6304988265037537, + "learning_rate": 2.8256983240223464e-05, + "loss": 0.1807, + "step": 8435 + }, + { + "epoch": 23.564245810055866, + "grad_norm": 0.40230169892311096, + "learning_rate": 2.826033519553073e-05, + "loss": 0.166, + "step": 8436 + }, + { + "epoch": 23.567039106145252, + "grad_norm": 1.4804346561431885, + "learning_rate": 2.826368715083799e-05, + "loss": 0.1585, + "step": 8437 + }, + { + "epoch": 23.56983240223464, + "grad_norm": 0.5572969913482666, + "learning_rate": 2.8267039106145255e-05, + "loss": 0.1806, + "step": 8438 + }, + { + "epoch": 23.57262569832402, + "grad_norm": 0.4724295735359192, + "learning_rate": 2.8270391061452516e-05, + "loss": 0.2028, + "step": 8439 + }, + { + "epoch": 23.575418994413408, + "grad_norm": 0.8079828023910522, + "learning_rate": 2.8273743016759778e-05, + "loss": 0.1438, + "step": 8440 + }, + { + "epoch": 23.578212290502794, + "grad_norm": 0.6676899790763855, + "learning_rate": 2.8277094972067042e-05, + "loss": 0.1486, + "step": 8441 + }, + { + "epoch": 23.58100558659218, + "grad_norm": 0.6974025368690491, + "learning_rate": 2.8280446927374304e-05, + "loss": 0.1933, + "step": 8442 + }, + { + "epoch": 23.583798882681563, + "grad_norm": 0.6108065247535706, + "learning_rate": 2.8283798882681565e-05, + "loss": 0.1655, + "step": 8443 + }, + { + "epoch": 23.58659217877095, + "grad_norm": 2.782561779022217, + "learning_rate": 2.828715083798883e-05, + "loss": 0.1894, + "step": 8444 + }, + { + "epoch": 23.589385474860336, + "grad_norm": 0.5515769124031067, + "learning_rate": 2.829050279329609e-05, + "loss": 0.1571, + "step": 8445 + }, + { + "epoch": 23.592178770949722, + "grad_norm": 0.6486025452613831, + "learning_rate": 2.8293854748603353e-05, + "loss": 0.1735, + "step": 8446 + }, + { + "epoch": 23.594972067039105, + "grad_norm": 0.6005235910415649, + "learning_rate": 2.8297206703910617e-05, + "loss": 0.1509, + "step": 8447 + }, + { + "epoch": 23.59776536312849, + "grad_norm": 0.6495254635810852, + "learning_rate": 2.830055865921788e-05, + "loss": 0.1604, + "step": 8448 + }, + { + "epoch": 23.600558659217878, + "grad_norm": 0.6412640810012817, + "learning_rate": 2.830391061452514e-05, + "loss": 0.1197, + "step": 8449 + }, + { + "epoch": 23.60335195530726, + "grad_norm": 1.076393485069275, + "learning_rate": 2.8307262569832405e-05, + "loss": 0.1402, + "step": 8450 + }, + { + "epoch": 23.606145251396647, + "grad_norm": 0.4434557259082794, + "learning_rate": 2.8310614525139666e-05, + "loss": 0.1538, + "step": 8451 + }, + { + "epoch": 23.608938547486034, + "grad_norm": 1.4378268718719482, + "learning_rate": 2.8313966480446927e-05, + "loss": 0.1668, + "step": 8452 + }, + { + "epoch": 23.61173184357542, + "grad_norm": 0.7364915609359741, + "learning_rate": 2.8317318435754192e-05, + "loss": 0.1428, + "step": 8453 + }, + { + "epoch": 23.614525139664803, + "grad_norm": 0.9414236545562744, + "learning_rate": 2.8320670391061454e-05, + "loss": 0.1609, + "step": 8454 + }, + { + "epoch": 23.61731843575419, + "grad_norm": 0.9828576445579529, + "learning_rate": 2.8324022346368715e-05, + "loss": 0.1989, + "step": 8455 + }, + { + "epoch": 23.620111731843576, + "grad_norm": 2.0491766929626465, + "learning_rate": 2.832737430167598e-05, + "loss": 0.139, + "step": 8456 + }, + { + "epoch": 23.622905027932962, + "grad_norm": 1.1560415029525757, + "learning_rate": 2.833072625698324e-05, + "loss": 0.1843, + "step": 8457 + }, + { + "epoch": 23.625698324022345, + "grad_norm": 1.1868535280227661, + "learning_rate": 2.8334078212290506e-05, + "loss": 0.1667, + "step": 8458 + }, + { + "epoch": 23.62849162011173, + "grad_norm": 1.644941806793213, + "learning_rate": 2.8337430167597767e-05, + "loss": 0.2431, + "step": 8459 + }, + { + "epoch": 23.631284916201118, + "grad_norm": 0.5112360119819641, + "learning_rate": 2.834078212290503e-05, + "loss": 0.1951, + "step": 8460 + }, + { + "epoch": 23.634078212290504, + "grad_norm": 0.46078723669052124, + "learning_rate": 2.8344134078212293e-05, + "loss": 0.1693, + "step": 8461 + }, + { + "epoch": 23.636871508379887, + "grad_norm": 0.539656400680542, + "learning_rate": 2.8347486033519554e-05, + "loss": 0.1799, + "step": 8462 + }, + { + "epoch": 23.639664804469273, + "grad_norm": 0.7588765025138855, + "learning_rate": 2.8350837988826816e-05, + "loss": 0.1703, + "step": 8463 + }, + { + "epoch": 23.64245810055866, + "grad_norm": 0.7561090588569641, + "learning_rate": 2.835418994413408e-05, + "loss": 0.2022, + "step": 8464 + }, + { + "epoch": 23.645251396648046, + "grad_norm": 1.2508628368377686, + "learning_rate": 2.8357541899441342e-05, + "loss": 0.1498, + "step": 8465 + }, + { + "epoch": 23.64804469273743, + "grad_norm": 0.5754123330116272, + "learning_rate": 2.8360893854748603e-05, + "loss": 0.1406, + "step": 8466 + }, + { + "epoch": 23.650837988826815, + "grad_norm": 0.42238515615463257, + "learning_rate": 2.8364245810055868e-05, + "loss": 0.161, + "step": 8467 + }, + { + "epoch": 23.6536312849162, + "grad_norm": 0.661138117313385, + "learning_rate": 2.836759776536313e-05, + "loss": 0.1539, + "step": 8468 + }, + { + "epoch": 23.656424581005588, + "grad_norm": 0.9941235184669495, + "learning_rate": 2.837094972067039e-05, + "loss": 0.19, + "step": 8469 + }, + { + "epoch": 23.65921787709497, + "grad_norm": 0.6868406534194946, + "learning_rate": 2.8374301675977655e-05, + "loss": 0.1439, + "step": 8470 + }, + { + "epoch": 23.662011173184357, + "grad_norm": 0.7833080887794495, + "learning_rate": 2.8377653631284917e-05, + "loss": 0.1551, + "step": 8471 + }, + { + "epoch": 23.664804469273744, + "grad_norm": 0.6937018632888794, + "learning_rate": 2.8381005586592178e-05, + "loss": 0.1491, + "step": 8472 + }, + { + "epoch": 23.66759776536313, + "grad_norm": 1.6566030979156494, + "learning_rate": 2.8384357541899443e-05, + "loss": 0.1312, + "step": 8473 + }, + { + "epoch": 23.670391061452513, + "grad_norm": 1.3836653232574463, + "learning_rate": 2.8387709497206704e-05, + "loss": 0.1886, + "step": 8474 + }, + { + "epoch": 23.6731843575419, + "grad_norm": 0.5535262823104858, + "learning_rate": 2.8391061452513965e-05, + "loss": 0.1423, + "step": 8475 + }, + { + "epoch": 23.675977653631286, + "grad_norm": 0.6766240000724792, + "learning_rate": 2.839441340782123e-05, + "loss": 0.1708, + "step": 8476 + }, + { + "epoch": 23.678770949720672, + "grad_norm": 1.1001018285751343, + "learning_rate": 2.839776536312849e-05, + "loss": 0.1638, + "step": 8477 + }, + { + "epoch": 23.681564245810055, + "grad_norm": 0.6722769141197205, + "learning_rate": 2.8401117318435756e-05, + "loss": 0.1718, + "step": 8478 + }, + { + "epoch": 23.68435754189944, + "grad_norm": 0.9078830480575562, + "learning_rate": 2.8404469273743018e-05, + "loss": 0.1557, + "step": 8479 + }, + { + "epoch": 23.687150837988828, + "grad_norm": 0.9059826731681824, + "learning_rate": 2.840782122905028e-05, + "loss": 0.1356, + "step": 8480 + }, + { + "epoch": 23.689944134078214, + "grad_norm": 0.9968307018280029, + "learning_rate": 2.8411173184357544e-05, + "loss": 0.1462, + "step": 8481 + }, + { + "epoch": 23.692737430167597, + "grad_norm": 0.603996753692627, + "learning_rate": 2.8414525139664805e-05, + "loss": 0.1534, + "step": 8482 + }, + { + "epoch": 23.695530726256983, + "grad_norm": 1.0691603422164917, + "learning_rate": 2.8417877094972066e-05, + "loss": 0.1915, + "step": 8483 + }, + { + "epoch": 23.69832402234637, + "grad_norm": 2.326662063598633, + "learning_rate": 2.842122905027933e-05, + "loss": 0.2435, + "step": 8484 + }, + { + "epoch": 23.701117318435756, + "grad_norm": 0.963589608669281, + "learning_rate": 2.8424581005586593e-05, + "loss": 0.2478, + "step": 8485 + }, + { + "epoch": 23.70391061452514, + "grad_norm": 1.1823402643203735, + "learning_rate": 2.8427932960893854e-05, + "loss": 0.1871, + "step": 8486 + }, + { + "epoch": 23.706703910614525, + "grad_norm": 0.44112396240234375, + "learning_rate": 2.843128491620112e-05, + "loss": 0.1822, + "step": 8487 + }, + { + "epoch": 23.70949720670391, + "grad_norm": 0.4484405517578125, + "learning_rate": 2.843463687150838e-05, + "loss": 0.1556, + "step": 8488 + }, + { + "epoch": 23.712290502793294, + "grad_norm": 0.5197588205337524, + "learning_rate": 2.843798882681564e-05, + "loss": 0.1464, + "step": 8489 + }, + { + "epoch": 23.71508379888268, + "grad_norm": 1.2803113460540771, + "learning_rate": 2.8441340782122906e-05, + "loss": 0.1729, + "step": 8490 + }, + { + "epoch": 23.717877094972067, + "grad_norm": 0.6767655611038208, + "learning_rate": 2.8444692737430167e-05, + "loss": 0.1676, + "step": 8491 + }, + { + "epoch": 23.720670391061454, + "grad_norm": 0.4887826442718506, + "learning_rate": 2.844804469273743e-05, + "loss": 0.1664, + "step": 8492 + }, + { + "epoch": 23.723463687150836, + "grad_norm": 0.47197747230529785, + "learning_rate": 2.8451396648044693e-05, + "loss": 0.1428, + "step": 8493 + }, + { + "epoch": 23.726256983240223, + "grad_norm": 0.45903104543685913, + "learning_rate": 2.8454748603351955e-05, + "loss": 0.1356, + "step": 8494 + }, + { + "epoch": 23.72905027932961, + "grad_norm": 0.49284160137176514, + "learning_rate": 2.845810055865922e-05, + "loss": 0.1526, + "step": 8495 + }, + { + "epoch": 23.731843575418996, + "grad_norm": 0.8832462430000305, + "learning_rate": 2.846145251396648e-05, + "loss": 0.1317, + "step": 8496 + }, + { + "epoch": 23.73463687150838, + "grad_norm": 0.6516457200050354, + "learning_rate": 2.8464804469273742e-05, + "loss": 0.1577, + "step": 8497 + }, + { + "epoch": 23.737430167597765, + "grad_norm": 0.4803456664085388, + "learning_rate": 2.8468156424581007e-05, + "loss": 0.1472, + "step": 8498 + }, + { + "epoch": 23.74022346368715, + "grad_norm": 0.7009912133216858, + "learning_rate": 2.8471508379888268e-05, + "loss": 0.1621, + "step": 8499 + }, + { + "epoch": 23.743016759776538, + "grad_norm": 0.8200787305831909, + "learning_rate": 2.847486033519553e-05, + "loss": 0.1491, + "step": 8500 + }, + { + "epoch": 23.74581005586592, + "grad_norm": 1.5500895977020264, + "learning_rate": 2.8478212290502794e-05, + "loss": 0.1433, + "step": 8501 + }, + { + "epoch": 23.748603351955307, + "grad_norm": 1.2663074731826782, + "learning_rate": 2.8481564245810056e-05, + "loss": 0.1398, + "step": 8502 + }, + { + "epoch": 23.751396648044693, + "grad_norm": 0.7359211444854736, + "learning_rate": 2.8484916201117317e-05, + "loss": 0.1316, + "step": 8503 + }, + { + "epoch": 23.75418994413408, + "grad_norm": 0.7820191979408264, + "learning_rate": 2.8488268156424582e-05, + "loss": 0.1763, + "step": 8504 + }, + { + "epoch": 23.756983240223462, + "grad_norm": 0.7646268606185913, + "learning_rate": 2.8491620111731843e-05, + "loss": 0.1492, + "step": 8505 + }, + { + "epoch": 23.75977653631285, + "grad_norm": 1.0323618650436401, + "learning_rate": 2.8494972067039104e-05, + "loss": 0.1571, + "step": 8506 + }, + { + "epoch": 23.762569832402235, + "grad_norm": 1.2276414632797241, + "learning_rate": 2.849832402234637e-05, + "loss": 0.1926, + "step": 8507 + }, + { + "epoch": 23.76536312849162, + "grad_norm": 0.7723167538642883, + "learning_rate": 2.850167597765363e-05, + "loss": 0.1543, + "step": 8508 + }, + { + "epoch": 23.768156424581004, + "grad_norm": 1.6178011894226074, + "learning_rate": 2.8505027932960892e-05, + "loss": 0.1903, + "step": 8509 + }, + { + "epoch": 23.77094972067039, + "grad_norm": 0.7513381242752075, + "learning_rate": 2.8508379888268157e-05, + "loss": 0.1768, + "step": 8510 + }, + { + "epoch": 23.773743016759777, + "grad_norm": 0.6123666167259216, + "learning_rate": 2.8511731843575418e-05, + "loss": 0.173, + "step": 8511 + }, + { + "epoch": 23.776536312849164, + "grad_norm": 0.5978202223777771, + "learning_rate": 2.851508379888268e-05, + "loss": 0.1513, + "step": 8512 + }, + { + "epoch": 23.779329608938546, + "grad_norm": 0.45150959491729736, + "learning_rate": 2.8518435754189944e-05, + "loss": 0.197, + "step": 8513 + }, + { + "epoch": 23.782122905027933, + "grad_norm": 0.9219286441802979, + "learning_rate": 2.8521787709497205e-05, + "loss": 0.1873, + "step": 8514 + }, + { + "epoch": 23.78491620111732, + "grad_norm": 0.45100724697113037, + "learning_rate": 2.852513966480447e-05, + "loss": 0.1783, + "step": 8515 + }, + { + "epoch": 23.787709497206706, + "grad_norm": 0.4630624055862427, + "learning_rate": 2.852849162011173e-05, + "loss": 0.1638, + "step": 8516 + }, + { + "epoch": 23.79050279329609, + "grad_norm": 0.5908763408660889, + "learning_rate": 2.8531843575418993e-05, + "loss": 0.1684, + "step": 8517 + }, + { + "epoch": 23.793296089385475, + "grad_norm": 0.5710327625274658, + "learning_rate": 2.8535195530726258e-05, + "loss": 0.1687, + "step": 8518 + }, + { + "epoch": 23.79608938547486, + "grad_norm": 0.7543351054191589, + "learning_rate": 2.853854748603352e-05, + "loss": 0.1713, + "step": 8519 + }, + { + "epoch": 23.798882681564244, + "grad_norm": 0.7097797989845276, + "learning_rate": 2.854189944134078e-05, + "loss": 0.1802, + "step": 8520 + }, + { + "epoch": 23.80167597765363, + "grad_norm": 0.7188547253608704, + "learning_rate": 2.8545251396648045e-05, + "loss": 0.1649, + "step": 8521 + }, + { + "epoch": 23.804469273743017, + "grad_norm": 0.6841999292373657, + "learning_rate": 2.854860335195531e-05, + "loss": 0.1307, + "step": 8522 + }, + { + "epoch": 23.807262569832403, + "grad_norm": 0.5422974228858948, + "learning_rate": 2.855195530726257e-05, + "loss": 0.156, + "step": 8523 + }, + { + "epoch": 23.810055865921786, + "grad_norm": 1.128111481666565, + "learning_rate": 2.8555307262569836e-05, + "loss": 0.1549, + "step": 8524 + }, + { + "epoch": 23.812849162011172, + "grad_norm": 0.8573923707008362, + "learning_rate": 2.8558659217877097e-05, + "loss": 0.16, + "step": 8525 + }, + { + "epoch": 23.81564245810056, + "grad_norm": 0.6000458002090454, + "learning_rate": 2.856201117318436e-05, + "loss": 0.1445, + "step": 8526 + }, + { + "epoch": 23.818435754189945, + "grad_norm": 0.5596504211425781, + "learning_rate": 2.8565363128491623e-05, + "loss": 0.1451, + "step": 8527 + }, + { + "epoch": 23.821229050279328, + "grad_norm": 1.2682924270629883, + "learning_rate": 2.8568715083798885e-05, + "loss": 0.1254, + "step": 8528 + }, + { + "epoch": 23.824022346368714, + "grad_norm": 0.5913763046264648, + "learning_rate": 2.8572067039106146e-05, + "loss": 0.1445, + "step": 8529 + }, + { + "epoch": 23.8268156424581, + "grad_norm": 0.8111798763275146, + "learning_rate": 2.857541899441341e-05, + "loss": 0.1388, + "step": 8530 + }, + { + "epoch": 23.829608938547487, + "grad_norm": 0.9486494660377502, + "learning_rate": 2.8578770949720672e-05, + "loss": 0.1535, + "step": 8531 + }, + { + "epoch": 23.83240223463687, + "grad_norm": 1.2603076696395874, + "learning_rate": 2.8582122905027937e-05, + "loss": 0.1721, + "step": 8532 + }, + { + "epoch": 23.835195530726256, + "grad_norm": 1.2245233058929443, + "learning_rate": 2.8585474860335198e-05, + "loss": 0.1823, + "step": 8533 + }, + { + "epoch": 23.837988826815643, + "grad_norm": 1.8480297327041626, + "learning_rate": 2.858882681564246e-05, + "loss": 0.1965, + "step": 8534 + }, + { + "epoch": 23.84078212290503, + "grad_norm": 0.5758872628211975, + "learning_rate": 2.8592178770949724e-05, + "loss": 0.1771, + "step": 8535 + }, + { + "epoch": 23.843575418994412, + "grad_norm": 0.49118363857269287, + "learning_rate": 2.8595530726256985e-05, + "loss": 0.149, + "step": 8536 + }, + { + "epoch": 23.8463687150838, + "grad_norm": 0.7360981702804565, + "learning_rate": 2.8598882681564247e-05, + "loss": 0.1718, + "step": 8537 + }, + { + "epoch": 23.849162011173185, + "grad_norm": 0.4460170269012451, + "learning_rate": 2.860223463687151e-05, + "loss": 0.1512, + "step": 8538 + }, + { + "epoch": 23.85195530726257, + "grad_norm": 0.8196663856506348, + "learning_rate": 2.8605586592178773e-05, + "loss": 0.169, + "step": 8539 + }, + { + "epoch": 23.854748603351954, + "grad_norm": 0.49374014139175415, + "learning_rate": 2.8608938547486034e-05, + "loss": 0.164, + "step": 8540 + }, + { + "epoch": 23.85754189944134, + "grad_norm": 0.7242518663406372, + "learning_rate": 2.86122905027933e-05, + "loss": 0.1985, + "step": 8541 + }, + { + "epoch": 23.860335195530727, + "grad_norm": 0.6739016771316528, + "learning_rate": 2.861564245810056e-05, + "loss": 0.1723, + "step": 8542 + }, + { + "epoch": 23.863128491620113, + "grad_norm": 0.4968107044696808, + "learning_rate": 2.861899441340782e-05, + "loss": 0.1654, + "step": 8543 + }, + { + "epoch": 23.865921787709496, + "grad_norm": 0.744697093963623, + "learning_rate": 2.8622346368715086e-05, + "loss": 0.1498, + "step": 8544 + }, + { + "epoch": 23.868715083798882, + "grad_norm": 1.1204935312271118, + "learning_rate": 2.8625698324022348e-05, + "loss": 0.1674, + "step": 8545 + }, + { + "epoch": 23.87150837988827, + "grad_norm": 0.686198890209198, + "learning_rate": 2.862905027932961e-05, + "loss": 0.1803, + "step": 8546 + }, + { + "epoch": 23.874301675977655, + "grad_norm": 1.2316946983337402, + "learning_rate": 2.8632402234636874e-05, + "loss": 0.1521, + "step": 8547 + }, + { + "epoch": 23.877094972067038, + "grad_norm": 0.7034990787506104, + "learning_rate": 2.8635754189944135e-05, + "loss": 0.1861, + "step": 8548 + }, + { + "epoch": 23.879888268156424, + "grad_norm": 0.68357914686203, + "learning_rate": 2.8639106145251397e-05, + "loss": 0.151, + "step": 8549 + }, + { + "epoch": 23.88268156424581, + "grad_norm": 0.6787564754486084, + "learning_rate": 2.864245810055866e-05, + "loss": 0.1285, + "step": 8550 + }, + { + "epoch": 23.885474860335197, + "grad_norm": 0.6386020183563232, + "learning_rate": 2.8645810055865923e-05, + "loss": 0.1604, + "step": 8551 + }, + { + "epoch": 23.88826815642458, + "grad_norm": 0.6550942063331604, + "learning_rate": 2.8649162011173187e-05, + "loss": 0.153, + "step": 8552 + }, + { + "epoch": 23.891061452513966, + "grad_norm": 0.7790955901145935, + "learning_rate": 2.865251396648045e-05, + "loss": 0.1704, + "step": 8553 + }, + { + "epoch": 23.893854748603353, + "grad_norm": 0.8555923104286194, + "learning_rate": 2.865586592178771e-05, + "loss": 0.1359, + "step": 8554 + }, + { + "epoch": 23.89664804469274, + "grad_norm": 0.9140127897262573, + "learning_rate": 2.8659217877094975e-05, + "loss": 0.1788, + "step": 8555 + }, + { + "epoch": 23.899441340782122, + "grad_norm": 2.0194571018218994, + "learning_rate": 2.8662569832402236e-05, + "loss": 0.1464, + "step": 8556 + }, + { + "epoch": 23.90223463687151, + "grad_norm": 1.135514259338379, + "learning_rate": 2.8665921787709497e-05, + "loss": 0.1289, + "step": 8557 + }, + { + "epoch": 23.905027932960895, + "grad_norm": 7.31110954284668, + "learning_rate": 2.8669273743016762e-05, + "loss": 0.166, + "step": 8558 + }, + { + "epoch": 23.907821229050278, + "grad_norm": 1.5454819202423096, + "learning_rate": 2.8672625698324024e-05, + "loss": 0.2129, + "step": 8559 + }, + { + "epoch": 23.910614525139664, + "grad_norm": 0.599492609500885, + "learning_rate": 2.8675977653631285e-05, + "loss": 0.1505, + "step": 8560 + }, + { + "epoch": 23.91340782122905, + "grad_norm": 0.5036564469337463, + "learning_rate": 2.867932960893855e-05, + "loss": 0.1635, + "step": 8561 + }, + { + "epoch": 23.916201117318437, + "grad_norm": 0.5207655429840088, + "learning_rate": 2.868268156424581e-05, + "loss": 0.1727, + "step": 8562 + }, + { + "epoch": 23.91899441340782, + "grad_norm": 0.6084097027778625, + "learning_rate": 2.8686033519553072e-05, + "loss": 0.1707, + "step": 8563 + }, + { + "epoch": 23.921787709497206, + "grad_norm": 0.6435678601264954, + "learning_rate": 2.8689385474860337e-05, + "loss": 0.1686, + "step": 8564 + }, + { + "epoch": 23.924581005586592, + "grad_norm": 0.6091101169586182, + "learning_rate": 2.86927374301676e-05, + "loss": 0.1574, + "step": 8565 + }, + { + "epoch": 23.92737430167598, + "grad_norm": 0.4999472498893738, + "learning_rate": 2.869608938547486e-05, + "loss": 0.187, + "step": 8566 + }, + { + "epoch": 23.93016759776536, + "grad_norm": 0.7779102921485901, + "learning_rate": 2.8699441340782124e-05, + "loss": 0.1273, + "step": 8567 + }, + { + "epoch": 23.932960893854748, + "grad_norm": 1.4887670278549194, + "learning_rate": 2.8702793296089386e-05, + "loss": 0.1538, + "step": 8568 + }, + { + "epoch": 23.935754189944134, + "grad_norm": 1.003900408744812, + "learning_rate": 2.8706145251396647e-05, + "loss": 0.1603, + "step": 8569 + }, + { + "epoch": 23.93854748603352, + "grad_norm": 0.8001010417938232, + "learning_rate": 2.8709497206703912e-05, + "loss": 0.1576, + "step": 8570 + }, + { + "epoch": 23.941340782122904, + "grad_norm": 0.6352811455726624, + "learning_rate": 2.8712849162011173e-05, + "loss": 0.1565, + "step": 8571 + }, + { + "epoch": 23.94413407821229, + "grad_norm": 0.5438004732131958, + "learning_rate": 2.8716201117318438e-05, + "loss": 0.1595, + "step": 8572 + }, + { + "epoch": 23.946927374301676, + "grad_norm": 0.868043839931488, + "learning_rate": 2.87195530726257e-05, + "loss": 0.1602, + "step": 8573 + }, + { + "epoch": 23.949720670391063, + "grad_norm": 1.21165132522583, + "learning_rate": 2.872290502793296e-05, + "loss": 0.159, + "step": 8574 + }, + { + "epoch": 23.952513966480446, + "grad_norm": 0.6765435934066772, + "learning_rate": 2.8726256983240225e-05, + "loss": 0.1505, + "step": 8575 + }, + { + "epoch": 23.955307262569832, + "grad_norm": 0.6370588541030884, + "learning_rate": 2.8729608938547487e-05, + "loss": 0.1458, + "step": 8576 + }, + { + "epoch": 23.95810055865922, + "grad_norm": 0.6963369846343994, + "learning_rate": 2.8732960893854748e-05, + "loss": 0.1295, + "step": 8577 + }, + { + "epoch": 23.960893854748605, + "grad_norm": 3.5155863761901855, + "learning_rate": 2.8736312849162013e-05, + "loss": 0.1551, + "step": 8578 + }, + { + "epoch": 23.963687150837988, + "grad_norm": 1.6986095905303955, + "learning_rate": 2.8739664804469274e-05, + "loss": 0.167, + "step": 8579 + }, + { + "epoch": 23.966480446927374, + "grad_norm": 0.7390466332435608, + "learning_rate": 2.8743016759776535e-05, + "loss": 0.1357, + "step": 8580 + }, + { + "epoch": 23.96927374301676, + "grad_norm": 1.2116914987564087, + "learning_rate": 2.87463687150838e-05, + "loss": 0.1855, + "step": 8581 + }, + { + "epoch": 23.972067039106147, + "grad_norm": 0.7110611796379089, + "learning_rate": 2.874972067039106e-05, + "loss": 0.1427, + "step": 8582 + }, + { + "epoch": 23.97486033519553, + "grad_norm": 1.404340386390686, + "learning_rate": 2.8753072625698323e-05, + "loss": 0.2226, + "step": 8583 + }, + { + "epoch": 23.977653631284916, + "grad_norm": 6.114239692687988, + "learning_rate": 2.8756424581005588e-05, + "loss": 0.2295, + "step": 8584 + }, + { + "epoch": 23.980446927374302, + "grad_norm": 0.50738525390625, + "learning_rate": 2.875977653631285e-05, + "loss": 0.1796, + "step": 8585 + }, + { + "epoch": 23.98324022346369, + "grad_norm": 0.649925708770752, + "learning_rate": 2.876312849162011e-05, + "loss": 0.1615, + "step": 8586 + }, + { + "epoch": 23.98603351955307, + "grad_norm": 0.9050689339637756, + "learning_rate": 2.8766480446927375e-05, + "loss": 0.1679, + "step": 8587 + }, + { + "epoch": 23.988826815642458, + "grad_norm": 0.6630035042762756, + "learning_rate": 2.8769832402234636e-05, + "loss": 0.1897, + "step": 8588 + }, + { + "epoch": 23.991620111731844, + "grad_norm": 0.6604133248329163, + "learning_rate": 2.87731843575419e-05, + "loss": 0.1583, + "step": 8589 + }, + { + "epoch": 23.994413407821227, + "grad_norm": 1.3836703300476074, + "learning_rate": 2.8776536312849163e-05, + "loss": 0.183, + "step": 8590 + }, + { + "epoch": 23.997206703910614, + "grad_norm": 0.9742031097412109, + "learning_rate": 2.8779888268156424e-05, + "loss": 0.1486, + "step": 8591 + }, + { + "epoch": 24.0, + "grad_norm": 1.0786552429199219, + "learning_rate": 2.878324022346369e-05, + "loss": 0.1723, + "step": 8592 + }, + { + "epoch": 24.002793296089386, + "grad_norm": 0.481562077999115, + "learning_rate": 2.878659217877095e-05, + "loss": 0.1481, + "step": 8593 + }, + { + "epoch": 24.00558659217877, + "grad_norm": 0.3857674300670624, + "learning_rate": 2.878994413407821e-05, + "loss": 0.1418, + "step": 8594 + }, + { + "epoch": 24.008379888268156, + "grad_norm": 0.4814370572566986, + "learning_rate": 2.8793296089385476e-05, + "loss": 0.1507, + "step": 8595 + }, + { + "epoch": 24.011173184357542, + "grad_norm": 0.4962961673736572, + "learning_rate": 2.8796648044692737e-05, + "loss": 0.1592, + "step": 8596 + }, + { + "epoch": 24.01396648044693, + "grad_norm": 0.6549685001373291, + "learning_rate": 2.88e-05, + "loss": 0.1546, + "step": 8597 + }, + { + "epoch": 24.01675977653631, + "grad_norm": 0.7389982342720032, + "learning_rate": 2.8803351955307263e-05, + "loss": 0.1945, + "step": 8598 + }, + { + "epoch": 24.019553072625698, + "grad_norm": 1.1544837951660156, + "learning_rate": 2.8806703910614525e-05, + "loss": 0.1701, + "step": 8599 + }, + { + "epoch": 24.022346368715084, + "grad_norm": 0.5143484473228455, + "learning_rate": 2.8810055865921786e-05, + "loss": 0.1414, + "step": 8600 + }, + { + "epoch": 24.02513966480447, + "grad_norm": 0.517044186592102, + "learning_rate": 2.881340782122905e-05, + "loss": 0.1348, + "step": 8601 + }, + { + "epoch": 24.027932960893853, + "grad_norm": 1.2998812198638916, + "learning_rate": 2.8816759776536312e-05, + "loss": 0.1586, + "step": 8602 + }, + { + "epoch": 24.03072625698324, + "grad_norm": 0.4938601851463318, + "learning_rate": 2.8820111731843574e-05, + "loss": 0.1559, + "step": 8603 + }, + { + "epoch": 24.033519553072626, + "grad_norm": 0.7520171403884888, + "learning_rate": 2.8823463687150838e-05, + "loss": 0.1563, + "step": 8604 + }, + { + "epoch": 24.036312849162012, + "grad_norm": 0.5938146114349365, + "learning_rate": 2.88268156424581e-05, + "loss": 0.1372, + "step": 8605 + }, + { + "epoch": 24.039106145251395, + "grad_norm": 0.8526359796524048, + "learning_rate": 2.883016759776536e-05, + "loss": 0.1631, + "step": 8606 + }, + { + "epoch": 24.04189944134078, + "grad_norm": 0.5920069813728333, + "learning_rate": 2.8833519553072626e-05, + "loss": 0.1118, + "step": 8607 + }, + { + "epoch": 24.044692737430168, + "grad_norm": 0.819749116897583, + "learning_rate": 2.8836871508379887e-05, + "loss": 0.1371, + "step": 8608 + }, + { + "epoch": 24.047486033519554, + "grad_norm": 1.8315272331237793, + "learning_rate": 2.8840223463687152e-05, + "loss": 0.1419, + "step": 8609 + }, + { + "epoch": 24.050279329608937, + "grad_norm": 0.772979199886322, + "learning_rate": 2.8843575418994413e-05, + "loss": 0.1365, + "step": 8610 + }, + { + "epoch": 24.053072625698324, + "grad_norm": 0.920295238494873, + "learning_rate": 2.8846927374301674e-05, + "loss": 0.16, + "step": 8611 + }, + { + "epoch": 24.05586592178771, + "grad_norm": 1.8534722328186035, + "learning_rate": 2.885027932960894e-05, + "loss": 0.1397, + "step": 8612 + }, + { + "epoch": 24.058659217877096, + "grad_norm": 0.8009552955627441, + "learning_rate": 2.88536312849162e-05, + "loss": 0.162, + "step": 8613 + }, + { + "epoch": 24.06145251396648, + "grad_norm": 1.5803658962249756, + "learning_rate": 2.8856983240223462e-05, + "loss": 0.1473, + "step": 8614 + }, + { + "epoch": 24.064245810055866, + "grad_norm": 1.3942879438400269, + "learning_rate": 2.8860335195530727e-05, + "loss": 0.2108, + "step": 8615 + }, + { + "epoch": 24.067039106145252, + "grad_norm": 0.9631518721580505, + "learning_rate": 2.8863687150837988e-05, + "loss": 0.1807, + "step": 8616 + }, + { + "epoch": 24.06983240223464, + "grad_norm": 1.9106346368789673, + "learning_rate": 2.886703910614525e-05, + "loss": 0.1899, + "step": 8617 + }, + { + "epoch": 24.07262569832402, + "grad_norm": 0.43440818786621094, + "learning_rate": 2.8870391061452514e-05, + "loss": 0.1632, + "step": 8618 + }, + { + "epoch": 24.075418994413408, + "grad_norm": 0.6677454113960266, + "learning_rate": 2.887374301675978e-05, + "loss": 0.182, + "step": 8619 + }, + { + "epoch": 24.078212290502794, + "grad_norm": 0.7159048318862915, + "learning_rate": 2.887709497206704e-05, + "loss": 0.1552, + "step": 8620 + }, + { + "epoch": 24.08100558659218, + "grad_norm": 0.7510294914245605, + "learning_rate": 2.8880446927374305e-05, + "loss": 0.1657, + "step": 8621 + }, + { + "epoch": 24.083798882681563, + "grad_norm": 0.5945203900337219, + "learning_rate": 2.8883798882681566e-05, + "loss": 0.1816, + "step": 8622 + }, + { + "epoch": 24.08659217877095, + "grad_norm": 0.6612391471862793, + "learning_rate": 2.8887150837988828e-05, + "loss": 0.1405, + "step": 8623 + }, + { + "epoch": 24.089385474860336, + "grad_norm": 0.5299933552742004, + "learning_rate": 2.8890502793296092e-05, + "loss": 0.1356, + "step": 8624 + }, + { + "epoch": 24.092178770949722, + "grad_norm": 1.4121366739273071, + "learning_rate": 2.8893854748603354e-05, + "loss": 0.1644, + "step": 8625 + }, + { + "epoch": 24.094972067039105, + "grad_norm": 0.5525647401809692, + "learning_rate": 2.889720670391062e-05, + "loss": 0.1594, + "step": 8626 + }, + { + "epoch": 24.09776536312849, + "grad_norm": 1.0915334224700928, + "learning_rate": 2.890055865921788e-05, + "loss": 0.1723, + "step": 8627 + }, + { + "epoch": 24.100558659217878, + "grad_norm": 1.0319507122039795, + "learning_rate": 2.890391061452514e-05, + "loss": 0.1672, + "step": 8628 + }, + { + "epoch": 24.10335195530726, + "grad_norm": 0.551236629486084, + "learning_rate": 2.8907262569832406e-05, + "loss": 0.1439, + "step": 8629 + }, + { + "epoch": 24.106145251396647, + "grad_norm": 0.7897946834564209, + "learning_rate": 2.8910614525139667e-05, + "loss": 0.1477, + "step": 8630 + }, + { + "epoch": 24.108938547486034, + "grad_norm": 0.8100122809410095, + "learning_rate": 2.891396648044693e-05, + "loss": 0.1353, + "step": 8631 + }, + { + "epoch": 24.11173184357542, + "grad_norm": 0.7977888584136963, + "learning_rate": 2.8917318435754193e-05, + "loss": 0.1421, + "step": 8632 + }, + { + "epoch": 24.114525139664803, + "grad_norm": 0.877657949924469, + "learning_rate": 2.8920670391061455e-05, + "loss": 0.1644, + "step": 8633 + }, + { + "epoch": 24.11731843575419, + "grad_norm": 0.8914011120796204, + "learning_rate": 2.8924022346368716e-05, + "loss": 0.1375, + "step": 8634 + }, + { + "epoch": 24.120111731843576, + "grad_norm": 1.037544846534729, + "learning_rate": 2.892737430167598e-05, + "loss": 0.1475, + "step": 8635 + }, + { + "epoch": 24.122905027932962, + "grad_norm": 1.5527957677841187, + "learning_rate": 2.8930726256983242e-05, + "loss": 0.1694, + "step": 8636 + }, + { + "epoch": 24.125698324022345, + "grad_norm": 0.9596747756004333, + "learning_rate": 2.8934078212290503e-05, + "loss": 0.1653, + "step": 8637 + }, + { + "epoch": 24.12849162011173, + "grad_norm": 1.2988542318344116, + "learning_rate": 2.8937430167597768e-05, + "loss": 0.1927, + "step": 8638 + }, + { + "epoch": 24.131284916201118, + "grad_norm": 1.0913902521133423, + "learning_rate": 2.894078212290503e-05, + "loss": 0.1387, + "step": 8639 + }, + { + "epoch": 24.134078212290504, + "grad_norm": 0.9617610573768616, + "learning_rate": 2.894413407821229e-05, + "loss": 0.171, + "step": 8640 + }, + { + "epoch": 24.136871508379887, + "grad_norm": 0.9915583729743958, + "learning_rate": 2.8947486033519555e-05, + "loss": 0.2022, + "step": 8641 + }, + { + "epoch": 24.139664804469273, + "grad_norm": 2.301741600036621, + "learning_rate": 2.8950837988826817e-05, + "loss": 0.2004, + "step": 8642 + }, + { + "epoch": 24.14245810055866, + "grad_norm": 0.6524962782859802, + "learning_rate": 2.8954189944134078e-05, + "loss": 0.177, + "step": 8643 + }, + { + "epoch": 24.145251396648046, + "grad_norm": 1.314521074295044, + "learning_rate": 2.8957541899441343e-05, + "loss": 0.1571, + "step": 8644 + }, + { + "epoch": 24.14804469273743, + "grad_norm": 0.5959486961364746, + "learning_rate": 2.8960893854748604e-05, + "loss": 0.1797, + "step": 8645 + }, + { + "epoch": 24.150837988826815, + "grad_norm": 0.5731748938560486, + "learning_rate": 2.896424581005587e-05, + "loss": 0.193, + "step": 8646 + }, + { + "epoch": 24.1536312849162, + "grad_norm": 0.8550352454185486, + "learning_rate": 2.896759776536313e-05, + "loss": 0.189, + "step": 8647 + }, + { + "epoch": 24.156424581005588, + "grad_norm": 0.5964093208312988, + "learning_rate": 2.897094972067039e-05, + "loss": 0.1541, + "step": 8648 + }, + { + "epoch": 24.15921787709497, + "grad_norm": 0.7276826500892639, + "learning_rate": 2.8974301675977656e-05, + "loss": 0.1691, + "step": 8649 + }, + { + "epoch": 24.162011173184357, + "grad_norm": 0.6179826855659485, + "learning_rate": 2.8977653631284918e-05, + "loss": 0.159, + "step": 8650 + }, + { + "epoch": 24.164804469273744, + "grad_norm": 0.4375210106372833, + "learning_rate": 2.898100558659218e-05, + "loss": 0.1511, + "step": 8651 + }, + { + "epoch": 24.16759776536313, + "grad_norm": 0.639118492603302, + "learning_rate": 2.8984357541899444e-05, + "loss": 0.1469, + "step": 8652 + }, + { + "epoch": 24.170391061452513, + "grad_norm": 1.836925745010376, + "learning_rate": 2.8987709497206705e-05, + "loss": 0.1664, + "step": 8653 + }, + { + "epoch": 24.1731843575419, + "grad_norm": 1.0663514137268066, + "learning_rate": 2.8991061452513967e-05, + "loss": 0.1695, + "step": 8654 + }, + { + "epoch": 24.175977653631286, + "grad_norm": 1.9546936750411987, + "learning_rate": 2.899441340782123e-05, + "loss": 0.1438, + "step": 8655 + }, + { + "epoch": 24.178770949720672, + "grad_norm": 1.5129022598266602, + "learning_rate": 2.8997765363128493e-05, + "loss": 0.1561, + "step": 8656 + }, + { + "epoch": 24.181564245810055, + "grad_norm": 0.77357017993927, + "learning_rate": 2.9001117318435754e-05, + "loss": 0.1709, + "step": 8657 + }, + { + "epoch": 24.18435754189944, + "grad_norm": 0.6973763704299927, + "learning_rate": 2.900446927374302e-05, + "loss": 0.1425, + "step": 8658 + }, + { + "epoch": 24.187150837988828, + "grad_norm": 1.409000039100647, + "learning_rate": 2.900782122905028e-05, + "loss": 0.1707, + "step": 8659 + }, + { + "epoch": 24.189944134078214, + "grad_norm": 1.0602530241012573, + "learning_rate": 2.901117318435754e-05, + "loss": 0.1309, + "step": 8660 + }, + { + "epoch": 24.192737430167597, + "grad_norm": 1.2794429063796997, + "learning_rate": 2.9014525139664806e-05, + "loss": 0.1486, + "step": 8661 + }, + { + "epoch": 24.195530726256983, + "grad_norm": 1.5396957397460938, + "learning_rate": 2.9017877094972067e-05, + "loss": 0.1339, + "step": 8662 + }, + { + "epoch": 24.19832402234637, + "grad_norm": 0.7343619465827942, + "learning_rate": 2.902122905027933e-05, + "loss": 0.149, + "step": 8663 + }, + { + "epoch": 24.201117318435756, + "grad_norm": 1.214141607284546, + "learning_rate": 2.9024581005586594e-05, + "loss": 0.1497, + "step": 8664 + }, + { + "epoch": 24.20391061452514, + "grad_norm": 0.9431669116020203, + "learning_rate": 2.9027932960893855e-05, + "loss": 0.1447, + "step": 8665 + }, + { + "epoch": 24.206703910614525, + "grad_norm": 1.3681031465530396, + "learning_rate": 2.903128491620112e-05, + "loss": 0.1923, + "step": 8666 + }, + { + "epoch": 24.20949720670391, + "grad_norm": 2.808117628097534, + "learning_rate": 2.903463687150838e-05, + "loss": 0.2238, + "step": 8667 + }, + { + "epoch": 24.212290502793294, + "grad_norm": 0.48470282554626465, + "learning_rate": 2.9037988826815642e-05, + "loss": 0.1802, + "step": 8668 + }, + { + "epoch": 24.21508379888268, + "grad_norm": 0.5129351615905762, + "learning_rate": 2.9041340782122907e-05, + "loss": 0.1625, + "step": 8669 + }, + { + "epoch": 24.217877094972067, + "grad_norm": 0.9107066988945007, + "learning_rate": 2.904469273743017e-05, + "loss": 0.1763, + "step": 8670 + }, + { + "epoch": 24.220670391061454, + "grad_norm": 0.5616417527198792, + "learning_rate": 2.904804469273743e-05, + "loss": 0.1871, + "step": 8671 + }, + { + "epoch": 24.223463687150836, + "grad_norm": 0.6072235703468323, + "learning_rate": 2.9051396648044694e-05, + "loss": 0.1689, + "step": 8672 + }, + { + "epoch": 24.226256983240223, + "grad_norm": 0.8418172001838684, + "learning_rate": 2.9054748603351956e-05, + "loss": 0.1821, + "step": 8673 + }, + { + "epoch": 24.22905027932961, + "grad_norm": 0.7524688243865967, + "learning_rate": 2.9058100558659217e-05, + "loss": 0.1738, + "step": 8674 + }, + { + "epoch": 24.231843575418996, + "grad_norm": 1.611892819404602, + "learning_rate": 2.9061452513966482e-05, + "loss": 0.1891, + "step": 8675 + }, + { + "epoch": 24.23463687150838, + "grad_norm": 0.5205845832824707, + "learning_rate": 2.9064804469273743e-05, + "loss": 0.1456, + "step": 8676 + }, + { + "epoch": 24.237430167597765, + "grad_norm": 0.5919631123542786, + "learning_rate": 2.9068156424581005e-05, + "loss": 0.1902, + "step": 8677 + }, + { + "epoch": 24.24022346368715, + "grad_norm": 0.5758237838745117, + "learning_rate": 2.907150837988827e-05, + "loss": 0.1198, + "step": 8678 + }, + { + "epoch": 24.243016759776538, + "grad_norm": 0.578046977519989, + "learning_rate": 2.907486033519553e-05, + "loss": 0.147, + "step": 8679 + }, + { + "epoch": 24.24581005586592, + "grad_norm": 0.555768609046936, + "learning_rate": 2.9078212290502792e-05, + "loss": 0.1355, + "step": 8680 + }, + { + "epoch": 24.248603351955307, + "grad_norm": 3.946075439453125, + "learning_rate": 2.9081564245810057e-05, + "loss": 0.1497, + "step": 8681 + }, + { + "epoch": 24.251396648044693, + "grad_norm": 0.5564936399459839, + "learning_rate": 2.9084916201117318e-05, + "loss": 0.1355, + "step": 8682 + }, + { + "epoch": 24.25418994413408, + "grad_norm": 1.029929280281067, + "learning_rate": 2.9088268156424583e-05, + "loss": 0.1688, + "step": 8683 + }, + { + "epoch": 24.256983240223462, + "grad_norm": 0.7525861859321594, + "learning_rate": 2.9091620111731844e-05, + "loss": 0.1728, + "step": 8684 + }, + { + "epoch": 24.25977653631285, + "grad_norm": 3.0512354373931885, + "learning_rate": 2.9094972067039106e-05, + "loss": 0.153, + "step": 8685 + }, + { + "epoch": 24.262569832402235, + "grad_norm": 1.5825406312942505, + "learning_rate": 2.909832402234637e-05, + "loss": 0.1659, + "step": 8686 + }, + { + "epoch": 24.26536312849162, + "grad_norm": 1.1344032287597656, + "learning_rate": 2.910167597765363e-05, + "loss": 0.2089, + "step": 8687 + }, + { + "epoch": 24.268156424581004, + "grad_norm": 2.7450168132781982, + "learning_rate": 2.9105027932960893e-05, + "loss": 0.147, + "step": 8688 + }, + { + "epoch": 24.27094972067039, + "grad_norm": 0.711538553237915, + "learning_rate": 2.9108379888268158e-05, + "loss": 0.1458, + "step": 8689 + }, + { + "epoch": 24.273743016759777, + "grad_norm": 1.6258749961853027, + "learning_rate": 2.911173184357542e-05, + "loss": 0.1407, + "step": 8690 + }, + { + "epoch": 24.276536312849164, + "grad_norm": 1.6167349815368652, + "learning_rate": 2.911508379888268e-05, + "loss": 0.1525, + "step": 8691 + }, + { + "epoch": 24.279329608938546, + "grad_norm": 4.259993076324463, + "learning_rate": 2.9118435754189945e-05, + "loss": 0.2245, + "step": 8692 + }, + { + "epoch": 24.282122905027933, + "grad_norm": 0.4286499619483948, + "learning_rate": 2.9121787709497206e-05, + "loss": 0.1838, + "step": 8693 + }, + { + "epoch": 24.28491620111732, + "grad_norm": 1.1599788665771484, + "learning_rate": 2.9125139664804468e-05, + "loss": 0.1624, + "step": 8694 + }, + { + "epoch": 24.287709497206706, + "grad_norm": 0.468360960483551, + "learning_rate": 2.9128491620111733e-05, + "loss": 0.1479, + "step": 8695 + }, + { + "epoch": 24.29050279329609, + "grad_norm": 0.6324694156646729, + "learning_rate": 2.9131843575418994e-05, + "loss": 0.1406, + "step": 8696 + }, + { + "epoch": 24.293296089385475, + "grad_norm": 0.685868501663208, + "learning_rate": 2.9135195530726255e-05, + "loss": 0.1626, + "step": 8697 + }, + { + "epoch": 24.29608938547486, + "grad_norm": 0.5987658500671387, + "learning_rate": 2.913854748603352e-05, + "loss": 0.1658, + "step": 8698 + }, + { + "epoch": 24.298882681564244, + "grad_norm": 0.8176558017730713, + "learning_rate": 2.914189944134078e-05, + "loss": 0.1937, + "step": 8699 + }, + { + "epoch": 24.30167597765363, + "grad_norm": 0.6278051733970642, + "learning_rate": 2.9145251396648043e-05, + "loss": 0.1761, + "step": 8700 + }, + { + "epoch": 24.304469273743017, + "grad_norm": 0.500758171081543, + "learning_rate": 2.9148603351955307e-05, + "loss": 0.1417, + "step": 8701 + }, + { + "epoch": 24.307262569832403, + "grad_norm": 0.7889264225959778, + "learning_rate": 2.915195530726257e-05, + "loss": 0.1498, + "step": 8702 + }, + { + "epoch": 24.310055865921786, + "grad_norm": 1.7457643747329712, + "learning_rate": 2.9155307262569833e-05, + "loss": 0.1912, + "step": 8703 + }, + { + "epoch": 24.312849162011172, + "grad_norm": 0.7052027583122253, + "learning_rate": 2.9158659217877095e-05, + "loss": 0.1679, + "step": 8704 + }, + { + "epoch": 24.31564245810056, + "grad_norm": 0.7781679034233093, + "learning_rate": 2.9162011173184356e-05, + "loss": 0.1627, + "step": 8705 + }, + { + "epoch": 24.318435754189945, + "grad_norm": 1.582472562789917, + "learning_rate": 2.916536312849162e-05, + "loss": 0.141, + "step": 8706 + }, + { + "epoch": 24.321229050279328, + "grad_norm": 0.9064821004867554, + "learning_rate": 2.9168715083798882e-05, + "loss": 0.1389, + "step": 8707 + }, + { + "epoch": 24.324022346368714, + "grad_norm": 0.6304162740707397, + "learning_rate": 2.9172067039106144e-05, + "loss": 0.1236, + "step": 8708 + }, + { + "epoch": 24.3268156424581, + "grad_norm": 0.8263901472091675, + "learning_rate": 2.9175418994413408e-05, + "loss": 0.1315, + "step": 8709 + }, + { + "epoch": 24.329608938547487, + "grad_norm": 0.954162061214447, + "learning_rate": 2.917877094972067e-05, + "loss": 0.1595, + "step": 8710 + }, + { + "epoch": 24.33240223463687, + "grad_norm": 1.96446692943573, + "learning_rate": 2.918212290502793e-05, + "loss": 0.1509, + "step": 8711 + }, + { + "epoch": 24.335195530726256, + "grad_norm": 1.0817787647247314, + "learning_rate": 2.9185474860335196e-05, + "loss": 0.1681, + "step": 8712 + }, + { + "epoch": 24.337988826815643, + "grad_norm": 1.2546696662902832, + "learning_rate": 2.9188826815642457e-05, + "loss": 0.1619, + "step": 8713 + }, + { + "epoch": 24.34078212290503, + "grad_norm": 1.3819695711135864, + "learning_rate": 2.9192178770949722e-05, + "loss": 0.1414, + "step": 8714 + }, + { + "epoch": 24.343575418994412, + "grad_norm": 1.1715528964996338, + "learning_rate": 2.9195530726256987e-05, + "loss": 0.1655, + "step": 8715 + }, + { + "epoch": 24.3463687150838, + "grad_norm": 1.8578327894210815, + "learning_rate": 2.9198882681564248e-05, + "loss": 0.1911, + "step": 8716 + }, + { + "epoch": 24.349162011173185, + "grad_norm": 2.3258707523345947, + "learning_rate": 2.920223463687151e-05, + "loss": 0.2202, + "step": 8717 + }, + { + "epoch": 24.35195530726257, + "grad_norm": 0.42142453789711, + "learning_rate": 2.9205586592178774e-05, + "loss": 0.1628, + "step": 8718 + }, + { + "epoch": 24.354748603351954, + "grad_norm": 0.5108674764633179, + "learning_rate": 2.9208938547486035e-05, + "loss": 0.176, + "step": 8719 + }, + { + "epoch": 24.35754189944134, + "grad_norm": 0.44519853591918945, + "learning_rate": 2.92122905027933e-05, + "loss": 0.162, + "step": 8720 + }, + { + "epoch": 24.360335195530727, + "grad_norm": 0.46030664443969727, + "learning_rate": 2.921564245810056e-05, + "loss": 0.175, + "step": 8721 + }, + { + "epoch": 24.363128491620113, + "grad_norm": 0.4687526226043701, + "learning_rate": 2.9218994413407823e-05, + "loss": 0.1535, + "step": 8722 + }, + { + "epoch": 24.365921787709496, + "grad_norm": 0.3969995975494385, + "learning_rate": 2.9222346368715087e-05, + "loss": 0.1396, + "step": 8723 + }, + { + "epoch": 24.368715083798882, + "grad_norm": 0.45602357387542725, + "learning_rate": 2.922569832402235e-05, + "loss": 0.1664, + "step": 8724 + }, + { + "epoch": 24.37150837988827, + "grad_norm": 0.803852379322052, + "learning_rate": 2.922905027932961e-05, + "loss": 0.1481, + "step": 8725 + }, + { + "epoch": 24.374301675977655, + "grad_norm": 0.5458903312683105, + "learning_rate": 2.9232402234636875e-05, + "loss": 0.1412, + "step": 8726 + }, + { + "epoch": 24.377094972067038, + "grad_norm": 0.48502835631370544, + "learning_rate": 2.9235754189944136e-05, + "loss": 0.144, + "step": 8727 + }, + { + "epoch": 24.379888268156424, + "grad_norm": 0.6083242893218994, + "learning_rate": 2.9239106145251398e-05, + "loss": 0.1997, + "step": 8728 + }, + { + "epoch": 24.38268156424581, + "grad_norm": 0.5887048840522766, + "learning_rate": 2.9242458100558662e-05, + "loss": 0.1487, + "step": 8729 + }, + { + "epoch": 24.385474860335197, + "grad_norm": 0.4933422803878784, + "learning_rate": 2.9245810055865924e-05, + "loss": 0.1269, + "step": 8730 + }, + { + "epoch": 24.38826815642458, + "grad_norm": 0.5134443640708923, + "learning_rate": 2.9249162011173185e-05, + "loss": 0.1313, + "step": 8731 + }, + { + "epoch": 24.391061452513966, + "grad_norm": 0.5495244264602661, + "learning_rate": 2.925251396648045e-05, + "loss": 0.1471, + "step": 8732 + }, + { + "epoch": 24.393854748603353, + "grad_norm": 0.7634639739990234, + "learning_rate": 2.925586592178771e-05, + "loss": 0.1353, + "step": 8733 + }, + { + "epoch": 24.39664804469274, + "grad_norm": 1.235334873199463, + "learning_rate": 2.9259217877094972e-05, + "loss": 0.1657, + "step": 8734 + }, + { + "epoch": 24.399441340782122, + "grad_norm": 0.5861709117889404, + "learning_rate": 2.9262569832402237e-05, + "loss": 0.1345, + "step": 8735 + }, + { + "epoch": 24.40223463687151, + "grad_norm": 0.8381406664848328, + "learning_rate": 2.92659217877095e-05, + "loss": 0.1768, + "step": 8736 + }, + { + "epoch": 24.405027932960895, + "grad_norm": 1.3816301822662354, + "learning_rate": 2.926927374301676e-05, + "loss": 0.1436, + "step": 8737 + }, + { + "epoch": 24.407821229050278, + "grad_norm": 0.963128387928009, + "learning_rate": 2.9272625698324025e-05, + "loss": 0.1514, + "step": 8738 + }, + { + "epoch": 24.410614525139664, + "grad_norm": 1.260421872138977, + "learning_rate": 2.9275977653631286e-05, + "loss": 0.1273, + "step": 8739 + }, + { + "epoch": 24.41340782122905, + "grad_norm": 0.8837178945541382, + "learning_rate": 2.927932960893855e-05, + "loss": 0.1505, + "step": 8740 + }, + { + "epoch": 24.416201117318437, + "grad_norm": 0.9451472759246826, + "learning_rate": 2.9282681564245812e-05, + "loss": 0.1626, + "step": 8741 + }, + { + "epoch": 24.41899441340782, + "grad_norm": 2.11775803565979, + "learning_rate": 2.9286033519553073e-05, + "loss": 0.1967, + "step": 8742 + }, + { + "epoch": 24.421787709497206, + "grad_norm": 0.43846651911735535, + "learning_rate": 2.9289385474860338e-05, + "loss": 0.1751, + "step": 8743 + }, + { + "epoch": 24.424581005586592, + "grad_norm": 0.5023484230041504, + "learning_rate": 2.92927374301676e-05, + "loss": 0.1509, + "step": 8744 + }, + { + "epoch": 24.42737430167598, + "grad_norm": 0.6122305393218994, + "learning_rate": 2.929608938547486e-05, + "loss": 0.1818, + "step": 8745 + }, + { + "epoch": 24.43016759776536, + "grad_norm": 0.676548421382904, + "learning_rate": 2.9299441340782125e-05, + "loss": 0.1662, + "step": 8746 + }, + { + "epoch": 24.432960893854748, + "grad_norm": 0.5590386390686035, + "learning_rate": 2.9302793296089387e-05, + "loss": 0.1698, + "step": 8747 + }, + { + "epoch": 24.435754189944134, + "grad_norm": 0.616129457950592, + "learning_rate": 2.9306145251396648e-05, + "loss": 0.1492, + "step": 8748 + }, + { + "epoch": 24.43854748603352, + "grad_norm": 0.5140752196311951, + "learning_rate": 2.9309497206703913e-05, + "loss": 0.1548, + "step": 8749 + }, + { + "epoch": 24.441340782122904, + "grad_norm": 0.48379719257354736, + "learning_rate": 2.9312849162011174e-05, + "loss": 0.1441, + "step": 8750 + }, + { + "epoch": 24.44413407821229, + "grad_norm": 0.7830666303634644, + "learning_rate": 2.9316201117318436e-05, + "loss": 0.1342, + "step": 8751 + }, + { + "epoch": 24.446927374301676, + "grad_norm": 0.458118736743927, + "learning_rate": 2.93195530726257e-05, + "loss": 0.1526, + "step": 8752 + }, + { + "epoch": 24.449720670391063, + "grad_norm": 0.7009314298629761, + "learning_rate": 2.932290502793296e-05, + "loss": 0.1499, + "step": 8753 + }, + { + "epoch": 24.452513966480446, + "grad_norm": 0.5757362246513367, + "learning_rate": 2.9326256983240223e-05, + "loss": 0.1394, + "step": 8754 + }, + { + "epoch": 24.455307262569832, + "grad_norm": 0.47178205847740173, + "learning_rate": 2.9329608938547488e-05, + "loss": 0.1626, + "step": 8755 + }, + { + "epoch": 24.45810055865922, + "grad_norm": 0.5503423810005188, + "learning_rate": 2.933296089385475e-05, + "loss": 0.147, + "step": 8756 + }, + { + "epoch": 24.460893854748605, + "grad_norm": 0.6040812730789185, + "learning_rate": 2.933631284916201e-05, + "loss": 0.1519, + "step": 8757 + }, + { + "epoch": 24.463687150837988, + "grad_norm": 0.7142344117164612, + "learning_rate": 2.9339664804469275e-05, + "loss": 0.1225, + "step": 8758 + }, + { + "epoch": 24.466480446927374, + "grad_norm": 0.5910365581512451, + "learning_rate": 2.9343016759776537e-05, + "loss": 0.1535, + "step": 8759 + }, + { + "epoch": 24.46927374301676, + "grad_norm": 0.5608110427856445, + "learning_rate": 2.93463687150838e-05, + "loss": 0.1426, + "step": 8760 + }, + { + "epoch": 24.472067039106147, + "grad_norm": 1.1458981037139893, + "learning_rate": 2.9349720670391063e-05, + "loss": 0.1592, + "step": 8761 + }, + { + "epoch": 24.47486033519553, + "grad_norm": 1.0237691402435303, + "learning_rate": 2.9353072625698324e-05, + "loss": 0.1509, + "step": 8762 + }, + { + "epoch": 24.477653631284916, + "grad_norm": 0.7380020022392273, + "learning_rate": 2.935642458100559e-05, + "loss": 0.1605, + "step": 8763 + }, + { + "epoch": 24.480446927374302, + "grad_norm": 3.7778961658477783, + "learning_rate": 2.935977653631285e-05, + "loss": 0.175, + "step": 8764 + }, + { + "epoch": 24.48324022346369, + "grad_norm": 1.1555372476577759, + "learning_rate": 2.936312849162011e-05, + "loss": 0.1386, + "step": 8765 + }, + { + "epoch": 24.48603351955307, + "grad_norm": 0.8670870065689087, + "learning_rate": 2.9366480446927376e-05, + "loss": 0.1598, + "step": 8766 + }, + { + "epoch": 24.488826815642458, + "grad_norm": 3.088646650314331, + "learning_rate": 2.9369832402234637e-05, + "loss": 0.2195, + "step": 8767 + }, + { + "epoch": 24.491620111731844, + "grad_norm": 0.42601341009140015, + "learning_rate": 2.93731843575419e-05, + "loss": 0.1576, + "step": 8768 + }, + { + "epoch": 24.49441340782123, + "grad_norm": 0.45518872141838074, + "learning_rate": 2.9376536312849164e-05, + "loss": 0.171, + "step": 8769 + }, + { + "epoch": 24.497206703910614, + "grad_norm": 0.826958954334259, + "learning_rate": 2.9379888268156425e-05, + "loss": 0.1729, + "step": 8770 + }, + { + "epoch": 24.5, + "grad_norm": 0.9718818664550781, + "learning_rate": 2.9383240223463686e-05, + "loss": 0.1475, + "step": 8771 + }, + { + "epoch": 24.502793296089386, + "grad_norm": 0.5186784267425537, + "learning_rate": 2.938659217877095e-05, + "loss": 0.17, + "step": 8772 + }, + { + "epoch": 24.505586592178773, + "grad_norm": 0.4647636115550995, + "learning_rate": 2.9389944134078212e-05, + "loss": 0.1771, + "step": 8773 + }, + { + "epoch": 24.508379888268156, + "grad_norm": 0.5396313071250916, + "learning_rate": 2.9393296089385474e-05, + "loss": 0.1639, + "step": 8774 + }, + { + "epoch": 24.511173184357542, + "grad_norm": 0.784922182559967, + "learning_rate": 2.939664804469274e-05, + "loss": 0.1669, + "step": 8775 + }, + { + "epoch": 24.51396648044693, + "grad_norm": 1.170357584953308, + "learning_rate": 2.94e-05, + "loss": 0.1648, + "step": 8776 + }, + { + "epoch": 24.51675977653631, + "grad_norm": 1.0031230449676514, + "learning_rate": 2.9403351955307264e-05, + "loss": 0.1671, + "step": 8777 + }, + { + "epoch": 24.519553072625698, + "grad_norm": 2.1590240001678467, + "learning_rate": 2.9406703910614526e-05, + "loss": 0.153, + "step": 8778 + }, + { + "epoch": 24.522346368715084, + "grad_norm": 0.8800271153450012, + "learning_rate": 2.9410055865921787e-05, + "loss": 0.1623, + "step": 8779 + }, + { + "epoch": 24.52513966480447, + "grad_norm": 0.6935554146766663, + "learning_rate": 2.9413407821229052e-05, + "loss": 0.1741, + "step": 8780 + }, + { + "epoch": 24.527932960893853, + "grad_norm": 0.6770201325416565, + "learning_rate": 2.9416759776536313e-05, + "loss": 0.1427, + "step": 8781 + }, + { + "epoch": 24.53072625698324, + "grad_norm": 0.5729080438613892, + "learning_rate": 2.9420111731843575e-05, + "loss": 0.1493, + "step": 8782 + }, + { + "epoch": 24.533519553072626, + "grad_norm": 0.5635392665863037, + "learning_rate": 2.942346368715084e-05, + "loss": 0.1406, + "step": 8783 + }, + { + "epoch": 24.536312849162012, + "grad_norm": 0.7195382714271545, + "learning_rate": 2.94268156424581e-05, + "loss": 0.1268, + "step": 8784 + }, + { + "epoch": 24.539106145251395, + "grad_norm": 0.6560285687446594, + "learning_rate": 2.9430167597765362e-05, + "loss": 0.1486, + "step": 8785 + }, + { + "epoch": 24.54189944134078, + "grad_norm": 0.6186663508415222, + "learning_rate": 2.9433519553072627e-05, + "loss": 0.1339, + "step": 8786 + }, + { + "epoch": 24.544692737430168, + "grad_norm": 1.1747981309890747, + "learning_rate": 2.9436871508379888e-05, + "loss": 0.1505, + "step": 8787 + }, + { + "epoch": 24.547486033519554, + "grad_norm": 1.3283145427703857, + "learning_rate": 2.944022346368715e-05, + "loss": 0.1549, + "step": 8788 + }, + { + "epoch": 24.550279329608937, + "grad_norm": 0.8280034065246582, + "learning_rate": 2.9443575418994414e-05, + "loss": 0.1506, + "step": 8789 + }, + { + "epoch": 24.553072625698324, + "grad_norm": 0.8426381945610046, + "learning_rate": 2.9446927374301676e-05, + "loss": 0.1688, + "step": 8790 + }, + { + "epoch": 24.55586592178771, + "grad_norm": 1.2370936870574951, + "learning_rate": 2.9450279329608937e-05, + "loss": 0.1572, + "step": 8791 + }, + { + "epoch": 24.558659217877096, + "grad_norm": 1.861057162284851, + "learning_rate": 2.94536312849162e-05, + "loss": 0.1869, + "step": 8792 + }, + { + "epoch": 24.56145251396648, + "grad_norm": 0.601140022277832, + "learning_rate": 2.9456983240223463e-05, + "loss": 0.2144, + "step": 8793 + }, + { + "epoch": 24.564245810055866, + "grad_norm": 0.5059716701507568, + "learning_rate": 2.9460335195530724e-05, + "loss": 0.1672, + "step": 8794 + }, + { + "epoch": 24.567039106145252, + "grad_norm": 0.4805697500705719, + "learning_rate": 2.946368715083799e-05, + "loss": 0.1495, + "step": 8795 + }, + { + "epoch": 24.56983240223464, + "grad_norm": 0.42694830894470215, + "learning_rate": 2.946703910614525e-05, + "loss": 0.1545, + "step": 8796 + }, + { + "epoch": 24.57262569832402, + "grad_norm": 0.3958050310611725, + "learning_rate": 2.9470391061452515e-05, + "loss": 0.1611, + "step": 8797 + }, + { + "epoch": 24.575418994413408, + "grad_norm": 0.774499773979187, + "learning_rate": 2.9473743016759776e-05, + "loss": 0.1831, + "step": 8798 + }, + { + "epoch": 24.578212290502794, + "grad_norm": 2.221261501312256, + "learning_rate": 2.9477094972067038e-05, + "loss": 0.1503, + "step": 8799 + }, + { + "epoch": 24.58100558659218, + "grad_norm": 0.6851657032966614, + "learning_rate": 2.9480446927374303e-05, + "loss": 0.1775, + "step": 8800 + }, + { + "epoch": 24.583798882681563, + "grad_norm": 0.6376696825027466, + "learning_rate": 2.9483798882681564e-05, + "loss": 0.1527, + "step": 8801 + }, + { + "epoch": 24.58659217877095, + "grad_norm": 0.9150886535644531, + "learning_rate": 2.9487150837988825e-05, + "loss": 0.1709, + "step": 8802 + }, + { + "epoch": 24.589385474860336, + "grad_norm": 0.43033096194267273, + "learning_rate": 2.949050279329609e-05, + "loss": 0.1331, + "step": 8803 + }, + { + "epoch": 24.592178770949722, + "grad_norm": 0.6573735475540161, + "learning_rate": 2.949385474860335e-05, + "loss": 0.1654, + "step": 8804 + }, + { + "epoch": 24.594972067039105, + "grad_norm": 0.7152764797210693, + "learning_rate": 2.9497206703910613e-05, + "loss": 0.1479, + "step": 8805 + }, + { + "epoch": 24.59776536312849, + "grad_norm": 0.7562827467918396, + "learning_rate": 2.9500558659217877e-05, + "loss": 0.1683, + "step": 8806 + }, + { + "epoch": 24.600558659217878, + "grad_norm": 0.6832447052001953, + "learning_rate": 2.950391061452514e-05, + "loss": 0.1266, + "step": 8807 + }, + { + "epoch": 24.60335195530726, + "grad_norm": 0.9578016400337219, + "learning_rate": 2.95072625698324e-05, + "loss": 0.1762, + "step": 8808 + }, + { + "epoch": 24.606145251396647, + "grad_norm": 1.4739888906478882, + "learning_rate": 2.9510614525139665e-05, + "loss": 0.1608, + "step": 8809 + }, + { + "epoch": 24.608938547486034, + "grad_norm": 1.1300346851348877, + "learning_rate": 2.9513966480446926e-05, + "loss": 0.1917, + "step": 8810 + }, + { + "epoch": 24.61173184357542, + "grad_norm": 1.1609266996383667, + "learning_rate": 2.951731843575419e-05, + "loss": 0.1405, + "step": 8811 + }, + { + "epoch": 24.614525139664803, + "grad_norm": 1.1488573551177979, + "learning_rate": 2.9520670391061456e-05, + "loss": 0.1445, + "step": 8812 + }, + { + "epoch": 24.61731843575419, + "grad_norm": 1.5745598077774048, + "learning_rate": 2.9524022346368717e-05, + "loss": 0.1788, + "step": 8813 + }, + { + "epoch": 24.620111731843576, + "grad_norm": 1.1697578430175781, + "learning_rate": 2.952737430167598e-05, + "loss": 0.1693, + "step": 8814 + }, + { + "epoch": 24.622905027932962, + "grad_norm": 1.100699543952942, + "learning_rate": 2.9530726256983243e-05, + "loss": 0.1587, + "step": 8815 + }, + { + "epoch": 24.625698324022345, + "grad_norm": 2.0486388206481934, + "learning_rate": 2.9534078212290504e-05, + "loss": 0.1619, + "step": 8816 + }, + { + "epoch": 24.62849162011173, + "grad_norm": 1.9657155275344849, + "learning_rate": 2.953743016759777e-05, + "loss": 0.2053, + "step": 8817 + }, + { + "epoch": 24.631284916201118, + "grad_norm": 0.6888479590415955, + "learning_rate": 2.954078212290503e-05, + "loss": 0.1765, + "step": 8818 + }, + { + "epoch": 24.634078212290504, + "grad_norm": 1.1807235479354858, + "learning_rate": 2.9544134078212292e-05, + "loss": 0.1556, + "step": 8819 + }, + { + "epoch": 24.636871508379887, + "grad_norm": 0.6090859770774841, + "learning_rate": 2.9547486033519557e-05, + "loss": 0.1838, + "step": 8820 + }, + { + "epoch": 24.639664804469273, + "grad_norm": 0.9757060408592224, + "learning_rate": 2.9550837988826818e-05, + "loss": 0.1595, + "step": 8821 + }, + { + "epoch": 24.64245810055866, + "grad_norm": 0.8062103986740112, + "learning_rate": 2.955418994413408e-05, + "loss": 0.1603, + "step": 8822 + }, + { + "epoch": 24.645251396648046, + "grad_norm": 0.5955566763877869, + "learning_rate": 2.9557541899441344e-05, + "loss": 0.1678, + "step": 8823 + }, + { + "epoch": 24.64804469273743, + "grad_norm": 0.5068458914756775, + "learning_rate": 2.9560893854748605e-05, + "loss": 0.1461, + "step": 8824 + }, + { + "epoch": 24.650837988826815, + "grad_norm": 0.8378484845161438, + "learning_rate": 2.9564245810055867e-05, + "loss": 0.1703, + "step": 8825 + }, + { + "epoch": 24.6536312849162, + "grad_norm": 0.6858642101287842, + "learning_rate": 2.956759776536313e-05, + "loss": 0.1586, + "step": 8826 + }, + { + "epoch": 24.656424581005588, + "grad_norm": 1.3608636856079102, + "learning_rate": 2.9570949720670393e-05, + "loss": 0.1499, + "step": 8827 + }, + { + "epoch": 24.65921787709497, + "grad_norm": 0.6801092028617859, + "learning_rate": 2.9574301675977654e-05, + "loss": 0.1934, + "step": 8828 + }, + { + "epoch": 24.662011173184357, + "grad_norm": 0.6128574013710022, + "learning_rate": 2.957765363128492e-05, + "loss": 0.1467, + "step": 8829 + }, + { + "epoch": 24.664804469273744, + "grad_norm": 0.733392596244812, + "learning_rate": 2.958100558659218e-05, + "loss": 0.1661, + "step": 8830 + }, + { + "epoch": 24.66759776536313, + "grad_norm": 1.0703340768814087, + "learning_rate": 2.958435754189944e-05, + "loss": 0.1556, + "step": 8831 + }, + { + "epoch": 24.670391061452513, + "grad_norm": 0.44056782126426697, + "learning_rate": 2.9587709497206706e-05, + "loss": 0.1362, + "step": 8832 + }, + { + "epoch": 24.6731843575419, + "grad_norm": 1.1497702598571777, + "learning_rate": 2.9591061452513968e-05, + "loss": 0.1652, + "step": 8833 + }, + { + "epoch": 24.675977653631286, + "grad_norm": 0.9903620481491089, + "learning_rate": 2.9594413407821232e-05, + "loss": 0.1366, + "step": 8834 + }, + { + "epoch": 24.678770949720672, + "grad_norm": 0.9160982370376587, + "learning_rate": 2.9597765363128494e-05, + "loss": 0.1609, + "step": 8835 + }, + { + "epoch": 24.681564245810055, + "grad_norm": 1.2652119398117065, + "learning_rate": 2.9601117318435755e-05, + "loss": 0.142, + "step": 8836 + }, + { + "epoch": 24.68435754189944, + "grad_norm": 0.5930799245834351, + "learning_rate": 2.960446927374302e-05, + "loss": 0.144, + "step": 8837 + }, + { + "epoch": 24.687150837988828, + "grad_norm": 0.7912847995758057, + "learning_rate": 2.960782122905028e-05, + "loss": 0.1346, + "step": 8838 + }, + { + "epoch": 24.689944134078214, + "grad_norm": 0.9316573739051819, + "learning_rate": 2.9611173184357542e-05, + "loss": 0.1657, + "step": 8839 + }, + { + "epoch": 24.692737430167597, + "grad_norm": 1.0616000890731812, + "learning_rate": 2.9614525139664807e-05, + "loss": 0.1536, + "step": 8840 + }, + { + "epoch": 24.695530726256983, + "grad_norm": 1.615958571434021, + "learning_rate": 2.961787709497207e-05, + "loss": 0.1835, + "step": 8841 + }, + { + "epoch": 24.69832402234637, + "grad_norm": 1.3294988870620728, + "learning_rate": 2.962122905027933e-05, + "loss": 0.2046, + "step": 8842 + }, + { + "epoch": 24.701117318435756, + "grad_norm": 0.5647497773170471, + "learning_rate": 2.9624581005586595e-05, + "loss": 0.155, + "step": 8843 + }, + { + "epoch": 24.70391061452514, + "grad_norm": 0.45876407623291016, + "learning_rate": 2.9627932960893856e-05, + "loss": 0.1408, + "step": 8844 + }, + { + "epoch": 24.706703910614525, + "grad_norm": 0.5550830960273743, + "learning_rate": 2.9631284916201117e-05, + "loss": 0.1775, + "step": 8845 + }, + { + "epoch": 24.70949720670391, + "grad_norm": 1.5683966875076294, + "learning_rate": 2.9634636871508382e-05, + "loss": 0.1558, + "step": 8846 + }, + { + "epoch": 24.712290502793294, + "grad_norm": 0.7632617950439453, + "learning_rate": 2.9637988826815643e-05, + "loss": 0.1808, + "step": 8847 + }, + { + "epoch": 24.71508379888268, + "grad_norm": 0.5362438559532166, + "learning_rate": 2.9641340782122905e-05, + "loss": 0.1628, + "step": 8848 + }, + { + "epoch": 24.717877094972067, + "grad_norm": 0.6616259217262268, + "learning_rate": 2.964469273743017e-05, + "loss": 0.1895, + "step": 8849 + }, + { + "epoch": 24.720670391061454, + "grad_norm": 0.6106072068214417, + "learning_rate": 2.964804469273743e-05, + "loss": 0.1461, + "step": 8850 + }, + { + "epoch": 24.723463687150836, + "grad_norm": 1.0441797971725464, + "learning_rate": 2.9651396648044692e-05, + "loss": 0.164, + "step": 8851 + }, + { + "epoch": 24.726256983240223, + "grad_norm": 7.742063522338867, + "learning_rate": 2.9654748603351957e-05, + "loss": 0.1733, + "step": 8852 + }, + { + "epoch": 24.72905027932961, + "grad_norm": 2.2140138149261475, + "learning_rate": 2.9658100558659218e-05, + "loss": 0.1922, + "step": 8853 + }, + { + "epoch": 24.731843575418996, + "grad_norm": 1.0894243717193604, + "learning_rate": 2.9661452513966483e-05, + "loss": 0.1697, + "step": 8854 + }, + { + "epoch": 24.73463687150838, + "grad_norm": 0.5802792310714722, + "learning_rate": 2.9664804469273744e-05, + "loss": 0.1769, + "step": 8855 + }, + { + "epoch": 24.737430167597765, + "grad_norm": 0.6731798648834229, + "learning_rate": 2.9668156424581006e-05, + "loss": 0.1698, + "step": 8856 + }, + { + "epoch": 24.74022346368715, + "grad_norm": 1.6078875064849854, + "learning_rate": 2.967150837988827e-05, + "loss": 0.153, + "step": 8857 + }, + { + "epoch": 24.743016759776538, + "grad_norm": 0.7859331965446472, + "learning_rate": 2.967486033519553e-05, + "loss": 0.1452, + "step": 8858 + }, + { + "epoch": 24.74581005586592, + "grad_norm": 1.100340723991394, + "learning_rate": 2.9678212290502793e-05, + "loss": 0.18, + "step": 8859 + }, + { + "epoch": 24.748603351955307, + "grad_norm": 0.8694734573364258, + "learning_rate": 2.9681564245810058e-05, + "loss": 0.1204, + "step": 8860 + }, + { + "epoch": 24.751396648044693, + "grad_norm": 6.404810428619385, + "learning_rate": 2.968491620111732e-05, + "loss": 0.1417, + "step": 8861 + }, + { + "epoch": 24.75418994413408, + "grad_norm": 0.7356884479522705, + "learning_rate": 2.968826815642458e-05, + "loss": 0.1385, + "step": 8862 + }, + { + "epoch": 24.756983240223462, + "grad_norm": 1.247222900390625, + "learning_rate": 2.9691620111731845e-05, + "loss": 0.1375, + "step": 8863 + }, + { + "epoch": 24.75977653631285, + "grad_norm": 1.6550911664962769, + "learning_rate": 2.9694972067039107e-05, + "loss": 0.1651, + "step": 8864 + }, + { + "epoch": 24.762569832402235, + "grad_norm": 1.1067975759506226, + "learning_rate": 2.9698324022346368e-05, + "loss": 0.1267, + "step": 8865 + }, + { + "epoch": 24.76536312849162, + "grad_norm": 0.8494910597801208, + "learning_rate": 2.9701675977653633e-05, + "loss": 0.1772, + "step": 8866 + }, + { + "epoch": 24.768156424581004, + "grad_norm": 1.9702320098876953, + "learning_rate": 2.9705027932960894e-05, + "loss": 0.1875, + "step": 8867 + }, + { + "epoch": 24.77094972067039, + "grad_norm": 1.9514011144638062, + "learning_rate": 2.9708379888268155e-05, + "loss": 0.1592, + "step": 8868 + }, + { + "epoch": 24.773743016759777, + "grad_norm": 0.7021152377128601, + "learning_rate": 2.971173184357542e-05, + "loss": 0.161, + "step": 8869 + }, + { + "epoch": 24.776536312849164, + "grad_norm": 0.6333081722259521, + "learning_rate": 2.971508379888268e-05, + "loss": 0.1687, + "step": 8870 + }, + { + "epoch": 24.779329608938546, + "grad_norm": 0.624336302280426, + "learning_rate": 2.9718435754189946e-05, + "loss": 0.1819, + "step": 8871 + }, + { + "epoch": 24.782122905027933, + "grad_norm": 0.6029762625694275, + "learning_rate": 2.9721787709497207e-05, + "loss": 0.1482, + "step": 8872 + }, + { + "epoch": 24.78491620111732, + "grad_norm": 1.101486086845398, + "learning_rate": 2.972513966480447e-05, + "loss": 0.1494, + "step": 8873 + }, + { + "epoch": 24.787709497206706, + "grad_norm": 0.5784600973129272, + "learning_rate": 2.9728491620111734e-05, + "loss": 0.1477, + "step": 8874 + }, + { + "epoch": 24.79050279329609, + "grad_norm": 1.2203857898712158, + "learning_rate": 2.9731843575418995e-05, + "loss": 0.1676, + "step": 8875 + }, + { + "epoch": 24.793296089385475, + "grad_norm": 1.4520492553710938, + "learning_rate": 2.9735195530726256e-05, + "loss": 0.1478, + "step": 8876 + }, + { + "epoch": 24.79608938547486, + "grad_norm": 0.7843199372291565, + "learning_rate": 2.973854748603352e-05, + "loss": 0.1757, + "step": 8877 + }, + { + "epoch": 24.798882681564244, + "grad_norm": 1.9374760389328003, + "learning_rate": 2.9741899441340782e-05, + "loss": 0.1615, + "step": 8878 + }, + { + "epoch": 24.80167597765363, + "grad_norm": 0.6162958145141602, + "learning_rate": 2.9745251396648044e-05, + "loss": 0.1519, + "step": 8879 + }, + { + "epoch": 24.804469273743017, + "grad_norm": 0.5332781672477722, + "learning_rate": 2.974860335195531e-05, + "loss": 0.1443, + "step": 8880 + }, + { + "epoch": 24.807262569832403, + "grad_norm": 0.6904085278511047, + "learning_rate": 2.975195530726257e-05, + "loss": 0.1792, + "step": 8881 + }, + { + "epoch": 24.810055865921786, + "grad_norm": 0.5481195449829102, + "learning_rate": 2.975530726256983e-05, + "loss": 0.1425, + "step": 8882 + }, + { + "epoch": 24.812849162011172, + "grad_norm": 1.0835037231445312, + "learning_rate": 2.9758659217877096e-05, + "loss": 0.1578, + "step": 8883 + }, + { + "epoch": 24.81564245810056, + "grad_norm": 0.6530140042304993, + "learning_rate": 2.9762011173184357e-05, + "loss": 0.1739, + "step": 8884 + }, + { + "epoch": 24.818435754189945, + "grad_norm": 0.7897777557373047, + "learning_rate": 2.976536312849162e-05, + "loss": 0.1517, + "step": 8885 + }, + { + "epoch": 24.821229050279328, + "grad_norm": 0.6960803866386414, + "learning_rate": 2.9768715083798883e-05, + "loss": 0.1244, + "step": 8886 + }, + { + "epoch": 24.824022346368714, + "grad_norm": 1.1538068056106567, + "learning_rate": 2.9772067039106145e-05, + "loss": 0.1689, + "step": 8887 + }, + { + "epoch": 24.8268156424581, + "grad_norm": 0.7732382416725159, + "learning_rate": 2.9775418994413406e-05, + "loss": 0.1351, + "step": 8888 + }, + { + "epoch": 24.829608938547487, + "grad_norm": 1.1636961698532104, + "learning_rate": 2.977877094972067e-05, + "loss": 0.1831, + "step": 8889 + }, + { + "epoch": 24.83240223463687, + "grad_norm": 1.1948699951171875, + "learning_rate": 2.9782122905027932e-05, + "loss": 0.1597, + "step": 8890 + }, + { + "epoch": 24.835195530726256, + "grad_norm": 1.0128284692764282, + "learning_rate": 2.9785474860335197e-05, + "loss": 0.1652, + "step": 8891 + }, + { + "epoch": 24.837988826815643, + "grad_norm": 2.335021495819092, + "learning_rate": 2.9788826815642458e-05, + "loss": 0.2045, + "step": 8892 + }, + { + "epoch": 24.84078212290503, + "grad_norm": 0.5534773468971252, + "learning_rate": 2.979217877094972e-05, + "loss": 0.1588, + "step": 8893 + }, + { + "epoch": 24.843575418994412, + "grad_norm": 0.5441893339157104, + "learning_rate": 2.9795530726256984e-05, + "loss": 0.179, + "step": 8894 + }, + { + "epoch": 24.8463687150838, + "grad_norm": 0.5431897044181824, + "learning_rate": 2.9798882681564246e-05, + "loss": 0.1758, + "step": 8895 + }, + { + "epoch": 24.849162011173185, + "grad_norm": 0.3716883957386017, + "learning_rate": 2.9802234636871507e-05, + "loss": 0.1345, + "step": 8896 + }, + { + "epoch": 24.85195530726257, + "grad_norm": 0.5610336661338806, + "learning_rate": 2.980558659217877e-05, + "loss": 0.169, + "step": 8897 + }, + { + "epoch": 24.854748603351954, + "grad_norm": 0.519856870174408, + "learning_rate": 2.9808938547486033e-05, + "loss": 0.1496, + "step": 8898 + }, + { + "epoch": 24.85754189944134, + "grad_norm": 0.9846234321594238, + "learning_rate": 2.9812290502793294e-05, + "loss": 0.1697, + "step": 8899 + }, + { + "epoch": 24.860335195530727, + "grad_norm": 1.121353030204773, + "learning_rate": 2.981564245810056e-05, + "loss": 0.1731, + "step": 8900 + }, + { + "epoch": 24.863128491620113, + "grad_norm": 1.3531159162521362, + "learning_rate": 2.981899441340782e-05, + "loss": 0.1523, + "step": 8901 + }, + { + "epoch": 24.865921787709496, + "grad_norm": 0.5906012058258057, + "learning_rate": 2.9822346368715082e-05, + "loss": 0.1474, + "step": 8902 + }, + { + "epoch": 24.868715083798882, + "grad_norm": 0.5210424065589905, + "learning_rate": 2.9825698324022346e-05, + "loss": 0.1707, + "step": 8903 + }, + { + "epoch": 24.87150837988827, + "grad_norm": 0.45649486780166626, + "learning_rate": 2.9829050279329608e-05, + "loss": 0.1351, + "step": 8904 + }, + { + "epoch": 24.874301675977655, + "grad_norm": 1.2539185285568237, + "learning_rate": 2.983240223463687e-05, + "loss": 0.1655, + "step": 8905 + }, + { + "epoch": 24.877094972067038, + "grad_norm": 0.7313699126243591, + "learning_rate": 2.9835754189944134e-05, + "loss": 0.1518, + "step": 8906 + }, + { + "epoch": 24.879888268156424, + "grad_norm": 0.8505001664161682, + "learning_rate": 2.98391061452514e-05, + "loss": 0.1285, + "step": 8907 + }, + { + "epoch": 24.88268156424581, + "grad_norm": 0.7679824233055115, + "learning_rate": 2.9842458100558663e-05, + "loss": 0.1386, + "step": 8908 + }, + { + "epoch": 24.885474860335197, + "grad_norm": 0.6724974513053894, + "learning_rate": 2.9845810055865925e-05, + "loss": 0.1502, + "step": 8909 + }, + { + "epoch": 24.88826815642458, + "grad_norm": 1.0421992540359497, + "learning_rate": 2.9849162011173186e-05, + "loss": 0.1423, + "step": 8910 + }, + { + "epoch": 24.891061452513966, + "grad_norm": 0.6626574993133545, + "learning_rate": 2.985251396648045e-05, + "loss": 0.1309, + "step": 8911 + }, + { + "epoch": 24.893854748603353, + "grad_norm": 1.106871485710144, + "learning_rate": 2.9855865921787712e-05, + "loss": 0.1599, + "step": 8912 + }, + { + "epoch": 24.89664804469274, + "grad_norm": 1.1171191930770874, + "learning_rate": 2.9859217877094973e-05, + "loss": 0.1415, + "step": 8913 + }, + { + "epoch": 24.899441340782122, + "grad_norm": 1.5205837488174438, + "learning_rate": 2.9862569832402238e-05, + "loss": 0.1444, + "step": 8914 + }, + { + "epoch": 24.90223463687151, + "grad_norm": 1.1348809003829956, + "learning_rate": 2.98659217877095e-05, + "loss": 0.1444, + "step": 8915 + }, + { + "epoch": 24.905027932960895, + "grad_norm": 1.3903964757919312, + "learning_rate": 2.986927374301676e-05, + "loss": 0.1724, + "step": 8916 + }, + { + "epoch": 24.907821229050278, + "grad_norm": 2.449398994445801, + "learning_rate": 2.9872625698324026e-05, + "loss": 0.2294, + "step": 8917 + }, + { + "epoch": 24.910614525139664, + "grad_norm": 0.6369403004646301, + "learning_rate": 2.9875977653631287e-05, + "loss": 0.2088, + "step": 8918 + }, + { + "epoch": 24.91340782122905, + "grad_norm": 0.42742404341697693, + "learning_rate": 2.9879329608938548e-05, + "loss": 0.1726, + "step": 8919 + }, + { + "epoch": 24.916201117318437, + "grad_norm": 0.5571948885917664, + "learning_rate": 2.9882681564245813e-05, + "loss": 0.1656, + "step": 8920 + }, + { + "epoch": 24.91899441340782, + "grad_norm": 0.5207695364952087, + "learning_rate": 2.9886033519553074e-05, + "loss": 0.1579, + "step": 8921 + }, + { + "epoch": 24.921787709497206, + "grad_norm": 0.7924759984016418, + "learning_rate": 2.9889385474860336e-05, + "loss": 0.1625, + "step": 8922 + }, + { + "epoch": 24.924581005586592, + "grad_norm": 1.7121155261993408, + "learning_rate": 2.98927374301676e-05, + "loss": 0.18, + "step": 8923 + }, + { + "epoch": 24.92737430167598, + "grad_norm": 0.5324057340621948, + "learning_rate": 2.9896089385474862e-05, + "loss": 0.1481, + "step": 8924 + }, + { + "epoch": 24.93016759776536, + "grad_norm": 0.6118828654289246, + "learning_rate": 2.9899441340782123e-05, + "loss": 0.164, + "step": 8925 + }, + { + "epoch": 24.932960893854748, + "grad_norm": 0.554201602935791, + "learning_rate": 2.9902793296089388e-05, + "loss": 0.1511, + "step": 8926 + }, + { + "epoch": 24.935754189944134, + "grad_norm": 0.4670882523059845, + "learning_rate": 2.990614525139665e-05, + "loss": 0.1573, + "step": 8927 + }, + { + "epoch": 24.93854748603352, + "grad_norm": 0.5055104494094849, + "learning_rate": 2.9909497206703914e-05, + "loss": 0.1082, + "step": 8928 + }, + { + "epoch": 24.941340782122904, + "grad_norm": 0.6629941463470459, + "learning_rate": 2.9912849162011175e-05, + "loss": 0.1509, + "step": 8929 + }, + { + "epoch": 24.94413407821229, + "grad_norm": 0.8676555752754211, + "learning_rate": 2.9916201117318437e-05, + "loss": 0.1378, + "step": 8930 + }, + { + "epoch": 24.946927374301676, + "grad_norm": 0.7783005237579346, + "learning_rate": 2.99195530726257e-05, + "loss": 0.1475, + "step": 8931 + }, + { + "epoch": 24.949720670391063, + "grad_norm": 1.0533597469329834, + "learning_rate": 2.9922905027932963e-05, + "loss": 0.1822, + "step": 8932 + }, + { + "epoch": 24.952513966480446, + "grad_norm": 0.6286926865577698, + "learning_rate": 2.9926256983240224e-05, + "loss": 0.1469, + "step": 8933 + }, + { + "epoch": 24.955307262569832, + "grad_norm": 1.200900912284851, + "learning_rate": 2.992960893854749e-05, + "loss": 0.1604, + "step": 8934 + }, + { + "epoch": 24.95810055865922, + "grad_norm": 0.9240032434463501, + "learning_rate": 2.993296089385475e-05, + "loss": 0.1485, + "step": 8935 + }, + { + "epoch": 24.960893854748605, + "grad_norm": 0.7974162697792053, + "learning_rate": 2.993631284916201e-05, + "loss": 0.1533, + "step": 8936 + }, + { + "epoch": 24.963687150837988, + "grad_norm": 2.1000633239746094, + "learning_rate": 2.9939664804469276e-05, + "loss": 0.1341, + "step": 8937 + }, + { + "epoch": 24.966480446927374, + "grad_norm": 1.1655693054199219, + "learning_rate": 2.9943016759776538e-05, + "loss": 0.1564, + "step": 8938 + }, + { + "epoch": 24.96927374301676, + "grad_norm": 1.0163437128067017, + "learning_rate": 2.99463687150838e-05, + "loss": 0.1722, + "step": 8939 + }, + { + "epoch": 24.972067039106147, + "grad_norm": 0.8153162598609924, + "learning_rate": 2.9949720670391064e-05, + "loss": 0.1472, + "step": 8940 + }, + { + "epoch": 24.97486033519553, + "grad_norm": 3.0208675861358643, + "learning_rate": 2.9953072625698325e-05, + "loss": 0.1995, + "step": 8941 + }, + { + "epoch": 24.977653631284916, + "grad_norm": 1.61673903465271, + "learning_rate": 2.9956424581005586e-05, + "loss": 0.2686, + "step": 8942 + }, + { + "epoch": 24.980446927374302, + "grad_norm": 1.1198093891143799, + "learning_rate": 2.995977653631285e-05, + "loss": 0.1987, + "step": 8943 + }, + { + "epoch": 24.98324022346369, + "grad_norm": 1.1633625030517578, + "learning_rate": 2.9963128491620112e-05, + "loss": 0.1992, + "step": 8944 + }, + { + "epoch": 24.98603351955307, + "grad_norm": 0.7779914736747742, + "learning_rate": 2.9966480446927374e-05, + "loss": 0.1609, + "step": 8945 + }, + { + "epoch": 24.988826815642458, + "grad_norm": 0.7949780225753784, + "learning_rate": 2.996983240223464e-05, + "loss": 0.1809, + "step": 8946 + }, + { + "epoch": 24.991620111731844, + "grad_norm": 0.6500733494758606, + "learning_rate": 2.99731843575419e-05, + "loss": 0.179, + "step": 8947 + }, + { + "epoch": 24.994413407821227, + "grad_norm": 2.389453172683716, + "learning_rate": 2.9976536312849165e-05, + "loss": 0.168, + "step": 8948 + }, + { + "epoch": 24.997206703910614, + "grad_norm": 0.6597934365272522, + "learning_rate": 2.9979888268156426e-05, + "loss": 0.1691, + "step": 8949 + }, + { + "epoch": 25.0, + "grad_norm": 1.14535653591156, + "learning_rate": 2.9983240223463687e-05, + "loss": 0.1967, + "step": 8950 + }, + { + "epoch": 25.002793296089386, + "grad_norm": 0.6736270785331726, + "learning_rate": 2.9986592178770952e-05, + "loss": 0.1919, + "step": 8951 + }, + { + "epoch": 25.00558659217877, + "grad_norm": 0.8841961026191711, + "learning_rate": 2.9989944134078213e-05, + "loss": 0.1551, + "step": 8952 + }, + { + "epoch": 25.008379888268156, + "grad_norm": 0.7234655618667603, + "learning_rate": 2.9993296089385475e-05, + "loss": 0.1712, + "step": 8953 + }, + { + "epoch": 25.011173184357542, + "grad_norm": 0.5566832423210144, + "learning_rate": 2.999664804469274e-05, + "loss": 0.1883, + "step": 8954 + }, + { + "epoch": 25.01396648044693, + "grad_norm": 1.0498926639556885, + "learning_rate": 3e-05, + "loss": 0.1541, + "step": 8955 + }, + { + "epoch": 25.01675977653631, + "grad_norm": 0.790528416633606, + "learning_rate": 2.9999627560521417e-05, + "loss": 0.1723, + "step": 8956 + }, + { + "epoch": 25.019553072625698, + "grad_norm": 0.8100999593734741, + "learning_rate": 2.9999255121042834e-05, + "loss": 0.1951, + "step": 8957 + }, + { + "epoch": 25.022346368715084, + "grad_norm": 0.7555283904075623, + "learning_rate": 2.9998882681564247e-05, + "loss": 0.1837, + "step": 8958 + }, + { + "epoch": 25.02513966480447, + "grad_norm": 0.6563220620155334, + "learning_rate": 2.999851024208566e-05, + "loss": 0.1656, + "step": 8959 + }, + { + "epoch": 25.027932960893853, + "grad_norm": 1.0344901084899902, + "learning_rate": 2.9998137802607077e-05, + "loss": 0.1493, + "step": 8960 + }, + { + "epoch": 25.03072625698324, + "grad_norm": 1.2399163246154785, + "learning_rate": 2.9997765363128493e-05, + "loss": 0.1704, + "step": 8961 + }, + { + "epoch": 25.033519553072626, + "grad_norm": 2.4205715656280518, + "learning_rate": 2.9997392923649906e-05, + "loss": 0.1499, + "step": 8962 + }, + { + "epoch": 25.036312849162012, + "grad_norm": 0.5811547040939331, + "learning_rate": 2.9997020484171323e-05, + "loss": 0.1603, + "step": 8963 + }, + { + "epoch": 25.039106145251395, + "grad_norm": 0.8624566197395325, + "learning_rate": 2.999664804469274e-05, + "loss": 0.1316, + "step": 8964 + }, + { + "epoch": 25.04189944134078, + "grad_norm": 1.0079821348190308, + "learning_rate": 2.9996275605214153e-05, + "loss": 0.1522, + "step": 8965 + }, + { + "epoch": 25.044692737430168, + "grad_norm": 0.7260400652885437, + "learning_rate": 2.999590316573557e-05, + "loss": 0.1693, + "step": 8966 + }, + { + "epoch": 25.047486033519554, + "grad_norm": 0.6331428289413452, + "learning_rate": 2.9995530726256982e-05, + "loss": 0.1286, + "step": 8967 + }, + { + "epoch": 25.050279329608937, + "grad_norm": 0.7783434987068176, + "learning_rate": 2.99951582867784e-05, + "loss": 0.1597, + "step": 8968 + }, + { + "epoch": 25.053072625698324, + "grad_norm": 0.7754214406013489, + "learning_rate": 2.9994785847299815e-05, + "loss": 0.1399, + "step": 8969 + }, + { + "epoch": 25.05586592178771, + "grad_norm": 0.7000446915626526, + "learning_rate": 2.9994413407821232e-05, + "loss": 0.1336, + "step": 8970 + }, + { + "epoch": 25.058659217877096, + "grad_norm": 1.0019088983535767, + "learning_rate": 2.9994040968342645e-05, + "loss": 0.1557, + "step": 8971 + }, + { + "epoch": 25.06145251396648, + "grad_norm": 0.8427886366844177, + "learning_rate": 2.9993668528864058e-05, + "loss": 0.1562, + "step": 8972 + }, + { + "epoch": 25.064245810055866, + "grad_norm": 0.7210360765457153, + "learning_rate": 2.9993296089385475e-05, + "loss": 0.1574, + "step": 8973 + }, + { + "epoch": 25.067039106145252, + "grad_norm": 1.3554476499557495, + "learning_rate": 2.999292364990689e-05, + "loss": 0.1952, + "step": 8974 + }, + { + "epoch": 25.06983240223464, + "grad_norm": 1.966795802116394, + "learning_rate": 2.9992551210428308e-05, + "loss": 0.2371, + "step": 8975 + }, + { + "epoch": 25.07262569832402, + "grad_norm": 0.5847614407539368, + "learning_rate": 2.999217877094972e-05, + "loss": 0.2209, + "step": 8976 + }, + { + "epoch": 25.075418994413408, + "grad_norm": 0.45562225580215454, + "learning_rate": 2.9991806331471137e-05, + "loss": 0.1471, + "step": 8977 + }, + { + "epoch": 25.078212290502794, + "grad_norm": 0.5547707676887512, + "learning_rate": 2.9991433891992554e-05, + "loss": 0.1309, + "step": 8978 + }, + { + "epoch": 25.08100558659218, + "grad_norm": 0.5664809346199036, + "learning_rate": 2.9991061452513967e-05, + "loss": 0.1557, + "step": 8979 + }, + { + "epoch": 25.083798882681563, + "grad_norm": 0.4757702052593231, + "learning_rate": 2.999068901303538e-05, + "loss": 0.1663, + "step": 8980 + }, + { + "epoch": 25.08659217877095, + "grad_norm": 0.7229875326156616, + "learning_rate": 2.9990316573556797e-05, + "loss": 0.1602, + "step": 8981 + }, + { + "epoch": 25.089385474860336, + "grad_norm": 0.647482693195343, + "learning_rate": 2.9989944134078213e-05, + "loss": 0.1579, + "step": 8982 + }, + { + "epoch": 25.092178770949722, + "grad_norm": 1.2052109241485596, + "learning_rate": 2.998957169459963e-05, + "loss": 0.1369, + "step": 8983 + }, + { + "epoch": 25.094972067039105, + "grad_norm": 0.8434433341026306, + "learning_rate": 2.9989199255121046e-05, + "loss": 0.126, + "step": 8984 + }, + { + "epoch": 25.09776536312849, + "grad_norm": 0.738816499710083, + "learning_rate": 2.9988826815642456e-05, + "loss": 0.1772, + "step": 8985 + }, + { + "epoch": 25.100558659217878, + "grad_norm": 0.43154260516166687, + "learning_rate": 2.9988454376163873e-05, + "loss": 0.1332, + "step": 8986 + }, + { + "epoch": 25.10335195530726, + "grad_norm": 1.0285452604293823, + "learning_rate": 2.998808193668529e-05, + "loss": 0.1404, + "step": 8987 + }, + { + "epoch": 25.106145251396647, + "grad_norm": 0.6873297095298767, + "learning_rate": 2.9987709497206706e-05, + "loss": 0.1473, + "step": 8988 + }, + { + "epoch": 25.108938547486034, + "grad_norm": 0.6712993383407593, + "learning_rate": 2.998733705772812e-05, + "loss": 0.163, + "step": 8989 + }, + { + "epoch": 25.11173184357542, + "grad_norm": 0.5585507750511169, + "learning_rate": 2.9986964618249535e-05, + "loss": 0.1142, + "step": 8990 + }, + { + "epoch": 25.114525139664803, + "grad_norm": 0.9009649753570557, + "learning_rate": 2.9986592178770952e-05, + "loss": 0.1254, + "step": 8991 + }, + { + "epoch": 25.11731843575419, + "grad_norm": 0.9423682689666748, + "learning_rate": 2.9986219739292365e-05, + "loss": 0.1503, + "step": 8992 + }, + { + "epoch": 25.120111731843576, + "grad_norm": 0.7262706160545349, + "learning_rate": 2.998584729981378e-05, + "loss": 0.1208, + "step": 8993 + }, + { + "epoch": 25.122905027932962, + "grad_norm": 1.4567815065383911, + "learning_rate": 2.9985474860335195e-05, + "loss": 0.1527, + "step": 8994 + }, + { + "epoch": 25.125698324022345, + "grad_norm": 0.8860154747962952, + "learning_rate": 2.998510242085661e-05, + "loss": 0.1303, + "step": 8995 + }, + { + "epoch": 25.12849162011173, + "grad_norm": 1.6551791429519653, + "learning_rate": 2.9984729981378028e-05, + "loss": 0.1312, + "step": 8996 + }, + { + "epoch": 25.131284916201118, + "grad_norm": 0.9151061773300171, + "learning_rate": 2.9984357541899444e-05, + "loss": 0.1395, + "step": 8997 + }, + { + "epoch": 25.134078212290504, + "grad_norm": 0.8704633712768555, + "learning_rate": 2.9983985102420858e-05, + "loss": 0.1338, + "step": 8998 + }, + { + "epoch": 25.136871508379887, + "grad_norm": 0.7345258593559265, + "learning_rate": 2.998361266294227e-05, + "loss": 0.1595, + "step": 8999 + }, + { + "epoch": 25.139664804469273, + "grad_norm": 2.0675971508026123, + "learning_rate": 2.9983240223463687e-05, + "loss": 0.1933, + "step": 9000 + }, + { + "epoch": 25.139664804469273, + "eval_cer": 0.19176677250062737, + "eval_loss": 0.2787339687347412, + "eval_runtime": 23.6177, + "eval_samples_per_second": 192.144, + "eval_steps_per_second": 0.635, + "eval_wer": 0.341303315549845, + "step": 9000 + }, + { + "epoch": 25.14245810055866, + "grad_norm": 0.3826466500759125, + "learning_rate": 2.9982867783985104e-05, + "loss": 0.1556, + "step": 9001 + }, + { + "epoch": 25.145251396648046, + "grad_norm": 0.6767383217811584, + "learning_rate": 2.998249534450652e-05, + "loss": 0.1596, + "step": 9002 + }, + { + "epoch": 25.14804469273743, + "grad_norm": 0.48674336075782776, + "learning_rate": 2.9982122905027933e-05, + "loss": 0.1734, + "step": 9003 + }, + { + "epoch": 25.150837988826815, + "grad_norm": 0.7811737060546875, + "learning_rate": 2.998175046554935e-05, + "loss": 0.1496, + "step": 9004 + }, + { + "epoch": 25.1536312849162, + "grad_norm": 0.6075583100318909, + "learning_rate": 2.9981378026070763e-05, + "loss": 0.1578, + "step": 9005 + }, + { + "epoch": 25.156424581005588, + "grad_norm": 0.6501827239990234, + "learning_rate": 2.998100558659218e-05, + "loss": 0.1669, + "step": 9006 + }, + { + "epoch": 25.15921787709497, + "grad_norm": 0.9368323683738708, + "learning_rate": 2.9980633147113593e-05, + "loss": 0.1617, + "step": 9007 + }, + { + "epoch": 25.162011173184357, + "grad_norm": 0.5408387780189514, + "learning_rate": 2.998026070763501e-05, + "loss": 0.1456, + "step": 9008 + }, + { + "epoch": 25.164804469273744, + "grad_norm": 1.1386147737503052, + "learning_rate": 2.9979888268156426e-05, + "loss": 0.1452, + "step": 9009 + }, + { + "epoch": 25.16759776536313, + "grad_norm": 0.4953697919845581, + "learning_rate": 2.9979515828677842e-05, + "loss": 0.1511, + "step": 9010 + }, + { + "epoch": 25.170391061452513, + "grad_norm": 0.5506260395050049, + "learning_rate": 2.9979143389199256e-05, + "loss": 0.1625, + "step": 9011 + }, + { + "epoch": 25.1731843575419, + "grad_norm": 0.4701654613018036, + "learning_rate": 2.997877094972067e-05, + "loss": 0.1572, + "step": 9012 + }, + { + "epoch": 25.175977653631286, + "grad_norm": 0.42667099833488464, + "learning_rate": 2.9978398510242085e-05, + "loss": 0.1286, + "step": 9013 + }, + { + "epoch": 25.178770949720672, + "grad_norm": 1.0562717914581299, + "learning_rate": 2.9978026070763502e-05, + "loss": 0.1262, + "step": 9014 + }, + { + "epoch": 25.181564245810055, + "grad_norm": 0.627861499786377, + "learning_rate": 2.997765363128492e-05, + "loss": 0.1532, + "step": 9015 + }, + { + "epoch": 25.18435754189944, + "grad_norm": 0.6640186309814453, + "learning_rate": 2.997728119180633e-05, + "loss": 0.1406, + "step": 9016 + }, + { + "epoch": 25.187150837988828, + "grad_norm": 0.6202899813652039, + "learning_rate": 2.9976908752327748e-05, + "loss": 0.143, + "step": 9017 + }, + { + "epoch": 25.189944134078214, + "grad_norm": 0.748394787311554, + "learning_rate": 2.9976536312849165e-05, + "loss": 0.1739, + "step": 9018 + }, + { + "epoch": 25.192737430167597, + "grad_norm": 0.7537829279899597, + "learning_rate": 2.9976163873370578e-05, + "loss": 0.1469, + "step": 9019 + }, + { + "epoch": 25.195530726256983, + "grad_norm": 1.1936919689178467, + "learning_rate": 2.997579143389199e-05, + "loss": 0.1379, + "step": 9020 + }, + { + "epoch": 25.19832402234637, + "grad_norm": 0.6338376402854919, + "learning_rate": 2.9975418994413407e-05, + "loss": 0.1314, + "step": 9021 + }, + { + "epoch": 25.201117318435756, + "grad_norm": 1.1217342615127563, + "learning_rate": 2.9975046554934824e-05, + "loss": 0.1457, + "step": 9022 + }, + { + "epoch": 25.20391061452514, + "grad_norm": 5.425338268280029, + "learning_rate": 2.997467411545624e-05, + "loss": 0.1549, + "step": 9023 + }, + { + "epoch": 25.206703910614525, + "grad_norm": 1.0825001001358032, + "learning_rate": 2.9974301675977657e-05, + "loss": 0.1333, + "step": 9024 + }, + { + "epoch": 25.20949720670391, + "grad_norm": 1.5950140953063965, + "learning_rate": 2.9973929236499067e-05, + "loss": 0.2079, + "step": 9025 + }, + { + "epoch": 25.212290502793294, + "grad_norm": 0.6276286840438843, + "learning_rate": 2.9973556797020483e-05, + "loss": 0.1544, + "step": 9026 + }, + { + "epoch": 25.21508379888268, + "grad_norm": 0.8727338314056396, + "learning_rate": 2.99731843575419e-05, + "loss": 0.185, + "step": 9027 + }, + { + "epoch": 25.217877094972067, + "grad_norm": 1.0023298263549805, + "learning_rate": 2.9972811918063316e-05, + "loss": 0.188, + "step": 9028 + }, + { + "epoch": 25.220670391061454, + "grad_norm": 0.7594616413116455, + "learning_rate": 2.997243947858473e-05, + "loss": 0.1547, + "step": 9029 + }, + { + "epoch": 25.223463687150836, + "grad_norm": 0.5084993839263916, + "learning_rate": 2.9972067039106146e-05, + "loss": 0.174, + "step": 9030 + }, + { + "epoch": 25.226256983240223, + "grad_norm": 0.5378205180168152, + "learning_rate": 2.9971694599627563e-05, + "loss": 0.1561, + "step": 9031 + }, + { + "epoch": 25.22905027932961, + "grad_norm": 0.7348217368125916, + "learning_rate": 2.9971322160148976e-05, + "loss": 0.167, + "step": 9032 + }, + { + "epoch": 25.231843575418996, + "grad_norm": 0.5308288335800171, + "learning_rate": 2.9970949720670392e-05, + "loss": 0.1474, + "step": 9033 + }, + { + "epoch": 25.23463687150838, + "grad_norm": 0.5044652223587036, + "learning_rate": 2.9970577281191805e-05, + "loss": 0.1671, + "step": 9034 + }, + { + "epoch": 25.237430167597765, + "grad_norm": 0.5767888426780701, + "learning_rate": 2.9970204841713222e-05, + "loss": 0.1637, + "step": 9035 + }, + { + "epoch": 25.24022346368715, + "grad_norm": 0.7339551448822021, + "learning_rate": 2.996983240223464e-05, + "loss": 0.1723, + "step": 9036 + }, + { + "epoch": 25.243016759776538, + "grad_norm": 0.8238263130187988, + "learning_rate": 2.9969459962756055e-05, + "loss": 0.1623, + "step": 9037 + }, + { + "epoch": 25.24581005586592, + "grad_norm": 1.2278550863265991, + "learning_rate": 2.9969087523277468e-05, + "loss": 0.1391, + "step": 9038 + }, + { + "epoch": 25.248603351955307, + "grad_norm": 1.4133626222610474, + "learning_rate": 2.996871508379888e-05, + "loss": 0.1592, + "step": 9039 + }, + { + "epoch": 25.251396648044693, + "grad_norm": 0.6277496814727783, + "learning_rate": 2.9968342644320298e-05, + "loss": 0.1319, + "step": 9040 + }, + { + "epoch": 25.25418994413408, + "grad_norm": 3.3552944660186768, + "learning_rate": 2.9967970204841714e-05, + "loss": 0.1342, + "step": 9041 + }, + { + "epoch": 25.256983240223462, + "grad_norm": 0.8441613912582397, + "learning_rate": 2.996759776536313e-05, + "loss": 0.1791, + "step": 9042 + }, + { + "epoch": 25.25977653631285, + "grad_norm": 0.9065885543823242, + "learning_rate": 2.9967225325884544e-05, + "loss": 0.129, + "step": 9043 + }, + { + "epoch": 25.262569832402235, + "grad_norm": 1.412946105003357, + "learning_rate": 2.996685288640596e-05, + "loss": 0.1593, + "step": 9044 + }, + { + "epoch": 25.26536312849162, + "grad_norm": 1.659420371055603, + "learning_rate": 2.9966480446927374e-05, + "loss": 0.1827, + "step": 9045 + }, + { + "epoch": 25.268156424581004, + "grad_norm": 1.7682029008865356, + "learning_rate": 2.996610800744879e-05, + "loss": 0.1519, + "step": 9046 + }, + { + "epoch": 25.27094972067039, + "grad_norm": 1.1385061740875244, + "learning_rate": 2.9965735567970203e-05, + "loss": 0.1614, + "step": 9047 + }, + { + "epoch": 25.273743016759777, + "grad_norm": 1.4121028184890747, + "learning_rate": 2.996536312849162e-05, + "loss": 0.1431, + "step": 9048 + }, + { + "epoch": 25.276536312849164, + "grad_norm": 2.202448844909668, + "learning_rate": 2.9964990689013037e-05, + "loss": 0.2204, + "step": 9049 + }, + { + "epoch": 25.279329608938546, + "grad_norm": 1.5324910879135132, + "learning_rate": 2.9964618249534453e-05, + "loss": 0.2033, + "step": 9050 + }, + { + "epoch": 25.282122905027933, + "grad_norm": 0.9025055170059204, + "learning_rate": 2.996424581005587e-05, + "loss": 0.1455, + "step": 9051 + }, + { + "epoch": 25.28491620111732, + "grad_norm": 0.6833521127700806, + "learning_rate": 2.996387337057728e-05, + "loss": 0.1635, + "step": 9052 + }, + { + "epoch": 25.287709497206706, + "grad_norm": 0.5386950373649597, + "learning_rate": 2.9963500931098696e-05, + "loss": 0.2113, + "step": 9053 + }, + { + "epoch": 25.29050279329609, + "grad_norm": 0.6680923104286194, + "learning_rate": 2.9963128491620112e-05, + "loss": 0.1535, + "step": 9054 + }, + { + "epoch": 25.293296089385475, + "grad_norm": 0.5185078382492065, + "learning_rate": 2.996275605214153e-05, + "loss": 0.1762, + "step": 9055 + }, + { + "epoch": 25.29608938547486, + "grad_norm": 0.6459643244743347, + "learning_rate": 2.9962383612662942e-05, + "loss": 0.1374, + "step": 9056 + }, + { + "epoch": 25.298882681564244, + "grad_norm": 0.6466339230537415, + "learning_rate": 2.996201117318436e-05, + "loss": 0.1465, + "step": 9057 + }, + { + "epoch": 25.30167597765363, + "grad_norm": 0.9083635210990906, + "learning_rate": 2.9961638733705775e-05, + "loss": 0.1567, + "step": 9058 + }, + { + "epoch": 25.304469273743017, + "grad_norm": 1.2669610977172852, + "learning_rate": 2.996126629422719e-05, + "loss": 0.1359, + "step": 9059 + }, + { + "epoch": 25.307262569832403, + "grad_norm": 0.49477341771125793, + "learning_rate": 2.9960893854748605e-05, + "loss": 0.1622, + "step": 9060 + }, + { + "epoch": 25.310055865921786, + "grad_norm": 1.9394882917404175, + "learning_rate": 2.9960521415270018e-05, + "loss": 0.1447, + "step": 9061 + }, + { + "epoch": 25.312849162011172, + "grad_norm": 0.6095957159996033, + "learning_rate": 2.9960148975791435e-05, + "loss": 0.1426, + "step": 9062 + }, + { + "epoch": 25.31564245810056, + "grad_norm": 0.5607377886772156, + "learning_rate": 2.995977653631285e-05, + "loss": 0.1464, + "step": 9063 + }, + { + "epoch": 25.318435754189945, + "grad_norm": 0.7486692667007446, + "learning_rate": 2.9959404096834268e-05, + "loss": 0.1544, + "step": 9064 + }, + { + "epoch": 25.321229050279328, + "grad_norm": 0.44044914841651917, + "learning_rate": 2.9959031657355677e-05, + "loss": 0.1311, + "step": 9065 + }, + { + "epoch": 25.324022346368714, + "grad_norm": 0.6411564350128174, + "learning_rate": 2.9958659217877094e-05, + "loss": 0.1342, + "step": 9066 + }, + { + "epoch": 25.3268156424581, + "grad_norm": 0.7480461001396179, + "learning_rate": 2.995828677839851e-05, + "loss": 0.1371, + "step": 9067 + }, + { + "epoch": 25.329608938547487, + "grad_norm": 0.8833625912666321, + "learning_rate": 2.9957914338919927e-05, + "loss": 0.1379, + "step": 9068 + }, + { + "epoch": 25.33240223463687, + "grad_norm": 1.2347311973571777, + "learning_rate": 2.9957541899441344e-05, + "loss": 0.1488, + "step": 9069 + }, + { + "epoch": 25.335195530726256, + "grad_norm": 0.6077716946601868, + "learning_rate": 2.9957169459962757e-05, + "loss": 0.1589, + "step": 9070 + }, + { + "epoch": 25.337988826815643, + "grad_norm": 0.8115511536598206, + "learning_rate": 2.9956797020484173e-05, + "loss": 0.1567, + "step": 9071 + }, + { + "epoch": 25.34078212290503, + "grad_norm": 1.0749098062515259, + "learning_rate": 2.9956424581005586e-05, + "loss": 0.1484, + "step": 9072 + }, + { + "epoch": 25.343575418994412, + "grad_norm": 0.775073766708374, + "learning_rate": 2.9956052141527003e-05, + "loss": 0.1491, + "step": 9073 + }, + { + "epoch": 25.3463687150838, + "grad_norm": 1.0473029613494873, + "learning_rate": 2.9955679702048416e-05, + "loss": 0.1978, + "step": 9074 + }, + { + "epoch": 25.349162011173185, + "grad_norm": 1.8152724504470825, + "learning_rate": 2.9955307262569833e-05, + "loss": 0.204, + "step": 9075 + }, + { + "epoch": 25.35195530726257, + "grad_norm": 0.642371416091919, + "learning_rate": 2.995493482309125e-05, + "loss": 0.1642, + "step": 9076 + }, + { + "epoch": 25.354748603351954, + "grad_norm": 0.5127820372581482, + "learning_rate": 2.9954562383612666e-05, + "loss": 0.1692, + "step": 9077 + }, + { + "epoch": 25.35754189944134, + "grad_norm": 0.42318177223205566, + "learning_rate": 2.9954189944134082e-05, + "loss": 0.1763, + "step": 9078 + }, + { + "epoch": 25.360335195530727, + "grad_norm": 0.8009966015815735, + "learning_rate": 2.9953817504655492e-05, + "loss": 0.1611, + "step": 9079 + }, + { + "epoch": 25.363128491620113, + "grad_norm": 2.261150360107422, + "learning_rate": 2.995344506517691e-05, + "loss": 0.1752, + "step": 9080 + }, + { + "epoch": 25.365921787709496, + "grad_norm": 0.7965113520622253, + "learning_rate": 2.9953072625698325e-05, + "loss": 0.1553, + "step": 9081 + }, + { + "epoch": 25.368715083798882, + "grad_norm": 0.7087690234184265, + "learning_rate": 2.995270018621974e-05, + "loss": 0.1611, + "step": 9082 + }, + { + "epoch": 25.37150837988827, + "grad_norm": 0.7068237066268921, + "learning_rate": 2.9952327746741155e-05, + "loss": 0.162, + "step": 9083 + }, + { + "epoch": 25.374301675977655, + "grad_norm": 0.5202464461326599, + "learning_rate": 2.995195530726257e-05, + "loss": 0.142, + "step": 9084 + }, + { + "epoch": 25.377094972067038, + "grad_norm": 0.7520607709884644, + "learning_rate": 2.9951582867783984e-05, + "loss": 0.1631, + "step": 9085 + }, + { + "epoch": 25.379888268156424, + "grad_norm": 0.7672001719474792, + "learning_rate": 2.99512104283054e-05, + "loss": 0.1951, + "step": 9086 + }, + { + "epoch": 25.38268156424581, + "grad_norm": 0.9711746573448181, + "learning_rate": 2.9950837988826817e-05, + "loss": 0.1508, + "step": 9087 + }, + { + "epoch": 25.385474860335197, + "grad_norm": 0.9048035144805908, + "learning_rate": 2.995046554934823e-05, + "loss": 0.1626, + "step": 9088 + }, + { + "epoch": 25.38826815642458, + "grad_norm": 0.5949077606201172, + "learning_rate": 2.9950093109869647e-05, + "loss": 0.1379, + "step": 9089 + }, + { + "epoch": 25.391061452513966, + "grad_norm": 0.6534552574157715, + "learning_rate": 2.9949720670391064e-05, + "loss": 0.1523, + "step": 9090 + }, + { + "epoch": 25.393854748603353, + "grad_norm": 0.7024869918823242, + "learning_rate": 2.994934823091248e-05, + "loss": 0.1644, + "step": 9091 + }, + { + "epoch": 25.39664804469274, + "grad_norm": 0.9009765982627869, + "learning_rate": 2.994897579143389e-05, + "loss": 0.1495, + "step": 9092 + }, + { + "epoch": 25.399441340782122, + "grad_norm": 0.7060795426368713, + "learning_rate": 2.9948603351955306e-05, + "loss": 0.135, + "step": 9093 + }, + { + "epoch": 25.40223463687151, + "grad_norm": 0.7270749807357788, + "learning_rate": 2.9948230912476723e-05, + "loss": 0.1496, + "step": 9094 + }, + { + "epoch": 25.405027932960895, + "grad_norm": 0.9576922059059143, + "learning_rate": 2.994785847299814e-05, + "loss": 0.1411, + "step": 9095 + }, + { + "epoch": 25.407821229050278, + "grad_norm": 0.5947214365005493, + "learning_rate": 2.9947486033519556e-05, + "loss": 0.1527, + "step": 9096 + }, + { + "epoch": 25.410614525139664, + "grad_norm": 1.0394736528396606, + "learning_rate": 2.994711359404097e-05, + "loss": 0.167, + "step": 9097 + }, + { + "epoch": 25.41340782122905, + "grad_norm": 0.8622907400131226, + "learning_rate": 2.9946741154562386e-05, + "loss": 0.1556, + "step": 9098 + }, + { + "epoch": 25.416201117318437, + "grad_norm": 0.8324684500694275, + "learning_rate": 2.99463687150838e-05, + "loss": 0.1551, + "step": 9099 + }, + { + "epoch": 25.41899441340782, + "grad_norm": 1.593802809715271, + "learning_rate": 2.9945996275605215e-05, + "loss": 0.1776, + "step": 9100 + }, + { + "epoch": 25.421787709497206, + "grad_norm": 0.496139794588089, + "learning_rate": 2.994562383612663e-05, + "loss": 0.1704, + "step": 9101 + }, + { + "epoch": 25.424581005586592, + "grad_norm": 0.5463382005691528, + "learning_rate": 2.9945251396648045e-05, + "loss": 0.1558, + "step": 9102 + }, + { + "epoch": 25.42737430167598, + "grad_norm": 0.7727879881858826, + "learning_rate": 2.994487895716946e-05, + "loss": 0.1601, + "step": 9103 + }, + { + "epoch": 25.43016759776536, + "grad_norm": 0.8856693506240845, + "learning_rate": 2.9944506517690878e-05, + "loss": 0.1572, + "step": 9104 + }, + { + "epoch": 25.432960893854748, + "grad_norm": 0.5650933384895325, + "learning_rate": 2.994413407821229e-05, + "loss": 0.1749, + "step": 9105 + }, + { + "epoch": 25.435754189944134, + "grad_norm": 0.6128475666046143, + "learning_rate": 2.9943761638733704e-05, + "loss": 0.1628, + "step": 9106 + }, + { + "epoch": 25.43854748603352, + "grad_norm": 0.6566910147666931, + "learning_rate": 2.994338919925512e-05, + "loss": 0.1889, + "step": 9107 + }, + { + "epoch": 25.441340782122904, + "grad_norm": 0.8629441857337952, + "learning_rate": 2.9943016759776538e-05, + "loss": 0.1478, + "step": 9108 + }, + { + "epoch": 25.44413407821229, + "grad_norm": 1.281890869140625, + "learning_rate": 2.9942644320297954e-05, + "loss": 0.1639, + "step": 9109 + }, + { + "epoch": 25.446927374301676, + "grad_norm": 0.7012952566146851, + "learning_rate": 2.9942271880819367e-05, + "loss": 0.1614, + "step": 9110 + }, + { + "epoch": 25.449720670391063, + "grad_norm": 0.662903368473053, + "learning_rate": 2.9941899441340784e-05, + "loss": 0.1528, + "step": 9111 + }, + { + "epoch": 25.452513966480446, + "grad_norm": 0.608973503112793, + "learning_rate": 2.9941527001862197e-05, + "loss": 0.1494, + "step": 9112 + }, + { + "epoch": 25.455307262569832, + "grad_norm": 0.6912122964859009, + "learning_rate": 2.9941154562383613e-05, + "loss": 0.1626, + "step": 9113 + }, + { + "epoch": 25.45810055865922, + "grad_norm": 0.8055962324142456, + "learning_rate": 2.9940782122905027e-05, + "loss": 0.1608, + "step": 9114 + }, + { + "epoch": 25.460893854748605, + "grad_norm": 0.9374245405197144, + "learning_rate": 2.9940409683426443e-05, + "loss": 0.1421, + "step": 9115 + }, + { + "epoch": 25.463687150837988, + "grad_norm": 0.5942680835723877, + "learning_rate": 2.994003724394786e-05, + "loss": 0.1394, + "step": 9116 + }, + { + "epoch": 25.466480446927374, + "grad_norm": 0.6186679601669312, + "learning_rate": 2.9939664804469276e-05, + "loss": 0.1489, + "step": 9117 + }, + { + "epoch": 25.46927374301676, + "grad_norm": 0.8338116407394409, + "learning_rate": 2.9939292364990693e-05, + "loss": 0.1308, + "step": 9118 + }, + { + "epoch": 25.472067039106147, + "grad_norm": 0.7920458912849426, + "learning_rate": 2.9938919925512103e-05, + "loss": 0.1853, + "step": 9119 + }, + { + "epoch": 25.47486033519553, + "grad_norm": 0.8683234453201294, + "learning_rate": 2.993854748603352e-05, + "loss": 0.1535, + "step": 9120 + }, + { + "epoch": 25.477653631284916, + "grad_norm": 1.70913565158844, + "learning_rate": 2.9938175046554936e-05, + "loss": 0.1462, + "step": 9121 + }, + { + "epoch": 25.480446927374302, + "grad_norm": 0.8282350897789001, + "learning_rate": 2.9937802607076352e-05, + "loss": 0.1575, + "step": 9122 + }, + { + "epoch": 25.48324022346369, + "grad_norm": 1.4825855493545532, + "learning_rate": 2.9937430167597765e-05, + "loss": 0.1595, + "step": 9123 + }, + { + "epoch": 25.48603351955307, + "grad_norm": 1.0166682004928589, + "learning_rate": 2.9937057728119182e-05, + "loss": 0.1495, + "step": 9124 + }, + { + "epoch": 25.488826815642458, + "grad_norm": 3.565109968185425, + "learning_rate": 2.9936685288640595e-05, + "loss": 0.195, + "step": 9125 + }, + { + "epoch": 25.491620111731844, + "grad_norm": 0.4094304144382477, + "learning_rate": 2.993631284916201e-05, + "loss": 0.1675, + "step": 9126 + }, + { + "epoch": 25.49441340782123, + "grad_norm": 0.5070602893829346, + "learning_rate": 2.9935940409683428e-05, + "loss": 0.144, + "step": 9127 + }, + { + "epoch": 25.497206703910614, + "grad_norm": 0.7675985097885132, + "learning_rate": 2.993556797020484e-05, + "loss": 0.1663, + "step": 9128 + }, + { + "epoch": 25.5, + "grad_norm": 0.5090705156326294, + "learning_rate": 2.9935195530726258e-05, + "loss": 0.1489, + "step": 9129 + }, + { + "epoch": 25.502793296089386, + "grad_norm": 0.7018420696258545, + "learning_rate": 2.9934823091247674e-05, + "loss": 0.1453, + "step": 9130 + }, + { + "epoch": 25.505586592178773, + "grad_norm": 0.4873214662075043, + "learning_rate": 2.993445065176909e-05, + "loss": 0.1544, + "step": 9131 + }, + { + "epoch": 25.508379888268156, + "grad_norm": 0.4692796468734741, + "learning_rate": 2.99340782122905e-05, + "loss": 0.1447, + "step": 9132 + }, + { + "epoch": 25.511173184357542, + "grad_norm": 0.6785284280776978, + "learning_rate": 2.9933705772811917e-05, + "loss": 0.1456, + "step": 9133 + }, + { + "epoch": 25.51396648044693, + "grad_norm": 0.6772041320800781, + "learning_rate": 2.9933333333333334e-05, + "loss": 0.1453, + "step": 9134 + }, + { + "epoch": 25.51675977653631, + "grad_norm": 0.507642924785614, + "learning_rate": 2.993296089385475e-05, + "loss": 0.1322, + "step": 9135 + }, + { + "epoch": 25.519553072625698, + "grad_norm": 1.0926706790924072, + "learning_rate": 2.9932588454376167e-05, + "loss": 0.1362, + "step": 9136 + }, + { + "epoch": 25.522346368715084, + "grad_norm": 0.8229749202728271, + "learning_rate": 2.993221601489758e-05, + "loss": 0.1654, + "step": 9137 + }, + { + "epoch": 25.52513966480447, + "grad_norm": 0.9163024425506592, + "learning_rate": 2.9931843575418996e-05, + "loss": 0.1603, + "step": 9138 + }, + { + "epoch": 25.527932960893853, + "grad_norm": 0.8509167432785034, + "learning_rate": 2.993147113594041e-05, + "loss": 0.1448, + "step": 9139 + }, + { + "epoch": 25.53072625698324, + "grad_norm": 0.7057985067367554, + "learning_rate": 2.9931098696461826e-05, + "loss": 0.1442, + "step": 9140 + }, + { + "epoch": 25.533519553072626, + "grad_norm": 0.6489566564559937, + "learning_rate": 2.993072625698324e-05, + "loss": 0.1371, + "step": 9141 + }, + { + "epoch": 25.536312849162012, + "grad_norm": 1.176023244857788, + "learning_rate": 2.9930353817504656e-05, + "loss": 0.1462, + "step": 9142 + }, + { + "epoch": 25.539106145251395, + "grad_norm": 0.9274229407310486, + "learning_rate": 2.9929981378026072e-05, + "loss": 0.1583, + "step": 9143 + }, + { + "epoch": 25.54189944134078, + "grad_norm": 0.9977539777755737, + "learning_rate": 2.992960893854749e-05, + "loss": 0.1463, + "step": 9144 + }, + { + "epoch": 25.544692737430168, + "grad_norm": 0.8830480575561523, + "learning_rate": 2.9929236499068902e-05, + "loss": 0.1694, + "step": 9145 + }, + { + "epoch": 25.547486033519554, + "grad_norm": 0.6843335032463074, + "learning_rate": 2.9928864059590315e-05, + "loss": 0.1295, + "step": 9146 + }, + { + "epoch": 25.550279329608937, + "grad_norm": 0.8812766671180725, + "learning_rate": 2.992849162011173e-05, + "loss": 0.1312, + "step": 9147 + }, + { + "epoch": 25.553072625698324, + "grad_norm": 1.6438214778900146, + "learning_rate": 2.9928119180633148e-05, + "loss": 0.1625, + "step": 9148 + }, + { + "epoch": 25.55586592178771, + "grad_norm": 1.010605812072754, + "learning_rate": 2.9927746741154565e-05, + "loss": 0.1555, + "step": 9149 + }, + { + "epoch": 25.558659217877096, + "grad_norm": 1.5892856121063232, + "learning_rate": 2.9927374301675978e-05, + "loss": 0.2032, + "step": 9150 + }, + { + "epoch": 25.56145251396648, + "grad_norm": 0.4437512457370758, + "learning_rate": 2.9927001862197394e-05, + "loss": 0.1758, + "step": 9151 + }, + { + "epoch": 25.564245810055866, + "grad_norm": 0.6676002740859985, + "learning_rate": 2.9926629422718808e-05, + "loss": 0.1398, + "step": 9152 + }, + { + "epoch": 25.567039106145252, + "grad_norm": 0.4370076656341553, + "learning_rate": 2.9926256983240224e-05, + "loss": 0.1287, + "step": 9153 + }, + { + "epoch": 25.56983240223464, + "grad_norm": 0.6429286003112793, + "learning_rate": 2.992588454376164e-05, + "loss": 0.1654, + "step": 9154 + }, + { + "epoch": 25.57262569832402, + "grad_norm": 1.9699103832244873, + "learning_rate": 2.9925512104283054e-05, + "loss": 0.1829, + "step": 9155 + }, + { + "epoch": 25.575418994413408, + "grad_norm": 1.940584659576416, + "learning_rate": 2.992513966480447e-05, + "loss": 0.1746, + "step": 9156 + }, + { + "epoch": 25.578212290502794, + "grad_norm": 0.5096760988235474, + "learning_rate": 2.9924767225325887e-05, + "loss": 0.1643, + "step": 9157 + }, + { + "epoch": 25.58100558659218, + "grad_norm": 0.9388226270675659, + "learning_rate": 2.9924394785847303e-05, + "loss": 0.1675, + "step": 9158 + }, + { + "epoch": 25.583798882681563, + "grad_norm": 0.657946765422821, + "learning_rate": 2.9924022346368713e-05, + "loss": 0.147, + "step": 9159 + }, + { + "epoch": 25.58659217877095, + "grad_norm": 0.5676870942115784, + "learning_rate": 2.992364990689013e-05, + "loss": 0.155, + "step": 9160 + }, + { + "epoch": 25.589385474860336, + "grad_norm": 0.5807114243507385, + "learning_rate": 2.9923277467411546e-05, + "loss": 0.1832, + "step": 9161 + }, + { + "epoch": 25.592178770949722, + "grad_norm": 0.7146822214126587, + "learning_rate": 2.9922905027932963e-05, + "loss": 0.143, + "step": 9162 + }, + { + "epoch": 25.594972067039105, + "grad_norm": 0.5566192865371704, + "learning_rate": 2.992253258845438e-05, + "loss": 0.1385, + "step": 9163 + }, + { + "epoch": 25.59776536312849, + "grad_norm": 0.8500552177429199, + "learning_rate": 2.9922160148975792e-05, + "loss": 0.1528, + "step": 9164 + }, + { + "epoch": 25.600558659217878, + "grad_norm": 0.954501748085022, + "learning_rate": 2.9921787709497206e-05, + "loss": 0.1427, + "step": 9165 + }, + { + "epoch": 25.60335195530726, + "grad_norm": 0.7576957941055298, + "learning_rate": 2.9921415270018622e-05, + "loss": 0.178, + "step": 9166 + }, + { + "epoch": 25.606145251396647, + "grad_norm": 1.166272521018982, + "learning_rate": 2.992104283054004e-05, + "loss": 0.1678, + "step": 9167 + }, + { + "epoch": 25.608938547486034, + "grad_norm": 1.188633680343628, + "learning_rate": 2.9920670391061452e-05, + "loss": 0.1353, + "step": 9168 + }, + { + "epoch": 25.61173184357542, + "grad_norm": 0.7721127867698669, + "learning_rate": 2.9920297951582868e-05, + "loss": 0.1676, + "step": 9169 + }, + { + "epoch": 25.614525139664803, + "grad_norm": 0.9475165009498596, + "learning_rate": 2.9919925512104285e-05, + "loss": 0.171, + "step": 9170 + }, + { + "epoch": 25.61731843575419, + "grad_norm": 1.0508520603179932, + "learning_rate": 2.99195530726257e-05, + "loss": 0.1553, + "step": 9171 + }, + { + "epoch": 25.620111731843576, + "grad_norm": 0.642099916934967, + "learning_rate": 2.9919180633147115e-05, + "loss": 0.1369, + "step": 9172 + }, + { + "epoch": 25.622905027932962, + "grad_norm": 0.9311045408248901, + "learning_rate": 2.9918808193668528e-05, + "loss": 0.1421, + "step": 9173 + }, + { + "epoch": 25.625698324022345, + "grad_norm": 1.1751644611358643, + "learning_rate": 2.9918435754189944e-05, + "loss": 0.1583, + "step": 9174 + }, + { + "epoch": 25.62849162011173, + "grad_norm": 1.5265007019042969, + "learning_rate": 2.991806331471136e-05, + "loss": 0.2094, + "step": 9175 + }, + { + "epoch": 25.631284916201118, + "grad_norm": 0.616005003452301, + "learning_rate": 2.9917690875232777e-05, + "loss": 0.1588, + "step": 9176 + }, + { + "epoch": 25.634078212290504, + "grad_norm": 0.4131748378276825, + "learning_rate": 2.991731843575419e-05, + "loss": 0.1488, + "step": 9177 + }, + { + "epoch": 25.636871508379887, + "grad_norm": 0.765606164932251, + "learning_rate": 2.9916945996275607e-05, + "loss": 0.1743, + "step": 9178 + }, + { + "epoch": 25.639664804469273, + "grad_norm": 0.8128997087478638, + "learning_rate": 2.991657355679702e-05, + "loss": 0.16, + "step": 9179 + }, + { + "epoch": 25.64245810055866, + "grad_norm": 0.4341287612915039, + "learning_rate": 2.9916201117318437e-05, + "loss": 0.1639, + "step": 9180 + }, + { + "epoch": 25.645251396648046, + "grad_norm": 0.6317914128303528, + "learning_rate": 2.9915828677839853e-05, + "loss": 0.1593, + "step": 9181 + }, + { + "epoch": 25.64804469273743, + "grad_norm": 0.4419688880443573, + "learning_rate": 2.9915456238361266e-05, + "loss": 0.1508, + "step": 9182 + }, + { + "epoch": 25.650837988826815, + "grad_norm": 0.5421468615531921, + "learning_rate": 2.9915083798882683e-05, + "loss": 0.1524, + "step": 9183 + }, + { + "epoch": 25.6536312849162, + "grad_norm": 1.14256751537323, + "learning_rate": 2.99147113594041e-05, + "loss": 0.1424, + "step": 9184 + }, + { + "epoch": 25.656424581005588, + "grad_norm": 0.7056170701980591, + "learning_rate": 2.9914338919925513e-05, + "loss": 0.1962, + "step": 9185 + }, + { + "epoch": 25.65921787709497, + "grad_norm": 0.39172130823135376, + "learning_rate": 2.9913966480446926e-05, + "loss": 0.1323, + "step": 9186 + }, + { + "epoch": 25.662011173184357, + "grad_norm": 0.6266517043113708, + "learning_rate": 2.9913594040968342e-05, + "loss": 0.1515, + "step": 9187 + }, + { + "epoch": 25.664804469273744, + "grad_norm": 0.9619919061660767, + "learning_rate": 2.991322160148976e-05, + "loss": 0.1365, + "step": 9188 + }, + { + "epoch": 25.66759776536313, + "grad_norm": 0.7127665281295776, + "learning_rate": 2.9912849162011175e-05, + "loss": 0.1676, + "step": 9189 + }, + { + "epoch": 25.670391061452513, + "grad_norm": 0.6382834911346436, + "learning_rate": 2.9912476722532592e-05, + "loss": 0.1293, + "step": 9190 + }, + { + "epoch": 25.6731843575419, + "grad_norm": 0.9053936004638672, + "learning_rate": 2.9912104283054005e-05, + "loss": 0.1367, + "step": 9191 + }, + { + "epoch": 25.675977653631286, + "grad_norm": 0.9375266432762146, + "learning_rate": 2.9911731843575418e-05, + "loss": 0.165, + "step": 9192 + }, + { + "epoch": 25.678770949720672, + "grad_norm": 0.7087886333465576, + "learning_rate": 2.9911359404096835e-05, + "loss": 0.1729, + "step": 9193 + }, + { + "epoch": 25.681564245810055, + "grad_norm": 0.8337056040763855, + "learning_rate": 2.991098696461825e-05, + "loss": 0.1554, + "step": 9194 + }, + { + "epoch": 25.68435754189944, + "grad_norm": 0.655512273311615, + "learning_rate": 2.9910614525139664e-05, + "loss": 0.1555, + "step": 9195 + }, + { + "epoch": 25.687150837988828, + "grad_norm": 0.8591168522834778, + "learning_rate": 2.991024208566108e-05, + "loss": 0.1808, + "step": 9196 + }, + { + "epoch": 25.689944134078214, + "grad_norm": 1.3877428770065308, + "learning_rate": 2.9909869646182497e-05, + "loss": 0.1705, + "step": 9197 + }, + { + "epoch": 25.692737430167597, + "grad_norm": 1.2391287088394165, + "learning_rate": 2.9909497206703914e-05, + "loss": 0.1801, + "step": 9198 + }, + { + "epoch": 25.695530726256983, + "grad_norm": 0.7876945734024048, + "learning_rate": 2.9909124767225327e-05, + "loss": 0.148, + "step": 9199 + }, + { + "epoch": 25.69832402234637, + "grad_norm": 2.5603137016296387, + "learning_rate": 2.990875232774674e-05, + "loss": 0.1964, + "step": 9200 + }, + { + "epoch": 25.701117318435756, + "grad_norm": 0.5715459585189819, + "learning_rate": 2.9908379888268157e-05, + "loss": 0.1706, + "step": 9201 + }, + { + "epoch": 25.70391061452514, + "grad_norm": 0.6153680086135864, + "learning_rate": 2.9908007448789573e-05, + "loss": 0.1672, + "step": 9202 + }, + { + "epoch": 25.706703910614525, + "grad_norm": 0.4989617168903351, + "learning_rate": 2.990763500931099e-05, + "loss": 0.1501, + "step": 9203 + }, + { + "epoch": 25.70949720670391, + "grad_norm": 0.6983944773674011, + "learning_rate": 2.9907262569832403e-05, + "loss": 0.1618, + "step": 9204 + }, + { + "epoch": 25.712290502793294, + "grad_norm": 0.584079384803772, + "learning_rate": 2.9906890130353816e-05, + "loss": 0.1651, + "step": 9205 + }, + { + "epoch": 25.71508379888268, + "grad_norm": 0.77744060754776, + "learning_rate": 2.9906517690875233e-05, + "loss": 0.1577, + "step": 9206 + }, + { + "epoch": 25.717877094972067, + "grad_norm": 0.480500727891922, + "learning_rate": 2.990614525139665e-05, + "loss": 0.1576, + "step": 9207 + }, + { + "epoch": 25.720670391061454, + "grad_norm": 0.7692481875419617, + "learning_rate": 2.9905772811918062e-05, + "loss": 0.1541, + "step": 9208 + }, + { + "epoch": 25.723463687150836, + "grad_norm": 1.1771327257156372, + "learning_rate": 2.990540037243948e-05, + "loss": 0.1419, + "step": 9209 + }, + { + "epoch": 25.726256983240223, + "grad_norm": 0.7748406529426575, + "learning_rate": 2.9905027932960895e-05, + "loss": 0.1734, + "step": 9210 + }, + { + "epoch": 25.72905027932961, + "grad_norm": 0.5273832082748413, + "learning_rate": 2.9904655493482312e-05, + "loss": 0.1695, + "step": 9211 + }, + { + "epoch": 25.731843575418996, + "grad_norm": 0.8127542734146118, + "learning_rate": 2.9904283054003725e-05, + "loss": 0.1646, + "step": 9212 + }, + { + "epoch": 25.73463687150838, + "grad_norm": 1.0701465606689453, + "learning_rate": 2.9903910614525138e-05, + "loss": 0.1572, + "step": 9213 + }, + { + "epoch": 25.737430167597765, + "grad_norm": 0.6163681745529175, + "learning_rate": 2.9903538175046555e-05, + "loss": 0.1697, + "step": 9214 + }, + { + "epoch": 25.74022346368715, + "grad_norm": 1.87282133102417, + "learning_rate": 2.990316573556797e-05, + "loss": 0.1365, + "step": 9215 + }, + { + "epoch": 25.743016759776538, + "grad_norm": 0.5407472848892212, + "learning_rate": 2.9902793296089388e-05, + "loss": 0.1315, + "step": 9216 + }, + { + "epoch": 25.74581005586592, + "grad_norm": 0.5475946068763733, + "learning_rate": 2.99024208566108e-05, + "loss": 0.1318, + "step": 9217 + }, + { + "epoch": 25.748603351955307, + "grad_norm": 0.9228734970092773, + "learning_rate": 2.9902048417132218e-05, + "loss": 0.1337, + "step": 9218 + }, + { + "epoch": 25.751396648044693, + "grad_norm": 0.6448841691017151, + "learning_rate": 2.990167597765363e-05, + "loss": 0.1215, + "step": 9219 + }, + { + "epoch": 25.75418994413408, + "grad_norm": 1.7959010601043701, + "learning_rate": 2.9901303538175047e-05, + "loss": 0.1604, + "step": 9220 + }, + { + "epoch": 25.756983240223462, + "grad_norm": 0.7665084004402161, + "learning_rate": 2.9900931098696464e-05, + "loss": 0.1346, + "step": 9221 + }, + { + "epoch": 25.75977653631285, + "grad_norm": 0.7633014917373657, + "learning_rate": 2.9900558659217877e-05, + "loss": 0.1392, + "step": 9222 + }, + { + "epoch": 25.762569832402235, + "grad_norm": 1.6353108882904053, + "learning_rate": 2.9900186219739293e-05, + "loss": 0.1568, + "step": 9223 + }, + { + "epoch": 25.76536312849162, + "grad_norm": 1.5431841611862183, + "learning_rate": 2.989981378026071e-05, + "loss": 0.1602, + "step": 9224 + }, + { + "epoch": 25.768156424581004, + "grad_norm": 1.467910647392273, + "learning_rate": 2.9899441340782123e-05, + "loss": 0.1976, + "step": 9225 + }, + { + "epoch": 25.77094972067039, + "grad_norm": 0.6133819222450256, + "learning_rate": 2.9899068901303536e-05, + "loss": 0.1814, + "step": 9226 + }, + { + "epoch": 25.773743016759777, + "grad_norm": 0.5043025612831116, + "learning_rate": 2.9898696461824953e-05, + "loss": 0.1519, + "step": 9227 + }, + { + "epoch": 25.776536312849164, + "grad_norm": 0.5529866218566895, + "learning_rate": 2.989832402234637e-05, + "loss": 0.1468, + "step": 9228 + }, + { + "epoch": 25.779329608938546, + "grad_norm": 0.4272448420524597, + "learning_rate": 2.9897951582867786e-05, + "loss": 0.158, + "step": 9229 + }, + { + "epoch": 25.782122905027933, + "grad_norm": 0.4957178831100464, + "learning_rate": 2.9897579143389202e-05, + "loss": 0.1547, + "step": 9230 + }, + { + "epoch": 25.78491620111732, + "grad_norm": 0.819429874420166, + "learning_rate": 2.9897206703910616e-05, + "loss": 0.1706, + "step": 9231 + }, + { + "epoch": 25.787709497206706, + "grad_norm": 0.6596844792366028, + "learning_rate": 2.989683426443203e-05, + "loss": 0.1629, + "step": 9232 + }, + { + "epoch": 25.79050279329609, + "grad_norm": 0.6922438144683838, + "learning_rate": 2.9896461824953445e-05, + "loss": 0.1663, + "step": 9233 + }, + { + "epoch": 25.793296089385475, + "grad_norm": 0.9159852862358093, + "learning_rate": 2.9896089385474862e-05, + "loss": 0.1446, + "step": 9234 + }, + { + "epoch": 25.79608938547486, + "grad_norm": 0.9201744794845581, + "learning_rate": 2.9895716945996275e-05, + "loss": 0.1444, + "step": 9235 + }, + { + "epoch": 25.798882681564244, + "grad_norm": 0.6729410290718079, + "learning_rate": 2.989534450651769e-05, + "loss": 0.1511, + "step": 9236 + }, + { + "epoch": 25.80167597765363, + "grad_norm": 0.7114450335502625, + "learning_rate": 2.9894972067039108e-05, + "loss": 0.1412, + "step": 9237 + }, + { + "epoch": 25.804469273743017, + "grad_norm": 4.596339702606201, + "learning_rate": 2.9894599627560525e-05, + "loss": 0.1291, + "step": 9238 + }, + { + "epoch": 25.807262569832403, + "grad_norm": 0.7156644463539124, + "learning_rate": 2.9894227188081938e-05, + "loss": 0.1479, + "step": 9239 + }, + { + "epoch": 25.810055865921786, + "grad_norm": 0.78841632604599, + "learning_rate": 2.989385474860335e-05, + "loss": 0.1411, + "step": 9240 + }, + { + "epoch": 25.812849162011172, + "grad_norm": 0.5820415019989014, + "learning_rate": 2.9893482309124767e-05, + "loss": 0.1517, + "step": 9241 + }, + { + "epoch": 25.81564245810056, + "grad_norm": 0.48676934838294983, + "learning_rate": 2.9893109869646184e-05, + "loss": 0.1256, + "step": 9242 + }, + { + "epoch": 25.818435754189945, + "grad_norm": 0.8799119591712952, + "learning_rate": 2.98927374301676e-05, + "loss": 0.1483, + "step": 9243 + }, + { + "epoch": 25.821229050279328, + "grad_norm": 0.8315162062644958, + "learning_rate": 2.9892364990689014e-05, + "loss": 0.1499, + "step": 9244 + }, + { + "epoch": 25.824022346368714, + "grad_norm": 0.5977195501327515, + "learning_rate": 2.9891992551210427e-05, + "loss": 0.128, + "step": 9245 + }, + { + "epoch": 25.8268156424581, + "grad_norm": 1.360857367515564, + "learning_rate": 2.9891620111731843e-05, + "loss": 0.1887, + "step": 9246 + }, + { + "epoch": 25.829608938547487, + "grad_norm": 0.6698628664016724, + "learning_rate": 2.989124767225326e-05, + "loss": 0.1255, + "step": 9247 + }, + { + "epoch": 25.83240223463687, + "grad_norm": 0.9907734394073486, + "learning_rate": 2.9890875232774676e-05, + "loss": 0.16, + "step": 9248 + }, + { + "epoch": 25.835195530726256, + "grad_norm": 1.1862497329711914, + "learning_rate": 2.989050279329609e-05, + "loss": 0.1771, + "step": 9249 + }, + { + "epoch": 25.837988826815643, + "grad_norm": 3.2848293781280518, + "learning_rate": 2.9890130353817506e-05, + "loss": 0.1855, + "step": 9250 + }, + { + "epoch": 25.84078212290503, + "grad_norm": 0.3807092308998108, + "learning_rate": 2.9889757914338923e-05, + "loss": 0.1557, + "step": 9251 + }, + { + "epoch": 25.843575418994412, + "grad_norm": 0.6104427576065063, + "learning_rate": 2.9889385474860336e-05, + "loss": 0.198, + "step": 9252 + }, + { + "epoch": 25.8463687150838, + "grad_norm": 0.44026124477386475, + "learning_rate": 2.988901303538175e-05, + "loss": 0.1532, + "step": 9253 + }, + { + "epoch": 25.849162011173185, + "grad_norm": 0.4410969614982605, + "learning_rate": 2.9888640595903165e-05, + "loss": 0.1751, + "step": 9254 + }, + { + "epoch": 25.85195530726257, + "grad_norm": 0.41022032499313354, + "learning_rate": 2.9888268156424582e-05, + "loss": 0.1343, + "step": 9255 + }, + { + "epoch": 25.854748603351954, + "grad_norm": 0.4330178499221802, + "learning_rate": 2.9887895716946e-05, + "loss": 0.1347, + "step": 9256 + }, + { + "epoch": 25.85754189944134, + "grad_norm": 0.583601176738739, + "learning_rate": 2.9887523277467415e-05, + "loss": 0.1637, + "step": 9257 + }, + { + "epoch": 25.860335195530727, + "grad_norm": 0.7644236087799072, + "learning_rate": 2.9887150837988828e-05, + "loss": 0.1668, + "step": 9258 + }, + { + "epoch": 25.863128491620113, + "grad_norm": 0.8699336051940918, + "learning_rate": 2.988677839851024e-05, + "loss": 0.1569, + "step": 9259 + }, + { + "epoch": 25.865921787709496, + "grad_norm": 0.9704641103744507, + "learning_rate": 2.9886405959031658e-05, + "loss": 0.1405, + "step": 9260 + }, + { + "epoch": 25.868715083798882, + "grad_norm": 0.5593857765197754, + "learning_rate": 2.9886033519553074e-05, + "loss": 0.1409, + "step": 9261 + }, + { + "epoch": 25.87150837988827, + "grad_norm": 0.5991430282592773, + "learning_rate": 2.9885661080074488e-05, + "loss": 0.162, + "step": 9262 + }, + { + "epoch": 25.874301675977655, + "grad_norm": 0.553438663482666, + "learning_rate": 2.9885288640595904e-05, + "loss": 0.1349, + "step": 9263 + }, + { + "epoch": 25.877094972067038, + "grad_norm": 0.5024096369743347, + "learning_rate": 2.988491620111732e-05, + "loss": 0.1397, + "step": 9264 + }, + { + "epoch": 25.879888268156424, + "grad_norm": 0.9585495591163635, + "learning_rate": 2.9884543761638734e-05, + "loss": 0.1527, + "step": 9265 + }, + { + "epoch": 25.88268156424581, + "grad_norm": 0.5867948532104492, + "learning_rate": 2.988417132216015e-05, + "loss": 0.1376, + "step": 9266 + }, + { + "epoch": 25.885474860335197, + "grad_norm": 0.6280497908592224, + "learning_rate": 2.9883798882681563e-05, + "loss": 0.1398, + "step": 9267 + }, + { + "epoch": 25.88826815642458, + "grad_norm": 0.6992106437683105, + "learning_rate": 2.988342644320298e-05, + "loss": 0.1328, + "step": 9268 + }, + { + "epoch": 25.891061452513966, + "grad_norm": 0.594268262386322, + "learning_rate": 2.9883054003724396e-05, + "loss": 0.1163, + "step": 9269 + }, + { + "epoch": 25.893854748603353, + "grad_norm": 0.6193984746932983, + "learning_rate": 2.9882681564245813e-05, + "loss": 0.137, + "step": 9270 + }, + { + "epoch": 25.89664804469274, + "grad_norm": 0.85435950756073, + "learning_rate": 2.9882309124767226e-05, + "loss": 0.1665, + "step": 9271 + }, + { + "epoch": 25.899441340782122, + "grad_norm": 1.1685895919799805, + "learning_rate": 2.988193668528864e-05, + "loss": 0.1487, + "step": 9272 + }, + { + "epoch": 25.90223463687151, + "grad_norm": 3.59941029548645, + "learning_rate": 2.9881564245810056e-05, + "loss": 0.1582, + "step": 9273 + }, + { + "epoch": 25.905027932960895, + "grad_norm": 2.1970393657684326, + "learning_rate": 2.9881191806331472e-05, + "loss": 0.1652, + "step": 9274 + }, + { + "epoch": 25.907821229050278, + "grad_norm": 1.9986233711242676, + "learning_rate": 2.988081936685289e-05, + "loss": 0.2251, + "step": 9275 + }, + { + "epoch": 25.910614525139664, + "grad_norm": 0.4925353229045868, + "learning_rate": 2.9880446927374302e-05, + "loss": 0.1713, + "step": 9276 + }, + { + "epoch": 25.91340782122905, + "grad_norm": 1.4965527057647705, + "learning_rate": 2.988007448789572e-05, + "loss": 0.165, + "step": 9277 + }, + { + "epoch": 25.916201117318437, + "grad_norm": 0.6155573725700378, + "learning_rate": 2.9879702048417135e-05, + "loss": 0.1597, + "step": 9278 + }, + { + "epoch": 25.91899441340782, + "grad_norm": 0.4757748246192932, + "learning_rate": 2.9879329608938548e-05, + "loss": 0.1526, + "step": 9279 + }, + { + "epoch": 25.921787709497206, + "grad_norm": 1.4732215404510498, + "learning_rate": 2.987895716945996e-05, + "loss": 0.1962, + "step": 9280 + }, + { + "epoch": 25.924581005586592, + "grad_norm": 0.5755227208137512, + "learning_rate": 2.9878584729981378e-05, + "loss": 0.1441, + "step": 9281 + }, + { + "epoch": 25.92737430167598, + "grad_norm": 0.5922356843948364, + "learning_rate": 2.9878212290502795e-05, + "loss": 0.1508, + "step": 9282 + }, + { + "epoch": 25.93016759776536, + "grad_norm": 0.5633593201637268, + "learning_rate": 2.987783985102421e-05, + "loss": 0.1495, + "step": 9283 + }, + { + "epoch": 25.932960893854748, + "grad_norm": 0.5811849236488342, + "learning_rate": 2.9877467411545628e-05, + "loss": 0.1275, + "step": 9284 + }, + { + "epoch": 25.935754189944134, + "grad_norm": 0.8659882545471191, + "learning_rate": 2.9877094972067037e-05, + "loss": 0.1573, + "step": 9285 + }, + { + "epoch": 25.93854748603352, + "grad_norm": 1.5863651037216187, + "learning_rate": 2.9876722532588454e-05, + "loss": 0.1483, + "step": 9286 + }, + { + "epoch": 25.941340782122904, + "grad_norm": 0.5941846370697021, + "learning_rate": 2.987635009310987e-05, + "loss": 0.166, + "step": 9287 + }, + { + "epoch": 25.94413407821229, + "grad_norm": 0.898478627204895, + "learning_rate": 2.9875977653631287e-05, + "loss": 0.1633, + "step": 9288 + }, + { + "epoch": 25.946927374301676, + "grad_norm": 0.7499394416809082, + "learning_rate": 2.98756052141527e-05, + "loss": 0.1523, + "step": 9289 + }, + { + "epoch": 25.949720670391063, + "grad_norm": 0.8334057927131653, + "learning_rate": 2.9875232774674117e-05, + "loss": 0.1282, + "step": 9290 + }, + { + "epoch": 25.952513966480446, + "grad_norm": 0.7250874042510986, + "learning_rate": 2.9874860335195533e-05, + "loss": 0.1381, + "step": 9291 + }, + { + "epoch": 25.955307262569832, + "grad_norm": 0.6293525099754333, + "learning_rate": 2.9874487895716946e-05, + "loss": 0.1244, + "step": 9292 + }, + { + "epoch": 25.95810055865922, + "grad_norm": 0.8840732574462891, + "learning_rate": 2.9874115456238363e-05, + "loss": 0.1666, + "step": 9293 + }, + { + "epoch": 25.960893854748605, + "grad_norm": 0.6722509860992432, + "learning_rate": 2.9873743016759776e-05, + "loss": 0.1478, + "step": 9294 + }, + { + "epoch": 25.963687150837988, + "grad_norm": 0.755993664264679, + "learning_rate": 2.9873370577281193e-05, + "loss": 0.1346, + "step": 9295 + }, + { + "epoch": 25.966480446927374, + "grad_norm": 0.5117778778076172, + "learning_rate": 2.987299813780261e-05, + "loss": 0.1347, + "step": 9296 + }, + { + "epoch": 25.96927374301676, + "grad_norm": 0.9691697359085083, + "learning_rate": 2.9872625698324026e-05, + "loss": 0.1532, + "step": 9297 + }, + { + "epoch": 25.972067039106147, + "grad_norm": 1.4053813219070435, + "learning_rate": 2.987225325884544e-05, + "loss": 0.1517, + "step": 9298 + }, + { + "epoch": 25.97486033519553, + "grad_norm": 2.332792043685913, + "learning_rate": 2.9871880819366852e-05, + "loss": 0.1628, + "step": 9299 + }, + { + "epoch": 25.977653631284916, + "grad_norm": 1.4979182481765747, + "learning_rate": 2.987150837988827e-05, + "loss": 0.2363, + "step": 9300 + }, + { + "epoch": 25.980446927374302, + "grad_norm": 0.6956382989883423, + "learning_rate": 2.9871135940409685e-05, + "loss": 0.157, + "step": 9301 + }, + { + "epoch": 25.98324022346369, + "grad_norm": 0.9133117198944092, + "learning_rate": 2.9870763500931098e-05, + "loss": 0.1676, + "step": 9302 + }, + { + "epoch": 25.98603351955307, + "grad_norm": 0.6853080987930298, + "learning_rate": 2.9870391061452515e-05, + "loss": 0.1593, + "step": 9303 + }, + { + "epoch": 25.988826815642458, + "grad_norm": 0.5224616527557373, + "learning_rate": 2.987001862197393e-05, + "loss": 0.1284, + "step": 9304 + }, + { + "epoch": 25.991620111731844, + "grad_norm": 1.5018715858459473, + "learning_rate": 2.9869646182495344e-05, + "loss": 0.161, + "step": 9305 + }, + { + "epoch": 25.994413407821227, + "grad_norm": 0.8679895997047424, + "learning_rate": 2.986927374301676e-05, + "loss": 0.1645, + "step": 9306 + }, + { + "epoch": 25.997206703910614, + "grad_norm": 0.6282352209091187, + "learning_rate": 2.9868901303538174e-05, + "loss": 0.1081, + "step": 9307 + }, + { + "epoch": 26.0, + "grad_norm": 4.041632175445557, + "learning_rate": 2.986852886405959e-05, + "loss": 0.1558, + "step": 9308 + }, + { + "epoch": 26.002793296089386, + "grad_norm": 0.5724602937698364, + "learning_rate": 2.9868156424581007e-05, + "loss": 0.1591, + "step": 9309 + }, + { + "epoch": 26.00558659217877, + "grad_norm": 0.4911389648914337, + "learning_rate": 2.9867783985102424e-05, + "loss": 0.1755, + "step": 9310 + }, + { + "epoch": 26.008379888268156, + "grad_norm": 0.5776236653327942, + "learning_rate": 2.9867411545623837e-05, + "loss": 0.15, + "step": 9311 + }, + { + "epoch": 26.011173184357542, + "grad_norm": 0.833506166934967, + "learning_rate": 2.986703910614525e-05, + "loss": 0.1577, + "step": 9312 + }, + { + "epoch": 26.01396648044693, + "grad_norm": 0.5729937553405762, + "learning_rate": 2.9866666666666666e-05, + "loss": 0.1866, + "step": 9313 + }, + { + "epoch": 26.01675977653631, + "grad_norm": 0.6370832920074463, + "learning_rate": 2.9866294227188083e-05, + "loss": 0.1408, + "step": 9314 + }, + { + "epoch": 26.019553072625698, + "grad_norm": 0.46381187438964844, + "learning_rate": 2.98659217877095e-05, + "loss": 0.1456, + "step": 9315 + }, + { + "epoch": 26.022346368715084, + "grad_norm": 1.06697678565979, + "learning_rate": 2.9865549348230913e-05, + "loss": 0.1528, + "step": 9316 + }, + { + "epoch": 26.02513966480447, + "grad_norm": 0.6908245086669922, + "learning_rate": 2.986517690875233e-05, + "loss": 0.1486, + "step": 9317 + }, + { + "epoch": 26.027932960893853, + "grad_norm": 0.558224618434906, + "learning_rate": 2.9864804469273746e-05, + "loss": 0.1391, + "step": 9318 + }, + { + "epoch": 26.03072625698324, + "grad_norm": 0.5563209056854248, + "learning_rate": 2.986443202979516e-05, + "loss": 0.1543, + "step": 9319 + }, + { + "epoch": 26.033519553072626, + "grad_norm": 0.6997672319412231, + "learning_rate": 2.9864059590316572e-05, + "loss": 0.178, + "step": 9320 + }, + { + "epoch": 26.036312849162012, + "grad_norm": 0.7333249449729919, + "learning_rate": 2.986368715083799e-05, + "loss": 0.1384, + "step": 9321 + }, + { + "epoch": 26.039106145251395, + "grad_norm": 1.3427629470825195, + "learning_rate": 2.9863314711359405e-05, + "loss": 0.1467, + "step": 9322 + }, + { + "epoch": 26.04189944134078, + "grad_norm": 0.7773298621177673, + "learning_rate": 2.986294227188082e-05, + "loss": 0.1132, + "step": 9323 + }, + { + "epoch": 26.044692737430168, + "grad_norm": 0.6317611932754517, + "learning_rate": 2.9862569832402238e-05, + "loss": 0.1501, + "step": 9324 + }, + { + "epoch": 26.047486033519554, + "grad_norm": 1.3426066637039185, + "learning_rate": 2.9862197392923648e-05, + "loss": 0.1385, + "step": 9325 + }, + { + "epoch": 26.050279329608937, + "grad_norm": 0.6434794664382935, + "learning_rate": 2.9861824953445064e-05, + "loss": 0.1449, + "step": 9326 + }, + { + "epoch": 26.053072625698324, + "grad_norm": 0.7645469903945923, + "learning_rate": 2.986145251396648e-05, + "loss": 0.137, + "step": 9327 + }, + { + "epoch": 26.05586592178771, + "grad_norm": 0.6543658375740051, + "learning_rate": 2.9861080074487898e-05, + "loss": 0.1482, + "step": 9328 + }, + { + "epoch": 26.058659217877096, + "grad_norm": 1.4632171392440796, + "learning_rate": 2.986070763500931e-05, + "loss": 0.1606, + "step": 9329 + }, + { + "epoch": 26.06145251396648, + "grad_norm": 0.9797531962394714, + "learning_rate": 2.9860335195530727e-05, + "loss": 0.1267, + "step": 9330 + }, + { + "epoch": 26.064245810055866, + "grad_norm": 0.7958671450614929, + "learning_rate": 2.9859962756052144e-05, + "loss": 0.1448, + "step": 9331 + }, + { + "epoch": 26.067039106145252, + "grad_norm": 0.9325397610664368, + "learning_rate": 2.9859590316573557e-05, + "loss": 0.1639, + "step": 9332 + }, + { + "epoch": 26.06983240223464, + "grad_norm": 1.8728950023651123, + "learning_rate": 2.9859217877094973e-05, + "loss": 0.1782, + "step": 9333 + }, + { + "epoch": 26.07262569832402, + "grad_norm": 0.49834826588630676, + "learning_rate": 2.9858845437616387e-05, + "loss": 0.1511, + "step": 9334 + }, + { + "epoch": 26.075418994413408, + "grad_norm": 0.4259965419769287, + "learning_rate": 2.9858472998137803e-05, + "loss": 0.1354, + "step": 9335 + }, + { + "epoch": 26.078212290502794, + "grad_norm": 1.1168043613433838, + "learning_rate": 2.985810055865922e-05, + "loss": 0.1744, + "step": 9336 + }, + { + "epoch": 26.08100558659218, + "grad_norm": 0.6833975315093994, + "learning_rate": 2.9857728119180636e-05, + "loss": 0.1469, + "step": 9337 + }, + { + "epoch": 26.083798882681563, + "grad_norm": 0.5970063805580139, + "learning_rate": 2.985735567970205e-05, + "loss": 0.1586, + "step": 9338 + }, + { + "epoch": 26.08659217877095, + "grad_norm": 0.521450936794281, + "learning_rate": 2.9856983240223462e-05, + "loss": 0.1678, + "step": 9339 + }, + { + "epoch": 26.089385474860336, + "grad_norm": 1.5499398708343506, + "learning_rate": 2.985661080074488e-05, + "loss": 0.1502, + "step": 9340 + }, + { + "epoch": 26.092178770949722, + "grad_norm": 0.5407489538192749, + "learning_rate": 2.9856238361266296e-05, + "loss": 0.1543, + "step": 9341 + }, + { + "epoch": 26.094972067039105, + "grad_norm": 0.7399012446403503, + "learning_rate": 2.9855865921787712e-05, + "loss": 0.1304, + "step": 9342 + }, + { + "epoch": 26.09776536312849, + "grad_norm": 0.5876880884170532, + "learning_rate": 2.9855493482309125e-05, + "loss": 0.153, + "step": 9343 + }, + { + "epoch": 26.100558659217878, + "grad_norm": 0.6244477033615112, + "learning_rate": 2.9855121042830542e-05, + "loss": 0.1625, + "step": 9344 + }, + { + "epoch": 26.10335195530726, + "grad_norm": 0.6264707446098328, + "learning_rate": 2.9854748603351955e-05, + "loss": 0.1428, + "step": 9345 + }, + { + "epoch": 26.106145251396647, + "grad_norm": 0.7058553099632263, + "learning_rate": 2.985437616387337e-05, + "loss": 0.1574, + "step": 9346 + }, + { + "epoch": 26.108938547486034, + "grad_norm": 1.4150114059448242, + "learning_rate": 2.9854003724394785e-05, + "loss": 0.1548, + "step": 9347 + }, + { + "epoch": 26.11173184357542, + "grad_norm": 0.866837203502655, + "learning_rate": 2.98536312849162e-05, + "loss": 0.1605, + "step": 9348 + }, + { + "epoch": 26.114525139664803, + "grad_norm": 0.5783774256706238, + "learning_rate": 2.9853258845437618e-05, + "loss": 0.142, + "step": 9349 + }, + { + "epoch": 26.11731843575419, + "grad_norm": 0.5454436540603638, + "learning_rate": 2.9852886405959034e-05, + "loss": 0.1263, + "step": 9350 + }, + { + "epoch": 26.120111731843576, + "grad_norm": 1.1593965291976929, + "learning_rate": 2.985251396648045e-05, + "loss": 0.1478, + "step": 9351 + }, + { + "epoch": 26.122905027932962, + "grad_norm": 1.1270662546157837, + "learning_rate": 2.985214152700186e-05, + "loss": 0.1401, + "step": 9352 + }, + { + "epoch": 26.125698324022345, + "grad_norm": 0.5751274228096008, + "learning_rate": 2.9851769087523277e-05, + "loss": 0.11, + "step": 9353 + }, + { + "epoch": 26.12849162011173, + "grad_norm": 0.7086458206176758, + "learning_rate": 2.9851396648044694e-05, + "loss": 0.1299, + "step": 9354 + }, + { + "epoch": 26.131284916201118, + "grad_norm": 1.0629396438598633, + "learning_rate": 2.985102420856611e-05, + "loss": 0.1141, + "step": 9355 + }, + { + "epoch": 26.134078212290504, + "grad_norm": 0.7715195417404175, + "learning_rate": 2.9850651769087523e-05, + "loss": 0.151, + "step": 9356 + }, + { + "epoch": 26.136871508379887, + "grad_norm": 1.7217628955841064, + "learning_rate": 2.985027932960894e-05, + "loss": 0.1909, + "step": 9357 + }, + { + "epoch": 26.139664804469273, + "grad_norm": 1.585845708847046, + "learning_rate": 2.9849906890130356e-05, + "loss": 0.1735, + "step": 9358 + }, + { + "epoch": 26.14245810055866, + "grad_norm": 0.6338599920272827, + "learning_rate": 2.984953445065177e-05, + "loss": 0.1753, + "step": 9359 + }, + { + "epoch": 26.145251396648046, + "grad_norm": 0.6912460923194885, + "learning_rate": 2.9849162011173186e-05, + "loss": 0.1624, + "step": 9360 + }, + { + "epoch": 26.14804469273743, + "grad_norm": 0.5436035394668579, + "learning_rate": 2.98487895716946e-05, + "loss": 0.1585, + "step": 9361 + }, + { + "epoch": 26.150837988826815, + "grad_norm": 0.3671044707298279, + "learning_rate": 2.9848417132216016e-05, + "loss": 0.1454, + "step": 9362 + }, + { + "epoch": 26.1536312849162, + "grad_norm": 0.6932704448699951, + "learning_rate": 2.9848044692737432e-05, + "loss": 0.1963, + "step": 9363 + }, + { + "epoch": 26.156424581005588, + "grad_norm": 0.5384007096290588, + "learning_rate": 2.984767225325885e-05, + "loss": 0.1354, + "step": 9364 + }, + { + "epoch": 26.15921787709497, + "grad_norm": 1.2964842319488525, + "learning_rate": 2.984729981378026e-05, + "loss": 0.1919, + "step": 9365 + }, + { + "epoch": 26.162011173184357, + "grad_norm": 0.6072390675544739, + "learning_rate": 2.9846927374301675e-05, + "loss": 0.1272, + "step": 9366 + }, + { + "epoch": 26.164804469273744, + "grad_norm": 0.508122444152832, + "learning_rate": 2.984655493482309e-05, + "loss": 0.1308, + "step": 9367 + }, + { + "epoch": 26.16759776536313, + "grad_norm": 0.9222935438156128, + "learning_rate": 2.9846182495344508e-05, + "loss": 0.1538, + "step": 9368 + }, + { + "epoch": 26.170391061452513, + "grad_norm": 0.4824483096599579, + "learning_rate": 2.9845810055865925e-05, + "loss": 0.1242, + "step": 9369 + }, + { + "epoch": 26.1731843575419, + "grad_norm": 0.5789571404457092, + "learning_rate": 2.9845437616387338e-05, + "loss": 0.1184, + "step": 9370 + }, + { + "epoch": 26.175977653631286, + "grad_norm": 0.5944896936416626, + "learning_rate": 2.9845065176908754e-05, + "loss": 0.1415, + "step": 9371 + }, + { + "epoch": 26.178770949720672, + "grad_norm": 1.4907135963439941, + "learning_rate": 2.9844692737430167e-05, + "loss": 0.1708, + "step": 9372 + }, + { + "epoch": 26.181564245810055, + "grad_norm": 0.548569917678833, + "learning_rate": 2.9844320297951584e-05, + "loss": 0.1418, + "step": 9373 + }, + { + "epoch": 26.18435754189944, + "grad_norm": 0.5022264122962952, + "learning_rate": 2.9843947858472997e-05, + "loss": 0.1099, + "step": 9374 + }, + { + "epoch": 26.187150837988828, + "grad_norm": 1.6007317304611206, + "learning_rate": 2.9843575418994414e-05, + "loss": 0.1502, + "step": 9375 + }, + { + "epoch": 26.189944134078214, + "grad_norm": 0.7463605999946594, + "learning_rate": 2.984320297951583e-05, + "loss": 0.1445, + "step": 9376 + }, + { + "epoch": 26.192737430167597, + "grad_norm": 0.9464161992073059, + "learning_rate": 2.9842830540037247e-05, + "loss": 0.1604, + "step": 9377 + }, + { + "epoch": 26.195530726256983, + "grad_norm": 0.9062737822532654, + "learning_rate": 2.9842458100558663e-05, + "loss": 0.1738, + "step": 9378 + }, + { + "epoch": 26.19832402234637, + "grad_norm": 0.5804096460342407, + "learning_rate": 2.9842085661080073e-05, + "loss": 0.135, + "step": 9379 + }, + { + "epoch": 26.201117318435756, + "grad_norm": 0.8583986163139343, + "learning_rate": 2.984171322160149e-05, + "loss": 0.1276, + "step": 9380 + }, + { + "epoch": 26.20391061452514, + "grad_norm": 1.0123659372329712, + "learning_rate": 2.9841340782122906e-05, + "loss": 0.1565, + "step": 9381 + }, + { + "epoch": 26.206703910614525, + "grad_norm": 1.7308744192123413, + "learning_rate": 2.9840968342644323e-05, + "loss": 0.1567, + "step": 9382 + }, + { + "epoch": 26.20949720670391, + "grad_norm": 1.5220764875411987, + "learning_rate": 2.9840595903165736e-05, + "loss": 0.1825, + "step": 9383 + }, + { + "epoch": 26.212290502793294, + "grad_norm": 0.4562910199165344, + "learning_rate": 2.9840223463687152e-05, + "loss": 0.1683, + "step": 9384 + }, + { + "epoch": 26.21508379888268, + "grad_norm": 0.40482819080352783, + "learning_rate": 2.9839851024208566e-05, + "loss": 0.1592, + "step": 9385 + }, + { + "epoch": 26.217877094972067, + "grad_norm": 0.8329761624336243, + "learning_rate": 2.9839478584729982e-05, + "loss": 0.1437, + "step": 9386 + }, + { + "epoch": 26.220670391061454, + "grad_norm": 0.5037257671356201, + "learning_rate": 2.98391061452514e-05, + "loss": 0.1569, + "step": 9387 + }, + { + "epoch": 26.223463687150836, + "grad_norm": 0.8361421823501587, + "learning_rate": 2.9838733705772812e-05, + "loss": 0.149, + "step": 9388 + }, + { + "epoch": 26.226256983240223, + "grad_norm": 0.5865645408630371, + "learning_rate": 2.9838361266294228e-05, + "loss": 0.1406, + "step": 9389 + }, + { + "epoch": 26.22905027932961, + "grad_norm": 0.5382375717163086, + "learning_rate": 2.9837988826815645e-05, + "loss": 0.1488, + "step": 9390 + }, + { + "epoch": 26.231843575418996, + "grad_norm": 0.4211346507072449, + "learning_rate": 2.983761638733706e-05, + "loss": 0.1626, + "step": 9391 + }, + { + "epoch": 26.23463687150838, + "grad_norm": 0.5926432609558105, + "learning_rate": 2.983724394785847e-05, + "loss": 0.1256, + "step": 9392 + }, + { + "epoch": 26.237430167597765, + "grad_norm": 0.4954787492752075, + "learning_rate": 2.9836871508379888e-05, + "loss": 0.1294, + "step": 9393 + }, + { + "epoch": 26.24022346368715, + "grad_norm": 1.3541737794876099, + "learning_rate": 2.9836499068901304e-05, + "loss": 0.1485, + "step": 9394 + }, + { + "epoch": 26.243016759776538, + "grad_norm": 0.8562806248664856, + "learning_rate": 2.983612662942272e-05, + "loss": 0.1193, + "step": 9395 + }, + { + "epoch": 26.24581005586592, + "grad_norm": 1.025938630104065, + "learning_rate": 2.9835754189944134e-05, + "loss": 0.167, + "step": 9396 + }, + { + "epoch": 26.248603351955307, + "grad_norm": 0.4916617274284363, + "learning_rate": 2.983538175046555e-05, + "loss": 0.1274, + "step": 9397 + }, + { + "epoch": 26.251396648044693, + "grad_norm": 0.964035153388977, + "learning_rate": 2.9835009310986964e-05, + "loss": 0.1224, + "step": 9398 + }, + { + "epoch": 26.25418994413408, + "grad_norm": 0.8861874341964722, + "learning_rate": 2.983463687150838e-05, + "loss": 0.1409, + "step": 9399 + }, + { + "epoch": 26.256983240223462, + "grad_norm": 0.8547370433807373, + "learning_rate": 2.9834264432029797e-05, + "loss": 0.1543, + "step": 9400 + }, + { + "epoch": 26.25977653631285, + "grad_norm": 0.5261646509170532, + "learning_rate": 2.983389199255121e-05, + "loss": 0.1324, + "step": 9401 + }, + { + "epoch": 26.262569832402235, + "grad_norm": 0.952120304107666, + "learning_rate": 2.9833519553072626e-05, + "loss": 0.1372, + "step": 9402 + }, + { + "epoch": 26.26536312849162, + "grad_norm": 0.5354689359664917, + "learning_rate": 2.9833147113594043e-05, + "loss": 0.1402, + "step": 9403 + }, + { + "epoch": 26.268156424581004, + "grad_norm": 1.6907576322555542, + "learning_rate": 2.983277467411546e-05, + "loss": 0.1113, + "step": 9404 + }, + { + "epoch": 26.27094972067039, + "grad_norm": 0.9588509798049927, + "learning_rate": 2.983240223463687e-05, + "loss": 0.1226, + "step": 9405 + }, + { + "epoch": 26.273743016759777, + "grad_norm": 0.7599055767059326, + "learning_rate": 2.9832029795158286e-05, + "loss": 0.1138, + "step": 9406 + }, + { + "epoch": 26.276536312849164, + "grad_norm": 1.325898289680481, + "learning_rate": 2.9831657355679702e-05, + "loss": 0.1649, + "step": 9407 + }, + { + "epoch": 26.279329608938546, + "grad_norm": 1.3904871940612793, + "learning_rate": 2.983128491620112e-05, + "loss": 0.205, + "step": 9408 + }, + { + "epoch": 26.282122905027933, + "grad_norm": 0.48487961292266846, + "learning_rate": 2.9830912476722535e-05, + "loss": 0.1516, + "step": 9409 + }, + { + "epoch": 26.28491620111732, + "grad_norm": 0.4065689444541931, + "learning_rate": 2.983054003724395e-05, + "loss": 0.1586, + "step": 9410 + }, + { + "epoch": 26.287709497206706, + "grad_norm": 0.4242178499698639, + "learning_rate": 2.9830167597765365e-05, + "loss": 0.1602, + "step": 9411 + }, + { + "epoch": 26.29050279329609, + "grad_norm": 0.5300463438034058, + "learning_rate": 2.9829795158286778e-05, + "loss": 0.1412, + "step": 9412 + }, + { + "epoch": 26.293296089385475, + "grad_norm": 0.6833614706993103, + "learning_rate": 2.9829422718808195e-05, + "loss": 0.1405, + "step": 9413 + }, + { + "epoch": 26.29608938547486, + "grad_norm": 0.4498862028121948, + "learning_rate": 2.9829050279329608e-05, + "loss": 0.1582, + "step": 9414 + }, + { + "epoch": 26.298882681564244, + "grad_norm": 0.5042440891265869, + "learning_rate": 2.9828677839851024e-05, + "loss": 0.1354, + "step": 9415 + }, + { + "epoch": 26.30167597765363, + "grad_norm": 0.4406660795211792, + "learning_rate": 2.982830540037244e-05, + "loss": 0.1351, + "step": 9416 + }, + { + "epoch": 26.304469273743017, + "grad_norm": 0.7120200991630554, + "learning_rate": 2.9827932960893857e-05, + "loss": 0.1413, + "step": 9417 + }, + { + "epoch": 26.307262569832403, + "grad_norm": 0.5179228782653809, + "learning_rate": 2.982756052141527e-05, + "loss": 0.1653, + "step": 9418 + }, + { + "epoch": 26.310055865921786, + "grad_norm": 0.9939771294593811, + "learning_rate": 2.9827188081936684e-05, + "loss": 0.1618, + "step": 9419 + }, + { + "epoch": 26.312849162011172, + "grad_norm": 0.7988784313201904, + "learning_rate": 2.98268156424581e-05, + "loss": 0.1359, + "step": 9420 + }, + { + "epoch": 26.31564245810056, + "grad_norm": 1.0193933248519897, + "learning_rate": 2.9826443202979517e-05, + "loss": 0.1345, + "step": 9421 + }, + { + "epoch": 26.318435754189945, + "grad_norm": 1.62282133102417, + "learning_rate": 2.9826070763500933e-05, + "loss": 0.1641, + "step": 9422 + }, + { + "epoch": 26.321229050279328, + "grad_norm": 0.844343364238739, + "learning_rate": 2.9825698324022346e-05, + "loss": 0.1608, + "step": 9423 + }, + { + "epoch": 26.324022346368714, + "grad_norm": 0.8903485536575317, + "learning_rate": 2.9825325884543763e-05, + "loss": 0.1342, + "step": 9424 + }, + { + "epoch": 26.3268156424581, + "grad_norm": 0.6465596556663513, + "learning_rate": 2.9824953445065176e-05, + "loss": 0.1257, + "step": 9425 + }, + { + "epoch": 26.329608938547487, + "grad_norm": 0.4592575430870056, + "learning_rate": 2.9824581005586593e-05, + "loss": 0.1369, + "step": 9426 + }, + { + "epoch": 26.33240223463687, + "grad_norm": 0.6561300754547119, + "learning_rate": 2.982420856610801e-05, + "loss": 0.1305, + "step": 9427 + }, + { + "epoch": 26.335195530726256, + "grad_norm": 0.65750652551651, + "learning_rate": 2.9823836126629422e-05, + "loss": 0.1678, + "step": 9428 + }, + { + "epoch": 26.337988826815643, + "grad_norm": 0.8124123811721802, + "learning_rate": 2.982346368715084e-05, + "loss": 0.1409, + "step": 9429 + }, + { + "epoch": 26.34078212290503, + "grad_norm": 1.7620444297790527, + "learning_rate": 2.9823091247672255e-05, + "loss": 0.1378, + "step": 9430 + }, + { + "epoch": 26.343575418994412, + "grad_norm": 0.5832147598266602, + "learning_rate": 2.9822718808193672e-05, + "loss": 0.1541, + "step": 9431 + }, + { + "epoch": 26.3463687150838, + "grad_norm": 1.234496831893921, + "learning_rate": 2.9822346368715082e-05, + "loss": 0.1952, + "step": 9432 + }, + { + "epoch": 26.349162011173185, + "grad_norm": 1.635608434677124, + "learning_rate": 2.9821973929236498e-05, + "loss": 0.212, + "step": 9433 + }, + { + "epoch": 26.35195530726257, + "grad_norm": 0.5835626721382141, + "learning_rate": 2.9821601489757915e-05, + "loss": 0.1486, + "step": 9434 + }, + { + "epoch": 26.354748603351954, + "grad_norm": 0.4507658779621124, + "learning_rate": 2.982122905027933e-05, + "loss": 0.16, + "step": 9435 + }, + { + "epoch": 26.35754189944134, + "grad_norm": 0.3969454765319824, + "learning_rate": 2.9820856610800748e-05, + "loss": 0.1319, + "step": 9436 + }, + { + "epoch": 26.360335195530727, + "grad_norm": 1.0673774480819702, + "learning_rate": 2.982048417132216e-05, + "loss": 0.1579, + "step": 9437 + }, + { + "epoch": 26.363128491620113, + "grad_norm": 0.6487478613853455, + "learning_rate": 2.9820111731843574e-05, + "loss": 0.1464, + "step": 9438 + }, + { + "epoch": 26.365921787709496, + "grad_norm": 0.6369600892066956, + "learning_rate": 2.981973929236499e-05, + "loss": 0.153, + "step": 9439 + }, + { + "epoch": 26.368715083798882, + "grad_norm": 0.5514166951179504, + "learning_rate": 2.9819366852886407e-05, + "loss": 0.1602, + "step": 9440 + }, + { + "epoch": 26.37150837988827, + "grad_norm": 0.7027620077133179, + "learning_rate": 2.981899441340782e-05, + "loss": 0.1572, + "step": 9441 + }, + { + "epoch": 26.374301675977655, + "grad_norm": 0.4553777575492859, + "learning_rate": 2.9818621973929237e-05, + "loss": 0.1407, + "step": 9442 + }, + { + "epoch": 26.377094972067038, + "grad_norm": 0.8435099720954895, + "learning_rate": 2.9818249534450653e-05, + "loss": 0.169, + "step": 9443 + }, + { + "epoch": 26.379888268156424, + "grad_norm": 0.49260208010673523, + "learning_rate": 2.981787709497207e-05, + "loss": 0.1314, + "step": 9444 + }, + { + "epoch": 26.38268156424581, + "grad_norm": 0.6398764848709106, + "learning_rate": 2.9817504655493483e-05, + "loss": 0.1361, + "step": 9445 + }, + { + "epoch": 26.385474860335197, + "grad_norm": 0.5376390218734741, + "learning_rate": 2.9817132216014896e-05, + "loss": 0.1381, + "step": 9446 + }, + { + "epoch": 26.38826815642458, + "grad_norm": 0.917216956615448, + "learning_rate": 2.9816759776536313e-05, + "loss": 0.1338, + "step": 9447 + }, + { + "epoch": 26.391061452513966, + "grad_norm": 0.7551066875457764, + "learning_rate": 2.981638733705773e-05, + "loss": 0.138, + "step": 9448 + }, + { + "epoch": 26.393854748603353, + "grad_norm": 0.6527129411697388, + "learning_rate": 2.9816014897579146e-05, + "loss": 0.1407, + "step": 9449 + }, + { + "epoch": 26.39664804469274, + "grad_norm": 1.1894927024841309, + "learning_rate": 2.981564245810056e-05, + "loss": 0.1212, + "step": 9450 + }, + { + "epoch": 26.399441340782122, + "grad_norm": 0.7208681702613831, + "learning_rate": 2.9815270018621976e-05, + "loss": 0.1269, + "step": 9451 + }, + { + "epoch": 26.40223463687151, + "grad_norm": 0.836277425289154, + "learning_rate": 2.981489757914339e-05, + "loss": 0.1367, + "step": 9452 + }, + { + "epoch": 26.405027932960895, + "grad_norm": 0.7971490621566772, + "learning_rate": 2.9814525139664805e-05, + "loss": 0.1323, + "step": 9453 + }, + { + "epoch": 26.407821229050278, + "grad_norm": 0.8257943391799927, + "learning_rate": 2.9814152700186222e-05, + "loss": 0.1365, + "step": 9454 + }, + { + "epoch": 26.410614525139664, + "grad_norm": 1.0504261255264282, + "learning_rate": 2.9813780260707635e-05, + "loss": 0.1469, + "step": 9455 + }, + { + "epoch": 26.41340782122905, + "grad_norm": 0.6743394732475281, + "learning_rate": 2.981340782122905e-05, + "loss": 0.1519, + "step": 9456 + }, + { + "epoch": 26.416201117318437, + "grad_norm": 1.2625887393951416, + "learning_rate": 2.9813035381750468e-05, + "loss": 0.1925, + "step": 9457 + }, + { + "epoch": 26.41899441340782, + "grad_norm": 1.4963619709014893, + "learning_rate": 2.981266294227188e-05, + "loss": 0.1802, + "step": 9458 + }, + { + "epoch": 26.421787709497206, + "grad_norm": 0.6221705675125122, + "learning_rate": 2.9812290502793294e-05, + "loss": 0.1973, + "step": 9459 + }, + { + "epoch": 26.424581005586592, + "grad_norm": 1.261310338973999, + "learning_rate": 2.981191806331471e-05, + "loss": 0.1634, + "step": 9460 + }, + { + "epoch": 26.42737430167598, + "grad_norm": 0.5994076728820801, + "learning_rate": 2.9811545623836127e-05, + "loss": 0.1684, + "step": 9461 + }, + { + "epoch": 26.43016759776536, + "grad_norm": 0.8657132983207703, + "learning_rate": 2.9811173184357544e-05, + "loss": 0.1577, + "step": 9462 + }, + { + "epoch": 26.432960893854748, + "grad_norm": 0.6957834362983704, + "learning_rate": 2.981080074487896e-05, + "loss": 0.1535, + "step": 9463 + }, + { + "epoch": 26.435754189944134, + "grad_norm": 0.5078385472297668, + "learning_rate": 2.9810428305400374e-05, + "loss": 0.1496, + "step": 9464 + }, + { + "epoch": 26.43854748603352, + "grad_norm": 0.5304924845695496, + "learning_rate": 2.9810055865921787e-05, + "loss": 0.1493, + "step": 9465 + }, + { + "epoch": 26.441340782122904, + "grad_norm": 0.4354632496833801, + "learning_rate": 2.9809683426443203e-05, + "loss": 0.1455, + "step": 9466 + }, + { + "epoch": 26.44413407821229, + "grad_norm": 0.612995445728302, + "learning_rate": 2.980931098696462e-05, + "loss": 0.1516, + "step": 9467 + }, + { + "epoch": 26.446927374301676, + "grad_norm": 0.7459174394607544, + "learning_rate": 2.9808938547486033e-05, + "loss": 0.1475, + "step": 9468 + }, + { + "epoch": 26.449720670391063, + "grad_norm": 0.848572850227356, + "learning_rate": 2.980856610800745e-05, + "loss": 0.1794, + "step": 9469 + }, + { + "epoch": 26.452513966480446, + "grad_norm": 0.7755463719367981, + "learning_rate": 2.9808193668528866e-05, + "loss": 0.1604, + "step": 9470 + }, + { + "epoch": 26.455307262569832, + "grad_norm": 0.5248885154724121, + "learning_rate": 2.9807821229050283e-05, + "loss": 0.1505, + "step": 9471 + }, + { + "epoch": 26.45810055865922, + "grad_norm": 0.5991409420967102, + "learning_rate": 2.9807448789571696e-05, + "loss": 0.1377, + "step": 9472 + }, + { + "epoch": 26.460893854748605, + "grad_norm": 0.6295472979545593, + "learning_rate": 2.980707635009311e-05, + "loss": 0.1411, + "step": 9473 + }, + { + "epoch": 26.463687150837988, + "grad_norm": 1.814648151397705, + "learning_rate": 2.9806703910614525e-05, + "loss": 0.139, + "step": 9474 + }, + { + "epoch": 26.466480446927374, + "grad_norm": 1.412001609802246, + "learning_rate": 2.9806331471135942e-05, + "loss": 0.1431, + "step": 9475 + }, + { + "epoch": 26.46927374301676, + "grad_norm": 0.7148787975311279, + "learning_rate": 2.980595903165736e-05, + "loss": 0.1495, + "step": 9476 + }, + { + "epoch": 26.472067039106147, + "grad_norm": 0.7522233724594116, + "learning_rate": 2.980558659217877e-05, + "loss": 0.1393, + "step": 9477 + }, + { + "epoch": 26.47486033519553, + "grad_norm": 0.7239769697189331, + "learning_rate": 2.9805214152700185e-05, + "loss": 0.1354, + "step": 9478 + }, + { + "epoch": 26.477653631284916, + "grad_norm": 1.7614487409591675, + "learning_rate": 2.98048417132216e-05, + "loss": 0.1438, + "step": 9479 + }, + { + "epoch": 26.480446927374302, + "grad_norm": 1.0193467140197754, + "learning_rate": 2.9804469273743018e-05, + "loss": 0.1485, + "step": 9480 + }, + { + "epoch": 26.48324022346369, + "grad_norm": 1.5893900394439697, + "learning_rate": 2.9804096834264434e-05, + "loss": 0.1488, + "step": 9481 + }, + { + "epoch": 26.48603351955307, + "grad_norm": 0.7154459953308105, + "learning_rate": 2.9803724394785847e-05, + "loss": 0.1287, + "step": 9482 + }, + { + "epoch": 26.488826815642458, + "grad_norm": 1.3972834348678589, + "learning_rate": 2.9803351955307264e-05, + "loss": 0.2112, + "step": 9483 + }, + { + "epoch": 26.491620111731844, + "grad_norm": 0.5065316557884216, + "learning_rate": 2.980297951582868e-05, + "loss": 0.1614, + "step": 9484 + }, + { + "epoch": 26.49441340782123, + "grad_norm": 0.533139705657959, + "learning_rate": 2.9802607076350094e-05, + "loss": 0.1498, + "step": 9485 + }, + { + "epoch": 26.497206703910614, + "grad_norm": 0.49991732835769653, + "learning_rate": 2.9802234636871507e-05, + "loss": 0.1669, + "step": 9486 + }, + { + "epoch": 26.5, + "grad_norm": 0.7595196962356567, + "learning_rate": 2.9801862197392923e-05, + "loss": 0.1264, + "step": 9487 + }, + { + "epoch": 26.502793296089386, + "grad_norm": 0.5659120678901672, + "learning_rate": 2.980148975791434e-05, + "loss": 0.1722, + "step": 9488 + }, + { + "epoch": 26.505586592178773, + "grad_norm": 1.35505211353302, + "learning_rate": 2.9801117318435756e-05, + "loss": 0.1606, + "step": 9489 + }, + { + "epoch": 26.508379888268156, + "grad_norm": 0.6923065185546875, + "learning_rate": 2.980074487895717e-05, + "loss": 0.1546, + "step": 9490 + }, + { + "epoch": 26.511173184357542, + "grad_norm": 0.45740383863449097, + "learning_rate": 2.9800372439478586e-05, + "loss": 0.1411, + "step": 9491 + }, + { + "epoch": 26.51396648044693, + "grad_norm": 0.8492443561553955, + "learning_rate": 2.98e-05, + "loss": 0.1681, + "step": 9492 + }, + { + "epoch": 26.51675977653631, + "grad_norm": 0.6114475131034851, + "learning_rate": 2.9799627560521416e-05, + "loss": 0.1608, + "step": 9493 + }, + { + "epoch": 26.519553072625698, + "grad_norm": 0.5532419681549072, + "learning_rate": 2.9799255121042832e-05, + "loss": 0.1364, + "step": 9494 + }, + { + "epoch": 26.522346368715084, + "grad_norm": 1.303935170173645, + "learning_rate": 2.9798882681564246e-05, + "loss": 0.1439, + "step": 9495 + }, + { + "epoch": 26.52513966480447, + "grad_norm": 0.8513576984405518, + "learning_rate": 2.9798510242085662e-05, + "loss": 0.1447, + "step": 9496 + }, + { + "epoch": 26.527932960893853, + "grad_norm": 0.5748317837715149, + "learning_rate": 2.979813780260708e-05, + "loss": 0.1346, + "step": 9497 + }, + { + "epoch": 26.53072625698324, + "grad_norm": 0.9459010362625122, + "learning_rate": 2.9797765363128492e-05, + "loss": 0.1329, + "step": 9498 + }, + { + "epoch": 26.533519553072626, + "grad_norm": 0.9295790195465088, + "learning_rate": 2.9797392923649905e-05, + "loss": 0.1259, + "step": 9499 + }, + { + "epoch": 26.536312849162012, + "grad_norm": 0.5307859182357788, + "learning_rate": 2.979702048417132e-05, + "loss": 0.1445, + "step": 9500 + }, + { + "epoch": 26.539106145251395, + "grad_norm": 1.5999679565429688, + "learning_rate": 2.9796648044692738e-05, + "loss": 0.1192, + "step": 9501 + }, + { + "epoch": 26.54189944134078, + "grad_norm": 0.9768978953361511, + "learning_rate": 2.9796275605214154e-05, + "loss": 0.149, + "step": 9502 + }, + { + "epoch": 26.544692737430168, + "grad_norm": 1.3280194997787476, + "learning_rate": 2.979590316573557e-05, + "loss": 0.1494, + "step": 9503 + }, + { + "epoch": 26.547486033519554, + "grad_norm": 0.8773160576820374, + "learning_rate": 2.9795530726256984e-05, + "loss": 0.1251, + "step": 9504 + }, + { + "epoch": 26.550279329608937, + "grad_norm": 0.9990923404693604, + "learning_rate": 2.9795158286778397e-05, + "loss": 0.135, + "step": 9505 + }, + { + "epoch": 26.553072625698324, + "grad_norm": 1.2805626392364502, + "learning_rate": 2.9794785847299814e-05, + "loss": 0.1393, + "step": 9506 + }, + { + "epoch": 26.55586592178771, + "grad_norm": 1.4124939441680908, + "learning_rate": 2.979441340782123e-05, + "loss": 0.1302, + "step": 9507 + }, + { + "epoch": 26.558659217877096, + "grad_norm": 2.5481340885162354, + "learning_rate": 2.9794040968342644e-05, + "loss": 0.1836, + "step": 9508 + }, + { + "epoch": 26.56145251396648, + "grad_norm": 0.4075491428375244, + "learning_rate": 2.979366852886406e-05, + "loss": 0.163, + "step": 9509 + }, + { + "epoch": 26.564245810055866, + "grad_norm": 0.3608132600784302, + "learning_rate": 2.9793296089385477e-05, + "loss": 0.1463, + "step": 9510 + }, + { + "epoch": 26.567039106145252, + "grad_norm": 0.8580267429351807, + "learning_rate": 2.9792923649906893e-05, + "loss": 0.1897, + "step": 9511 + }, + { + "epoch": 26.56983240223464, + "grad_norm": 0.5543921589851379, + "learning_rate": 2.9792551210428306e-05, + "loss": 0.1541, + "step": 9512 + }, + { + "epoch": 26.57262569832402, + "grad_norm": 0.5502670407295227, + "learning_rate": 2.979217877094972e-05, + "loss": 0.1436, + "step": 9513 + }, + { + "epoch": 26.575418994413408, + "grad_norm": 0.9712799191474915, + "learning_rate": 2.9791806331471136e-05, + "loss": 0.1541, + "step": 9514 + }, + { + "epoch": 26.578212290502794, + "grad_norm": 0.6206946969032288, + "learning_rate": 2.9791433891992552e-05, + "loss": 0.1665, + "step": 9515 + }, + { + "epoch": 26.58100558659218, + "grad_norm": 0.627593457698822, + "learning_rate": 2.979106145251397e-05, + "loss": 0.1619, + "step": 9516 + }, + { + "epoch": 26.583798882681563, + "grad_norm": 0.8098658323287964, + "learning_rate": 2.9790689013035382e-05, + "loss": 0.1279, + "step": 9517 + }, + { + "epoch": 26.58659217877095, + "grad_norm": 1.7921704053878784, + "learning_rate": 2.9790316573556795e-05, + "loss": 0.1481, + "step": 9518 + }, + { + "epoch": 26.589385474860336, + "grad_norm": 0.6893184781074524, + "learning_rate": 2.9789944134078212e-05, + "loss": 0.1695, + "step": 9519 + }, + { + "epoch": 26.592178770949722, + "grad_norm": 0.716614305973053, + "learning_rate": 2.978957169459963e-05, + "loss": 0.1338, + "step": 9520 + }, + { + "epoch": 26.594972067039105, + "grad_norm": 0.8705875277519226, + "learning_rate": 2.9789199255121045e-05, + "loss": 0.14, + "step": 9521 + }, + { + "epoch": 26.59776536312849, + "grad_norm": 0.6066675186157227, + "learning_rate": 2.9788826815642458e-05, + "loss": 0.1527, + "step": 9522 + }, + { + "epoch": 26.600558659217878, + "grad_norm": 0.5573334693908691, + "learning_rate": 2.9788454376163875e-05, + "loss": 0.1357, + "step": 9523 + }, + { + "epoch": 26.60335195530726, + "grad_norm": 1.9442429542541504, + "learning_rate": 2.978808193668529e-05, + "loss": 0.1255, + "step": 9524 + }, + { + "epoch": 26.606145251396647, + "grad_norm": 1.1305309534072876, + "learning_rate": 2.9787709497206704e-05, + "loss": 0.1635, + "step": 9525 + }, + { + "epoch": 26.608938547486034, + "grad_norm": 1.1939131021499634, + "learning_rate": 2.9787337057728117e-05, + "loss": 0.1568, + "step": 9526 + }, + { + "epoch": 26.61173184357542, + "grad_norm": 0.9040828943252563, + "learning_rate": 2.9786964618249534e-05, + "loss": 0.1296, + "step": 9527 + }, + { + "epoch": 26.614525139664803, + "grad_norm": 1.0297651290893555, + "learning_rate": 2.978659217877095e-05, + "loss": 0.1087, + "step": 9528 + }, + { + "epoch": 26.61731843575419, + "grad_norm": 0.6894010901451111, + "learning_rate": 2.9786219739292367e-05, + "loss": 0.1544, + "step": 9529 + }, + { + "epoch": 26.620111731843576, + "grad_norm": 3.921915292739868, + "learning_rate": 2.9785847299813784e-05, + "loss": 0.1502, + "step": 9530 + }, + { + "epoch": 26.622905027932962, + "grad_norm": 0.8665048480033875, + "learning_rate": 2.9785474860335197e-05, + "loss": 0.1332, + "step": 9531 + }, + { + "epoch": 26.625698324022345, + "grad_norm": 1.0457220077514648, + "learning_rate": 2.978510242085661e-05, + "loss": 0.1356, + "step": 9532 + }, + { + "epoch": 26.62849162011173, + "grad_norm": 3.749147653579712, + "learning_rate": 2.9784729981378026e-05, + "loss": 0.1937, + "step": 9533 + }, + { + "epoch": 26.631284916201118, + "grad_norm": 0.8716644048690796, + "learning_rate": 2.9784357541899443e-05, + "loss": 0.1687, + "step": 9534 + }, + { + "epoch": 26.634078212290504, + "grad_norm": 1.3061587810516357, + "learning_rate": 2.9783985102420856e-05, + "loss": 0.1645, + "step": 9535 + }, + { + "epoch": 26.636871508379887, + "grad_norm": 0.5376194715499878, + "learning_rate": 2.9783612662942273e-05, + "loss": 0.1514, + "step": 9536 + }, + { + "epoch": 26.639664804469273, + "grad_norm": 0.9851638078689575, + "learning_rate": 2.978324022346369e-05, + "loss": 0.1466, + "step": 9537 + }, + { + "epoch": 26.64245810055866, + "grad_norm": 0.9837214350700378, + "learning_rate": 2.9782867783985102e-05, + "loss": 0.1582, + "step": 9538 + }, + { + "epoch": 26.645251396648046, + "grad_norm": 0.553628146648407, + "learning_rate": 2.978249534450652e-05, + "loss": 0.1375, + "step": 9539 + }, + { + "epoch": 26.64804469273743, + "grad_norm": 0.6782757639884949, + "learning_rate": 2.9782122905027932e-05, + "loss": 0.193, + "step": 9540 + }, + { + "epoch": 26.650837988826815, + "grad_norm": 0.9298056364059448, + "learning_rate": 2.978175046554935e-05, + "loss": 0.1419, + "step": 9541 + }, + { + "epoch": 26.6536312849162, + "grad_norm": 1.302604079246521, + "learning_rate": 2.9781378026070765e-05, + "loss": 0.1539, + "step": 9542 + }, + { + "epoch": 26.656424581005588, + "grad_norm": 0.5764168500900269, + "learning_rate": 2.978100558659218e-05, + "loss": 0.172, + "step": 9543 + }, + { + "epoch": 26.65921787709497, + "grad_norm": 0.8224489688873291, + "learning_rate": 2.9780633147113595e-05, + "loss": 0.184, + "step": 9544 + }, + { + "epoch": 26.662011173184357, + "grad_norm": 0.6118678450584412, + "learning_rate": 2.9780260707635008e-05, + "loss": 0.1518, + "step": 9545 + }, + { + "epoch": 26.664804469273744, + "grad_norm": 0.5809131264686584, + "learning_rate": 2.9779888268156424e-05, + "loss": 0.1313, + "step": 9546 + }, + { + "epoch": 26.66759776536313, + "grad_norm": 0.6408783197402954, + "learning_rate": 2.977951582867784e-05, + "loss": 0.149, + "step": 9547 + }, + { + "epoch": 26.670391061452513, + "grad_norm": 0.6579083800315857, + "learning_rate": 2.9779143389199258e-05, + "loss": 0.1441, + "step": 9548 + }, + { + "epoch": 26.6731843575419, + "grad_norm": 0.7900832295417786, + "learning_rate": 2.977877094972067e-05, + "loss": 0.164, + "step": 9549 + }, + { + "epoch": 26.675977653631286, + "grad_norm": 0.5329010486602783, + "learning_rate": 2.9778398510242087e-05, + "loss": 0.1454, + "step": 9550 + }, + { + "epoch": 26.678770949720672, + "grad_norm": 0.6584247946739197, + "learning_rate": 2.9778026070763504e-05, + "loss": 0.1419, + "step": 9551 + }, + { + "epoch": 26.681564245810055, + "grad_norm": 0.8427168130874634, + "learning_rate": 2.9777653631284917e-05, + "loss": 0.1212, + "step": 9552 + }, + { + "epoch": 26.68435754189944, + "grad_norm": 1.0184451341629028, + "learning_rate": 2.977728119180633e-05, + "loss": 0.1507, + "step": 9553 + }, + { + "epoch": 26.687150837988828, + "grad_norm": 1.0332744121551514, + "learning_rate": 2.9776908752327747e-05, + "loss": 0.1619, + "step": 9554 + }, + { + "epoch": 26.689944134078214, + "grad_norm": 0.9853688478469849, + "learning_rate": 2.9776536312849163e-05, + "loss": 0.1585, + "step": 9555 + }, + { + "epoch": 26.692737430167597, + "grad_norm": 0.9591356515884399, + "learning_rate": 2.977616387337058e-05, + "loss": 0.1262, + "step": 9556 + }, + { + "epoch": 26.695530726256983, + "grad_norm": 1.9863977432250977, + "learning_rate": 2.9775791433891996e-05, + "loss": 0.1842, + "step": 9557 + }, + { + "epoch": 26.69832402234637, + "grad_norm": 3.9271912574768066, + "learning_rate": 2.9775418994413406e-05, + "loss": 0.208, + "step": 9558 + }, + { + "epoch": 26.701117318435756, + "grad_norm": 0.7929865121841431, + "learning_rate": 2.9775046554934822e-05, + "loss": 0.196, + "step": 9559 + }, + { + "epoch": 26.70391061452514, + "grad_norm": 0.8710619807243347, + "learning_rate": 2.977467411545624e-05, + "loss": 0.162, + "step": 9560 + }, + { + "epoch": 26.706703910614525, + "grad_norm": 1.1520907878875732, + "learning_rate": 2.9774301675977656e-05, + "loss": 0.1589, + "step": 9561 + }, + { + "epoch": 26.70949720670391, + "grad_norm": 1.1823104619979858, + "learning_rate": 2.977392923649907e-05, + "loss": 0.1795, + "step": 9562 + }, + { + "epoch": 26.712290502793294, + "grad_norm": 2.4228153228759766, + "learning_rate": 2.9773556797020485e-05, + "loss": 0.1845, + "step": 9563 + }, + { + "epoch": 26.71508379888268, + "grad_norm": 1.438799262046814, + "learning_rate": 2.9773184357541902e-05, + "loss": 0.1985, + "step": 9564 + }, + { + "epoch": 26.717877094972067, + "grad_norm": 0.7204163670539856, + "learning_rate": 2.9772811918063315e-05, + "loss": 0.1453, + "step": 9565 + }, + { + "epoch": 26.720670391061454, + "grad_norm": 1.3482325077056885, + "learning_rate": 2.977243947858473e-05, + "loss": 0.1596, + "step": 9566 + }, + { + "epoch": 26.723463687150836, + "grad_norm": 0.8454867005348206, + "learning_rate": 2.9772067039106145e-05, + "loss": 0.1394, + "step": 9567 + }, + { + "epoch": 26.726256983240223, + "grad_norm": 0.8021855354309082, + "learning_rate": 2.977169459962756e-05, + "loss": 0.142, + "step": 9568 + }, + { + "epoch": 26.72905027932961, + "grad_norm": 0.6792356371879578, + "learning_rate": 2.9771322160148978e-05, + "loss": 0.1519, + "step": 9569 + }, + { + "epoch": 26.731843575418996, + "grad_norm": 0.8543950915336609, + "learning_rate": 2.9770949720670394e-05, + "loss": 0.1737, + "step": 9570 + }, + { + "epoch": 26.73463687150838, + "grad_norm": 0.5087469220161438, + "learning_rate": 2.9770577281191807e-05, + "loss": 0.1369, + "step": 9571 + }, + { + "epoch": 26.737430167597765, + "grad_norm": 0.9969617128372192, + "learning_rate": 2.977020484171322e-05, + "loss": 0.1433, + "step": 9572 + }, + { + "epoch": 26.74022346368715, + "grad_norm": 0.790826141834259, + "learning_rate": 2.9769832402234637e-05, + "loss": 0.1366, + "step": 9573 + }, + { + "epoch": 26.743016759776538, + "grad_norm": 0.7096070051193237, + "learning_rate": 2.9769459962756054e-05, + "loss": 0.1546, + "step": 9574 + }, + { + "epoch": 26.74581005586592, + "grad_norm": 0.7104173302650452, + "learning_rate": 2.976908752327747e-05, + "loss": 0.167, + "step": 9575 + }, + { + "epoch": 26.748603351955307, + "grad_norm": 0.7902045249938965, + "learning_rate": 2.9768715083798883e-05, + "loss": 0.1254, + "step": 9576 + }, + { + "epoch": 26.751396648044693, + "grad_norm": 0.6848521828651428, + "learning_rate": 2.97683426443203e-05, + "loss": 0.1585, + "step": 9577 + }, + { + "epoch": 26.75418994413408, + "grad_norm": 0.927600085735321, + "learning_rate": 2.9767970204841713e-05, + "loss": 0.1478, + "step": 9578 + }, + { + "epoch": 26.756983240223462, + "grad_norm": 0.9982954263687134, + "learning_rate": 2.976759776536313e-05, + "loss": 0.1199, + "step": 9579 + }, + { + "epoch": 26.75977653631285, + "grad_norm": 0.8809913396835327, + "learning_rate": 2.9767225325884543e-05, + "loss": 0.1735, + "step": 9580 + }, + { + "epoch": 26.762569832402235, + "grad_norm": 1.5299177169799805, + "learning_rate": 2.976685288640596e-05, + "loss": 0.1444, + "step": 9581 + }, + { + "epoch": 26.76536312849162, + "grad_norm": 1.5736596584320068, + "learning_rate": 2.9766480446927376e-05, + "loss": 0.1463, + "step": 9582 + }, + { + "epoch": 26.768156424581004, + "grad_norm": 2.2555880546569824, + "learning_rate": 2.9766108007448792e-05, + "loss": 0.1668, + "step": 9583 + }, + { + "epoch": 26.77094972067039, + "grad_norm": 0.5647091865539551, + "learning_rate": 2.9765735567970205e-05, + "loss": 0.1612, + "step": 9584 + }, + { + "epoch": 26.773743016759777, + "grad_norm": 0.9517591595649719, + "learning_rate": 2.976536312849162e-05, + "loss": 0.1501, + "step": 9585 + }, + { + "epoch": 26.776536312849164, + "grad_norm": 0.6099241375923157, + "learning_rate": 2.9764990689013035e-05, + "loss": 0.1466, + "step": 9586 + }, + { + "epoch": 26.779329608938546, + "grad_norm": 0.7670639157295227, + "learning_rate": 2.976461824953445e-05, + "loss": 0.1689, + "step": 9587 + }, + { + "epoch": 26.782122905027933, + "grad_norm": 0.5416477918624878, + "learning_rate": 2.9764245810055868e-05, + "loss": 0.1618, + "step": 9588 + }, + { + "epoch": 26.78491620111732, + "grad_norm": 1.0611006021499634, + "learning_rate": 2.976387337057728e-05, + "loss": 0.1408, + "step": 9589 + }, + { + "epoch": 26.787709497206706, + "grad_norm": 0.6528387665748596, + "learning_rate": 2.9763500931098698e-05, + "loss": 0.1541, + "step": 9590 + }, + { + "epoch": 26.79050279329609, + "grad_norm": 0.6154284477233887, + "learning_rate": 2.9763128491620114e-05, + "loss": 0.1435, + "step": 9591 + }, + { + "epoch": 26.793296089385475, + "grad_norm": 0.9024016261100769, + "learning_rate": 2.9762756052141527e-05, + "loss": 0.1558, + "step": 9592 + }, + { + "epoch": 26.79608938547486, + "grad_norm": 1.0119229555130005, + "learning_rate": 2.976238361266294e-05, + "loss": 0.143, + "step": 9593 + }, + { + "epoch": 26.798882681564244, + "grad_norm": 1.006234884262085, + "learning_rate": 2.9762011173184357e-05, + "loss": 0.1447, + "step": 9594 + }, + { + "epoch": 26.80167597765363, + "grad_norm": 0.770735502243042, + "learning_rate": 2.9761638733705774e-05, + "loss": 0.1338, + "step": 9595 + }, + { + "epoch": 26.804469273743017, + "grad_norm": 4.272714614868164, + "learning_rate": 2.976126629422719e-05, + "loss": 0.1315, + "step": 9596 + }, + { + "epoch": 26.807262569832403, + "grad_norm": 0.9351842999458313, + "learning_rate": 2.9760893854748607e-05, + "loss": 0.1697, + "step": 9597 + }, + { + "epoch": 26.810055865921786, + "grad_norm": 0.9285427927970886, + "learning_rate": 2.9760521415270017e-05, + "loss": 0.1382, + "step": 9598 + }, + { + "epoch": 26.812849162011172, + "grad_norm": 1.3186737298965454, + "learning_rate": 2.9760148975791433e-05, + "loss": 0.132, + "step": 9599 + }, + { + "epoch": 26.81564245810056, + "grad_norm": 0.7670881152153015, + "learning_rate": 2.975977653631285e-05, + "loss": 0.1472, + "step": 9600 + }, + { + "epoch": 26.818435754189945, + "grad_norm": 0.9415439367294312, + "learning_rate": 2.9759404096834266e-05, + "loss": 0.1362, + "step": 9601 + }, + { + "epoch": 26.821229050279328, + "grad_norm": 0.7100020051002502, + "learning_rate": 2.975903165735568e-05, + "loss": 0.1865, + "step": 9602 + }, + { + "epoch": 26.824022346368714, + "grad_norm": 1.0365676879882812, + "learning_rate": 2.9758659217877096e-05, + "loss": 0.1349, + "step": 9603 + }, + { + "epoch": 26.8268156424581, + "grad_norm": 2.03767991065979, + "learning_rate": 2.9758286778398512e-05, + "loss": 0.1721, + "step": 9604 + }, + { + "epoch": 26.829608938547487, + "grad_norm": 0.9582839608192444, + "learning_rate": 2.9757914338919925e-05, + "loss": 0.1774, + "step": 9605 + }, + { + "epoch": 26.83240223463687, + "grad_norm": 1.2460533380508423, + "learning_rate": 2.9757541899441342e-05, + "loss": 0.1399, + "step": 9606 + }, + { + "epoch": 26.835195530726256, + "grad_norm": 1.3431968688964844, + "learning_rate": 2.9757169459962755e-05, + "loss": 0.1551, + "step": 9607 + }, + { + "epoch": 26.837988826815643, + "grad_norm": 1.1531323194503784, + "learning_rate": 2.9756797020484172e-05, + "loss": 0.2371, + "step": 9608 + }, + { + "epoch": 26.84078212290503, + "grad_norm": 0.599652111530304, + "learning_rate": 2.9756424581005588e-05, + "loss": 0.1692, + "step": 9609 + }, + { + "epoch": 26.843575418994412, + "grad_norm": 0.583005964756012, + "learning_rate": 2.9756052141527005e-05, + "loss": 0.162, + "step": 9610 + }, + { + "epoch": 26.8463687150838, + "grad_norm": 0.6162207126617432, + "learning_rate": 2.9755679702048418e-05, + "loss": 0.1425, + "step": 9611 + }, + { + "epoch": 26.849162011173185, + "grad_norm": 0.4902907609939575, + "learning_rate": 2.975530726256983e-05, + "loss": 0.1742, + "step": 9612 + }, + { + "epoch": 26.85195530726257, + "grad_norm": 0.5708051919937134, + "learning_rate": 2.9754934823091248e-05, + "loss": 0.1426, + "step": 9613 + }, + { + "epoch": 26.854748603351954, + "grad_norm": 0.6496902108192444, + "learning_rate": 2.9754562383612664e-05, + "loss": 0.1848, + "step": 9614 + }, + { + "epoch": 26.85754189944134, + "grad_norm": 0.5860134959220886, + "learning_rate": 2.975418994413408e-05, + "loss": 0.1677, + "step": 9615 + }, + { + "epoch": 26.860335195530727, + "grad_norm": 1.5943944454193115, + "learning_rate": 2.9753817504655494e-05, + "loss": 0.1449, + "step": 9616 + }, + { + "epoch": 26.863128491620113, + "grad_norm": 0.8872518539428711, + "learning_rate": 2.975344506517691e-05, + "loss": 0.1448, + "step": 9617 + }, + { + "epoch": 26.865921787709496, + "grad_norm": 2.673450231552124, + "learning_rate": 2.9753072625698324e-05, + "loss": 0.1657, + "step": 9618 + }, + { + "epoch": 26.868715083798882, + "grad_norm": 0.6665160059928894, + "learning_rate": 2.975270018621974e-05, + "loss": 0.1737, + "step": 9619 + }, + { + "epoch": 26.87150837988827, + "grad_norm": 0.6020649075508118, + "learning_rate": 2.9752327746741153e-05, + "loss": 0.1428, + "step": 9620 + }, + { + "epoch": 26.874301675977655, + "grad_norm": 1.2780423164367676, + "learning_rate": 2.975195530726257e-05, + "loss": 0.1431, + "step": 9621 + }, + { + "epoch": 26.877094972067038, + "grad_norm": 1.6568130254745483, + "learning_rate": 2.9751582867783986e-05, + "loss": 0.1659, + "step": 9622 + }, + { + "epoch": 26.879888268156424, + "grad_norm": 0.6817127466201782, + "learning_rate": 2.9751210428305403e-05, + "loss": 0.1386, + "step": 9623 + }, + { + "epoch": 26.88268156424581, + "grad_norm": 0.9145358800888062, + "learning_rate": 2.975083798882682e-05, + "loss": 0.1245, + "step": 9624 + }, + { + "epoch": 26.885474860335197, + "grad_norm": 0.9433081150054932, + "learning_rate": 2.975046554934823e-05, + "loss": 0.1392, + "step": 9625 + }, + { + "epoch": 26.88826815642458, + "grad_norm": 0.846739649772644, + "learning_rate": 2.9750093109869646e-05, + "loss": 0.1171, + "step": 9626 + }, + { + "epoch": 26.891061452513966, + "grad_norm": 1.099108338356018, + "learning_rate": 2.9749720670391062e-05, + "loss": 0.1665, + "step": 9627 + }, + { + "epoch": 26.893854748603353, + "grad_norm": 1.2712678909301758, + "learning_rate": 2.974934823091248e-05, + "loss": 0.1422, + "step": 9628 + }, + { + "epoch": 26.89664804469274, + "grad_norm": 0.6048637628555298, + "learning_rate": 2.9748975791433892e-05, + "loss": 0.1324, + "step": 9629 + }, + { + "epoch": 26.899441340782122, + "grad_norm": 1.115857720375061, + "learning_rate": 2.974860335195531e-05, + "loss": 0.1443, + "step": 9630 + }, + { + "epoch": 26.90223463687151, + "grad_norm": 1.5740293264389038, + "learning_rate": 2.9748230912476725e-05, + "loss": 0.1698, + "step": 9631 + }, + { + "epoch": 26.905027932960895, + "grad_norm": 2.5726218223571777, + "learning_rate": 2.9747858472998138e-05, + "loss": 0.144, + "step": 9632 + }, + { + "epoch": 26.907821229050278, + "grad_norm": 1.9781173467636108, + "learning_rate": 2.9747486033519555e-05, + "loss": 0.2703, + "step": 9633 + }, + { + "epoch": 26.910614525139664, + "grad_norm": 0.670721709728241, + "learning_rate": 2.9747113594040968e-05, + "loss": 0.1487, + "step": 9634 + }, + { + "epoch": 26.91340782122905, + "grad_norm": 0.6197160482406616, + "learning_rate": 2.9746741154562384e-05, + "loss": 0.1928, + "step": 9635 + }, + { + "epoch": 26.916201117318437, + "grad_norm": 0.7581655383110046, + "learning_rate": 2.97463687150838e-05, + "loss": 0.1614, + "step": 9636 + }, + { + "epoch": 26.91899441340782, + "grad_norm": 0.7843906879425049, + "learning_rate": 2.9745996275605217e-05, + "loss": 0.1878, + "step": 9637 + }, + { + "epoch": 26.921787709497206, + "grad_norm": 0.5005249381065369, + "learning_rate": 2.9745623836126627e-05, + "loss": 0.1689, + "step": 9638 + }, + { + "epoch": 26.924581005586592, + "grad_norm": 0.7426415085792542, + "learning_rate": 2.9745251396648044e-05, + "loss": 0.1769, + "step": 9639 + }, + { + "epoch": 26.92737430167598, + "grad_norm": 0.8407852649688721, + "learning_rate": 2.974487895716946e-05, + "loss": 0.1632, + "step": 9640 + }, + { + "epoch": 26.93016759776536, + "grad_norm": 1.1377462148666382, + "learning_rate": 2.9744506517690877e-05, + "loss": 0.1848, + "step": 9641 + }, + { + "epoch": 26.932960893854748, + "grad_norm": 0.44610950350761414, + "learning_rate": 2.9744134078212293e-05, + "loss": 0.1559, + "step": 9642 + }, + { + "epoch": 26.935754189944134, + "grad_norm": 0.5610420107841492, + "learning_rate": 2.9743761638733706e-05, + "loss": 0.1386, + "step": 9643 + }, + { + "epoch": 26.93854748603352, + "grad_norm": 0.9036240577697754, + "learning_rate": 2.9743389199255123e-05, + "loss": 0.1813, + "step": 9644 + }, + { + "epoch": 26.941340782122904, + "grad_norm": 0.5979248881340027, + "learning_rate": 2.9743016759776536e-05, + "loss": 0.1664, + "step": 9645 + }, + { + "epoch": 26.94413407821229, + "grad_norm": 0.7333125472068787, + "learning_rate": 2.9742644320297953e-05, + "loss": 0.1447, + "step": 9646 + }, + { + "epoch": 26.946927374301676, + "grad_norm": 0.6925642490386963, + "learning_rate": 2.9742271880819366e-05, + "loss": 0.1491, + "step": 9647 + }, + { + "epoch": 26.949720670391063, + "grad_norm": 0.6725579500198364, + "learning_rate": 2.9741899441340782e-05, + "loss": 0.1431, + "step": 9648 + }, + { + "epoch": 26.952513966480446, + "grad_norm": 2.59598708152771, + "learning_rate": 2.97415270018622e-05, + "loss": 0.1757, + "step": 9649 + }, + { + "epoch": 26.955307262569832, + "grad_norm": 1.0436394214630127, + "learning_rate": 2.9741154562383615e-05, + "loss": 0.1436, + "step": 9650 + }, + { + "epoch": 26.95810055865922, + "grad_norm": 1.0654914379119873, + "learning_rate": 2.9740782122905032e-05, + "loss": 0.1732, + "step": 9651 + }, + { + "epoch": 26.960893854748605, + "grad_norm": 0.59696364402771, + "learning_rate": 2.974040968342644e-05, + "loss": 0.1337, + "step": 9652 + }, + { + "epoch": 26.963687150837988, + "grad_norm": 0.6496047973632812, + "learning_rate": 2.9740037243947858e-05, + "loss": 0.1711, + "step": 9653 + }, + { + "epoch": 26.966480446927374, + "grad_norm": 1.7387843132019043, + "learning_rate": 2.9739664804469275e-05, + "loss": 0.1395, + "step": 9654 + }, + { + "epoch": 26.96927374301676, + "grad_norm": 0.7157253623008728, + "learning_rate": 2.973929236499069e-05, + "loss": 0.1586, + "step": 9655 + }, + { + "epoch": 26.972067039106147, + "grad_norm": 1.555635690689087, + "learning_rate": 2.9738919925512104e-05, + "loss": 0.1675, + "step": 9656 + }, + { + "epoch": 26.97486033519553, + "grad_norm": 1.543807029724121, + "learning_rate": 2.973854748603352e-05, + "loss": 0.1541, + "step": 9657 + }, + { + "epoch": 26.977653631284916, + "grad_norm": 2.032933473587036, + "learning_rate": 2.9738175046554934e-05, + "loss": 0.1953, + "step": 9658 + }, + { + "epoch": 26.980446927374302, + "grad_norm": 0.5715081691741943, + "learning_rate": 2.973780260707635e-05, + "loss": 0.1787, + "step": 9659 + }, + { + "epoch": 26.98324022346369, + "grad_norm": 0.5808831453323364, + "learning_rate": 2.9737430167597767e-05, + "loss": 0.1697, + "step": 9660 + }, + { + "epoch": 26.98603351955307, + "grad_norm": 0.9140205979347229, + "learning_rate": 2.973705772811918e-05, + "loss": 0.1458, + "step": 9661 + }, + { + "epoch": 26.988826815642458, + "grad_norm": 0.42216572165489197, + "learning_rate": 2.9736685288640597e-05, + "loss": 0.1314, + "step": 9662 + }, + { + "epoch": 26.991620111731844, + "grad_norm": 1.357410192489624, + "learning_rate": 2.9736312849162013e-05, + "loss": 0.1171, + "step": 9663 + }, + { + "epoch": 26.994413407821227, + "grad_norm": 1.313704490661621, + "learning_rate": 2.973594040968343e-05, + "loss": 0.1313, + "step": 9664 + }, + { + "epoch": 26.997206703910614, + "grad_norm": 1.1594548225402832, + "learning_rate": 2.973556797020484e-05, + "loss": 0.131, + "step": 9665 + }, + { + "epoch": 27.0, + "grad_norm": 2.256600856781006, + "learning_rate": 2.9735195530726256e-05, + "loss": 0.1632, + "step": 9666 + }, + { + "epoch": 27.002793296089386, + "grad_norm": 0.5457233786582947, + "learning_rate": 2.9734823091247673e-05, + "loss": 0.1578, + "step": 9667 + }, + { + "epoch": 27.00558659217877, + "grad_norm": 0.6228638887405396, + "learning_rate": 2.973445065176909e-05, + "loss": 0.1591, + "step": 9668 + }, + { + "epoch": 27.008379888268156, + "grad_norm": 0.5940813422203064, + "learning_rate": 2.9734078212290506e-05, + "loss": 0.1454, + "step": 9669 + }, + { + "epoch": 27.011173184357542, + "grad_norm": 1.0394841432571411, + "learning_rate": 2.973370577281192e-05, + "loss": 0.1628, + "step": 9670 + }, + { + "epoch": 27.01396648044693, + "grad_norm": 0.7822326421737671, + "learning_rate": 2.9733333333333336e-05, + "loss": 0.2099, + "step": 9671 + }, + { + "epoch": 27.01675977653631, + "grad_norm": 0.5524436235427856, + "learning_rate": 2.973296089385475e-05, + "loss": 0.1382, + "step": 9672 + }, + { + "epoch": 27.019553072625698, + "grad_norm": 3.894541025161743, + "learning_rate": 2.9732588454376165e-05, + "loss": 0.1627, + "step": 9673 + }, + { + "epoch": 27.022346368715084, + "grad_norm": 0.8705493211746216, + "learning_rate": 2.973221601489758e-05, + "loss": 0.1506, + "step": 9674 + }, + { + "epoch": 27.02513966480447, + "grad_norm": 0.6809353828430176, + "learning_rate": 2.9731843575418995e-05, + "loss": 0.1314, + "step": 9675 + }, + { + "epoch": 27.027932960893853, + "grad_norm": 0.524966299533844, + "learning_rate": 2.973147113594041e-05, + "loss": 0.1146, + "step": 9676 + }, + { + "epoch": 27.03072625698324, + "grad_norm": 0.5623594522476196, + "learning_rate": 2.9731098696461828e-05, + "loss": 0.1268, + "step": 9677 + }, + { + "epoch": 27.033519553072626, + "grad_norm": 0.7997907996177673, + "learning_rate": 2.973072625698324e-05, + "loss": 0.1537, + "step": 9678 + }, + { + "epoch": 27.036312849162012, + "grad_norm": 0.4810328781604767, + "learning_rate": 2.9730353817504654e-05, + "loss": 0.151, + "step": 9679 + }, + { + "epoch": 27.039106145251395, + "grad_norm": 0.8831645250320435, + "learning_rate": 2.972998137802607e-05, + "loss": 0.1129, + "step": 9680 + }, + { + "epoch": 27.04189944134078, + "grad_norm": 0.928955078125, + "learning_rate": 2.9729608938547487e-05, + "loss": 0.1401, + "step": 9681 + }, + { + "epoch": 27.044692737430168, + "grad_norm": 1.7560443878173828, + "learning_rate": 2.9729236499068904e-05, + "loss": 0.1422, + "step": 9682 + }, + { + "epoch": 27.047486033519554, + "grad_norm": 0.6022185683250427, + "learning_rate": 2.9728864059590317e-05, + "loss": 0.1414, + "step": 9683 + }, + { + "epoch": 27.050279329608937, + "grad_norm": 0.4997418224811554, + "learning_rate": 2.9728491620111734e-05, + "loss": 0.1127, + "step": 9684 + }, + { + "epoch": 27.053072625698324, + "grad_norm": 0.9356067776679993, + "learning_rate": 2.9728119180633147e-05, + "loss": 0.119, + "step": 9685 + }, + { + "epoch": 27.05586592178771, + "grad_norm": 0.8934686183929443, + "learning_rate": 2.9727746741154563e-05, + "loss": 0.1257, + "step": 9686 + }, + { + "epoch": 27.058659217877096, + "grad_norm": 0.8169559836387634, + "learning_rate": 2.9727374301675976e-05, + "loss": 0.1443, + "step": 9687 + }, + { + "epoch": 27.06145251396648, + "grad_norm": 0.572685956954956, + "learning_rate": 2.9727001862197393e-05, + "loss": 0.1241, + "step": 9688 + }, + { + "epoch": 27.064245810055866, + "grad_norm": 1.8428634405136108, + "learning_rate": 2.972662942271881e-05, + "loss": 0.1464, + "step": 9689 + }, + { + "epoch": 27.067039106145252, + "grad_norm": 0.9990090131759644, + "learning_rate": 2.9726256983240226e-05, + "loss": 0.1393, + "step": 9690 + }, + { + "epoch": 27.06983240223464, + "grad_norm": 1.5566617250442505, + "learning_rate": 2.9725884543761643e-05, + "loss": 0.1655, + "step": 9691 + }, + { + "epoch": 27.07262569832402, + "grad_norm": 1.181634783744812, + "learning_rate": 2.9725512104283052e-05, + "loss": 0.1639, + "step": 9692 + }, + { + "epoch": 27.075418994413408, + "grad_norm": 0.6957917809486389, + "learning_rate": 2.972513966480447e-05, + "loss": 0.1689, + "step": 9693 + }, + { + "epoch": 27.078212290502794, + "grad_norm": 0.49984198808670044, + "learning_rate": 2.9724767225325885e-05, + "loss": 0.1621, + "step": 9694 + }, + { + "epoch": 27.08100558659218, + "grad_norm": 0.8004072904586792, + "learning_rate": 2.9724394785847302e-05, + "loss": 0.1603, + "step": 9695 + }, + { + "epoch": 27.083798882681563, + "grad_norm": 0.38992589712142944, + "learning_rate": 2.9724022346368715e-05, + "loss": 0.1413, + "step": 9696 + }, + { + "epoch": 27.08659217877095, + "grad_norm": 0.3917827308177948, + "learning_rate": 2.972364990689013e-05, + "loss": 0.1458, + "step": 9697 + }, + { + "epoch": 27.089385474860336, + "grad_norm": 0.7969008088111877, + "learning_rate": 2.9723277467411545e-05, + "loss": 0.1295, + "step": 9698 + }, + { + "epoch": 27.092178770949722, + "grad_norm": 0.6650158166885376, + "learning_rate": 2.972290502793296e-05, + "loss": 0.1571, + "step": 9699 + }, + { + "epoch": 27.094972067039105, + "grad_norm": 0.4364674687385559, + "learning_rate": 2.9722532588454378e-05, + "loss": 0.148, + "step": 9700 + }, + { + "epoch": 27.09776536312849, + "grad_norm": 0.6544800996780396, + "learning_rate": 2.972216014897579e-05, + "loss": 0.1426, + "step": 9701 + }, + { + "epoch": 27.100558659217878, + "grad_norm": 2.3464009761810303, + "learning_rate": 2.9721787709497207e-05, + "loss": 0.1545, + "step": 9702 + }, + { + "epoch": 27.10335195530726, + "grad_norm": 2.7267463207244873, + "learning_rate": 2.9721415270018624e-05, + "loss": 0.173, + "step": 9703 + }, + { + "epoch": 27.106145251396647, + "grad_norm": 0.4943252205848694, + "learning_rate": 2.972104283054004e-05, + "loss": 0.1532, + "step": 9704 + }, + { + "epoch": 27.108938547486034, + "grad_norm": 0.6837851405143738, + "learning_rate": 2.972067039106145e-05, + "loss": 0.1435, + "step": 9705 + }, + { + "epoch": 27.11173184357542, + "grad_norm": 0.4842035472393036, + "learning_rate": 2.9720297951582867e-05, + "loss": 0.122, + "step": 9706 + }, + { + "epoch": 27.114525139664803, + "grad_norm": 1.2329391241073608, + "learning_rate": 2.9719925512104283e-05, + "loss": 0.1414, + "step": 9707 + }, + { + "epoch": 27.11731843575419, + "grad_norm": 0.9303504824638367, + "learning_rate": 2.97195530726257e-05, + "loss": 0.1114, + "step": 9708 + }, + { + "epoch": 27.120111731843576, + "grad_norm": 2.2419488430023193, + "learning_rate": 2.9719180633147116e-05, + "loss": 0.1485, + "step": 9709 + }, + { + "epoch": 27.122905027932962, + "grad_norm": 4.812856674194336, + "learning_rate": 2.971880819366853e-05, + "loss": 0.1272, + "step": 9710 + }, + { + "epoch": 27.125698324022345, + "grad_norm": 0.8677388429641724, + "learning_rate": 2.9718435754189946e-05, + "loss": 0.1303, + "step": 9711 + }, + { + "epoch": 27.12849162011173, + "grad_norm": 1.1128137111663818, + "learning_rate": 2.971806331471136e-05, + "loss": 0.1312, + "step": 9712 + }, + { + "epoch": 27.131284916201118, + "grad_norm": 1.7881814241409302, + "learning_rate": 2.9717690875232776e-05, + "loss": 0.1061, + "step": 9713 + }, + { + "epoch": 27.134078212290504, + "grad_norm": 1.039788842201233, + "learning_rate": 2.971731843575419e-05, + "loss": 0.1416, + "step": 9714 + }, + { + "epoch": 27.136871508379887, + "grad_norm": 0.8863144516944885, + "learning_rate": 2.9716945996275605e-05, + "loss": 0.1591, + "step": 9715 + }, + { + "epoch": 27.139664804469273, + "grad_norm": 3.36336088180542, + "learning_rate": 2.9716573556797022e-05, + "loss": 0.1816, + "step": 9716 + }, + { + "epoch": 27.14245810055866, + "grad_norm": 0.6269659399986267, + "learning_rate": 2.971620111731844e-05, + "loss": 0.1597, + "step": 9717 + }, + { + "epoch": 27.145251396648046, + "grad_norm": 0.55548095703125, + "learning_rate": 2.9715828677839852e-05, + "loss": 0.1629, + "step": 9718 + }, + { + "epoch": 27.14804469273743, + "grad_norm": 0.4928908050060272, + "learning_rate": 2.9715456238361265e-05, + "loss": 0.1409, + "step": 9719 + }, + { + "epoch": 27.150837988826815, + "grad_norm": 0.4318718910217285, + "learning_rate": 2.971508379888268e-05, + "loss": 0.1527, + "step": 9720 + }, + { + "epoch": 27.1536312849162, + "grad_norm": 1.652867317199707, + "learning_rate": 2.9714711359404098e-05, + "loss": 0.158, + "step": 9721 + }, + { + "epoch": 27.156424581005588, + "grad_norm": 0.5640828609466553, + "learning_rate": 2.9714338919925514e-05, + "loss": 0.1713, + "step": 9722 + }, + { + "epoch": 27.15921787709497, + "grad_norm": 0.6203129291534424, + "learning_rate": 2.9713966480446928e-05, + "loss": 0.1609, + "step": 9723 + }, + { + "epoch": 27.162011173184357, + "grad_norm": 0.7132992744445801, + "learning_rate": 2.9713594040968344e-05, + "loss": 0.1363, + "step": 9724 + }, + { + "epoch": 27.164804469273744, + "grad_norm": 0.869511067867279, + "learning_rate": 2.9713221601489757e-05, + "loss": 0.1368, + "step": 9725 + }, + { + "epoch": 27.16759776536313, + "grad_norm": 0.4737740755081177, + "learning_rate": 2.9712849162011174e-05, + "loss": 0.128, + "step": 9726 + }, + { + "epoch": 27.170391061452513, + "grad_norm": 0.6076182723045349, + "learning_rate": 2.971247672253259e-05, + "loss": 0.1441, + "step": 9727 + }, + { + "epoch": 27.1731843575419, + "grad_norm": 0.5042256712913513, + "learning_rate": 2.9712104283054003e-05, + "loss": 0.1279, + "step": 9728 + }, + { + "epoch": 27.175977653631286, + "grad_norm": 0.7080655694007874, + "learning_rate": 2.971173184357542e-05, + "loss": 0.1403, + "step": 9729 + }, + { + "epoch": 27.178770949720672, + "grad_norm": 0.6493347883224487, + "learning_rate": 2.9711359404096837e-05, + "loss": 0.1552, + "step": 9730 + }, + { + "epoch": 27.181564245810055, + "grad_norm": 0.4605763852596283, + "learning_rate": 2.9710986964618253e-05, + "loss": 0.1149, + "step": 9731 + }, + { + "epoch": 27.18435754189944, + "grad_norm": 1.712839961051941, + "learning_rate": 2.9710614525139663e-05, + "loss": 0.1333, + "step": 9732 + }, + { + "epoch": 27.187150837988828, + "grad_norm": 1.1106209754943848, + "learning_rate": 2.971024208566108e-05, + "loss": 0.1507, + "step": 9733 + }, + { + "epoch": 27.189944134078214, + "grad_norm": 0.6741195321083069, + "learning_rate": 2.9709869646182496e-05, + "loss": 0.1359, + "step": 9734 + }, + { + "epoch": 27.192737430167597, + "grad_norm": 0.8075659871101379, + "learning_rate": 2.9709497206703912e-05, + "loss": 0.1282, + "step": 9735 + }, + { + "epoch": 27.195530726256983, + "grad_norm": 0.9342575669288635, + "learning_rate": 2.970912476722533e-05, + "loss": 0.134, + "step": 9736 + }, + { + "epoch": 27.19832402234637, + "grad_norm": 1.6321957111358643, + "learning_rate": 2.9708752327746742e-05, + "loss": 0.1752, + "step": 9737 + }, + { + "epoch": 27.201117318435756, + "grad_norm": 2.2559149265289307, + "learning_rate": 2.9708379888268155e-05, + "loss": 0.1409, + "step": 9738 + }, + { + "epoch": 27.20391061452514, + "grad_norm": 2.483125925064087, + "learning_rate": 2.9708007448789572e-05, + "loss": 0.1595, + "step": 9739 + }, + { + "epoch": 27.206703910614525, + "grad_norm": 1.1943247318267822, + "learning_rate": 2.970763500931099e-05, + "loss": 0.1125, + "step": 9740 + }, + { + "epoch": 27.20949720670391, + "grad_norm": 1.3249475955963135, + "learning_rate": 2.97072625698324e-05, + "loss": 0.1916, + "step": 9741 + }, + { + "epoch": 27.212290502793294, + "grad_norm": 0.5891750454902649, + "learning_rate": 2.9706890130353818e-05, + "loss": 0.1675, + "step": 9742 + }, + { + "epoch": 27.21508379888268, + "grad_norm": 0.5157848000526428, + "learning_rate": 2.9706517690875235e-05, + "loss": 0.1477, + "step": 9743 + }, + { + "epoch": 27.217877094972067, + "grad_norm": 0.9940093755722046, + "learning_rate": 2.970614525139665e-05, + "loss": 0.1386, + "step": 9744 + }, + { + "epoch": 27.220670391061454, + "grad_norm": 0.3654722273349762, + "learning_rate": 2.9705772811918064e-05, + "loss": 0.1319, + "step": 9745 + }, + { + "epoch": 27.223463687150836, + "grad_norm": 1.2151212692260742, + "learning_rate": 2.9705400372439477e-05, + "loss": 0.1627, + "step": 9746 + }, + { + "epoch": 27.226256983240223, + "grad_norm": 0.67387855052948, + "learning_rate": 2.9705027932960894e-05, + "loss": 0.1622, + "step": 9747 + }, + { + "epoch": 27.22905027932961, + "grad_norm": 0.83343905210495, + "learning_rate": 2.970465549348231e-05, + "loss": 0.1489, + "step": 9748 + }, + { + "epoch": 27.231843575418996, + "grad_norm": 0.5367786884307861, + "learning_rate": 2.9704283054003727e-05, + "loss": 0.1178, + "step": 9749 + }, + { + "epoch": 27.23463687150838, + "grad_norm": 0.5655161142349243, + "learning_rate": 2.970391061452514e-05, + "loss": 0.1336, + "step": 9750 + }, + { + "epoch": 27.237430167597765, + "grad_norm": 0.5666723251342773, + "learning_rate": 2.9703538175046557e-05, + "loss": 0.1484, + "step": 9751 + }, + { + "epoch": 27.24022346368715, + "grad_norm": 0.6150602698326111, + "learning_rate": 2.970316573556797e-05, + "loss": 0.1542, + "step": 9752 + }, + { + "epoch": 27.243016759776538, + "grad_norm": 0.46587055921554565, + "learning_rate": 2.9702793296089386e-05, + "loss": 0.1328, + "step": 9753 + }, + { + "epoch": 27.24581005586592, + "grad_norm": 1.1982585191726685, + "learning_rate": 2.9702420856610803e-05, + "loss": 0.1522, + "step": 9754 + }, + { + "epoch": 27.248603351955307, + "grad_norm": 1.0264418125152588, + "learning_rate": 2.9702048417132216e-05, + "loss": 0.1335, + "step": 9755 + }, + { + "epoch": 27.251396648044693, + "grad_norm": 0.6208540201187134, + "learning_rate": 2.9701675977653633e-05, + "loss": 0.1211, + "step": 9756 + }, + { + "epoch": 27.25418994413408, + "grad_norm": 0.9917865991592407, + "learning_rate": 2.970130353817505e-05, + "loss": 0.159, + "step": 9757 + }, + { + "epoch": 27.256983240223462, + "grad_norm": 0.5670569539070129, + "learning_rate": 2.9700931098696462e-05, + "loss": 0.1182, + "step": 9758 + }, + { + "epoch": 27.25977653631285, + "grad_norm": 0.5705154538154602, + "learning_rate": 2.9700558659217875e-05, + "loss": 0.1298, + "step": 9759 + }, + { + "epoch": 27.262569832402235, + "grad_norm": 0.790009081363678, + "learning_rate": 2.9700186219739292e-05, + "loss": 0.1542, + "step": 9760 + }, + { + "epoch": 27.26536312849162, + "grad_norm": 1.6098508834838867, + "learning_rate": 2.969981378026071e-05, + "loss": 0.1412, + "step": 9761 + }, + { + "epoch": 27.268156424581004, + "grad_norm": 0.8757769465446472, + "learning_rate": 2.9699441340782125e-05, + "loss": 0.1342, + "step": 9762 + }, + { + "epoch": 27.27094972067039, + "grad_norm": 1.083573818206787, + "learning_rate": 2.969906890130354e-05, + "loss": 0.138, + "step": 9763 + }, + { + "epoch": 27.273743016759777, + "grad_norm": 3.6451988220214844, + "learning_rate": 2.9698696461824955e-05, + "loss": 0.153, + "step": 9764 + }, + { + "epoch": 27.276536312849164, + "grad_norm": 0.7798306345939636, + "learning_rate": 2.9698324022346368e-05, + "loss": 0.159, + "step": 9765 + }, + { + "epoch": 27.279329608938546, + "grad_norm": 1.2600971460342407, + "learning_rate": 2.9697951582867784e-05, + "loss": 0.1764, + "step": 9766 + }, + { + "epoch": 27.282122905027933, + "grad_norm": 0.5858533382415771, + "learning_rate": 2.96975791433892e-05, + "loss": 0.1496, + "step": 9767 + }, + { + "epoch": 27.28491620111732, + "grad_norm": 1.6721446514129639, + "learning_rate": 2.9697206703910614e-05, + "loss": 0.1715, + "step": 9768 + }, + { + "epoch": 27.287709497206706, + "grad_norm": 0.4008203446865082, + "learning_rate": 2.969683426443203e-05, + "loss": 0.1473, + "step": 9769 + }, + { + "epoch": 27.29050279329609, + "grad_norm": 0.3902397155761719, + "learning_rate": 2.9696461824953447e-05, + "loss": 0.1385, + "step": 9770 + }, + { + "epoch": 27.293296089385475, + "grad_norm": 0.7123595476150513, + "learning_rate": 2.9696089385474864e-05, + "loss": 0.1729, + "step": 9771 + }, + { + "epoch": 27.29608938547486, + "grad_norm": 0.6042717099189758, + "learning_rate": 2.9695716945996277e-05, + "loss": 0.1643, + "step": 9772 + }, + { + "epoch": 27.298882681564244, + "grad_norm": 0.5185795426368713, + "learning_rate": 2.969534450651769e-05, + "loss": 0.1616, + "step": 9773 + }, + { + "epoch": 27.30167597765363, + "grad_norm": 0.36752116680145264, + "learning_rate": 2.9694972067039107e-05, + "loss": 0.1506, + "step": 9774 + }, + { + "epoch": 27.304469273743017, + "grad_norm": 0.5950642824172974, + "learning_rate": 2.9694599627560523e-05, + "loss": 0.1522, + "step": 9775 + }, + { + "epoch": 27.307262569832403, + "grad_norm": 0.6275365948677063, + "learning_rate": 2.969422718808194e-05, + "loss": 0.1463, + "step": 9776 + }, + { + "epoch": 27.310055865921786, + "grad_norm": 0.564717173576355, + "learning_rate": 2.9693854748603353e-05, + "loss": 0.1487, + "step": 9777 + }, + { + "epoch": 27.312849162011172, + "grad_norm": 0.6907269358634949, + "learning_rate": 2.9693482309124766e-05, + "loss": 0.1356, + "step": 9778 + }, + { + "epoch": 27.31564245810056, + "grad_norm": 0.9630100131034851, + "learning_rate": 2.9693109869646182e-05, + "loss": 0.1557, + "step": 9779 + }, + { + "epoch": 27.318435754189945, + "grad_norm": 0.7857763171195984, + "learning_rate": 2.96927374301676e-05, + "loss": 0.1245, + "step": 9780 + }, + { + "epoch": 27.321229050279328, + "grad_norm": 0.6600329875946045, + "learning_rate": 2.9692364990689012e-05, + "loss": 0.1119, + "step": 9781 + }, + { + "epoch": 27.324022346368714, + "grad_norm": 0.8472611308097839, + "learning_rate": 2.969199255121043e-05, + "loss": 0.1212, + "step": 9782 + }, + { + "epoch": 27.3268156424581, + "grad_norm": 0.7676531076431274, + "learning_rate": 2.9691620111731845e-05, + "loss": 0.1417, + "step": 9783 + }, + { + "epoch": 27.329608938547487, + "grad_norm": 0.9655988216400146, + "learning_rate": 2.9691247672253262e-05, + "loss": 0.1346, + "step": 9784 + }, + { + "epoch": 27.33240223463687, + "grad_norm": 0.8987203240394592, + "learning_rate": 2.9690875232774675e-05, + "loss": 0.122, + "step": 9785 + }, + { + "epoch": 27.335195530726256, + "grad_norm": 0.590457022190094, + "learning_rate": 2.9690502793296088e-05, + "loss": 0.1299, + "step": 9786 + }, + { + "epoch": 27.337988826815643, + "grad_norm": 0.7016376256942749, + "learning_rate": 2.9690130353817505e-05, + "loss": 0.1409, + "step": 9787 + }, + { + "epoch": 27.34078212290503, + "grad_norm": 0.6536491513252258, + "learning_rate": 2.968975791433892e-05, + "loss": 0.1256, + "step": 9788 + }, + { + "epoch": 27.343575418994412, + "grad_norm": 0.798127293586731, + "learning_rate": 2.9689385474860338e-05, + "loss": 0.1478, + "step": 9789 + }, + { + "epoch": 27.3463687150838, + "grad_norm": 1.3350133895874023, + "learning_rate": 2.968901303538175e-05, + "loss": 0.1682, + "step": 9790 + }, + { + "epoch": 27.349162011173185, + "grad_norm": 1.1777451038360596, + "learning_rate": 2.9688640595903167e-05, + "loss": 0.1498, + "step": 9791 + }, + { + "epoch": 27.35195530726257, + "grad_norm": 0.5037779808044434, + "learning_rate": 2.968826815642458e-05, + "loss": 0.1773, + "step": 9792 + }, + { + "epoch": 27.354748603351954, + "grad_norm": 0.40441226959228516, + "learning_rate": 2.9687895716945997e-05, + "loss": 0.1447, + "step": 9793 + }, + { + "epoch": 27.35754189944134, + "grad_norm": 0.778721809387207, + "learning_rate": 2.9687523277467414e-05, + "loss": 0.1798, + "step": 9794 + }, + { + "epoch": 27.360335195530727, + "grad_norm": 0.38455408811569214, + "learning_rate": 2.9687150837988827e-05, + "loss": 0.1464, + "step": 9795 + }, + { + "epoch": 27.363128491620113, + "grad_norm": 2.8897387981414795, + "learning_rate": 2.9686778398510243e-05, + "loss": 0.1448, + "step": 9796 + }, + { + "epoch": 27.365921787709496, + "grad_norm": 0.48637792468070984, + "learning_rate": 2.968640595903166e-05, + "loss": 0.1455, + "step": 9797 + }, + { + "epoch": 27.368715083798882, + "grad_norm": 0.5309453010559082, + "learning_rate": 2.9686033519553073e-05, + "loss": 0.1344, + "step": 9798 + }, + { + "epoch": 27.37150837988827, + "grad_norm": 0.8383184671401978, + "learning_rate": 2.9685661080074486e-05, + "loss": 0.1676, + "step": 9799 + }, + { + "epoch": 27.374301675977655, + "grad_norm": 1.4831129312515259, + "learning_rate": 2.9685288640595903e-05, + "loss": 0.1417, + "step": 9800 + }, + { + "epoch": 27.377094972067038, + "grad_norm": 0.5000360012054443, + "learning_rate": 2.968491620111732e-05, + "loss": 0.159, + "step": 9801 + }, + { + "epoch": 27.379888268156424, + "grad_norm": 1.0643459558486938, + "learning_rate": 2.9684543761638736e-05, + "loss": 0.142, + "step": 9802 + }, + { + "epoch": 27.38268156424581, + "grad_norm": 1.0993967056274414, + "learning_rate": 2.9684171322160152e-05, + "loss": 0.1448, + "step": 9803 + }, + { + "epoch": 27.385474860335197, + "grad_norm": 0.5307853817939758, + "learning_rate": 2.9683798882681565e-05, + "loss": 0.1362, + "step": 9804 + }, + { + "epoch": 27.38826815642458, + "grad_norm": 0.6193860173225403, + "learning_rate": 2.968342644320298e-05, + "loss": 0.1491, + "step": 9805 + }, + { + "epoch": 27.391061452513966, + "grad_norm": 0.6494642496109009, + "learning_rate": 2.9683054003724395e-05, + "loss": 0.1343, + "step": 9806 + }, + { + "epoch": 27.393854748603353, + "grad_norm": 0.9190837740898132, + "learning_rate": 2.968268156424581e-05, + "loss": 0.1515, + "step": 9807 + }, + { + "epoch": 27.39664804469274, + "grad_norm": 0.5161193609237671, + "learning_rate": 2.9682309124767225e-05, + "loss": 0.1205, + "step": 9808 + }, + { + "epoch": 27.399441340782122, + "grad_norm": 1.181720495223999, + "learning_rate": 2.968193668528864e-05, + "loss": 0.1174, + "step": 9809 + }, + { + "epoch": 27.40223463687151, + "grad_norm": 1.93496572971344, + "learning_rate": 2.9681564245810058e-05, + "loss": 0.1305, + "step": 9810 + }, + { + "epoch": 27.405027932960895, + "grad_norm": 0.5896297097206116, + "learning_rate": 2.968119180633147e-05, + "loss": 0.1367, + "step": 9811 + }, + { + "epoch": 27.407821229050278, + "grad_norm": 0.6743430495262146, + "learning_rate": 2.9680819366852887e-05, + "loss": 0.1309, + "step": 9812 + }, + { + "epoch": 27.410614525139664, + "grad_norm": 1.1747642755508423, + "learning_rate": 2.96804469273743e-05, + "loss": 0.1562, + "step": 9813 + }, + { + "epoch": 27.41340782122905, + "grad_norm": 0.6798600554466248, + "learning_rate": 2.9680074487895717e-05, + "loss": 0.1239, + "step": 9814 + }, + { + "epoch": 27.416201117318437, + "grad_norm": 1.055080771446228, + "learning_rate": 2.9679702048417134e-05, + "loss": 0.1824, + "step": 9815 + }, + { + "epoch": 27.41899441340782, + "grad_norm": 4.23495626449585, + "learning_rate": 2.967932960893855e-05, + "loss": 0.1801, + "step": 9816 + }, + { + "epoch": 27.421787709497206, + "grad_norm": 0.4669424295425415, + "learning_rate": 2.9678957169459963e-05, + "loss": 0.1613, + "step": 9817 + }, + { + "epoch": 27.424581005586592, + "grad_norm": 1.194547414779663, + "learning_rate": 2.9678584729981376e-05, + "loss": 0.182, + "step": 9818 + }, + { + "epoch": 27.42737430167598, + "grad_norm": 0.6379705667495728, + "learning_rate": 2.9678212290502793e-05, + "loss": 0.1785, + "step": 9819 + }, + { + "epoch": 27.43016759776536, + "grad_norm": 0.5442931652069092, + "learning_rate": 2.967783985102421e-05, + "loss": 0.1565, + "step": 9820 + }, + { + "epoch": 27.432960893854748, + "grad_norm": 0.4237368106842041, + "learning_rate": 2.9677467411545626e-05, + "loss": 0.1725, + "step": 9821 + }, + { + "epoch": 27.435754189944134, + "grad_norm": 0.5082820653915405, + "learning_rate": 2.967709497206704e-05, + "loss": 0.1601, + "step": 9822 + }, + { + "epoch": 27.43854748603352, + "grad_norm": 0.6949779987335205, + "learning_rate": 2.9676722532588456e-05, + "loss": 0.1715, + "step": 9823 + }, + { + "epoch": 27.441340782122904, + "grad_norm": 0.6238696575164795, + "learning_rate": 2.9676350093109872e-05, + "loss": 0.1521, + "step": 9824 + }, + { + "epoch": 27.44413407821229, + "grad_norm": 0.3873690664768219, + "learning_rate": 2.9675977653631285e-05, + "loss": 0.1306, + "step": 9825 + }, + { + "epoch": 27.446927374301676, + "grad_norm": 0.6172774434089661, + "learning_rate": 2.96756052141527e-05, + "loss": 0.1738, + "step": 9826 + }, + { + "epoch": 27.449720670391063, + "grad_norm": 0.7542896270751953, + "learning_rate": 2.9675232774674115e-05, + "loss": 0.1523, + "step": 9827 + }, + { + "epoch": 27.452513966480446, + "grad_norm": 0.6453026533126831, + "learning_rate": 2.967486033519553e-05, + "loss": 0.1334, + "step": 9828 + }, + { + "epoch": 27.455307262569832, + "grad_norm": 0.6445602178573608, + "learning_rate": 2.9674487895716948e-05, + "loss": 0.127, + "step": 9829 + }, + { + "epoch": 27.45810055865922, + "grad_norm": 0.5242617726325989, + "learning_rate": 2.9674115456238365e-05, + "loss": 0.1337, + "step": 9830 + }, + { + "epoch": 27.460893854748605, + "grad_norm": 0.6733556389808655, + "learning_rate": 2.9673743016759775e-05, + "loss": 0.1477, + "step": 9831 + }, + { + "epoch": 27.463687150837988, + "grad_norm": 0.39112892746925354, + "learning_rate": 2.967337057728119e-05, + "loss": 0.1356, + "step": 9832 + }, + { + "epoch": 27.466480446927374, + "grad_norm": 1.0950182676315308, + "learning_rate": 2.9672998137802608e-05, + "loss": 0.1529, + "step": 9833 + }, + { + "epoch": 27.46927374301676, + "grad_norm": 0.560577392578125, + "learning_rate": 2.9672625698324024e-05, + "loss": 0.1266, + "step": 9834 + }, + { + "epoch": 27.472067039106147, + "grad_norm": 0.6146395206451416, + "learning_rate": 2.9672253258845437e-05, + "loss": 0.1201, + "step": 9835 + }, + { + "epoch": 27.47486033519553, + "grad_norm": 0.5775479674339294, + "learning_rate": 2.9671880819366854e-05, + "loss": 0.142, + "step": 9836 + }, + { + "epoch": 27.477653631284916, + "grad_norm": 9.488886833190918, + "learning_rate": 2.967150837988827e-05, + "loss": 0.1431, + "step": 9837 + }, + { + "epoch": 27.480446927374302, + "grad_norm": 1.2783479690551758, + "learning_rate": 2.9671135940409683e-05, + "loss": 0.1159, + "step": 9838 + }, + { + "epoch": 27.48324022346369, + "grad_norm": 1.0887709856033325, + "learning_rate": 2.96707635009311e-05, + "loss": 0.1101, + "step": 9839 + }, + { + "epoch": 27.48603351955307, + "grad_norm": 0.5748500227928162, + "learning_rate": 2.9670391061452513e-05, + "loss": 0.1077, + "step": 9840 + }, + { + "epoch": 27.488826815642458, + "grad_norm": 1.5758888721466064, + "learning_rate": 2.967001862197393e-05, + "loss": 0.2016, + "step": 9841 + }, + { + "epoch": 27.491620111731844, + "grad_norm": 0.3659570813179016, + "learning_rate": 2.9669646182495346e-05, + "loss": 0.1471, + "step": 9842 + }, + { + "epoch": 27.49441340782123, + "grad_norm": 0.6599975824356079, + "learning_rate": 2.9669273743016763e-05, + "loss": 0.1402, + "step": 9843 + }, + { + "epoch": 27.497206703910614, + "grad_norm": 0.5299981236457825, + "learning_rate": 2.9668901303538176e-05, + "loss": 0.1762, + "step": 9844 + }, + { + "epoch": 27.5, + "grad_norm": 0.4255751073360443, + "learning_rate": 2.966852886405959e-05, + "loss": 0.1566, + "step": 9845 + }, + { + "epoch": 27.502793296089386, + "grad_norm": 0.7346686124801636, + "learning_rate": 2.9668156424581006e-05, + "loss": 0.147, + "step": 9846 + }, + { + "epoch": 27.505586592178773, + "grad_norm": 0.4661877155303955, + "learning_rate": 2.9667783985102422e-05, + "loss": 0.1362, + "step": 9847 + }, + { + "epoch": 27.508379888268156, + "grad_norm": 0.8656718134880066, + "learning_rate": 2.966741154562384e-05, + "loss": 0.171, + "step": 9848 + }, + { + "epoch": 27.511173184357542, + "grad_norm": 0.49153634905815125, + "learning_rate": 2.9667039106145252e-05, + "loss": 0.1385, + "step": 9849 + }, + { + "epoch": 27.51396648044693, + "grad_norm": 0.6261147260665894, + "learning_rate": 2.966666666666667e-05, + "loss": 0.1511, + "step": 9850 + }, + { + "epoch": 27.51675977653631, + "grad_norm": 0.6251435875892639, + "learning_rate": 2.966629422718808e-05, + "loss": 0.1535, + "step": 9851 + }, + { + "epoch": 27.519553072625698, + "grad_norm": 0.6880931854248047, + "learning_rate": 2.9665921787709498e-05, + "loss": 0.1324, + "step": 9852 + }, + { + "epoch": 27.522346368715084, + "grad_norm": 0.599320113658905, + "learning_rate": 2.966554934823091e-05, + "loss": 0.1218, + "step": 9853 + }, + { + "epoch": 27.52513966480447, + "grad_norm": 0.5531005859375, + "learning_rate": 2.9665176908752328e-05, + "loss": 0.1139, + "step": 9854 + }, + { + "epoch": 27.527932960893853, + "grad_norm": 1.7052927017211914, + "learning_rate": 2.9664804469273744e-05, + "loss": 0.1307, + "step": 9855 + }, + { + "epoch": 27.53072625698324, + "grad_norm": 0.8484750986099243, + "learning_rate": 2.966443202979516e-05, + "loss": 0.1081, + "step": 9856 + }, + { + "epoch": 27.533519553072626, + "grad_norm": 0.5047814249992371, + "learning_rate": 2.9664059590316577e-05, + "loss": 0.1318, + "step": 9857 + }, + { + "epoch": 27.536312849162012, + "grad_norm": 0.5988427996635437, + "learning_rate": 2.9663687150837987e-05, + "loss": 0.1382, + "step": 9858 + }, + { + "epoch": 27.539106145251395, + "grad_norm": 0.8027397990226746, + "learning_rate": 2.9663314711359404e-05, + "loss": 0.1507, + "step": 9859 + }, + { + "epoch": 27.54189944134078, + "grad_norm": 0.6703481078147888, + "learning_rate": 2.966294227188082e-05, + "loss": 0.1246, + "step": 9860 + }, + { + "epoch": 27.544692737430168, + "grad_norm": 1.041808843612671, + "learning_rate": 2.9662569832402237e-05, + "loss": 0.1415, + "step": 9861 + }, + { + "epoch": 27.547486033519554, + "grad_norm": 0.5612601041793823, + "learning_rate": 2.966219739292365e-05, + "loss": 0.1339, + "step": 9862 + }, + { + "epoch": 27.550279329608937, + "grad_norm": 2.487323045730591, + "learning_rate": 2.9661824953445066e-05, + "loss": 0.124, + "step": 9863 + }, + { + "epoch": 27.553072625698324, + "grad_norm": 0.9529021382331848, + "learning_rate": 2.9661452513966483e-05, + "loss": 0.1334, + "step": 9864 + }, + { + "epoch": 27.55586592178771, + "grad_norm": 1.3877876996994019, + "learning_rate": 2.9661080074487896e-05, + "loss": 0.1754, + "step": 9865 + }, + { + "epoch": 27.558659217877096, + "grad_norm": 1.365715503692627, + "learning_rate": 2.9660707635009313e-05, + "loss": 0.2256, + "step": 9866 + }, + { + "epoch": 27.56145251396648, + "grad_norm": 0.5926316380500793, + "learning_rate": 2.9660335195530726e-05, + "loss": 0.1513, + "step": 9867 + }, + { + "epoch": 27.564245810055866, + "grad_norm": 0.40436023473739624, + "learning_rate": 2.9659962756052142e-05, + "loss": 0.1282, + "step": 9868 + }, + { + "epoch": 27.567039106145252, + "grad_norm": 0.5319579243659973, + "learning_rate": 2.965959031657356e-05, + "loss": 0.1495, + "step": 9869 + }, + { + "epoch": 27.56983240223464, + "grad_norm": 0.6117899417877197, + "learning_rate": 2.9659217877094975e-05, + "loss": 0.1407, + "step": 9870 + }, + { + "epoch": 27.57262569832402, + "grad_norm": 0.5061189532279968, + "learning_rate": 2.9658845437616385e-05, + "loss": 0.1579, + "step": 9871 + }, + { + "epoch": 27.575418994413408, + "grad_norm": 0.7233691811561584, + "learning_rate": 2.96584729981378e-05, + "loss": 0.1593, + "step": 9872 + }, + { + "epoch": 27.578212290502794, + "grad_norm": 0.48618024587631226, + "learning_rate": 2.9658100558659218e-05, + "loss": 0.1364, + "step": 9873 + }, + { + "epoch": 27.58100558659218, + "grad_norm": 0.6157983541488647, + "learning_rate": 2.9657728119180635e-05, + "loss": 0.1475, + "step": 9874 + }, + { + "epoch": 27.583798882681563, + "grad_norm": 0.7704933285713196, + "learning_rate": 2.9657355679702048e-05, + "loss": 0.1238, + "step": 9875 + }, + { + "epoch": 27.58659217877095, + "grad_norm": 0.8893761038780212, + "learning_rate": 2.9656983240223464e-05, + "loss": 0.1623, + "step": 9876 + }, + { + "epoch": 27.589385474860336, + "grad_norm": 0.8426198959350586, + "learning_rate": 2.965661080074488e-05, + "loss": 0.1473, + "step": 9877 + }, + { + "epoch": 27.592178770949722, + "grad_norm": 0.8414502739906311, + "learning_rate": 2.9656238361266294e-05, + "loss": 0.1401, + "step": 9878 + }, + { + "epoch": 27.594972067039105, + "grad_norm": 0.5981059670448303, + "learning_rate": 2.965586592178771e-05, + "loss": 0.1254, + "step": 9879 + }, + { + "epoch": 27.59776536312849, + "grad_norm": 0.8142374157905579, + "learning_rate": 2.9655493482309124e-05, + "loss": 0.138, + "step": 9880 + }, + { + "epoch": 27.600558659217878, + "grad_norm": 1.1160221099853516, + "learning_rate": 2.965512104283054e-05, + "loss": 0.1289, + "step": 9881 + }, + { + "epoch": 27.60335195530726, + "grad_norm": 1.2933000326156616, + "learning_rate": 2.9654748603351957e-05, + "loss": 0.1261, + "step": 9882 + }, + { + "epoch": 27.606145251396647, + "grad_norm": 0.8113556504249573, + "learning_rate": 2.9654376163873373e-05, + "loss": 0.1405, + "step": 9883 + }, + { + "epoch": 27.608938547486034, + "grad_norm": 0.8618024587631226, + "learning_rate": 2.9654003724394787e-05, + "loss": 0.1662, + "step": 9884 + }, + { + "epoch": 27.61173184357542, + "grad_norm": 0.8047616481781006, + "learning_rate": 2.96536312849162e-05, + "loss": 0.1268, + "step": 9885 + }, + { + "epoch": 27.614525139664803, + "grad_norm": 1.8191349506378174, + "learning_rate": 2.9653258845437616e-05, + "loss": 0.1327, + "step": 9886 + }, + { + "epoch": 27.61731843575419, + "grad_norm": 0.7650716304779053, + "learning_rate": 2.9652886405959033e-05, + "loss": 0.1507, + "step": 9887 + }, + { + "epoch": 27.620111731843576, + "grad_norm": 1.8824890851974487, + "learning_rate": 2.965251396648045e-05, + "loss": 0.1256, + "step": 9888 + }, + { + "epoch": 27.622905027932962, + "grad_norm": 1.160725712776184, + "learning_rate": 2.9652141527001862e-05, + "loss": 0.1431, + "step": 9889 + }, + { + "epoch": 27.625698324022345, + "grad_norm": 1.4293848276138306, + "learning_rate": 2.965176908752328e-05, + "loss": 0.1481, + "step": 9890 + }, + { + "epoch": 27.62849162011173, + "grad_norm": 2.0164854526519775, + "learning_rate": 2.9651396648044692e-05, + "loss": 0.2148, + "step": 9891 + }, + { + "epoch": 27.631284916201118, + "grad_norm": 0.5731896162033081, + "learning_rate": 2.965102420856611e-05, + "loss": 0.1578, + "step": 9892 + }, + { + "epoch": 27.634078212290504, + "grad_norm": 0.6887102127075195, + "learning_rate": 2.9650651769087522e-05, + "loss": 0.1514, + "step": 9893 + }, + { + "epoch": 27.636871508379887, + "grad_norm": 0.828866720199585, + "learning_rate": 2.965027932960894e-05, + "loss": 0.1577, + "step": 9894 + }, + { + "epoch": 27.639664804469273, + "grad_norm": 0.43193167448043823, + "learning_rate": 2.9649906890130355e-05, + "loss": 0.1594, + "step": 9895 + }, + { + "epoch": 27.64245810055866, + "grad_norm": 0.8484553694725037, + "learning_rate": 2.964953445065177e-05, + "loss": 0.1446, + "step": 9896 + }, + { + "epoch": 27.645251396648046, + "grad_norm": 0.6234706044197083, + "learning_rate": 2.9649162011173188e-05, + "loss": 0.1759, + "step": 9897 + }, + { + "epoch": 27.64804469273743, + "grad_norm": 0.8827102780342102, + "learning_rate": 2.9648789571694598e-05, + "loss": 0.1476, + "step": 9898 + }, + { + "epoch": 27.650837988826815, + "grad_norm": 1.3931434154510498, + "learning_rate": 2.9648417132216014e-05, + "loss": 0.1448, + "step": 9899 + }, + { + "epoch": 27.6536312849162, + "grad_norm": 0.525802493095398, + "learning_rate": 2.964804469273743e-05, + "loss": 0.1221, + "step": 9900 + }, + { + "epoch": 27.656424581005588, + "grad_norm": 0.8921838998794556, + "learning_rate": 2.9647672253258847e-05, + "loss": 0.202, + "step": 9901 + }, + { + "epoch": 27.65921787709497, + "grad_norm": 1.12272047996521, + "learning_rate": 2.964729981378026e-05, + "loss": 0.1562, + "step": 9902 + }, + { + "epoch": 27.662011173184357, + "grad_norm": 0.5924628973007202, + "learning_rate": 2.9646927374301677e-05, + "loss": 0.1291, + "step": 9903 + }, + { + "epoch": 27.664804469273744, + "grad_norm": 0.5899050831794739, + "learning_rate": 2.9646554934823094e-05, + "loss": 0.152, + "step": 9904 + }, + { + "epoch": 27.66759776536313, + "grad_norm": 0.8182613253593445, + "learning_rate": 2.9646182495344507e-05, + "loss": 0.1355, + "step": 9905 + }, + { + "epoch": 27.670391061452513, + "grad_norm": 0.8269279599189758, + "learning_rate": 2.9645810055865923e-05, + "loss": 0.1295, + "step": 9906 + }, + { + "epoch": 27.6731843575419, + "grad_norm": 0.8590107560157776, + "learning_rate": 2.9645437616387336e-05, + "loss": 0.1304, + "step": 9907 + }, + { + "epoch": 27.675977653631286, + "grad_norm": 0.6573381423950195, + "learning_rate": 2.9645065176908753e-05, + "loss": 0.1384, + "step": 9908 + }, + { + "epoch": 27.678770949720672, + "grad_norm": 0.7335901856422424, + "learning_rate": 2.964469273743017e-05, + "loss": 0.1453, + "step": 9909 + }, + { + "epoch": 27.681564245810055, + "grad_norm": 1.0012896060943604, + "learning_rate": 2.9644320297951586e-05, + "loss": 0.144, + "step": 9910 + }, + { + "epoch": 27.68435754189944, + "grad_norm": 0.9690141081809998, + "learning_rate": 2.9643947858472996e-05, + "loss": 0.1389, + "step": 9911 + }, + { + "epoch": 27.687150837988828, + "grad_norm": 0.8410539031028748, + "learning_rate": 2.9643575418994412e-05, + "loss": 0.1192, + "step": 9912 + }, + { + "epoch": 27.689944134078214, + "grad_norm": 0.9019057154655457, + "learning_rate": 2.964320297951583e-05, + "loss": 0.1668, + "step": 9913 + }, + { + "epoch": 27.692737430167597, + "grad_norm": 3.8439741134643555, + "learning_rate": 2.9642830540037245e-05, + "loss": 0.1434, + "step": 9914 + }, + { + "epoch": 27.695530726256983, + "grad_norm": 1.0622528791427612, + "learning_rate": 2.9642458100558662e-05, + "loss": 0.136, + "step": 9915 + }, + { + "epoch": 27.69832402234637, + "grad_norm": 1.3088053464889526, + "learning_rate": 2.9642085661080075e-05, + "loss": 0.2123, + "step": 9916 + }, + { + "epoch": 27.701117318435756, + "grad_norm": 0.5496951341629028, + "learning_rate": 2.964171322160149e-05, + "loss": 0.1683, + "step": 9917 + }, + { + "epoch": 27.70391061452514, + "grad_norm": 0.550130307674408, + "learning_rate": 2.9641340782122905e-05, + "loss": 0.1689, + "step": 9918 + }, + { + "epoch": 27.706703910614525, + "grad_norm": 0.5701974630355835, + "learning_rate": 2.964096834264432e-05, + "loss": 0.1475, + "step": 9919 + }, + { + "epoch": 27.70949720670391, + "grad_norm": 1.037457823753357, + "learning_rate": 2.9640595903165734e-05, + "loss": 0.1544, + "step": 9920 + }, + { + "epoch": 27.712290502793294, + "grad_norm": 0.8996815085411072, + "learning_rate": 2.964022346368715e-05, + "loss": 0.1458, + "step": 9921 + }, + { + "epoch": 27.71508379888268, + "grad_norm": 0.5748428106307983, + "learning_rate": 2.9639851024208567e-05, + "loss": 0.1294, + "step": 9922 + }, + { + "epoch": 27.717877094972067, + "grad_norm": 0.49800485372543335, + "learning_rate": 2.9639478584729984e-05, + "loss": 0.1512, + "step": 9923 + }, + { + "epoch": 27.720670391061454, + "grad_norm": 0.6813650727272034, + "learning_rate": 2.96391061452514e-05, + "loss": 0.1539, + "step": 9924 + }, + { + "epoch": 27.723463687150836, + "grad_norm": 0.6486937999725342, + "learning_rate": 2.963873370577281e-05, + "loss": 0.145, + "step": 9925 + }, + { + "epoch": 27.726256983240223, + "grad_norm": 0.5629799962043762, + "learning_rate": 2.9638361266294227e-05, + "loss": 0.1355, + "step": 9926 + }, + { + "epoch": 27.72905027932961, + "grad_norm": 0.7466745376586914, + "learning_rate": 2.9637988826815643e-05, + "loss": 0.1376, + "step": 9927 + }, + { + "epoch": 27.731843575418996, + "grad_norm": 0.7181533575057983, + "learning_rate": 2.963761638733706e-05, + "loss": 0.1549, + "step": 9928 + }, + { + "epoch": 27.73463687150838, + "grad_norm": 0.698631763458252, + "learning_rate": 2.9637243947858473e-05, + "loss": 0.1518, + "step": 9929 + }, + { + "epoch": 27.737430167597765, + "grad_norm": 1.3914713859558105, + "learning_rate": 2.963687150837989e-05, + "loss": 0.1385, + "step": 9930 + }, + { + "epoch": 27.74022346368715, + "grad_norm": 1.163758397102356, + "learning_rate": 2.9636499068901303e-05, + "loss": 0.1187, + "step": 9931 + }, + { + "epoch": 27.743016759776538, + "grad_norm": 0.7985554933547974, + "learning_rate": 2.963612662942272e-05, + "loss": 0.1416, + "step": 9932 + }, + { + "epoch": 27.74581005586592, + "grad_norm": 0.49560630321502686, + "learning_rate": 2.9635754189944136e-05, + "loss": 0.1199, + "step": 9933 + }, + { + "epoch": 27.748603351955307, + "grad_norm": 2.000105142593384, + "learning_rate": 2.963538175046555e-05, + "loss": 0.139, + "step": 9934 + }, + { + "epoch": 27.751396648044693, + "grad_norm": 0.8870655298233032, + "learning_rate": 2.9635009310986965e-05, + "loss": 0.1645, + "step": 9935 + }, + { + "epoch": 27.75418994413408, + "grad_norm": 0.7599616050720215, + "learning_rate": 2.9634636871508382e-05, + "loss": 0.1351, + "step": 9936 + }, + { + "epoch": 27.756983240223462, + "grad_norm": 1.07082200050354, + "learning_rate": 2.96342644320298e-05, + "loss": 0.1437, + "step": 9937 + }, + { + "epoch": 27.75977653631285, + "grad_norm": 1.307384967803955, + "learning_rate": 2.9633891992551208e-05, + "loss": 0.1362, + "step": 9938 + }, + { + "epoch": 27.762569832402235, + "grad_norm": 2.1692354679107666, + "learning_rate": 2.9633519553072625e-05, + "loss": 0.1603, + "step": 9939 + }, + { + "epoch": 27.76536312849162, + "grad_norm": 1.290369987487793, + "learning_rate": 2.963314711359404e-05, + "loss": 0.1659, + "step": 9940 + }, + { + "epoch": 27.768156424581004, + "grad_norm": 1.1998696327209473, + "learning_rate": 2.9632774674115458e-05, + "loss": 0.1648, + "step": 9941 + }, + { + "epoch": 27.77094972067039, + "grad_norm": 0.5542314052581787, + "learning_rate": 2.9632402234636874e-05, + "loss": 0.1529, + "step": 9942 + }, + { + "epoch": 27.773743016759777, + "grad_norm": 1.0239657163619995, + "learning_rate": 2.9632029795158288e-05, + "loss": 0.1417, + "step": 9943 + }, + { + "epoch": 27.776536312849164, + "grad_norm": 1.5461112260818481, + "learning_rate": 2.9631657355679704e-05, + "loss": 0.1542, + "step": 9944 + }, + { + "epoch": 27.779329608938546, + "grad_norm": 0.967809796333313, + "learning_rate": 2.9631284916201117e-05, + "loss": 0.1627, + "step": 9945 + }, + { + "epoch": 27.782122905027933, + "grad_norm": 0.534260094165802, + "learning_rate": 2.9630912476722534e-05, + "loss": 0.1492, + "step": 9946 + }, + { + "epoch": 27.78491620111732, + "grad_norm": 0.5734409689903259, + "learning_rate": 2.9630540037243947e-05, + "loss": 0.1392, + "step": 9947 + }, + { + "epoch": 27.787709497206706, + "grad_norm": 0.5683371424674988, + "learning_rate": 2.9630167597765363e-05, + "loss": 0.1415, + "step": 9948 + }, + { + "epoch": 27.79050279329609, + "grad_norm": 1.0908176898956299, + "learning_rate": 2.962979515828678e-05, + "loss": 0.1479, + "step": 9949 + }, + { + "epoch": 27.793296089385475, + "grad_norm": 1.6067003011703491, + "learning_rate": 2.9629422718808197e-05, + "loss": 0.1401, + "step": 9950 + }, + { + "epoch": 27.79608938547486, + "grad_norm": 1.186672329902649, + "learning_rate": 2.962905027932961e-05, + "loss": 0.1392, + "step": 9951 + }, + { + "epoch": 27.798882681564244, + "grad_norm": 0.9979287385940552, + "learning_rate": 2.9628677839851023e-05, + "loss": 0.1219, + "step": 9952 + }, + { + "epoch": 27.80167597765363, + "grad_norm": 1.018131136894226, + "learning_rate": 2.962830540037244e-05, + "loss": 0.1309, + "step": 9953 + }, + { + "epoch": 27.804469273743017, + "grad_norm": 0.8389286398887634, + "learning_rate": 2.9627932960893856e-05, + "loss": 0.1411, + "step": 9954 + }, + { + "epoch": 27.807262569832403, + "grad_norm": 0.751658022403717, + "learning_rate": 2.9627560521415272e-05, + "loss": 0.1633, + "step": 9955 + }, + { + "epoch": 27.810055865921786, + "grad_norm": 0.7597814202308655, + "learning_rate": 2.9627188081936686e-05, + "loss": 0.1848, + "step": 9956 + }, + { + "epoch": 27.812849162011172, + "grad_norm": 1.4574906826019287, + "learning_rate": 2.9626815642458102e-05, + "loss": 0.127, + "step": 9957 + }, + { + "epoch": 27.81564245810056, + "grad_norm": 0.7773324251174927, + "learning_rate": 2.9626443202979515e-05, + "loss": 0.144, + "step": 9958 + }, + { + "epoch": 27.818435754189945, + "grad_norm": 1.0102843046188354, + "learning_rate": 2.9626070763500932e-05, + "loss": 0.1201, + "step": 9959 + }, + { + "epoch": 27.821229050279328, + "grad_norm": 1.5127524137496948, + "learning_rate": 2.962569832402235e-05, + "loss": 0.1331, + "step": 9960 + }, + { + "epoch": 27.824022346368714, + "grad_norm": 0.7649465799331665, + "learning_rate": 2.962532588454376e-05, + "loss": 0.1381, + "step": 9961 + }, + { + "epoch": 27.8268156424581, + "grad_norm": 3.042299270629883, + "learning_rate": 2.9624953445065178e-05, + "loss": 0.1444, + "step": 9962 + }, + { + "epoch": 27.829608938547487, + "grad_norm": 0.9207214117050171, + "learning_rate": 2.9624581005586595e-05, + "loss": 0.1516, + "step": 9963 + }, + { + "epoch": 27.83240223463687, + "grad_norm": 0.9397889375686646, + "learning_rate": 2.962420856610801e-05, + "loss": 0.1534, + "step": 9964 + }, + { + "epoch": 27.835195530726256, + "grad_norm": 0.7277540564537048, + "learning_rate": 2.962383612662942e-05, + "loss": 0.1451, + "step": 9965 + }, + { + "epoch": 27.837988826815643, + "grad_norm": 1.1739156246185303, + "learning_rate": 2.9623463687150837e-05, + "loss": 0.2069, + "step": 9966 + }, + { + "epoch": 27.84078212290503, + "grad_norm": 0.6383875012397766, + "learning_rate": 2.9623091247672254e-05, + "loss": 0.1696, + "step": 9967 + }, + { + "epoch": 27.843575418994412, + "grad_norm": 0.8241717219352722, + "learning_rate": 2.962271880819367e-05, + "loss": 0.1619, + "step": 9968 + }, + { + "epoch": 27.8463687150838, + "grad_norm": 0.462087482213974, + "learning_rate": 2.9622346368715084e-05, + "loss": 0.1657, + "step": 9969 + }, + { + "epoch": 27.849162011173185, + "grad_norm": 1.0009483098983765, + "learning_rate": 2.96219739292365e-05, + "loss": 0.1506, + "step": 9970 + }, + { + "epoch": 27.85195530726257, + "grad_norm": 0.49439239501953125, + "learning_rate": 2.9621601489757913e-05, + "loss": 0.134, + "step": 9971 + }, + { + "epoch": 27.854748603351954, + "grad_norm": 7.371710300445557, + "learning_rate": 2.962122905027933e-05, + "loss": 0.1514, + "step": 9972 + }, + { + "epoch": 27.85754189944134, + "grad_norm": 0.7341774106025696, + "learning_rate": 2.9620856610800746e-05, + "loss": 0.1747, + "step": 9973 + }, + { + "epoch": 27.860335195530727, + "grad_norm": 0.5609772205352783, + "learning_rate": 2.962048417132216e-05, + "loss": 0.1559, + "step": 9974 + }, + { + "epoch": 27.863128491620113, + "grad_norm": 0.768280565738678, + "learning_rate": 2.9620111731843576e-05, + "loss": 0.1408, + "step": 9975 + }, + { + "epoch": 27.865921787709496, + "grad_norm": 0.7014777660369873, + "learning_rate": 2.9619739292364993e-05, + "loss": 0.1558, + "step": 9976 + }, + { + "epoch": 27.868715083798882, + "grad_norm": 0.49809977412223816, + "learning_rate": 2.961936685288641e-05, + "loss": 0.1561, + "step": 9977 + }, + { + "epoch": 27.87150837988827, + "grad_norm": 0.6352024078369141, + "learning_rate": 2.961899441340782e-05, + "loss": 0.1457, + "step": 9978 + }, + { + "epoch": 27.874301675977655, + "grad_norm": 0.48892855644226074, + "learning_rate": 2.9618621973929235e-05, + "loss": 0.1282, + "step": 9979 + }, + { + "epoch": 27.877094972067038, + "grad_norm": 0.4475594162940979, + "learning_rate": 2.9618249534450652e-05, + "loss": 0.1497, + "step": 9980 + }, + { + "epoch": 27.879888268156424, + "grad_norm": 0.7913422584533691, + "learning_rate": 2.961787709497207e-05, + "loss": 0.1738, + "step": 9981 + }, + { + "epoch": 27.88268156424581, + "grad_norm": 1.9327009916305542, + "learning_rate": 2.9617504655493485e-05, + "loss": 0.1421, + "step": 9982 + }, + { + "epoch": 27.885474860335197, + "grad_norm": 2.8794071674346924, + "learning_rate": 2.9617132216014898e-05, + "loss": 0.1442, + "step": 9983 + }, + { + "epoch": 27.88826815642458, + "grad_norm": 0.6334337592124939, + "learning_rate": 2.9616759776536315e-05, + "loss": 0.1281, + "step": 9984 + }, + { + "epoch": 27.891061452513966, + "grad_norm": 4.637495517730713, + "learning_rate": 2.9616387337057728e-05, + "loss": 0.1614, + "step": 9985 + }, + { + "epoch": 27.893854748603353, + "grad_norm": 0.6283245086669922, + "learning_rate": 2.9616014897579144e-05, + "loss": 0.151, + "step": 9986 + }, + { + "epoch": 27.89664804469274, + "grad_norm": 0.8660380244255066, + "learning_rate": 2.9615642458100558e-05, + "loss": 0.172, + "step": 9987 + }, + { + "epoch": 27.899441340782122, + "grad_norm": 1.6173322200775146, + "learning_rate": 2.9615270018621974e-05, + "loss": 0.1562, + "step": 9988 + }, + { + "epoch": 27.90223463687151, + "grad_norm": 1.933774709701538, + "learning_rate": 2.961489757914339e-05, + "loss": 0.1456, + "step": 9989 + }, + { + "epoch": 27.905027932960895, + "grad_norm": 1.2209080457687378, + "learning_rate": 2.9614525139664807e-05, + "loss": 0.1562, + "step": 9990 + }, + { + "epoch": 27.907821229050278, + "grad_norm": 2.5312745571136475, + "learning_rate": 2.961415270018622e-05, + "loss": 0.2229, + "step": 9991 + }, + { + "epoch": 27.910614525139664, + "grad_norm": 0.5760404467582703, + "learning_rate": 2.9613780260707633e-05, + "loss": 0.1634, + "step": 9992 + }, + { + "epoch": 27.91340782122905, + "grad_norm": 0.5713164210319519, + "learning_rate": 2.961340782122905e-05, + "loss": 0.1587, + "step": 9993 + }, + { + "epoch": 27.916201117318437, + "grad_norm": 0.5153211355209351, + "learning_rate": 2.9613035381750467e-05, + "loss": 0.1431, + "step": 9994 + }, + { + "epoch": 27.91899441340782, + "grad_norm": 0.4577149450778961, + "learning_rate": 2.9612662942271883e-05, + "loss": 0.1649, + "step": 9995 + }, + { + "epoch": 27.921787709497206, + "grad_norm": 0.5290731191635132, + "learning_rate": 2.9612290502793296e-05, + "loss": 0.1594, + "step": 9996 + }, + { + "epoch": 27.924581005586592, + "grad_norm": 0.48113977909088135, + "learning_rate": 2.9611918063314713e-05, + "loss": 0.1429, + "step": 9997 + }, + { + "epoch": 27.92737430167598, + "grad_norm": 0.8133769035339355, + "learning_rate": 2.9611545623836126e-05, + "loss": 0.1553, + "step": 9998 + }, + { + "epoch": 27.93016759776536, + "grad_norm": 0.5677641034126282, + "learning_rate": 2.9611173184357542e-05, + "loss": 0.129, + "step": 9999 + }, + { + "epoch": 27.932960893854748, + "grad_norm": 0.509327232837677, + "learning_rate": 2.961080074487896e-05, + "loss": 0.144, + "step": 10000 + }, + { + "epoch": 27.932960893854748, + "eval_cer": 0.1915540135074684, + "eval_loss": 0.2650785446166992, + "eval_runtime": 23.0788, + "eval_samples_per_second": 196.631, + "eval_steps_per_second": 0.65, + "eval_wer": 0.33965531689058964, + "step": 10000 + }, + { + "epoch": 27.935754189944134, + "grad_norm": 3.4556398391723633, + "learning_rate": 2.9610428305400372e-05, + "loss": 0.1396, + "step": 10001 + }, + { + "epoch": 27.93854748603352, + "grad_norm": 0.505800724029541, + "learning_rate": 2.961005586592179e-05, + "loss": 0.1647, + "step": 10002 + }, + { + "epoch": 27.941340782122904, + "grad_norm": 0.6111153364181519, + "learning_rate": 2.9609683426443205e-05, + "loss": 0.1491, + "step": 10003 + }, + { + "epoch": 27.94413407821229, + "grad_norm": 0.850070059299469, + "learning_rate": 2.9609310986964622e-05, + "loss": 0.1463, + "step": 10004 + }, + { + "epoch": 27.946927374301676, + "grad_norm": 1.1689777374267578, + "learning_rate": 2.960893854748603e-05, + "loss": 0.1325, + "step": 10005 + }, + { + "epoch": 27.949720670391063, + "grad_norm": 0.7154826521873474, + "learning_rate": 2.9608566108007448e-05, + "loss": 0.1419, + "step": 10006 + }, + { + "epoch": 27.952513966480446, + "grad_norm": 0.7328900694847107, + "learning_rate": 2.9608193668528865e-05, + "loss": 0.1216, + "step": 10007 + }, + { + "epoch": 27.955307262569832, + "grad_norm": 0.9057138562202454, + "learning_rate": 2.960782122905028e-05, + "loss": 0.1346, + "step": 10008 + }, + { + "epoch": 27.95810055865922, + "grad_norm": 0.975844144821167, + "learning_rate": 2.9607448789571698e-05, + "loss": 0.1491, + "step": 10009 + }, + { + "epoch": 27.960893854748605, + "grad_norm": 0.6220122575759888, + "learning_rate": 2.960707635009311e-05, + "loss": 0.1336, + "step": 10010 + }, + { + "epoch": 27.963687150837988, + "grad_norm": 1.7246402502059937, + "learning_rate": 2.9606703910614524e-05, + "loss": 0.1413, + "step": 10011 + }, + { + "epoch": 27.966480446927374, + "grad_norm": 0.7029556632041931, + "learning_rate": 2.960633147113594e-05, + "loss": 0.1292, + "step": 10012 + }, + { + "epoch": 27.96927374301676, + "grad_norm": 0.9540677070617676, + "learning_rate": 2.9605959031657357e-05, + "loss": 0.127, + "step": 10013 + }, + { + "epoch": 27.972067039106147, + "grad_norm": 0.9542888402938843, + "learning_rate": 2.960558659217877e-05, + "loss": 0.1437, + "step": 10014 + }, + { + "epoch": 27.97486033519553, + "grad_norm": 1.3869632482528687, + "learning_rate": 2.9605214152700187e-05, + "loss": 0.1722, + "step": 10015 + }, + { + "epoch": 27.977653631284916, + "grad_norm": 33.08216094970703, + "learning_rate": 2.9604841713221603e-05, + "loss": 0.212, + "step": 10016 + }, + { + "epoch": 27.980446927374302, + "grad_norm": 1.1831276416778564, + "learning_rate": 2.960446927374302e-05, + "loss": 0.1807, + "step": 10017 + }, + { + "epoch": 27.98324022346369, + "grad_norm": 0.4835861325263977, + "learning_rate": 2.9604096834264433e-05, + "loss": 0.151, + "step": 10018 + }, + { + "epoch": 27.98603351955307, + "grad_norm": 0.5728359818458557, + "learning_rate": 2.9603724394785846e-05, + "loss": 0.137, + "step": 10019 + }, + { + "epoch": 27.988826815642458, + "grad_norm": 0.762850821018219, + "learning_rate": 2.9603351955307263e-05, + "loss": 0.1655, + "step": 10020 + }, + { + "epoch": 27.991620111731844, + "grad_norm": 0.6157329678535461, + "learning_rate": 2.960297951582868e-05, + "loss": 0.1459, + "step": 10021 + }, + { + "epoch": 27.994413407821227, + "grad_norm": 0.7178266048431396, + "learning_rate": 2.9602607076350096e-05, + "loss": 0.1318, + "step": 10022 + }, + { + "epoch": 27.997206703910614, + "grad_norm": 0.6186360120773315, + "learning_rate": 2.960223463687151e-05, + "loss": 0.1213, + "step": 10023 + }, + { + "epoch": 28.0, + "grad_norm": 0.8501427173614502, + "learning_rate": 2.9601862197392925e-05, + "loss": 0.1369, + "step": 10024 + }, + { + "epoch": 28.002793296089386, + "grad_norm": 0.5345637798309326, + "learning_rate": 2.960148975791434e-05, + "loss": 0.162, + "step": 10025 + }, + { + "epoch": 28.00558659217877, + "grad_norm": 0.5997533202171326, + "learning_rate": 2.9601117318435755e-05, + "loss": 0.1631, + "step": 10026 + }, + { + "epoch": 28.008379888268156, + "grad_norm": 0.40839308500289917, + "learning_rate": 2.960074487895717e-05, + "loss": 0.1574, + "step": 10027 + }, + { + "epoch": 28.011173184357542, + "grad_norm": 0.5557401180267334, + "learning_rate": 2.9600372439478585e-05, + "loss": 0.1517, + "step": 10028 + }, + { + "epoch": 28.01396648044693, + "grad_norm": 0.5408746600151062, + "learning_rate": 2.96e-05, + "loss": 0.139, + "step": 10029 + }, + { + "epoch": 28.01675977653631, + "grad_norm": 0.4471789002418518, + "learning_rate": 2.9599627560521418e-05, + "loss": 0.133, + "step": 10030 + }, + { + "epoch": 28.019553072625698, + "grad_norm": 0.9463058710098267, + "learning_rate": 2.959925512104283e-05, + "loss": 0.1297, + "step": 10031 + }, + { + "epoch": 28.022346368715084, + "grad_norm": 0.8242188692092896, + "learning_rate": 2.9598882681564244e-05, + "loss": 0.1311, + "step": 10032 + }, + { + "epoch": 28.02513966480447, + "grad_norm": 0.42533430457115173, + "learning_rate": 2.959851024208566e-05, + "loss": 0.1323, + "step": 10033 + }, + { + "epoch": 28.027932960893853, + "grad_norm": 1.2213718891143799, + "learning_rate": 2.9598137802607077e-05, + "loss": 0.1392, + "step": 10034 + }, + { + "epoch": 28.03072625698324, + "grad_norm": 0.4316934645175934, + "learning_rate": 2.9597765363128494e-05, + "loss": 0.163, + "step": 10035 + }, + { + "epoch": 28.033519553072626, + "grad_norm": 1.422203779220581, + "learning_rate": 2.959739292364991e-05, + "loss": 0.1713, + "step": 10036 + }, + { + "epoch": 28.036312849162012, + "grad_norm": 0.7455061078071594, + "learning_rate": 2.9597020484171323e-05, + "loss": 0.127, + "step": 10037 + }, + { + "epoch": 28.039106145251395, + "grad_norm": 1.5783143043518066, + "learning_rate": 2.9596648044692736e-05, + "loss": 0.1158, + "step": 10038 + }, + { + "epoch": 28.04189944134078, + "grad_norm": 0.5236300230026245, + "learning_rate": 2.9596275605214153e-05, + "loss": 0.1233, + "step": 10039 + }, + { + "epoch": 28.044692737430168, + "grad_norm": 0.7085267305374146, + "learning_rate": 2.959590316573557e-05, + "loss": 0.1242, + "step": 10040 + }, + { + "epoch": 28.047486033519554, + "grad_norm": 3.086510419845581, + "learning_rate": 2.9595530726256983e-05, + "loss": 0.1292, + "step": 10041 + }, + { + "epoch": 28.050279329608937, + "grad_norm": 0.6791415810585022, + "learning_rate": 2.95951582867784e-05, + "loss": 0.1464, + "step": 10042 + }, + { + "epoch": 28.053072625698324, + "grad_norm": 0.5826219916343689, + "learning_rate": 2.9594785847299816e-05, + "loss": 0.1314, + "step": 10043 + }, + { + "epoch": 28.05586592178771, + "grad_norm": 0.6934998035430908, + "learning_rate": 2.9594413407821232e-05, + "loss": 0.1529, + "step": 10044 + }, + { + "epoch": 28.058659217877096, + "grad_norm": 1.2280902862548828, + "learning_rate": 2.9594040968342645e-05, + "loss": 0.1286, + "step": 10045 + }, + { + "epoch": 28.06145251396648, + "grad_norm": 2.031154155731201, + "learning_rate": 2.959366852886406e-05, + "loss": 0.1105, + "step": 10046 + }, + { + "epoch": 28.064245810055866, + "grad_norm": 0.7969191074371338, + "learning_rate": 2.9593296089385475e-05, + "loss": 0.1276, + "step": 10047 + }, + { + "epoch": 28.067039106145252, + "grad_norm": 1.2643804550170898, + "learning_rate": 2.959292364990689e-05, + "loss": 0.1473, + "step": 10048 + }, + { + "epoch": 28.06983240223464, + "grad_norm": 1.5694897174835205, + "learning_rate": 2.9592551210428308e-05, + "loss": 0.1462, + "step": 10049 + }, + { + "epoch": 28.07262569832402, + "grad_norm": 0.46465951204299927, + "learning_rate": 2.959217877094972e-05, + "loss": 0.1756, + "step": 10050 + }, + { + "epoch": 28.075418994413408, + "grad_norm": 0.5716705918312073, + "learning_rate": 2.9591806331471134e-05, + "loss": 0.1761, + "step": 10051 + }, + { + "epoch": 28.078212290502794, + "grad_norm": 0.46439334750175476, + "learning_rate": 2.959143389199255e-05, + "loss": 0.1675, + "step": 10052 + }, + { + "epoch": 28.08100558659218, + "grad_norm": 0.37713220715522766, + "learning_rate": 2.9591061452513968e-05, + "loss": 0.1373, + "step": 10053 + }, + { + "epoch": 28.083798882681563, + "grad_norm": 1.996329426765442, + "learning_rate": 2.9590689013035384e-05, + "loss": 0.1446, + "step": 10054 + }, + { + "epoch": 28.08659217877095, + "grad_norm": 0.5521246194839478, + "learning_rate": 2.9590316573556797e-05, + "loss": 0.1343, + "step": 10055 + }, + { + "epoch": 28.089385474860336, + "grad_norm": 0.5536831617355347, + "learning_rate": 2.9589944134078214e-05, + "loss": 0.1293, + "step": 10056 + }, + { + "epoch": 28.092178770949722, + "grad_norm": 0.3886401057243347, + "learning_rate": 2.958957169459963e-05, + "loss": 0.1483, + "step": 10057 + }, + { + "epoch": 28.094972067039105, + "grad_norm": 0.5239360928535461, + "learning_rate": 2.9589199255121043e-05, + "loss": 0.1366, + "step": 10058 + }, + { + "epoch": 28.09776536312849, + "grad_norm": 0.9474433064460754, + "learning_rate": 2.9588826815642457e-05, + "loss": 0.1497, + "step": 10059 + }, + { + "epoch": 28.100558659217878, + "grad_norm": 0.4574422240257263, + "learning_rate": 2.9588454376163873e-05, + "loss": 0.1424, + "step": 10060 + }, + { + "epoch": 28.10335195530726, + "grad_norm": 0.6315430402755737, + "learning_rate": 2.958808193668529e-05, + "loss": 0.1313, + "step": 10061 + }, + { + "epoch": 28.106145251396647, + "grad_norm": 0.5310167074203491, + "learning_rate": 2.9587709497206706e-05, + "loss": 0.1321, + "step": 10062 + }, + { + "epoch": 28.108938547486034, + "grad_norm": 1.2159026861190796, + "learning_rate": 2.958733705772812e-05, + "loss": 0.1406, + "step": 10063 + }, + { + "epoch": 28.11173184357542, + "grad_norm": 1.5165271759033203, + "learning_rate": 2.9586964618249536e-05, + "loss": 0.1176, + "step": 10064 + }, + { + "epoch": 28.114525139664803, + "grad_norm": 0.909514307975769, + "learning_rate": 2.958659217877095e-05, + "loss": 0.1273, + "step": 10065 + }, + { + "epoch": 28.11731843575419, + "grad_norm": 0.5209634900093079, + "learning_rate": 2.9586219739292366e-05, + "loss": 0.1336, + "step": 10066 + }, + { + "epoch": 28.120111731843576, + "grad_norm": 0.7456034421920776, + "learning_rate": 2.9585847299813782e-05, + "loss": 0.1458, + "step": 10067 + }, + { + "epoch": 28.122905027932962, + "grad_norm": 0.6901075839996338, + "learning_rate": 2.9585474860335195e-05, + "loss": 0.1394, + "step": 10068 + }, + { + "epoch": 28.125698324022345, + "grad_norm": 0.5676772594451904, + "learning_rate": 2.9585102420856612e-05, + "loss": 0.1365, + "step": 10069 + }, + { + "epoch": 28.12849162011173, + "grad_norm": 0.8845539689064026, + "learning_rate": 2.958472998137803e-05, + "loss": 0.1524, + "step": 10070 + }, + { + "epoch": 28.131284916201118, + "grad_norm": 0.7871348261833191, + "learning_rate": 2.958435754189944e-05, + "loss": 0.1164, + "step": 10071 + }, + { + "epoch": 28.134078212290504, + "grad_norm": 1.1737688779830933, + "learning_rate": 2.9583985102420855e-05, + "loss": 0.1222, + "step": 10072 + }, + { + "epoch": 28.136871508379887, + "grad_norm": 0.756400465965271, + "learning_rate": 2.958361266294227e-05, + "loss": 0.1215, + "step": 10073 + }, + { + "epoch": 28.139664804469273, + "grad_norm": 1.2613722085952759, + "learning_rate": 2.9583240223463688e-05, + "loss": 0.1484, + "step": 10074 + }, + { + "epoch": 28.14245810055866, + "grad_norm": 0.38578253984451294, + "learning_rate": 2.9582867783985104e-05, + "loss": 0.1498, + "step": 10075 + }, + { + "epoch": 28.145251396648046, + "grad_norm": 0.6082139015197754, + "learning_rate": 2.958249534450652e-05, + "loss": 0.1413, + "step": 10076 + }, + { + "epoch": 28.14804469273743, + "grad_norm": 0.4586232304573059, + "learning_rate": 2.9582122905027934e-05, + "loss": 0.181, + "step": 10077 + }, + { + "epoch": 28.150837988826815, + "grad_norm": 0.6845012307167053, + "learning_rate": 2.9581750465549347e-05, + "loss": 0.1599, + "step": 10078 + }, + { + "epoch": 28.1536312849162, + "grad_norm": 0.5114811062812805, + "learning_rate": 2.9581378026070764e-05, + "loss": 0.152, + "step": 10079 + }, + { + "epoch": 28.156424581005588, + "grad_norm": 0.4526713788509369, + "learning_rate": 2.958100558659218e-05, + "loss": 0.1299, + "step": 10080 + }, + { + "epoch": 28.15921787709497, + "grad_norm": 0.3703678250312805, + "learning_rate": 2.9580633147113593e-05, + "loss": 0.1305, + "step": 10081 + }, + { + "epoch": 28.162011173184357, + "grad_norm": 0.42771145701408386, + "learning_rate": 2.958026070763501e-05, + "loss": 0.134, + "step": 10082 + }, + { + "epoch": 28.164804469273744, + "grad_norm": 0.93843013048172, + "learning_rate": 2.9579888268156426e-05, + "loss": 0.1509, + "step": 10083 + }, + { + "epoch": 28.16759776536313, + "grad_norm": 0.672741174697876, + "learning_rate": 2.9579515828677843e-05, + "loss": 0.1434, + "step": 10084 + }, + { + "epoch": 28.170391061452513, + "grad_norm": 0.6066272258758545, + "learning_rate": 2.9579143389199256e-05, + "loss": 0.1272, + "step": 10085 + }, + { + "epoch": 28.1731843575419, + "grad_norm": 0.9732383489608765, + "learning_rate": 2.957877094972067e-05, + "loss": 0.1093, + "step": 10086 + }, + { + "epoch": 28.175977653631286, + "grad_norm": 0.7073915600776672, + "learning_rate": 2.9578398510242086e-05, + "loss": 0.122, + "step": 10087 + }, + { + "epoch": 28.178770949720672, + "grad_norm": 1.5857399702072144, + "learning_rate": 2.9578026070763502e-05, + "loss": 0.1168, + "step": 10088 + }, + { + "epoch": 28.181564245810055, + "grad_norm": 1.35296630859375, + "learning_rate": 2.957765363128492e-05, + "loss": 0.1155, + "step": 10089 + }, + { + "epoch": 28.18435754189944, + "grad_norm": 0.5695287585258484, + "learning_rate": 2.9577281191806332e-05, + "loss": 0.132, + "step": 10090 + }, + { + "epoch": 28.187150837988828, + "grad_norm": 0.5216072201728821, + "learning_rate": 2.9576908752327745e-05, + "loss": 0.1366, + "step": 10091 + }, + { + "epoch": 28.189944134078214, + "grad_norm": 0.6675209403038025, + "learning_rate": 2.957653631284916e-05, + "loss": 0.1391, + "step": 10092 + }, + { + "epoch": 28.192737430167597, + "grad_norm": 0.9874434471130371, + "learning_rate": 2.9576163873370578e-05, + "loss": 0.1418, + "step": 10093 + }, + { + "epoch": 28.195530726256983, + "grad_norm": 1.2951745986938477, + "learning_rate": 2.9575791433891995e-05, + "loss": 0.1583, + "step": 10094 + }, + { + "epoch": 28.19832402234637, + "grad_norm": 1.1041152477264404, + "learning_rate": 2.9575418994413408e-05, + "loss": 0.137, + "step": 10095 + }, + { + "epoch": 28.201117318435756, + "grad_norm": 0.863351583480835, + "learning_rate": 2.9575046554934824e-05, + "loss": 0.1407, + "step": 10096 + }, + { + "epoch": 28.20391061452514, + "grad_norm": 0.6614184975624084, + "learning_rate": 2.957467411545624e-05, + "loss": 0.126, + "step": 10097 + }, + { + "epoch": 28.206703910614525, + "grad_norm": 2.293856143951416, + "learning_rate": 2.9574301675977654e-05, + "loss": 0.1748, + "step": 10098 + }, + { + "epoch": 28.20949720670391, + "grad_norm": 2.167203664779663, + "learning_rate": 2.9573929236499067e-05, + "loss": 0.1581, + "step": 10099 + }, + { + "epoch": 28.212290502793294, + "grad_norm": 0.8436817526817322, + "learning_rate": 2.9573556797020484e-05, + "loss": 0.1628, + "step": 10100 + }, + { + "epoch": 28.21508379888268, + "grad_norm": 0.5757008194923401, + "learning_rate": 2.95731843575419e-05, + "loss": 0.1401, + "step": 10101 + }, + { + "epoch": 28.217877094972067, + "grad_norm": 0.6769471764564514, + "learning_rate": 2.9572811918063317e-05, + "loss": 0.1472, + "step": 10102 + }, + { + "epoch": 28.220670391061454, + "grad_norm": 1.0440351963043213, + "learning_rate": 2.9572439478584733e-05, + "loss": 0.1755, + "step": 10103 + }, + { + "epoch": 28.223463687150836, + "grad_norm": 0.4111757278442383, + "learning_rate": 2.9572067039106146e-05, + "loss": 0.1377, + "step": 10104 + }, + { + "epoch": 28.226256983240223, + "grad_norm": 0.6306076645851135, + "learning_rate": 2.957169459962756e-05, + "loss": 0.1805, + "step": 10105 + }, + { + "epoch": 28.22905027932961, + "grad_norm": 0.5419257879257202, + "learning_rate": 2.9571322160148976e-05, + "loss": 0.1623, + "step": 10106 + }, + { + "epoch": 28.231843575418996, + "grad_norm": 0.6016229391098022, + "learning_rate": 2.9570949720670393e-05, + "loss": 0.1179, + "step": 10107 + }, + { + "epoch": 28.23463687150838, + "grad_norm": 1.2356760501861572, + "learning_rate": 2.9570577281191806e-05, + "loss": 0.12, + "step": 10108 + }, + { + "epoch": 28.237430167597765, + "grad_norm": 1.3644688129425049, + "learning_rate": 2.9570204841713222e-05, + "loss": 0.1494, + "step": 10109 + }, + { + "epoch": 28.24022346368715, + "grad_norm": 0.6140114068984985, + "learning_rate": 2.956983240223464e-05, + "loss": 0.1558, + "step": 10110 + }, + { + "epoch": 28.243016759776538, + "grad_norm": 0.41832050681114197, + "learning_rate": 2.9569459962756052e-05, + "loss": 0.1344, + "step": 10111 + }, + { + "epoch": 28.24581005586592, + "grad_norm": 0.6785983443260193, + "learning_rate": 2.956908752327747e-05, + "loss": 0.1324, + "step": 10112 + }, + { + "epoch": 28.248603351955307, + "grad_norm": 0.8406807780265808, + "learning_rate": 2.9568715083798882e-05, + "loss": 0.1695, + "step": 10113 + }, + { + "epoch": 28.251396648044693, + "grad_norm": 0.46861863136291504, + "learning_rate": 2.9568342644320298e-05, + "loss": 0.1331, + "step": 10114 + }, + { + "epoch": 28.25418994413408, + "grad_norm": 4.656551837921143, + "learning_rate": 2.9567970204841715e-05, + "loss": 0.1326, + "step": 10115 + }, + { + "epoch": 28.256983240223462, + "grad_norm": 0.892148494720459, + "learning_rate": 2.956759776536313e-05, + "loss": 0.1235, + "step": 10116 + }, + { + "epoch": 28.25977653631285, + "grad_norm": 0.5329297184944153, + "learning_rate": 2.9567225325884545e-05, + "loss": 0.1347, + "step": 10117 + }, + { + "epoch": 28.262569832402235, + "grad_norm": 0.6994174122810364, + "learning_rate": 2.9566852886405958e-05, + "loss": 0.1276, + "step": 10118 + }, + { + "epoch": 28.26536312849162, + "grad_norm": 0.655913233757019, + "learning_rate": 2.9566480446927374e-05, + "loss": 0.1428, + "step": 10119 + }, + { + "epoch": 28.268156424581004, + "grad_norm": 0.9514911770820618, + "learning_rate": 2.956610800744879e-05, + "loss": 0.1236, + "step": 10120 + }, + { + "epoch": 28.27094972067039, + "grad_norm": 0.5544720888137817, + "learning_rate": 2.9565735567970207e-05, + "loss": 0.1185, + "step": 10121 + }, + { + "epoch": 28.273743016759777, + "grad_norm": 1.8054792881011963, + "learning_rate": 2.956536312849162e-05, + "loss": 0.1343, + "step": 10122 + }, + { + "epoch": 28.276536312849164, + "grad_norm": 1.3808226585388184, + "learning_rate": 2.9564990689013037e-05, + "loss": 0.1552, + "step": 10123 + }, + { + "epoch": 28.279329608938546, + "grad_norm": 1.1294400691986084, + "learning_rate": 2.9564618249534453e-05, + "loss": 0.1679, + "step": 10124 + }, + { + "epoch": 28.282122905027933, + "grad_norm": 0.6196085810661316, + "learning_rate": 2.9564245810055867e-05, + "loss": 0.15, + "step": 10125 + }, + { + "epoch": 28.28491620111732, + "grad_norm": 0.6601141691207886, + "learning_rate": 2.956387337057728e-05, + "loss": 0.1705, + "step": 10126 + }, + { + "epoch": 28.287709497206706, + "grad_norm": 0.9696556329727173, + "learning_rate": 2.9563500931098696e-05, + "loss": 0.1559, + "step": 10127 + }, + { + "epoch": 28.29050279329609, + "grad_norm": 0.5571941137313843, + "learning_rate": 2.9563128491620113e-05, + "loss": 0.1597, + "step": 10128 + }, + { + "epoch": 28.293296089385475, + "grad_norm": 0.47618407011032104, + "learning_rate": 2.956275605214153e-05, + "loss": 0.1506, + "step": 10129 + }, + { + "epoch": 28.29608938547486, + "grad_norm": 0.43927431106567383, + "learning_rate": 2.9562383612662946e-05, + "loss": 0.1318, + "step": 10130 + }, + { + "epoch": 28.298882681564244, + "grad_norm": 0.5514461994171143, + "learning_rate": 2.9562011173184356e-05, + "loss": 0.173, + "step": 10131 + }, + { + "epoch": 28.30167597765363, + "grad_norm": 0.49970605969429016, + "learning_rate": 2.9561638733705772e-05, + "loss": 0.1283, + "step": 10132 + }, + { + "epoch": 28.304469273743017, + "grad_norm": 1.0261073112487793, + "learning_rate": 2.956126629422719e-05, + "loss": 0.1391, + "step": 10133 + }, + { + "epoch": 28.307262569832403, + "grad_norm": 1.3348052501678467, + "learning_rate": 2.9560893854748605e-05, + "loss": 0.1297, + "step": 10134 + }, + { + "epoch": 28.310055865921786, + "grad_norm": 0.49302977323532104, + "learning_rate": 2.956052141527002e-05, + "loss": 0.1392, + "step": 10135 + }, + { + "epoch": 28.312849162011172, + "grad_norm": 0.5596616864204407, + "learning_rate": 2.9560148975791435e-05, + "loss": 0.1113, + "step": 10136 + }, + { + "epoch": 28.31564245810056, + "grad_norm": 0.7297159433364868, + "learning_rate": 2.955977653631285e-05, + "loss": 0.154, + "step": 10137 + }, + { + "epoch": 28.318435754189945, + "grad_norm": 0.5797239542007446, + "learning_rate": 2.9559404096834265e-05, + "loss": 0.1216, + "step": 10138 + }, + { + "epoch": 28.321229050279328, + "grad_norm": 0.6064387559890747, + "learning_rate": 2.955903165735568e-05, + "loss": 0.14, + "step": 10139 + }, + { + "epoch": 28.324022346368714, + "grad_norm": 1.4730796813964844, + "learning_rate": 2.9558659217877094e-05, + "loss": 0.1307, + "step": 10140 + }, + { + "epoch": 28.3268156424581, + "grad_norm": 0.9733483195304871, + "learning_rate": 2.955828677839851e-05, + "loss": 0.1438, + "step": 10141 + }, + { + "epoch": 28.329608938547487, + "grad_norm": 0.6298713088035583, + "learning_rate": 2.9557914338919927e-05, + "loss": 0.1134, + "step": 10142 + }, + { + "epoch": 28.33240223463687, + "grad_norm": 0.6097264289855957, + "learning_rate": 2.9557541899441344e-05, + "loss": 0.1026, + "step": 10143 + }, + { + "epoch": 28.335195530726256, + "grad_norm": 0.634994387626648, + "learning_rate": 2.9557169459962757e-05, + "loss": 0.1125, + "step": 10144 + }, + { + "epoch": 28.337988826815643, + "grad_norm": 1.5880661010742188, + "learning_rate": 2.955679702048417e-05, + "loss": 0.1806, + "step": 10145 + }, + { + "epoch": 28.34078212290503, + "grad_norm": 1.046897530555725, + "learning_rate": 2.9556424581005587e-05, + "loss": 0.1251, + "step": 10146 + }, + { + "epoch": 28.343575418994412, + "grad_norm": 0.9302058219909668, + "learning_rate": 2.9556052141527003e-05, + "loss": 0.1354, + "step": 10147 + }, + { + "epoch": 28.3463687150838, + "grad_norm": 0.9427276849746704, + "learning_rate": 2.955567970204842e-05, + "loss": 0.1266, + "step": 10148 + }, + { + "epoch": 28.349162011173185, + "grad_norm": 1.0121973752975464, + "learning_rate": 2.9555307262569833e-05, + "loss": 0.1592, + "step": 10149 + }, + { + "epoch": 28.35195530726257, + "grad_norm": 0.5147683024406433, + "learning_rate": 2.955493482309125e-05, + "loss": 0.1759, + "step": 10150 + }, + { + "epoch": 28.354748603351954, + "grad_norm": 0.3949033319950104, + "learning_rate": 2.9554562383612663e-05, + "loss": 0.1424, + "step": 10151 + }, + { + "epoch": 28.35754189944134, + "grad_norm": 1.5173237323760986, + "learning_rate": 2.955418994413408e-05, + "loss": 0.1398, + "step": 10152 + }, + { + "epoch": 28.360335195530727, + "grad_norm": 0.5059146285057068, + "learning_rate": 2.9553817504655492e-05, + "loss": 0.1415, + "step": 10153 + }, + { + "epoch": 28.363128491620113, + "grad_norm": 0.8825187087059021, + "learning_rate": 2.955344506517691e-05, + "loss": 0.1371, + "step": 10154 + }, + { + "epoch": 28.365921787709496, + "grad_norm": 1.0010086297988892, + "learning_rate": 2.9553072625698325e-05, + "loss": 0.1506, + "step": 10155 + }, + { + "epoch": 28.368715083798882, + "grad_norm": 0.7214546799659729, + "learning_rate": 2.9552700186219742e-05, + "loss": 0.1551, + "step": 10156 + }, + { + "epoch": 28.37150837988827, + "grad_norm": 0.5314242839813232, + "learning_rate": 2.9552327746741155e-05, + "loss": 0.1767, + "step": 10157 + }, + { + "epoch": 28.374301675977655, + "grad_norm": 0.9250205159187317, + "learning_rate": 2.9551955307262568e-05, + "loss": 0.1203, + "step": 10158 + }, + { + "epoch": 28.377094972067038, + "grad_norm": 1.0268011093139648, + "learning_rate": 2.9551582867783985e-05, + "loss": 0.122, + "step": 10159 + }, + { + "epoch": 28.379888268156424, + "grad_norm": 0.5631047487258911, + "learning_rate": 2.95512104283054e-05, + "loss": 0.1261, + "step": 10160 + }, + { + "epoch": 28.38268156424581, + "grad_norm": 0.7613141536712646, + "learning_rate": 2.9550837988826818e-05, + "loss": 0.1195, + "step": 10161 + }, + { + "epoch": 28.385474860335197, + "grad_norm": 1.4568493366241455, + "learning_rate": 2.955046554934823e-05, + "loss": 0.145, + "step": 10162 + }, + { + "epoch": 28.38826815642458, + "grad_norm": 0.5283349752426147, + "learning_rate": 2.9550093109869648e-05, + "loss": 0.1175, + "step": 10163 + }, + { + "epoch": 28.391061452513966, + "grad_norm": 0.8016228079795837, + "learning_rate": 2.9549720670391064e-05, + "loss": 0.1297, + "step": 10164 + }, + { + "epoch": 28.393854748603353, + "grad_norm": 0.6492945551872253, + "learning_rate": 2.9549348230912477e-05, + "loss": 0.1115, + "step": 10165 + }, + { + "epoch": 28.39664804469274, + "grad_norm": 0.8857135772705078, + "learning_rate": 2.954897579143389e-05, + "loss": 0.1302, + "step": 10166 + }, + { + "epoch": 28.399441340782122, + "grad_norm": 1.6111117601394653, + "learning_rate": 2.9548603351955307e-05, + "loss": 0.1428, + "step": 10167 + }, + { + "epoch": 28.40223463687151, + "grad_norm": 2.57065749168396, + "learning_rate": 2.9548230912476723e-05, + "loss": 0.1409, + "step": 10168 + }, + { + "epoch": 28.405027932960895, + "grad_norm": 2.2291314601898193, + "learning_rate": 2.954785847299814e-05, + "loss": 0.1478, + "step": 10169 + }, + { + "epoch": 28.407821229050278, + "grad_norm": 0.719061553478241, + "learning_rate": 2.9547486033519557e-05, + "loss": 0.1349, + "step": 10170 + }, + { + "epoch": 28.410614525139664, + "grad_norm": 2.95825457572937, + "learning_rate": 2.9547113594040966e-05, + "loss": 0.1328, + "step": 10171 + }, + { + "epoch": 28.41340782122905, + "grad_norm": 1.5634002685546875, + "learning_rate": 2.9546741154562383e-05, + "loss": 0.1656, + "step": 10172 + }, + { + "epoch": 28.416201117318437, + "grad_norm": 0.8446335792541504, + "learning_rate": 2.95463687150838e-05, + "loss": 0.1512, + "step": 10173 + }, + { + "epoch": 28.41899441340782, + "grad_norm": 0.9896931648254395, + "learning_rate": 2.9545996275605216e-05, + "loss": 0.2269, + "step": 10174 + }, + { + "epoch": 28.421787709497206, + "grad_norm": 0.7375166416168213, + "learning_rate": 2.954562383612663e-05, + "loss": 0.1479, + "step": 10175 + }, + { + "epoch": 28.424581005586592, + "grad_norm": 0.5271154046058655, + "learning_rate": 2.9545251396648046e-05, + "loss": 0.1561, + "step": 10176 + }, + { + "epoch": 28.42737430167598, + "grad_norm": 0.5730719566345215, + "learning_rate": 2.9544878957169462e-05, + "loss": 0.1479, + "step": 10177 + }, + { + "epoch": 28.43016759776536, + "grad_norm": 0.43782976269721985, + "learning_rate": 2.9544506517690875e-05, + "loss": 0.1674, + "step": 10178 + }, + { + "epoch": 28.432960893854748, + "grad_norm": 0.5662083625793457, + "learning_rate": 2.9544134078212292e-05, + "loss": 0.1482, + "step": 10179 + }, + { + "epoch": 28.435754189944134, + "grad_norm": 0.7226879000663757, + "learning_rate": 2.9543761638733705e-05, + "loss": 0.1564, + "step": 10180 + }, + { + "epoch": 28.43854748603352, + "grad_norm": 0.8816201686859131, + "learning_rate": 2.954338919925512e-05, + "loss": 0.1599, + "step": 10181 + }, + { + "epoch": 28.441340782122904, + "grad_norm": 0.6536687612533569, + "learning_rate": 2.9543016759776538e-05, + "loss": 0.1572, + "step": 10182 + }, + { + "epoch": 28.44413407821229, + "grad_norm": 0.5501503348350525, + "learning_rate": 2.9542644320297955e-05, + "loss": 0.1463, + "step": 10183 + }, + { + "epoch": 28.446927374301676, + "grad_norm": 0.7549936771392822, + "learning_rate": 2.9542271880819368e-05, + "loss": 0.1461, + "step": 10184 + }, + { + "epoch": 28.449720670391063, + "grad_norm": 0.6263824701309204, + "learning_rate": 2.954189944134078e-05, + "loss": 0.1533, + "step": 10185 + }, + { + "epoch": 28.452513966480446, + "grad_norm": 1.269284963607788, + "learning_rate": 2.9541527001862197e-05, + "loss": 0.1623, + "step": 10186 + }, + { + "epoch": 28.455307262569832, + "grad_norm": 0.7112523913383484, + "learning_rate": 2.9541154562383614e-05, + "loss": 0.1423, + "step": 10187 + }, + { + "epoch": 28.45810055865922, + "grad_norm": 0.5335250496864319, + "learning_rate": 2.954078212290503e-05, + "loss": 0.1443, + "step": 10188 + }, + { + "epoch": 28.460893854748605, + "grad_norm": 0.7642614841461182, + "learning_rate": 2.9540409683426444e-05, + "loss": 0.132, + "step": 10189 + }, + { + "epoch": 28.463687150837988, + "grad_norm": 0.5610061287879944, + "learning_rate": 2.954003724394786e-05, + "loss": 0.1412, + "step": 10190 + }, + { + "epoch": 28.466480446927374, + "grad_norm": 0.7585299015045166, + "learning_rate": 2.9539664804469273e-05, + "loss": 0.1199, + "step": 10191 + }, + { + "epoch": 28.46927374301676, + "grad_norm": 0.5922784209251404, + "learning_rate": 2.953929236499069e-05, + "loss": 0.1471, + "step": 10192 + }, + { + "epoch": 28.472067039106147, + "grad_norm": 0.6069446206092834, + "learning_rate": 2.9538919925512103e-05, + "loss": 0.1289, + "step": 10193 + }, + { + "epoch": 28.47486033519553, + "grad_norm": 0.5626804828643799, + "learning_rate": 2.953854748603352e-05, + "loss": 0.1052, + "step": 10194 + }, + { + "epoch": 28.477653631284916, + "grad_norm": 0.805557370185852, + "learning_rate": 2.9538175046554936e-05, + "loss": 0.1406, + "step": 10195 + }, + { + "epoch": 28.480446927374302, + "grad_norm": 1.558178424835205, + "learning_rate": 2.9537802607076353e-05, + "loss": 0.1185, + "step": 10196 + }, + { + "epoch": 28.48324022346369, + "grad_norm": 1.2119213342666626, + "learning_rate": 2.953743016759777e-05, + "loss": 0.1635, + "step": 10197 + }, + { + "epoch": 28.48603351955307, + "grad_norm": 1.6988393068313599, + "learning_rate": 2.953705772811918e-05, + "loss": 0.165, + "step": 10198 + }, + { + "epoch": 28.488826815642458, + "grad_norm": 1.2043678760528564, + "learning_rate": 2.9536685288640595e-05, + "loss": 0.1713, + "step": 10199 + }, + { + "epoch": 28.491620111731844, + "grad_norm": 0.5750095248222351, + "learning_rate": 2.9536312849162012e-05, + "loss": 0.1799, + "step": 10200 + }, + { + "epoch": 28.49441340782123, + "grad_norm": 0.4626733660697937, + "learning_rate": 2.953594040968343e-05, + "loss": 0.1392, + "step": 10201 + }, + { + "epoch": 28.497206703910614, + "grad_norm": 0.46261531114578247, + "learning_rate": 2.953556797020484e-05, + "loss": 0.1406, + "step": 10202 + }, + { + "epoch": 28.5, + "grad_norm": 0.9302212595939636, + "learning_rate": 2.9535195530726258e-05, + "loss": 0.1508, + "step": 10203 + }, + { + "epoch": 28.502793296089386, + "grad_norm": 1.1684811115264893, + "learning_rate": 2.9534823091247675e-05, + "loss": 0.2024, + "step": 10204 + }, + { + "epoch": 28.505586592178773, + "grad_norm": 0.4615778625011444, + "learning_rate": 2.9534450651769088e-05, + "loss": 0.128, + "step": 10205 + }, + { + "epoch": 28.508379888268156, + "grad_norm": 0.46602413058280945, + "learning_rate": 2.9534078212290504e-05, + "loss": 0.1316, + "step": 10206 + }, + { + "epoch": 28.511173184357542, + "grad_norm": 0.6439805030822754, + "learning_rate": 2.9533705772811918e-05, + "loss": 0.1563, + "step": 10207 + }, + { + "epoch": 28.51396648044693, + "grad_norm": 0.7161436080932617, + "learning_rate": 2.9533333333333334e-05, + "loss": 0.1492, + "step": 10208 + }, + { + "epoch": 28.51675977653631, + "grad_norm": 0.5262812376022339, + "learning_rate": 2.953296089385475e-05, + "loss": 0.1037, + "step": 10209 + }, + { + "epoch": 28.519553072625698, + "grad_norm": 0.7831761837005615, + "learning_rate": 2.9532588454376167e-05, + "loss": 0.1504, + "step": 10210 + }, + { + "epoch": 28.522346368715084, + "grad_norm": 1.475426435470581, + "learning_rate": 2.9532216014897577e-05, + "loss": 0.1725, + "step": 10211 + }, + { + "epoch": 28.52513966480447, + "grad_norm": 0.8725184798240662, + "learning_rate": 2.9531843575418993e-05, + "loss": 0.1276, + "step": 10212 + }, + { + "epoch": 28.527932960893853, + "grad_norm": 0.5029345154762268, + "learning_rate": 2.953147113594041e-05, + "loss": 0.1378, + "step": 10213 + }, + { + "epoch": 28.53072625698324, + "grad_norm": 0.8958428502082825, + "learning_rate": 2.9531098696461826e-05, + "loss": 0.1326, + "step": 10214 + }, + { + "epoch": 28.533519553072626, + "grad_norm": 0.9463875889778137, + "learning_rate": 2.9530726256983243e-05, + "loss": 0.1283, + "step": 10215 + }, + { + "epoch": 28.536312849162012, + "grad_norm": 0.9381642937660217, + "learning_rate": 2.9530353817504656e-05, + "loss": 0.1244, + "step": 10216 + }, + { + "epoch": 28.539106145251395, + "grad_norm": 1.7516433000564575, + "learning_rate": 2.9529981378026073e-05, + "loss": 0.133, + "step": 10217 + }, + { + "epoch": 28.54189944134078, + "grad_norm": 0.5721677541732788, + "learning_rate": 2.9529608938547486e-05, + "loss": 0.1419, + "step": 10218 + }, + { + "epoch": 28.544692737430168, + "grad_norm": 0.6777116656303406, + "learning_rate": 2.9529236499068902e-05, + "loss": 0.1388, + "step": 10219 + }, + { + "epoch": 28.547486033519554, + "grad_norm": 1.5049278736114502, + "learning_rate": 2.9528864059590316e-05, + "loss": 0.136, + "step": 10220 + }, + { + "epoch": 28.550279329608937, + "grad_norm": 0.9147613048553467, + "learning_rate": 2.9528491620111732e-05, + "loss": 0.1515, + "step": 10221 + }, + { + "epoch": 28.553072625698324, + "grad_norm": 0.6670833230018616, + "learning_rate": 2.952811918063315e-05, + "loss": 0.1414, + "step": 10222 + }, + { + "epoch": 28.55586592178771, + "grad_norm": 1.419734001159668, + "learning_rate": 2.9527746741154565e-05, + "loss": 0.1378, + "step": 10223 + }, + { + "epoch": 28.558659217877096, + "grad_norm": 4.56435489654541, + "learning_rate": 2.952737430167598e-05, + "loss": 0.1983, + "step": 10224 + }, + { + "epoch": 28.56145251396648, + "grad_norm": 0.4282947778701782, + "learning_rate": 2.952700186219739e-05, + "loss": 0.1612, + "step": 10225 + }, + { + "epoch": 28.564245810055866, + "grad_norm": 0.9588721990585327, + "learning_rate": 2.9526629422718808e-05, + "loss": 0.1386, + "step": 10226 + }, + { + "epoch": 28.567039106145252, + "grad_norm": 0.6453548669815063, + "learning_rate": 2.9526256983240225e-05, + "loss": 0.1551, + "step": 10227 + }, + { + "epoch": 28.56983240223464, + "grad_norm": 1.0761858224868774, + "learning_rate": 2.952588454376164e-05, + "loss": 0.1573, + "step": 10228 + }, + { + "epoch": 28.57262569832402, + "grad_norm": 0.5267358422279358, + "learning_rate": 2.9525512104283054e-05, + "loss": 0.144, + "step": 10229 + }, + { + "epoch": 28.575418994413408, + "grad_norm": 0.5664010047912598, + "learning_rate": 2.952513966480447e-05, + "loss": 0.1521, + "step": 10230 + }, + { + "epoch": 28.578212290502794, + "grad_norm": 0.6915371417999268, + "learning_rate": 2.9524767225325884e-05, + "loss": 0.1387, + "step": 10231 + }, + { + "epoch": 28.58100558659218, + "grad_norm": 0.4634493589401245, + "learning_rate": 2.95243947858473e-05, + "loss": 0.1452, + "step": 10232 + }, + { + "epoch": 28.583798882681563, + "grad_norm": 0.6776282787322998, + "learning_rate": 2.9524022346368717e-05, + "loss": 0.135, + "step": 10233 + }, + { + "epoch": 28.58659217877095, + "grad_norm": 0.38759151101112366, + "learning_rate": 2.952364990689013e-05, + "loss": 0.13, + "step": 10234 + }, + { + "epoch": 28.589385474860336, + "grad_norm": 0.7721878290176392, + "learning_rate": 2.9523277467411547e-05, + "loss": 0.1481, + "step": 10235 + }, + { + "epoch": 28.592178770949722, + "grad_norm": 0.5642040967941284, + "learning_rate": 2.9522905027932963e-05, + "loss": 0.1386, + "step": 10236 + }, + { + "epoch": 28.594972067039105, + "grad_norm": 1.1349173784255981, + "learning_rate": 2.952253258845438e-05, + "loss": 0.1243, + "step": 10237 + }, + { + "epoch": 28.59776536312849, + "grad_norm": 2.0046565532684326, + "learning_rate": 2.952216014897579e-05, + "loss": 0.1376, + "step": 10238 + }, + { + "epoch": 28.600558659217878, + "grad_norm": 0.5481731295585632, + "learning_rate": 2.9521787709497206e-05, + "loss": 0.1297, + "step": 10239 + }, + { + "epoch": 28.60335195530726, + "grad_norm": 1.435805082321167, + "learning_rate": 2.9521415270018623e-05, + "loss": 0.1031, + "step": 10240 + }, + { + "epoch": 28.606145251396647, + "grad_norm": 0.842072069644928, + "learning_rate": 2.952104283054004e-05, + "loss": 0.1372, + "step": 10241 + }, + { + "epoch": 28.608938547486034, + "grad_norm": 1.8337695598602295, + "learning_rate": 2.9520670391061456e-05, + "loss": 0.1323, + "step": 10242 + }, + { + "epoch": 28.61173184357542, + "grad_norm": 0.5836145281791687, + "learning_rate": 2.952029795158287e-05, + "loss": 0.1276, + "step": 10243 + }, + { + "epoch": 28.614525139664803, + "grad_norm": 0.8080611228942871, + "learning_rate": 2.9519925512104282e-05, + "loss": 0.1239, + "step": 10244 + }, + { + "epoch": 28.61731843575419, + "grad_norm": 1.05278742313385, + "learning_rate": 2.95195530726257e-05, + "loss": 0.1356, + "step": 10245 + }, + { + "epoch": 28.620111731843576, + "grad_norm": 1.1756367683410645, + "learning_rate": 2.9519180633147115e-05, + "loss": 0.155, + "step": 10246 + }, + { + "epoch": 28.622905027932962, + "grad_norm": 0.9928547739982605, + "learning_rate": 2.9518808193668528e-05, + "loss": 0.1204, + "step": 10247 + }, + { + "epoch": 28.625698324022345, + "grad_norm": 1.1023895740509033, + "learning_rate": 2.9518435754189945e-05, + "loss": 0.155, + "step": 10248 + }, + { + "epoch": 28.62849162011173, + "grad_norm": 8.347577095031738, + "learning_rate": 2.951806331471136e-05, + "loss": 0.1424, + "step": 10249 + }, + { + "epoch": 28.631284916201118, + "grad_norm": 0.452975332736969, + "learning_rate": 2.9517690875232778e-05, + "loss": 0.1452, + "step": 10250 + }, + { + "epoch": 28.634078212290504, + "grad_norm": 0.5034193992614746, + "learning_rate": 2.951731843575419e-05, + "loss": 0.176, + "step": 10251 + }, + { + "epoch": 28.636871508379887, + "grad_norm": 0.4744928777217865, + "learning_rate": 2.9516945996275604e-05, + "loss": 0.1636, + "step": 10252 + }, + { + "epoch": 28.639664804469273, + "grad_norm": 0.4677009880542755, + "learning_rate": 2.951657355679702e-05, + "loss": 0.1478, + "step": 10253 + }, + { + "epoch": 28.64245810055866, + "grad_norm": 0.7895140647888184, + "learning_rate": 2.9516201117318437e-05, + "loss": 0.1675, + "step": 10254 + }, + { + "epoch": 28.645251396648046, + "grad_norm": 0.6981340646743774, + "learning_rate": 2.9515828677839854e-05, + "loss": 0.165, + "step": 10255 + }, + { + "epoch": 28.64804469273743, + "grad_norm": 0.5273815393447876, + "learning_rate": 2.9515456238361267e-05, + "loss": 0.1704, + "step": 10256 + }, + { + "epoch": 28.650837988826815, + "grad_norm": 0.40680208802223206, + "learning_rate": 2.9515083798882683e-05, + "loss": 0.1173, + "step": 10257 + }, + { + "epoch": 28.6536312849162, + "grad_norm": 0.5272083878517151, + "learning_rate": 2.9514711359404096e-05, + "loss": 0.1499, + "step": 10258 + }, + { + "epoch": 28.656424581005588, + "grad_norm": 0.5854878425598145, + "learning_rate": 2.9514338919925513e-05, + "loss": 0.1373, + "step": 10259 + }, + { + "epoch": 28.65921787709497, + "grad_norm": 0.7564172148704529, + "learning_rate": 2.9513966480446926e-05, + "loss": 0.1591, + "step": 10260 + }, + { + "epoch": 28.662011173184357, + "grad_norm": 0.5609917640686035, + "learning_rate": 2.9513594040968343e-05, + "loss": 0.136, + "step": 10261 + }, + { + "epoch": 28.664804469273744, + "grad_norm": 0.9410874247550964, + "learning_rate": 2.951322160148976e-05, + "loss": 0.1309, + "step": 10262 + }, + { + "epoch": 28.66759776536313, + "grad_norm": 0.5666813850402832, + "learning_rate": 2.9512849162011176e-05, + "loss": 0.1353, + "step": 10263 + }, + { + "epoch": 28.670391061452513, + "grad_norm": 0.5571916103363037, + "learning_rate": 2.951247672253259e-05, + "loss": 0.1232, + "step": 10264 + }, + { + "epoch": 28.6731843575419, + "grad_norm": 0.7614725232124329, + "learning_rate": 2.9512104283054002e-05, + "loss": 0.1352, + "step": 10265 + }, + { + "epoch": 28.675977653631286, + "grad_norm": 0.47381919622421265, + "learning_rate": 2.951173184357542e-05, + "loss": 0.1228, + "step": 10266 + }, + { + "epoch": 28.678770949720672, + "grad_norm": 0.684596598148346, + "learning_rate": 2.9511359404096835e-05, + "loss": 0.1178, + "step": 10267 + }, + { + "epoch": 28.681564245810055, + "grad_norm": 0.8044804334640503, + "learning_rate": 2.951098696461825e-05, + "loss": 0.1117, + "step": 10268 + }, + { + "epoch": 28.68435754189944, + "grad_norm": 0.661723792552948, + "learning_rate": 2.9510614525139665e-05, + "loss": 0.1349, + "step": 10269 + }, + { + "epoch": 28.687150837988828, + "grad_norm": 0.7884909510612488, + "learning_rate": 2.951024208566108e-05, + "loss": 0.1203, + "step": 10270 + }, + { + "epoch": 28.689944134078214, + "grad_norm": 1.4866797924041748, + "learning_rate": 2.9509869646182494e-05, + "loss": 0.1381, + "step": 10271 + }, + { + "epoch": 28.692737430167597, + "grad_norm": 0.9290087223052979, + "learning_rate": 2.950949720670391e-05, + "loss": 0.1298, + "step": 10272 + }, + { + "epoch": 28.695530726256983, + "grad_norm": 0.87422114610672, + "learning_rate": 2.9509124767225328e-05, + "loss": 0.1463, + "step": 10273 + }, + { + "epoch": 28.69832402234637, + "grad_norm": 1.0770938396453857, + "learning_rate": 2.950875232774674e-05, + "loss": 0.1727, + "step": 10274 + }, + { + "epoch": 28.701117318435756, + "grad_norm": 0.47939372062683105, + "learning_rate": 2.9508379888268157e-05, + "loss": 0.1483, + "step": 10275 + }, + { + "epoch": 28.70391061452514, + "grad_norm": 1.3682209253311157, + "learning_rate": 2.9508007448789574e-05, + "loss": 0.1637, + "step": 10276 + }, + { + "epoch": 28.706703910614525, + "grad_norm": 0.5597934722900391, + "learning_rate": 2.950763500931099e-05, + "loss": 0.1433, + "step": 10277 + }, + { + "epoch": 28.70949720670391, + "grad_norm": 0.5782397985458374, + "learning_rate": 2.95072625698324e-05, + "loss": 0.1464, + "step": 10278 + }, + { + "epoch": 28.712290502793294, + "grad_norm": 0.5501046180725098, + "learning_rate": 2.9506890130353817e-05, + "loss": 0.1588, + "step": 10279 + }, + { + "epoch": 28.71508379888268, + "grad_norm": 0.6879538893699646, + "learning_rate": 2.9506517690875233e-05, + "loss": 0.1634, + "step": 10280 + }, + { + "epoch": 28.717877094972067, + "grad_norm": 0.575212836265564, + "learning_rate": 2.950614525139665e-05, + "loss": 0.1559, + "step": 10281 + }, + { + "epoch": 28.720670391061454, + "grad_norm": 0.4788237512111664, + "learning_rate": 2.9505772811918066e-05, + "loss": 0.1328, + "step": 10282 + }, + { + "epoch": 28.723463687150836, + "grad_norm": 2.3124842643737793, + "learning_rate": 2.950540037243948e-05, + "loss": 0.1283, + "step": 10283 + }, + { + "epoch": 28.726256983240223, + "grad_norm": 0.502138614654541, + "learning_rate": 2.9505027932960892e-05, + "loss": 0.1649, + "step": 10284 + }, + { + "epoch": 28.72905027932961, + "grad_norm": 0.47984594106674194, + "learning_rate": 2.950465549348231e-05, + "loss": 0.1574, + "step": 10285 + }, + { + "epoch": 28.731843575418996, + "grad_norm": 0.8810647130012512, + "learning_rate": 2.9504283054003726e-05, + "loss": 0.1519, + "step": 10286 + }, + { + "epoch": 28.73463687150838, + "grad_norm": 0.4731312394142151, + "learning_rate": 2.950391061452514e-05, + "loss": 0.1234, + "step": 10287 + }, + { + "epoch": 28.737430167597765, + "grad_norm": 1.5101064443588257, + "learning_rate": 2.9503538175046555e-05, + "loss": 0.1572, + "step": 10288 + }, + { + "epoch": 28.74022346368715, + "grad_norm": 1.1682133674621582, + "learning_rate": 2.9503165735567972e-05, + "loss": 0.1412, + "step": 10289 + }, + { + "epoch": 28.743016759776538, + "grad_norm": 0.5376226305961609, + "learning_rate": 2.950279329608939e-05, + "loss": 0.143, + "step": 10290 + }, + { + "epoch": 28.74581005586592, + "grad_norm": 0.7914925217628479, + "learning_rate": 2.95024208566108e-05, + "loss": 0.1241, + "step": 10291 + }, + { + "epoch": 28.748603351955307, + "grad_norm": 0.7695657014846802, + "learning_rate": 2.9502048417132215e-05, + "loss": 0.1425, + "step": 10292 + }, + { + "epoch": 28.751396648044693, + "grad_norm": 1.4194029569625854, + "learning_rate": 2.950167597765363e-05, + "loss": 0.1145, + "step": 10293 + }, + { + "epoch": 28.75418994413408, + "grad_norm": 1.0597848892211914, + "learning_rate": 2.9501303538175048e-05, + "loss": 0.1262, + "step": 10294 + }, + { + "epoch": 28.756983240223462, + "grad_norm": 0.5770659446716309, + "learning_rate": 2.9500931098696464e-05, + "loss": 0.115, + "step": 10295 + }, + { + "epoch": 28.75977653631285, + "grad_norm": 1.6571372747421265, + "learning_rate": 2.9500558659217877e-05, + "loss": 0.1428, + "step": 10296 + }, + { + "epoch": 28.762569832402235, + "grad_norm": 0.7750247120857239, + "learning_rate": 2.9500186219739294e-05, + "loss": 0.127, + "step": 10297 + }, + { + "epoch": 28.76536312849162, + "grad_norm": 1.028030514717102, + "learning_rate": 2.9499813780260707e-05, + "loss": 0.1349, + "step": 10298 + }, + { + "epoch": 28.768156424581004, + "grad_norm": 1.4402050971984863, + "learning_rate": 2.9499441340782124e-05, + "loss": 0.2029, + "step": 10299 + }, + { + "epoch": 28.77094972067039, + "grad_norm": 0.5793277025222778, + "learning_rate": 2.949906890130354e-05, + "loss": 0.159, + "step": 10300 + }, + { + "epoch": 28.773743016759777, + "grad_norm": 0.9445043802261353, + "learning_rate": 2.9498696461824953e-05, + "loss": 0.1641, + "step": 10301 + }, + { + "epoch": 28.776536312849164, + "grad_norm": 0.4224027693271637, + "learning_rate": 2.949832402234637e-05, + "loss": 0.1591, + "step": 10302 + }, + { + "epoch": 28.779329608938546, + "grad_norm": 0.4538438320159912, + "learning_rate": 2.9497951582867786e-05, + "loss": 0.1451, + "step": 10303 + }, + { + "epoch": 28.782122905027933, + "grad_norm": 0.5686238408088684, + "learning_rate": 2.94975791433892e-05, + "loss": 0.1604, + "step": 10304 + }, + { + "epoch": 28.78491620111732, + "grad_norm": 0.6568220257759094, + "learning_rate": 2.9497206703910613e-05, + "loss": 0.1478, + "step": 10305 + }, + { + "epoch": 28.787709497206706, + "grad_norm": 0.7868368625640869, + "learning_rate": 2.949683426443203e-05, + "loss": 0.1446, + "step": 10306 + }, + { + "epoch": 28.79050279329609, + "grad_norm": 2.0749025344848633, + "learning_rate": 2.9496461824953446e-05, + "loss": 0.1658, + "step": 10307 + }, + { + "epoch": 28.793296089385475, + "grad_norm": 0.5366005897521973, + "learning_rate": 2.9496089385474862e-05, + "loss": 0.1443, + "step": 10308 + }, + { + "epoch": 28.79608938547486, + "grad_norm": 0.8822179436683655, + "learning_rate": 2.949571694599628e-05, + "loss": 0.1581, + "step": 10309 + }, + { + "epoch": 28.798882681564244, + "grad_norm": 0.4116005003452301, + "learning_rate": 2.9495344506517692e-05, + "loss": 0.1366, + "step": 10310 + }, + { + "epoch": 28.80167597765363, + "grad_norm": 0.6376591920852661, + "learning_rate": 2.9494972067039105e-05, + "loss": 0.1206, + "step": 10311 + }, + { + "epoch": 28.804469273743017, + "grad_norm": 1.0408825874328613, + "learning_rate": 2.949459962756052e-05, + "loss": 0.1183, + "step": 10312 + }, + { + "epoch": 28.807262569832403, + "grad_norm": 0.6636953949928284, + "learning_rate": 2.9494227188081938e-05, + "loss": 0.1438, + "step": 10313 + }, + { + "epoch": 28.810055865921786, + "grad_norm": 0.7860391139984131, + "learning_rate": 2.949385474860335e-05, + "loss": 0.1226, + "step": 10314 + }, + { + "epoch": 28.812849162011172, + "grad_norm": 1.1945557594299316, + "learning_rate": 2.9493482309124768e-05, + "loss": 0.1458, + "step": 10315 + }, + { + "epoch": 28.81564245810056, + "grad_norm": 1.2006547451019287, + "learning_rate": 2.9493109869646184e-05, + "loss": 0.1677, + "step": 10316 + }, + { + "epoch": 28.818435754189945, + "grad_norm": 0.5961652398109436, + "learning_rate": 2.94927374301676e-05, + "loss": 0.1091, + "step": 10317 + }, + { + "epoch": 28.821229050279328, + "grad_norm": 2.041992425918579, + "learning_rate": 2.9492364990689014e-05, + "loss": 0.1686, + "step": 10318 + }, + { + "epoch": 28.824022346368714, + "grad_norm": 0.8380600810050964, + "learning_rate": 2.9491992551210427e-05, + "loss": 0.1356, + "step": 10319 + }, + { + "epoch": 28.8268156424581, + "grad_norm": 2.5354561805725098, + "learning_rate": 2.9491620111731844e-05, + "loss": 0.1528, + "step": 10320 + }, + { + "epoch": 28.829608938547487, + "grad_norm": 1.4748072624206543, + "learning_rate": 2.949124767225326e-05, + "loss": 0.1378, + "step": 10321 + }, + { + "epoch": 28.83240223463687, + "grad_norm": 0.6131622195243835, + "learning_rate": 2.9490875232774677e-05, + "loss": 0.1342, + "step": 10322 + }, + { + "epoch": 28.835195530726256, + "grad_norm": 1.1532524824142456, + "learning_rate": 2.949050279329609e-05, + "loss": 0.1589, + "step": 10323 + }, + { + "epoch": 28.837988826815643, + "grad_norm": 1.4136549234390259, + "learning_rate": 2.9490130353817503e-05, + "loss": 0.1953, + "step": 10324 + }, + { + "epoch": 28.84078212290503, + "grad_norm": 0.3305223286151886, + "learning_rate": 2.948975791433892e-05, + "loss": 0.147, + "step": 10325 + }, + { + "epoch": 28.843575418994412, + "grad_norm": 0.43520230054855347, + "learning_rate": 2.9489385474860336e-05, + "loss": 0.1438, + "step": 10326 + }, + { + "epoch": 28.8463687150838, + "grad_norm": 0.4971548914909363, + "learning_rate": 2.9489013035381753e-05, + "loss": 0.1566, + "step": 10327 + }, + { + "epoch": 28.849162011173185, + "grad_norm": 0.6115166544914246, + "learning_rate": 2.9488640595903166e-05, + "loss": 0.1411, + "step": 10328 + }, + { + "epoch": 28.85195530726257, + "grad_norm": 0.5286633968353271, + "learning_rate": 2.9488268156424582e-05, + "loss": 0.1601, + "step": 10329 + }, + { + "epoch": 28.854748603351954, + "grad_norm": 0.49194470047950745, + "learning_rate": 2.9487895716946e-05, + "loss": 0.14, + "step": 10330 + }, + { + "epoch": 28.85754189944134, + "grad_norm": 0.4243936836719513, + "learning_rate": 2.9487523277467412e-05, + "loss": 0.154, + "step": 10331 + }, + { + "epoch": 28.860335195530727, + "grad_norm": 0.4914831221103668, + "learning_rate": 2.9487150837988825e-05, + "loss": 0.1505, + "step": 10332 + }, + { + "epoch": 28.863128491620113, + "grad_norm": 0.517983615398407, + "learning_rate": 2.9486778398510242e-05, + "loss": 0.1384, + "step": 10333 + }, + { + "epoch": 28.865921787709496, + "grad_norm": 0.672977089881897, + "learning_rate": 2.9486405959031658e-05, + "loss": 0.1545, + "step": 10334 + }, + { + "epoch": 28.868715083798882, + "grad_norm": 0.6163352727890015, + "learning_rate": 2.9486033519553075e-05, + "loss": 0.1268, + "step": 10335 + }, + { + "epoch": 28.87150837988827, + "grad_norm": 0.5915514230728149, + "learning_rate": 2.948566108007449e-05, + "loss": 0.1359, + "step": 10336 + }, + { + "epoch": 28.874301675977655, + "grad_norm": 0.5258777141571045, + "learning_rate": 2.9485288640595904e-05, + "loss": 0.1208, + "step": 10337 + }, + { + "epoch": 28.877094972067038, + "grad_norm": 1.1718382835388184, + "learning_rate": 2.9484916201117318e-05, + "loss": 0.128, + "step": 10338 + }, + { + "epoch": 28.879888268156424, + "grad_norm": 0.6044517159461975, + "learning_rate": 2.9484543761638734e-05, + "loss": 0.1451, + "step": 10339 + }, + { + "epoch": 28.88268156424581, + "grad_norm": 0.613288938999176, + "learning_rate": 2.948417132216015e-05, + "loss": 0.1412, + "step": 10340 + }, + { + "epoch": 28.885474860335197, + "grad_norm": 0.7448855638504028, + "learning_rate": 2.9483798882681564e-05, + "loss": 0.1523, + "step": 10341 + }, + { + "epoch": 28.88826815642458, + "grad_norm": 1.2593276500701904, + "learning_rate": 2.948342644320298e-05, + "loss": 0.1329, + "step": 10342 + }, + { + "epoch": 28.891061452513966, + "grad_norm": 0.865999698638916, + "learning_rate": 2.9483054003724397e-05, + "loss": 0.1368, + "step": 10343 + }, + { + "epoch": 28.893854748603353, + "grad_norm": 0.5864124298095703, + "learning_rate": 2.948268156424581e-05, + "loss": 0.133, + "step": 10344 + }, + { + "epoch": 28.89664804469274, + "grad_norm": 0.9003466963768005, + "learning_rate": 2.9482309124767227e-05, + "loss": 0.13, + "step": 10345 + }, + { + "epoch": 28.899441340782122, + "grad_norm": 0.5973618626594543, + "learning_rate": 2.948193668528864e-05, + "loss": 0.1328, + "step": 10346 + }, + { + "epoch": 28.90223463687151, + "grad_norm": 0.7774901986122131, + "learning_rate": 2.9481564245810056e-05, + "loss": 0.1556, + "step": 10347 + }, + { + "epoch": 28.905027932960895, + "grad_norm": 0.6658979654312134, + "learning_rate": 2.9481191806331473e-05, + "loss": 0.1353, + "step": 10348 + }, + { + "epoch": 28.907821229050278, + "grad_norm": 1.5821760892868042, + "learning_rate": 2.948081936685289e-05, + "loss": 0.2037, + "step": 10349 + }, + { + "epoch": 28.910614525139664, + "grad_norm": 0.4488055109977722, + "learning_rate": 2.9480446927374303e-05, + "loss": 0.1631, + "step": 10350 + }, + { + "epoch": 28.91340782122905, + "grad_norm": 0.33243170380592346, + "learning_rate": 2.9480074487895716e-05, + "loss": 0.1336, + "step": 10351 + }, + { + "epoch": 28.916201117318437, + "grad_norm": 0.814609706401825, + "learning_rate": 2.9479702048417132e-05, + "loss": 0.1342, + "step": 10352 + }, + { + "epoch": 28.91899441340782, + "grad_norm": 0.45232507586479187, + "learning_rate": 2.947932960893855e-05, + "loss": 0.1514, + "step": 10353 + }, + { + "epoch": 28.921787709497206, + "grad_norm": 0.4895811378955841, + "learning_rate": 2.9478957169459962e-05, + "loss": 0.1452, + "step": 10354 + }, + { + "epoch": 28.924581005586592, + "grad_norm": 0.669699490070343, + "learning_rate": 2.947858472998138e-05, + "loss": 0.1674, + "step": 10355 + }, + { + "epoch": 28.92737430167598, + "grad_norm": 0.7535391449928284, + "learning_rate": 2.9478212290502795e-05, + "loss": 0.1564, + "step": 10356 + }, + { + "epoch": 28.93016759776536, + "grad_norm": 0.5268202424049377, + "learning_rate": 2.947783985102421e-05, + "loss": 0.1412, + "step": 10357 + }, + { + "epoch": 28.932960893854748, + "grad_norm": 0.5883979797363281, + "learning_rate": 2.9477467411545625e-05, + "loss": 0.1302, + "step": 10358 + }, + { + "epoch": 28.935754189944134, + "grad_norm": 0.724997878074646, + "learning_rate": 2.9477094972067038e-05, + "loss": 0.1602, + "step": 10359 + }, + { + "epoch": 28.93854748603352, + "grad_norm": 0.7068378925323486, + "learning_rate": 2.9476722532588454e-05, + "loss": 0.1506, + "step": 10360 + }, + { + "epoch": 28.941340782122904, + "grad_norm": 0.5711111426353455, + "learning_rate": 2.947635009310987e-05, + "loss": 0.1379, + "step": 10361 + }, + { + "epoch": 28.94413407821229, + "grad_norm": 0.7274206876754761, + "learning_rate": 2.9475977653631287e-05, + "loss": 0.1646, + "step": 10362 + }, + { + "epoch": 28.946927374301676, + "grad_norm": 0.5935824513435364, + "learning_rate": 2.94756052141527e-05, + "loss": 0.1276, + "step": 10363 + }, + { + "epoch": 28.949720670391063, + "grad_norm": 0.43059441447257996, + "learning_rate": 2.9475232774674114e-05, + "loss": 0.1123, + "step": 10364 + }, + { + "epoch": 28.952513966480446, + "grad_norm": 0.6307093501091003, + "learning_rate": 2.947486033519553e-05, + "loss": 0.1354, + "step": 10365 + }, + { + "epoch": 28.955307262569832, + "grad_norm": 0.5249803066253662, + "learning_rate": 2.9474487895716947e-05, + "loss": 0.121, + "step": 10366 + }, + { + "epoch": 28.95810055865922, + "grad_norm": 0.7905276417732239, + "learning_rate": 2.9474115456238363e-05, + "loss": 0.1301, + "step": 10367 + }, + { + "epoch": 28.960893854748605, + "grad_norm": 0.8280601501464844, + "learning_rate": 2.9473743016759776e-05, + "loss": 0.1361, + "step": 10368 + }, + { + "epoch": 28.963687150837988, + "grad_norm": 0.8811527490615845, + "learning_rate": 2.9473370577281193e-05, + "loss": 0.1264, + "step": 10369 + }, + { + "epoch": 28.966480446927374, + "grad_norm": 0.9165343046188354, + "learning_rate": 2.947299813780261e-05, + "loss": 0.1381, + "step": 10370 + }, + { + "epoch": 28.96927374301676, + "grad_norm": 1.6066055297851562, + "learning_rate": 2.9472625698324023e-05, + "loss": 0.1234, + "step": 10371 + }, + { + "epoch": 28.972067039106147, + "grad_norm": 1.0050272941589355, + "learning_rate": 2.9472253258845436e-05, + "loss": 0.1532, + "step": 10372 + }, + { + "epoch": 28.97486033519553, + "grad_norm": 1.2838335037231445, + "learning_rate": 2.9471880819366852e-05, + "loss": 0.1643, + "step": 10373 + }, + { + "epoch": 28.977653631284916, + "grad_norm": 1.0902066230773926, + "learning_rate": 2.947150837988827e-05, + "loss": 0.1659, + "step": 10374 + }, + { + "epoch": 28.980446927374302, + "grad_norm": 5.145961761474609, + "learning_rate": 2.9471135940409685e-05, + "loss": 0.1495, + "step": 10375 + }, + { + "epoch": 28.98324022346369, + "grad_norm": 0.4343758225440979, + "learning_rate": 2.9470763500931102e-05, + "loss": 0.1422, + "step": 10376 + }, + { + "epoch": 28.98603351955307, + "grad_norm": 0.6984539031982422, + "learning_rate": 2.9470391061452515e-05, + "loss": 0.1434, + "step": 10377 + }, + { + "epoch": 28.988826815642458, + "grad_norm": 0.5893527865409851, + "learning_rate": 2.9470018621973928e-05, + "loss": 0.1296, + "step": 10378 + }, + { + "epoch": 28.991620111731844, + "grad_norm": 0.5456815958023071, + "learning_rate": 2.9469646182495345e-05, + "loss": 0.1354, + "step": 10379 + }, + { + "epoch": 28.994413407821227, + "grad_norm": 1.3092902898788452, + "learning_rate": 2.946927374301676e-05, + "loss": 0.1136, + "step": 10380 + }, + { + "epoch": 28.997206703910614, + "grad_norm": 1.9374045133590698, + "learning_rate": 2.9468901303538174e-05, + "loss": 0.1166, + "step": 10381 + }, + { + "epoch": 29.0, + "grad_norm": 1.551430106163025, + "learning_rate": 2.946852886405959e-05, + "loss": 0.1885, + "step": 10382 + }, + { + "epoch": 29.002793296089386, + "grad_norm": 0.30405333638191223, + "learning_rate": 2.9468156424581008e-05, + "loss": 0.1328, + "step": 10383 + }, + { + "epoch": 29.00558659217877, + "grad_norm": 0.8938578367233276, + "learning_rate": 2.946778398510242e-05, + "loss": 0.1572, + "step": 10384 + }, + { + "epoch": 29.008379888268156, + "grad_norm": 0.5532771944999695, + "learning_rate": 2.9467411545623837e-05, + "loss": 0.1506, + "step": 10385 + }, + { + "epoch": 29.011173184357542, + "grad_norm": 0.38460251688957214, + "learning_rate": 2.946703910614525e-05, + "loss": 0.1219, + "step": 10386 + }, + { + "epoch": 29.01396648044693, + "grad_norm": 0.4696371853351593, + "learning_rate": 2.9466666666666667e-05, + "loss": 0.1395, + "step": 10387 + }, + { + "epoch": 29.01675977653631, + "grad_norm": 0.837220311164856, + "learning_rate": 2.9466294227188083e-05, + "loss": 0.1526, + "step": 10388 + }, + { + "epoch": 29.019553072625698, + "grad_norm": 0.490579754114151, + "learning_rate": 2.94659217877095e-05, + "loss": 0.1424, + "step": 10389 + }, + { + "epoch": 29.022346368715084, + "grad_norm": 0.7525264024734497, + "learning_rate": 2.9465549348230913e-05, + "loss": 0.109, + "step": 10390 + }, + { + "epoch": 29.02513966480447, + "grad_norm": 0.3585844933986664, + "learning_rate": 2.9465176908752326e-05, + "loss": 0.1159, + "step": 10391 + }, + { + "epoch": 29.027932960893853, + "grad_norm": 0.7512149214744568, + "learning_rate": 2.9464804469273743e-05, + "loss": 0.1334, + "step": 10392 + }, + { + "epoch": 29.03072625698324, + "grad_norm": 0.4921032190322876, + "learning_rate": 2.946443202979516e-05, + "loss": 0.1437, + "step": 10393 + }, + { + "epoch": 29.033519553072626, + "grad_norm": 0.6376190185546875, + "learning_rate": 2.9464059590316576e-05, + "loss": 0.1452, + "step": 10394 + }, + { + "epoch": 29.036312849162012, + "grad_norm": 0.6728608012199402, + "learning_rate": 2.946368715083799e-05, + "loss": 0.1155, + "step": 10395 + }, + { + "epoch": 29.039106145251395, + "grad_norm": 0.4851863384246826, + "learning_rate": 2.9463314711359406e-05, + "loss": 0.1339, + "step": 10396 + }, + { + "epoch": 29.04189944134078, + "grad_norm": 0.6244794726371765, + "learning_rate": 2.9462942271880822e-05, + "loss": 0.1159, + "step": 10397 + }, + { + "epoch": 29.044692737430168, + "grad_norm": 0.8516942262649536, + "learning_rate": 2.9462569832402235e-05, + "loss": 0.1253, + "step": 10398 + }, + { + "epoch": 29.047486033519554, + "grad_norm": 1.070511817932129, + "learning_rate": 2.946219739292365e-05, + "loss": 0.1545, + "step": 10399 + }, + { + "epoch": 29.050279329608937, + "grad_norm": 0.7163169384002686, + "learning_rate": 2.9461824953445065e-05, + "loss": 0.1332, + "step": 10400 + }, + { + "epoch": 29.053072625698324, + "grad_norm": 0.9092696309089661, + "learning_rate": 2.946145251396648e-05, + "loss": 0.1142, + "step": 10401 + }, + { + "epoch": 29.05586592178771, + "grad_norm": 0.5673463940620422, + "learning_rate": 2.9461080074487898e-05, + "loss": 0.1137, + "step": 10402 + }, + { + "epoch": 29.058659217877096, + "grad_norm": 1.0644291639328003, + "learning_rate": 2.9460707635009315e-05, + "loss": 0.1155, + "step": 10403 + }, + { + "epoch": 29.06145251396648, + "grad_norm": 3.5724754333496094, + "learning_rate": 2.9460335195530724e-05, + "loss": 0.1115, + "step": 10404 + }, + { + "epoch": 29.064245810055866, + "grad_norm": 0.9343311786651611, + "learning_rate": 2.945996275605214e-05, + "loss": 0.1626, + "step": 10405 + }, + { + "epoch": 29.067039106145252, + "grad_norm": 13.728198051452637, + "learning_rate": 2.9459590316573557e-05, + "loss": 0.1382, + "step": 10406 + }, + { + "epoch": 29.06983240223464, + "grad_norm": 1.2744945287704468, + "learning_rate": 2.9459217877094974e-05, + "loss": 0.1544, + "step": 10407 + }, + { + "epoch": 29.07262569832402, + "grad_norm": 0.38809922337532043, + "learning_rate": 2.9458845437616387e-05, + "loss": 0.1452, + "step": 10408 + }, + { + "epoch": 29.075418994413408, + "grad_norm": 0.811974048614502, + "learning_rate": 2.9458472998137804e-05, + "loss": 0.1382, + "step": 10409 + }, + { + "epoch": 29.078212290502794, + "grad_norm": 0.516060471534729, + "learning_rate": 2.945810055865922e-05, + "loss": 0.1452, + "step": 10410 + }, + { + "epoch": 29.08100558659218, + "grad_norm": 2.638909339904785, + "learning_rate": 2.9457728119180633e-05, + "loss": 0.1502, + "step": 10411 + }, + { + "epoch": 29.083798882681563, + "grad_norm": 0.5373002290725708, + "learning_rate": 2.945735567970205e-05, + "loss": 0.1362, + "step": 10412 + }, + { + "epoch": 29.08659217877095, + "grad_norm": 0.5613301396369934, + "learning_rate": 2.9456983240223463e-05, + "loss": 0.1525, + "step": 10413 + }, + { + "epoch": 29.089385474860336, + "grad_norm": 0.4493902325630188, + "learning_rate": 2.945661080074488e-05, + "loss": 0.1464, + "step": 10414 + }, + { + "epoch": 29.092178770949722, + "grad_norm": 1.3013583421707153, + "learning_rate": 2.9456238361266296e-05, + "loss": 0.1349, + "step": 10415 + }, + { + "epoch": 29.094972067039105, + "grad_norm": 0.4774348735809326, + "learning_rate": 2.9455865921787713e-05, + "loss": 0.1476, + "step": 10416 + }, + { + "epoch": 29.09776536312849, + "grad_norm": 0.5906072854995728, + "learning_rate": 2.9455493482309126e-05, + "loss": 0.1564, + "step": 10417 + }, + { + "epoch": 29.100558659217878, + "grad_norm": 0.5502637624740601, + "learning_rate": 2.945512104283054e-05, + "loss": 0.1237, + "step": 10418 + }, + { + "epoch": 29.10335195530726, + "grad_norm": 0.4481019675731659, + "learning_rate": 2.9454748603351955e-05, + "loss": 0.116, + "step": 10419 + }, + { + "epoch": 29.106145251396647, + "grad_norm": 0.5967357754707336, + "learning_rate": 2.9454376163873372e-05, + "loss": 0.1451, + "step": 10420 + }, + { + "epoch": 29.108938547486034, + "grad_norm": 0.6565940380096436, + "learning_rate": 2.945400372439479e-05, + "loss": 0.13, + "step": 10421 + }, + { + "epoch": 29.11173184357542, + "grad_norm": 1.0652837753295898, + "learning_rate": 2.94536312849162e-05, + "loss": 0.1177, + "step": 10422 + }, + { + "epoch": 29.114525139664803, + "grad_norm": 0.5982703566551208, + "learning_rate": 2.9453258845437618e-05, + "loss": 0.1251, + "step": 10423 + }, + { + "epoch": 29.11731843575419, + "grad_norm": 0.8810552358627319, + "learning_rate": 2.945288640595903e-05, + "loss": 0.1257, + "step": 10424 + }, + { + "epoch": 29.120111731843576, + "grad_norm": 0.637971818447113, + "learning_rate": 2.9452513966480448e-05, + "loss": 0.155, + "step": 10425 + }, + { + "epoch": 29.122905027932962, + "grad_norm": 1.129172921180725, + "learning_rate": 2.945214152700186e-05, + "loss": 0.1072, + "step": 10426 + }, + { + "epoch": 29.125698324022345, + "grad_norm": 0.7237590551376343, + "learning_rate": 2.9451769087523277e-05, + "loss": 0.1376, + "step": 10427 + }, + { + "epoch": 29.12849162011173, + "grad_norm": 0.6782786846160889, + "learning_rate": 2.9451396648044694e-05, + "loss": 0.1384, + "step": 10428 + }, + { + "epoch": 29.131284916201118, + "grad_norm": 0.954495906829834, + "learning_rate": 2.945102420856611e-05, + "loss": 0.1503, + "step": 10429 + }, + { + "epoch": 29.134078212290504, + "grad_norm": 3.077610969543457, + "learning_rate": 2.9450651769087527e-05, + "loss": 0.1396, + "step": 10430 + }, + { + "epoch": 29.136871508379887, + "grad_norm": 1.211439847946167, + "learning_rate": 2.9450279329608937e-05, + "loss": 0.1443, + "step": 10431 + }, + { + "epoch": 29.139664804469273, + "grad_norm": 1.9293723106384277, + "learning_rate": 2.9449906890130353e-05, + "loss": 0.1764, + "step": 10432 + }, + { + "epoch": 29.14245810055866, + "grad_norm": 0.44011390209198, + "learning_rate": 2.944953445065177e-05, + "loss": 0.1627, + "step": 10433 + }, + { + "epoch": 29.145251396648046, + "grad_norm": 0.6133772730827332, + "learning_rate": 2.9449162011173186e-05, + "loss": 0.1389, + "step": 10434 + }, + { + "epoch": 29.14804469273743, + "grad_norm": 0.5923448801040649, + "learning_rate": 2.94487895716946e-05, + "loss": 0.1507, + "step": 10435 + }, + { + "epoch": 29.150837988826815, + "grad_norm": 0.4464341700077057, + "learning_rate": 2.9448417132216016e-05, + "loss": 0.1445, + "step": 10436 + }, + { + "epoch": 29.1536312849162, + "grad_norm": 0.3693169057369232, + "learning_rate": 2.9448044692737433e-05, + "loss": 0.1227, + "step": 10437 + }, + { + "epoch": 29.156424581005588, + "grad_norm": 0.531552791595459, + "learning_rate": 2.9447672253258846e-05, + "loss": 0.1582, + "step": 10438 + }, + { + "epoch": 29.15921787709497, + "grad_norm": 0.5403973460197449, + "learning_rate": 2.9447299813780262e-05, + "loss": 0.1324, + "step": 10439 + }, + { + "epoch": 29.162011173184357, + "grad_norm": 0.7266975045204163, + "learning_rate": 2.9446927374301676e-05, + "loss": 0.1422, + "step": 10440 + }, + { + "epoch": 29.164804469273744, + "grad_norm": 0.43875765800476074, + "learning_rate": 2.9446554934823092e-05, + "loss": 0.1309, + "step": 10441 + }, + { + "epoch": 29.16759776536313, + "grad_norm": 0.463179349899292, + "learning_rate": 2.944618249534451e-05, + "loss": 0.1539, + "step": 10442 + }, + { + "epoch": 29.170391061452513, + "grad_norm": 1.0357532501220703, + "learning_rate": 2.9445810055865925e-05, + "loss": 0.1462, + "step": 10443 + }, + { + "epoch": 29.1731843575419, + "grad_norm": 0.5153865814208984, + "learning_rate": 2.9445437616387335e-05, + "loss": 0.1366, + "step": 10444 + }, + { + "epoch": 29.175977653631286, + "grad_norm": 0.6177476048469543, + "learning_rate": 2.944506517690875e-05, + "loss": 0.1316, + "step": 10445 + }, + { + "epoch": 29.178770949720672, + "grad_norm": 0.6592856049537659, + "learning_rate": 2.9444692737430168e-05, + "loss": 0.1797, + "step": 10446 + }, + { + "epoch": 29.181564245810055, + "grad_norm": 0.5870290994644165, + "learning_rate": 2.9444320297951584e-05, + "loss": 0.0991, + "step": 10447 + }, + { + "epoch": 29.18435754189944, + "grad_norm": 0.8290870189666748, + "learning_rate": 2.9443947858472998e-05, + "loss": 0.1198, + "step": 10448 + }, + { + "epoch": 29.187150837988828, + "grad_norm": 3.4357829093933105, + "learning_rate": 2.9443575418994414e-05, + "loss": 0.1192, + "step": 10449 + }, + { + "epoch": 29.189944134078214, + "grad_norm": 0.7428920269012451, + "learning_rate": 2.944320297951583e-05, + "loss": 0.1407, + "step": 10450 + }, + { + "epoch": 29.192737430167597, + "grad_norm": 0.7368698716163635, + "learning_rate": 2.9442830540037244e-05, + "loss": 0.1295, + "step": 10451 + }, + { + "epoch": 29.195530726256983, + "grad_norm": 0.9260104298591614, + "learning_rate": 2.944245810055866e-05, + "loss": 0.0996, + "step": 10452 + }, + { + "epoch": 29.19832402234637, + "grad_norm": 0.8167164921760559, + "learning_rate": 2.9442085661080074e-05, + "loss": 0.115, + "step": 10453 + }, + { + "epoch": 29.201117318435756, + "grad_norm": 0.6991883516311646, + "learning_rate": 2.944171322160149e-05, + "loss": 0.1051, + "step": 10454 + }, + { + "epoch": 29.20391061452514, + "grad_norm": 0.8750757575035095, + "learning_rate": 2.9441340782122907e-05, + "loss": 0.1152, + "step": 10455 + }, + { + "epoch": 29.206703910614525, + "grad_norm": 0.7621267437934875, + "learning_rate": 2.9440968342644323e-05, + "loss": 0.116, + "step": 10456 + }, + { + "epoch": 29.20949720670391, + "grad_norm": 1.1134202480316162, + "learning_rate": 2.9440595903165736e-05, + "loss": 0.1826, + "step": 10457 + }, + { + "epoch": 29.212290502793294, + "grad_norm": 0.7415518760681152, + "learning_rate": 2.944022346368715e-05, + "loss": 0.1604, + "step": 10458 + }, + { + "epoch": 29.21508379888268, + "grad_norm": 0.5301984548568726, + "learning_rate": 2.9439851024208566e-05, + "loss": 0.1662, + "step": 10459 + }, + { + "epoch": 29.217877094972067, + "grad_norm": 0.4061637818813324, + "learning_rate": 2.9439478584729982e-05, + "loss": 0.1492, + "step": 10460 + }, + { + "epoch": 29.220670391061454, + "grad_norm": 0.3766739070415497, + "learning_rate": 2.94391061452514e-05, + "loss": 0.1328, + "step": 10461 + }, + { + "epoch": 29.223463687150836, + "grad_norm": 0.9172808527946472, + "learning_rate": 2.9438733705772812e-05, + "loss": 0.1822, + "step": 10462 + }, + { + "epoch": 29.226256983240223, + "grad_norm": 0.5038400888442993, + "learning_rate": 2.943836126629423e-05, + "loss": 0.15, + "step": 10463 + }, + { + "epoch": 29.22905027932961, + "grad_norm": 0.6789467930793762, + "learning_rate": 2.9437988826815642e-05, + "loss": 0.1478, + "step": 10464 + }, + { + "epoch": 29.231843575418996, + "grad_norm": 0.41191813349723816, + "learning_rate": 2.943761638733706e-05, + "loss": 0.1437, + "step": 10465 + }, + { + "epoch": 29.23463687150838, + "grad_norm": 1.0910618305206299, + "learning_rate": 2.943724394785847e-05, + "loss": 0.1198, + "step": 10466 + }, + { + "epoch": 29.237430167597765, + "grad_norm": 0.7163441777229309, + "learning_rate": 2.9436871508379888e-05, + "loss": 0.1617, + "step": 10467 + }, + { + "epoch": 29.24022346368715, + "grad_norm": 0.5814156532287598, + "learning_rate": 2.9436499068901305e-05, + "loss": 0.1295, + "step": 10468 + }, + { + "epoch": 29.243016759776538, + "grad_norm": 0.5598593354225159, + "learning_rate": 2.943612662942272e-05, + "loss": 0.1407, + "step": 10469 + }, + { + "epoch": 29.24581005586592, + "grad_norm": 0.6585691571235657, + "learning_rate": 2.9435754189944138e-05, + "loss": 0.1239, + "step": 10470 + }, + { + "epoch": 29.248603351955307, + "grad_norm": 0.6769425272941589, + "learning_rate": 2.9435381750465547e-05, + "loss": 0.118, + "step": 10471 + }, + { + "epoch": 29.251396648044693, + "grad_norm": 0.5737789273262024, + "learning_rate": 2.9435009310986964e-05, + "loss": 0.1268, + "step": 10472 + }, + { + "epoch": 29.25418994413408, + "grad_norm": 0.944153904914856, + "learning_rate": 2.943463687150838e-05, + "loss": 0.1372, + "step": 10473 + }, + { + "epoch": 29.256983240223462, + "grad_norm": 0.565727949142456, + "learning_rate": 2.9434264432029797e-05, + "loss": 0.1281, + "step": 10474 + }, + { + "epoch": 29.25977653631285, + "grad_norm": 0.48327621817588806, + "learning_rate": 2.943389199255121e-05, + "loss": 0.1124, + "step": 10475 + }, + { + "epoch": 29.262569832402235, + "grad_norm": 0.6199228763580322, + "learning_rate": 2.9433519553072627e-05, + "loss": 0.1282, + "step": 10476 + }, + { + "epoch": 29.26536312849162, + "grad_norm": 0.8130072951316833, + "learning_rate": 2.9433147113594043e-05, + "loss": 0.1171, + "step": 10477 + }, + { + "epoch": 29.268156424581004, + "grad_norm": 1.173012137413025, + "learning_rate": 2.9432774674115456e-05, + "loss": 0.1579, + "step": 10478 + }, + { + "epoch": 29.27094972067039, + "grad_norm": 1.0165550708770752, + "learning_rate": 2.9432402234636873e-05, + "loss": 0.1373, + "step": 10479 + }, + { + "epoch": 29.273743016759777, + "grad_norm": 0.9537060260772705, + "learning_rate": 2.9432029795158286e-05, + "loss": 0.1151, + "step": 10480 + }, + { + "epoch": 29.276536312849164, + "grad_norm": 0.8519675731658936, + "learning_rate": 2.9431657355679703e-05, + "loss": 0.144, + "step": 10481 + }, + { + "epoch": 29.279329608938546, + "grad_norm": 0.8550373911857605, + "learning_rate": 2.943128491620112e-05, + "loss": 0.1468, + "step": 10482 + }, + { + "epoch": 29.282122905027933, + "grad_norm": 0.4374224841594696, + "learning_rate": 2.9430912476722536e-05, + "loss": 0.1576, + "step": 10483 + }, + { + "epoch": 29.28491620111732, + "grad_norm": 1.0766063928604126, + "learning_rate": 2.9430540037243945e-05, + "loss": 0.1346, + "step": 10484 + }, + { + "epoch": 29.287709497206706, + "grad_norm": 0.5032468438148499, + "learning_rate": 2.9430167597765362e-05, + "loss": 0.1601, + "step": 10485 + }, + { + "epoch": 29.29050279329609, + "grad_norm": 0.48072418570518494, + "learning_rate": 2.942979515828678e-05, + "loss": 0.1552, + "step": 10486 + }, + { + "epoch": 29.293296089385475, + "grad_norm": 0.4345191717147827, + "learning_rate": 2.9429422718808195e-05, + "loss": 0.1322, + "step": 10487 + }, + { + "epoch": 29.29608938547486, + "grad_norm": 0.5512000918388367, + "learning_rate": 2.942905027932961e-05, + "loss": 0.1512, + "step": 10488 + }, + { + "epoch": 29.298882681564244, + "grad_norm": 0.4929177463054657, + "learning_rate": 2.9428677839851025e-05, + "loss": 0.1417, + "step": 10489 + }, + { + "epoch": 29.30167597765363, + "grad_norm": 0.4389182925224304, + "learning_rate": 2.942830540037244e-05, + "loss": 0.1397, + "step": 10490 + }, + { + "epoch": 29.304469273743017, + "grad_norm": 0.8300992250442505, + "learning_rate": 2.9427932960893854e-05, + "loss": 0.1375, + "step": 10491 + }, + { + "epoch": 29.307262569832403, + "grad_norm": 0.5978677272796631, + "learning_rate": 2.942756052141527e-05, + "loss": 0.123, + "step": 10492 + }, + { + "epoch": 29.310055865921786, + "grad_norm": 0.5969186425209045, + "learning_rate": 2.9427188081936684e-05, + "loss": 0.1444, + "step": 10493 + }, + { + "epoch": 29.312849162011172, + "grad_norm": 1.6380987167358398, + "learning_rate": 2.94268156424581e-05, + "loss": 0.1498, + "step": 10494 + }, + { + "epoch": 29.31564245810056, + "grad_norm": 0.638852596282959, + "learning_rate": 2.9426443202979517e-05, + "loss": 0.1204, + "step": 10495 + }, + { + "epoch": 29.318435754189945, + "grad_norm": 0.8109268546104431, + "learning_rate": 2.9426070763500934e-05, + "loss": 0.1261, + "step": 10496 + }, + { + "epoch": 29.321229050279328, + "grad_norm": 0.7935317158699036, + "learning_rate": 2.942569832402235e-05, + "loss": 0.1311, + "step": 10497 + }, + { + "epoch": 29.324022346368714, + "grad_norm": 0.9897924661636353, + "learning_rate": 2.942532588454376e-05, + "loss": 0.1337, + "step": 10498 + }, + { + "epoch": 29.3268156424581, + "grad_norm": 0.8058103919029236, + "learning_rate": 2.9424953445065177e-05, + "loss": 0.1451, + "step": 10499 + }, + { + "epoch": 29.329608938547487, + "grad_norm": 0.919959306716919, + "learning_rate": 2.9424581005586593e-05, + "loss": 0.1148, + "step": 10500 + }, + { + "epoch": 29.33240223463687, + "grad_norm": 0.5707392692565918, + "learning_rate": 2.942420856610801e-05, + "loss": 0.1249, + "step": 10501 + }, + { + "epoch": 29.335195530726256, + "grad_norm": 0.5946323871612549, + "learning_rate": 2.9423836126629423e-05, + "loss": 0.1058, + "step": 10502 + }, + { + "epoch": 29.337988826815643, + "grad_norm": 0.616500735282898, + "learning_rate": 2.942346368715084e-05, + "loss": 0.1333, + "step": 10503 + }, + { + "epoch": 29.34078212290503, + "grad_norm": 0.9339783191680908, + "learning_rate": 2.9423091247672252e-05, + "loss": 0.1266, + "step": 10504 + }, + { + "epoch": 29.343575418994412, + "grad_norm": 1.265979290008545, + "learning_rate": 2.942271880819367e-05, + "loss": 0.1362, + "step": 10505 + }, + { + "epoch": 29.3463687150838, + "grad_norm": 1.1186059713363647, + "learning_rate": 2.9422346368715086e-05, + "loss": 0.1464, + "step": 10506 + }, + { + "epoch": 29.349162011173185, + "grad_norm": 2.1735808849334717, + "learning_rate": 2.94219739292365e-05, + "loss": 0.1849, + "step": 10507 + }, + { + "epoch": 29.35195530726257, + "grad_norm": 0.8704769611358643, + "learning_rate": 2.9421601489757915e-05, + "loss": 0.1671, + "step": 10508 + }, + { + "epoch": 29.354748603351954, + "grad_norm": 1.0171657800674438, + "learning_rate": 2.9421229050279332e-05, + "loss": 0.1579, + "step": 10509 + }, + { + "epoch": 29.35754189944134, + "grad_norm": 0.578435480594635, + "learning_rate": 2.9420856610800748e-05, + "loss": 0.1436, + "step": 10510 + }, + { + "epoch": 29.360335195530727, + "grad_norm": 0.6933241486549377, + "learning_rate": 2.9420484171322158e-05, + "loss": 0.1523, + "step": 10511 + }, + { + "epoch": 29.363128491620113, + "grad_norm": 0.5105825066566467, + "learning_rate": 2.9420111731843575e-05, + "loss": 0.1733, + "step": 10512 + }, + { + "epoch": 29.365921787709496, + "grad_norm": 0.44320404529571533, + "learning_rate": 2.941973929236499e-05, + "loss": 0.1377, + "step": 10513 + }, + { + "epoch": 29.368715083798882, + "grad_norm": 0.4951266944408417, + "learning_rate": 2.9419366852886408e-05, + "loss": 0.1435, + "step": 10514 + }, + { + "epoch": 29.37150837988827, + "grad_norm": 0.4189750552177429, + "learning_rate": 2.9418994413407824e-05, + "loss": 0.1463, + "step": 10515 + }, + { + "epoch": 29.374301675977655, + "grad_norm": 0.33483046293258667, + "learning_rate": 2.9418621973929237e-05, + "loss": 0.1341, + "step": 10516 + }, + { + "epoch": 29.377094972067038, + "grad_norm": 0.7134982347488403, + "learning_rate": 2.9418249534450654e-05, + "loss": 0.1623, + "step": 10517 + }, + { + "epoch": 29.379888268156424, + "grad_norm": 0.846622109413147, + "learning_rate": 2.9417877094972067e-05, + "loss": 0.1276, + "step": 10518 + }, + { + "epoch": 29.38268156424581, + "grad_norm": 0.605177640914917, + "learning_rate": 2.9417504655493484e-05, + "loss": 0.1302, + "step": 10519 + }, + { + "epoch": 29.385474860335197, + "grad_norm": 0.8363968133926392, + "learning_rate": 2.9417132216014897e-05, + "loss": 0.1405, + "step": 10520 + }, + { + "epoch": 29.38826815642458, + "grad_norm": 0.5143560171127319, + "learning_rate": 2.9416759776536313e-05, + "loss": 0.1433, + "step": 10521 + }, + { + "epoch": 29.391061452513966, + "grad_norm": 0.6231856942176819, + "learning_rate": 2.941638733705773e-05, + "loss": 0.0991, + "step": 10522 + }, + { + "epoch": 29.393854748603353, + "grad_norm": 0.485272616147995, + "learning_rate": 2.9416014897579146e-05, + "loss": 0.1189, + "step": 10523 + }, + { + "epoch": 29.39664804469274, + "grad_norm": 0.5193650722503662, + "learning_rate": 2.941564245810056e-05, + "loss": 0.1193, + "step": 10524 + }, + { + "epoch": 29.399441340782122, + "grad_norm": 0.9793933033943176, + "learning_rate": 2.9415270018621973e-05, + "loss": 0.1154, + "step": 10525 + }, + { + "epoch": 29.40223463687151, + "grad_norm": 0.7582762837409973, + "learning_rate": 2.941489757914339e-05, + "loss": 0.1136, + "step": 10526 + }, + { + "epoch": 29.405027932960895, + "grad_norm": 0.5359487533569336, + "learning_rate": 2.9414525139664806e-05, + "loss": 0.1469, + "step": 10527 + }, + { + "epoch": 29.407821229050278, + "grad_norm": 0.8032533526420593, + "learning_rate": 2.9414152700186222e-05, + "loss": 0.1297, + "step": 10528 + }, + { + "epoch": 29.410614525139664, + "grad_norm": 0.5733479261398315, + "learning_rate": 2.9413780260707635e-05, + "loss": 0.1243, + "step": 10529 + }, + { + "epoch": 29.41340782122905, + "grad_norm": 0.6183791756629944, + "learning_rate": 2.9413407821229052e-05, + "loss": 0.1234, + "step": 10530 + }, + { + "epoch": 29.416201117318437, + "grad_norm": 1.2271229028701782, + "learning_rate": 2.9413035381750465e-05, + "loss": 0.1722, + "step": 10531 + }, + { + "epoch": 29.41899441340782, + "grad_norm": 1.599050760269165, + "learning_rate": 2.941266294227188e-05, + "loss": 0.171, + "step": 10532 + }, + { + "epoch": 29.421787709497206, + "grad_norm": 0.34258487820625305, + "learning_rate": 2.9412290502793298e-05, + "loss": 0.1568, + "step": 10533 + }, + { + "epoch": 29.424581005586592, + "grad_norm": 0.4740338623523712, + "learning_rate": 2.941191806331471e-05, + "loss": 0.1601, + "step": 10534 + }, + { + "epoch": 29.42737430167598, + "grad_norm": 0.45685815811157227, + "learning_rate": 2.9411545623836128e-05, + "loss": 0.1486, + "step": 10535 + }, + { + "epoch": 29.43016759776536, + "grad_norm": 0.3503548204898834, + "learning_rate": 2.9411173184357544e-05, + "loss": 0.1371, + "step": 10536 + }, + { + "epoch": 29.432960893854748, + "grad_norm": 0.9786669015884399, + "learning_rate": 2.941080074487896e-05, + "loss": 0.1324, + "step": 10537 + }, + { + "epoch": 29.435754189944134, + "grad_norm": 0.5977602601051331, + "learning_rate": 2.941042830540037e-05, + "loss": 0.1445, + "step": 10538 + }, + { + "epoch": 29.43854748603352, + "grad_norm": 0.711875855922699, + "learning_rate": 2.9410055865921787e-05, + "loss": 0.1301, + "step": 10539 + }, + { + "epoch": 29.441340782122904, + "grad_norm": 1.2337687015533447, + "learning_rate": 2.9409683426443204e-05, + "loss": 0.1384, + "step": 10540 + }, + { + "epoch": 29.44413407821229, + "grad_norm": 0.34358519315719604, + "learning_rate": 2.940931098696462e-05, + "loss": 0.1083, + "step": 10541 + }, + { + "epoch": 29.446927374301676, + "grad_norm": 0.854315996170044, + "learning_rate": 2.9408938547486033e-05, + "loss": 0.1458, + "step": 10542 + }, + { + "epoch": 29.449720670391063, + "grad_norm": 0.6628797054290771, + "learning_rate": 2.940856610800745e-05, + "loss": 0.1327, + "step": 10543 + }, + { + "epoch": 29.452513966480446, + "grad_norm": 0.7053240537643433, + "learning_rate": 2.9408193668528863e-05, + "loss": 0.1427, + "step": 10544 + }, + { + "epoch": 29.455307262569832, + "grad_norm": 0.5971826314926147, + "learning_rate": 2.940782122905028e-05, + "loss": 0.1297, + "step": 10545 + }, + { + "epoch": 29.45810055865922, + "grad_norm": 0.5458451509475708, + "learning_rate": 2.9407448789571696e-05, + "loss": 0.1194, + "step": 10546 + }, + { + "epoch": 29.460893854748605, + "grad_norm": 0.6103687882423401, + "learning_rate": 2.940707635009311e-05, + "loss": 0.1481, + "step": 10547 + }, + { + "epoch": 29.463687150837988, + "grad_norm": 6.27344274520874, + "learning_rate": 2.9406703910614526e-05, + "loss": 0.1327, + "step": 10548 + }, + { + "epoch": 29.466480446927374, + "grad_norm": 0.746681809425354, + "learning_rate": 2.9406331471135942e-05, + "loss": 0.1276, + "step": 10549 + }, + { + "epoch": 29.46927374301676, + "grad_norm": 0.46948087215423584, + "learning_rate": 2.940595903165736e-05, + "loss": 0.1101, + "step": 10550 + }, + { + "epoch": 29.472067039106147, + "grad_norm": 0.7582101225852966, + "learning_rate": 2.940558659217877e-05, + "loss": 0.113, + "step": 10551 + }, + { + "epoch": 29.47486033519553, + "grad_norm": 1.6060010194778442, + "learning_rate": 2.9405214152700185e-05, + "loss": 0.134, + "step": 10552 + }, + { + "epoch": 29.477653631284916, + "grad_norm": 0.7300018072128296, + "learning_rate": 2.9404841713221602e-05, + "loss": 0.13, + "step": 10553 + }, + { + "epoch": 29.480446927374302, + "grad_norm": 0.9380890130996704, + "learning_rate": 2.9404469273743018e-05, + "loss": 0.1155, + "step": 10554 + }, + { + "epoch": 29.48324022346369, + "grad_norm": 0.8584399819374084, + "learning_rate": 2.9404096834264435e-05, + "loss": 0.1227, + "step": 10555 + }, + { + "epoch": 29.48603351955307, + "grad_norm": 1.3268927335739136, + "learning_rate": 2.9403724394785848e-05, + "loss": 0.1413, + "step": 10556 + }, + { + "epoch": 29.488826815642458, + "grad_norm": 1.1468411684036255, + "learning_rate": 2.9403351955307264e-05, + "loss": 0.1897, + "step": 10557 + }, + { + "epoch": 29.491620111731844, + "grad_norm": 0.3983496427536011, + "learning_rate": 2.9402979515828678e-05, + "loss": 0.1627, + "step": 10558 + }, + { + "epoch": 29.49441340782123, + "grad_norm": 0.5051746964454651, + "learning_rate": 2.9402607076350094e-05, + "loss": 0.1465, + "step": 10559 + }, + { + "epoch": 29.497206703910614, + "grad_norm": 0.5331940054893494, + "learning_rate": 2.9402234636871507e-05, + "loss": 0.1415, + "step": 10560 + }, + { + "epoch": 29.5, + "grad_norm": 0.4462803900241852, + "learning_rate": 2.9401862197392924e-05, + "loss": 0.1359, + "step": 10561 + }, + { + "epoch": 29.502793296089386, + "grad_norm": 0.8652021884918213, + "learning_rate": 2.940148975791434e-05, + "loss": 0.1486, + "step": 10562 + }, + { + "epoch": 29.505586592178773, + "grad_norm": 0.371593713760376, + "learning_rate": 2.9401117318435757e-05, + "loss": 0.1498, + "step": 10563 + }, + { + "epoch": 29.508379888268156, + "grad_norm": 0.5405426621437073, + "learning_rate": 2.940074487895717e-05, + "loss": 0.1175, + "step": 10564 + }, + { + "epoch": 29.511173184357542, + "grad_norm": 0.4678661823272705, + "learning_rate": 2.9400372439478583e-05, + "loss": 0.1439, + "step": 10565 + }, + { + "epoch": 29.51396648044693, + "grad_norm": 0.47997739911079407, + "learning_rate": 2.94e-05, + "loss": 0.1114, + "step": 10566 + }, + { + "epoch": 29.51675977653631, + "grad_norm": 1.0294495820999146, + "learning_rate": 2.9399627560521416e-05, + "loss": 0.1266, + "step": 10567 + }, + { + "epoch": 29.519553072625698, + "grad_norm": 0.7888020873069763, + "learning_rate": 2.9399255121042833e-05, + "loss": 0.1241, + "step": 10568 + }, + { + "epoch": 29.522346368715084, + "grad_norm": 0.5707956552505493, + "learning_rate": 2.9398882681564246e-05, + "loss": 0.1458, + "step": 10569 + }, + { + "epoch": 29.52513966480447, + "grad_norm": 0.62593674659729, + "learning_rate": 2.9398510242085662e-05, + "loss": 0.1329, + "step": 10570 + }, + { + "epoch": 29.527932960893853, + "grad_norm": 0.3958328068256378, + "learning_rate": 2.9398137802607076e-05, + "loss": 0.1144, + "step": 10571 + }, + { + "epoch": 29.53072625698324, + "grad_norm": 0.6685848832130432, + "learning_rate": 2.9397765363128492e-05, + "loss": 0.1203, + "step": 10572 + }, + { + "epoch": 29.533519553072626, + "grad_norm": 0.8417068123817444, + "learning_rate": 2.939739292364991e-05, + "loss": 0.1266, + "step": 10573 + }, + { + "epoch": 29.536312849162012, + "grad_norm": 0.6941617131233215, + "learning_rate": 2.9397020484171322e-05, + "loss": 0.11, + "step": 10574 + }, + { + "epoch": 29.539106145251395, + "grad_norm": 0.7907447814941406, + "learning_rate": 2.939664804469274e-05, + "loss": 0.1277, + "step": 10575 + }, + { + "epoch": 29.54189944134078, + "grad_norm": 0.842665433883667, + "learning_rate": 2.9396275605214155e-05, + "loss": 0.0961, + "step": 10576 + }, + { + "epoch": 29.544692737430168, + "grad_norm": 0.8059181571006775, + "learning_rate": 2.939590316573557e-05, + "loss": 0.1245, + "step": 10577 + }, + { + "epoch": 29.547486033519554, + "grad_norm": 0.7097334861755371, + "learning_rate": 2.939553072625698e-05, + "loss": 0.1166, + "step": 10578 + }, + { + "epoch": 29.550279329608937, + "grad_norm": 0.7580041289329529, + "learning_rate": 2.9395158286778398e-05, + "loss": 0.1138, + "step": 10579 + }, + { + "epoch": 29.553072625698324, + "grad_norm": 1.1183631420135498, + "learning_rate": 2.9394785847299814e-05, + "loss": 0.139, + "step": 10580 + }, + { + "epoch": 29.55586592178771, + "grad_norm": 1.0369480848312378, + "learning_rate": 2.939441340782123e-05, + "loss": 0.1233, + "step": 10581 + }, + { + "epoch": 29.558659217877096, + "grad_norm": 1.5841624736785889, + "learning_rate": 2.9394040968342647e-05, + "loss": 0.1548, + "step": 10582 + }, + { + "epoch": 29.56145251396648, + "grad_norm": 0.5917270183563232, + "learning_rate": 2.939366852886406e-05, + "loss": 0.1418, + "step": 10583 + }, + { + "epoch": 29.564245810055866, + "grad_norm": 0.49922508001327515, + "learning_rate": 2.9393296089385474e-05, + "loss": 0.1408, + "step": 10584 + }, + { + "epoch": 29.567039106145252, + "grad_norm": 0.5298774242401123, + "learning_rate": 2.939292364990689e-05, + "loss": 0.1573, + "step": 10585 + }, + { + "epoch": 29.56983240223464, + "grad_norm": 0.7163188457489014, + "learning_rate": 2.9392551210428307e-05, + "loss": 0.1434, + "step": 10586 + }, + { + "epoch": 29.57262569832402, + "grad_norm": 0.5576711893081665, + "learning_rate": 2.939217877094972e-05, + "loss": 0.1757, + "step": 10587 + }, + { + "epoch": 29.575418994413408, + "grad_norm": 0.6797800064086914, + "learning_rate": 2.9391806331471136e-05, + "loss": 0.1272, + "step": 10588 + }, + { + "epoch": 29.578212290502794, + "grad_norm": 0.5996577739715576, + "learning_rate": 2.9391433891992553e-05, + "loss": 0.1284, + "step": 10589 + }, + { + "epoch": 29.58100558659218, + "grad_norm": 0.5282605886459351, + "learning_rate": 2.939106145251397e-05, + "loss": 0.1257, + "step": 10590 + }, + { + "epoch": 29.583798882681563, + "grad_norm": 0.8012151122093201, + "learning_rate": 2.9390689013035383e-05, + "loss": 0.1413, + "step": 10591 + }, + { + "epoch": 29.58659217877095, + "grad_norm": 1.0409581661224365, + "learning_rate": 2.9390316573556796e-05, + "loss": 0.12, + "step": 10592 + }, + { + "epoch": 29.589385474860336, + "grad_norm": 1.0612879991531372, + "learning_rate": 2.9389944134078212e-05, + "loss": 0.1697, + "step": 10593 + }, + { + "epoch": 29.592178770949722, + "grad_norm": 0.5875619053840637, + "learning_rate": 2.938957169459963e-05, + "loss": 0.1111, + "step": 10594 + }, + { + "epoch": 29.594972067039105, + "grad_norm": 0.5456404685974121, + "learning_rate": 2.9389199255121045e-05, + "loss": 0.124, + "step": 10595 + }, + { + "epoch": 29.59776536312849, + "grad_norm": 0.5516364574432373, + "learning_rate": 2.938882681564246e-05, + "loss": 0.1314, + "step": 10596 + }, + { + "epoch": 29.600558659217878, + "grad_norm": 1.4851598739624023, + "learning_rate": 2.9388454376163875e-05, + "loss": 0.1127, + "step": 10597 + }, + { + "epoch": 29.60335195530726, + "grad_norm": 0.6057856678962708, + "learning_rate": 2.9388081936685288e-05, + "loss": 0.1262, + "step": 10598 + }, + { + "epoch": 29.606145251396647, + "grad_norm": 0.5095430612564087, + "learning_rate": 2.9387709497206705e-05, + "loss": 0.1285, + "step": 10599 + }, + { + "epoch": 29.608938547486034, + "grad_norm": 0.9096830487251282, + "learning_rate": 2.938733705772812e-05, + "loss": 0.1253, + "step": 10600 + }, + { + "epoch": 29.61173184357542, + "grad_norm": 0.6844052672386169, + "learning_rate": 2.9386964618249534e-05, + "loss": 0.1271, + "step": 10601 + }, + { + "epoch": 29.614525139664803, + "grad_norm": 0.6677237749099731, + "learning_rate": 2.938659217877095e-05, + "loss": 0.1356, + "step": 10602 + }, + { + "epoch": 29.61731843575419, + "grad_norm": 0.7124325037002563, + "learning_rate": 2.9386219739292367e-05, + "loss": 0.1499, + "step": 10603 + }, + { + "epoch": 29.620111731843576, + "grad_norm": 0.7017527222633362, + "learning_rate": 2.938584729981378e-05, + "loss": 0.135, + "step": 10604 + }, + { + "epoch": 29.622905027932962, + "grad_norm": 0.8095778226852417, + "learning_rate": 2.9385474860335194e-05, + "loss": 0.1416, + "step": 10605 + }, + { + "epoch": 29.625698324022345, + "grad_norm": 0.8359748721122742, + "learning_rate": 2.938510242085661e-05, + "loss": 0.1227, + "step": 10606 + }, + { + "epoch": 29.62849162011173, + "grad_norm": 1.0952584743499756, + "learning_rate": 2.9384729981378027e-05, + "loss": 0.1675, + "step": 10607 + }, + { + "epoch": 29.631284916201118, + "grad_norm": 0.5011212825775146, + "learning_rate": 2.9384357541899443e-05, + "loss": 0.1578, + "step": 10608 + }, + { + "epoch": 29.634078212290504, + "grad_norm": 0.44427192211151123, + "learning_rate": 2.938398510242086e-05, + "loss": 0.1606, + "step": 10609 + }, + { + "epoch": 29.636871508379887, + "grad_norm": 0.5942420363426208, + "learning_rate": 2.9383612662942273e-05, + "loss": 0.1603, + "step": 10610 + }, + { + "epoch": 29.639664804469273, + "grad_norm": 0.4994801878929138, + "learning_rate": 2.9383240223463686e-05, + "loss": 0.1483, + "step": 10611 + }, + { + "epoch": 29.64245810055866, + "grad_norm": 0.46086230874061584, + "learning_rate": 2.9382867783985103e-05, + "loss": 0.1405, + "step": 10612 + }, + { + "epoch": 29.645251396648046, + "grad_norm": 2.5640363693237305, + "learning_rate": 2.938249534450652e-05, + "loss": 0.1598, + "step": 10613 + }, + { + "epoch": 29.64804469273743, + "grad_norm": 0.46979275345802307, + "learning_rate": 2.9382122905027932e-05, + "loss": 0.1631, + "step": 10614 + }, + { + "epoch": 29.650837988826815, + "grad_norm": 0.6739952564239502, + "learning_rate": 2.938175046554935e-05, + "loss": 0.1574, + "step": 10615 + }, + { + "epoch": 29.6536312849162, + "grad_norm": 0.6682666540145874, + "learning_rate": 2.9381378026070766e-05, + "loss": 0.1279, + "step": 10616 + }, + { + "epoch": 29.656424581005588, + "grad_norm": 0.48962533473968506, + "learning_rate": 2.9381005586592182e-05, + "loss": 0.1361, + "step": 10617 + }, + { + "epoch": 29.65921787709497, + "grad_norm": 0.7188294529914856, + "learning_rate": 2.9380633147113595e-05, + "loss": 0.1296, + "step": 10618 + }, + { + "epoch": 29.662011173184357, + "grad_norm": 0.9147038459777832, + "learning_rate": 2.938026070763501e-05, + "loss": 0.1363, + "step": 10619 + }, + { + "epoch": 29.664804469273744, + "grad_norm": 1.1441426277160645, + "learning_rate": 2.9379888268156425e-05, + "loss": 0.1343, + "step": 10620 + }, + { + "epoch": 29.66759776536313, + "grad_norm": 1.0553549528121948, + "learning_rate": 2.937951582867784e-05, + "loss": 0.1191, + "step": 10621 + }, + { + "epoch": 29.670391061452513, + "grad_norm": 0.43613293766975403, + "learning_rate": 2.9379143389199258e-05, + "loss": 0.1186, + "step": 10622 + }, + { + "epoch": 29.6731843575419, + "grad_norm": 0.6146807670593262, + "learning_rate": 2.937877094972067e-05, + "loss": 0.1368, + "step": 10623 + }, + { + "epoch": 29.675977653631286, + "grad_norm": 0.6435033679008484, + "learning_rate": 2.9378398510242084e-05, + "loss": 0.1284, + "step": 10624 + }, + { + "epoch": 29.678770949720672, + "grad_norm": 1.1289857625961304, + "learning_rate": 2.93780260707635e-05, + "loss": 0.1075, + "step": 10625 + }, + { + "epoch": 29.681564245810055, + "grad_norm": 0.7356782555580139, + "learning_rate": 2.9377653631284917e-05, + "loss": 0.1502, + "step": 10626 + }, + { + "epoch": 29.68435754189944, + "grad_norm": 0.717106819152832, + "learning_rate": 2.9377281191806334e-05, + "loss": 0.1015, + "step": 10627 + }, + { + "epoch": 29.687150837988828, + "grad_norm": 1.5027852058410645, + "learning_rate": 2.9376908752327747e-05, + "loss": 0.1066, + "step": 10628 + }, + { + "epoch": 29.689944134078214, + "grad_norm": 0.6935351490974426, + "learning_rate": 2.9376536312849164e-05, + "loss": 0.1283, + "step": 10629 + }, + { + "epoch": 29.692737430167597, + "grad_norm": 1.5517454147338867, + "learning_rate": 2.937616387337058e-05, + "loss": 0.1301, + "step": 10630 + }, + { + "epoch": 29.695530726256983, + "grad_norm": 0.736274242401123, + "learning_rate": 2.9375791433891993e-05, + "loss": 0.1113, + "step": 10631 + }, + { + "epoch": 29.69832402234637, + "grad_norm": 1.0854341983795166, + "learning_rate": 2.9375418994413406e-05, + "loss": 0.1895, + "step": 10632 + }, + { + "epoch": 29.701117318435756, + "grad_norm": 0.7550715804100037, + "learning_rate": 2.9375046554934823e-05, + "loss": 0.1696, + "step": 10633 + }, + { + "epoch": 29.70391061452514, + "grad_norm": 0.4506986439228058, + "learning_rate": 2.937467411545624e-05, + "loss": 0.1466, + "step": 10634 + }, + { + "epoch": 29.706703910614525, + "grad_norm": 0.5066136717796326, + "learning_rate": 2.9374301675977656e-05, + "loss": 0.1256, + "step": 10635 + }, + { + "epoch": 29.70949720670391, + "grad_norm": 0.6008812785148621, + "learning_rate": 2.937392923649907e-05, + "loss": 0.141, + "step": 10636 + }, + { + "epoch": 29.712290502793294, + "grad_norm": 0.6040401458740234, + "learning_rate": 2.9373556797020486e-05, + "loss": 0.1407, + "step": 10637 + }, + { + "epoch": 29.71508379888268, + "grad_norm": 0.5958768725395203, + "learning_rate": 2.93731843575419e-05, + "loss": 0.1299, + "step": 10638 + }, + { + "epoch": 29.717877094972067, + "grad_norm": 0.68144291639328, + "learning_rate": 2.9372811918063315e-05, + "loss": 0.153, + "step": 10639 + }, + { + "epoch": 29.720670391061454, + "grad_norm": 0.6175395846366882, + "learning_rate": 2.9372439478584732e-05, + "loss": 0.1374, + "step": 10640 + }, + { + "epoch": 29.723463687150836, + "grad_norm": 0.7193018794059753, + "learning_rate": 2.9372067039106145e-05, + "loss": 0.1362, + "step": 10641 + }, + { + "epoch": 29.726256983240223, + "grad_norm": 0.6670440435409546, + "learning_rate": 2.937169459962756e-05, + "loss": 0.1617, + "step": 10642 + }, + { + "epoch": 29.72905027932961, + "grad_norm": 1.0487710237503052, + "learning_rate": 2.9371322160148978e-05, + "loss": 0.1497, + "step": 10643 + }, + { + "epoch": 29.731843575418996, + "grad_norm": 0.5141791701316833, + "learning_rate": 2.937094972067039e-05, + "loss": 0.128, + "step": 10644 + }, + { + "epoch": 29.73463687150838, + "grad_norm": 0.484594464302063, + "learning_rate": 2.9370577281191804e-05, + "loss": 0.1404, + "step": 10645 + }, + { + "epoch": 29.737430167597765, + "grad_norm": 1.7349698543548584, + "learning_rate": 2.937020484171322e-05, + "loss": 0.13, + "step": 10646 + }, + { + "epoch": 29.74022346368715, + "grad_norm": 0.591040849685669, + "learning_rate": 2.9369832402234637e-05, + "loss": 0.1337, + "step": 10647 + }, + { + "epoch": 29.743016759776538, + "grad_norm": 0.5441344976425171, + "learning_rate": 2.9369459962756054e-05, + "loss": 0.1264, + "step": 10648 + }, + { + "epoch": 29.74581005586592, + "grad_norm": 0.5198759436607361, + "learning_rate": 2.936908752327747e-05, + "loss": 0.1374, + "step": 10649 + }, + { + "epoch": 29.748603351955307, + "grad_norm": 0.8003717064857483, + "learning_rate": 2.9368715083798884e-05, + "loss": 0.1284, + "step": 10650 + }, + { + "epoch": 29.751396648044693, + "grad_norm": 0.9710149765014648, + "learning_rate": 2.9368342644320297e-05, + "loss": 0.1229, + "step": 10651 + }, + { + "epoch": 29.75418994413408, + "grad_norm": 0.5438783168792725, + "learning_rate": 2.9367970204841713e-05, + "loss": 0.1144, + "step": 10652 + }, + { + "epoch": 29.756983240223462, + "grad_norm": 0.7861520648002625, + "learning_rate": 2.936759776536313e-05, + "loss": 0.1099, + "step": 10653 + }, + { + "epoch": 29.75977653631285, + "grad_norm": 0.6704384684562683, + "learning_rate": 2.9367225325884543e-05, + "loss": 0.1167, + "step": 10654 + }, + { + "epoch": 29.762569832402235, + "grad_norm": 1.163910984992981, + "learning_rate": 2.936685288640596e-05, + "loss": 0.1137, + "step": 10655 + }, + { + "epoch": 29.76536312849162, + "grad_norm": 1.2814276218414307, + "learning_rate": 2.9366480446927376e-05, + "loss": 0.1293, + "step": 10656 + }, + { + "epoch": 29.768156424581004, + "grad_norm": 1.9360063076019287, + "learning_rate": 2.936610800744879e-05, + "loss": 0.1964, + "step": 10657 + }, + { + "epoch": 29.77094972067039, + "grad_norm": 0.5944881439208984, + "learning_rate": 2.9365735567970206e-05, + "loss": 0.169, + "step": 10658 + }, + { + "epoch": 29.773743016759777, + "grad_norm": 0.6936995983123779, + "learning_rate": 2.936536312849162e-05, + "loss": 0.1385, + "step": 10659 + }, + { + "epoch": 29.776536312849164, + "grad_norm": 0.45505762100219727, + "learning_rate": 2.9364990689013035e-05, + "loss": 0.1509, + "step": 10660 + }, + { + "epoch": 29.779329608938546, + "grad_norm": 1.1024165153503418, + "learning_rate": 2.9364618249534452e-05, + "loss": 0.1481, + "step": 10661 + }, + { + "epoch": 29.782122905027933, + "grad_norm": 0.6289942860603333, + "learning_rate": 2.936424581005587e-05, + "loss": 0.1475, + "step": 10662 + }, + { + "epoch": 29.78491620111732, + "grad_norm": 0.5570648908615112, + "learning_rate": 2.9363873370577282e-05, + "loss": 0.1267, + "step": 10663 + }, + { + "epoch": 29.787709497206706, + "grad_norm": 0.8213441967964172, + "learning_rate": 2.9363500931098695e-05, + "loss": 0.138, + "step": 10664 + }, + { + "epoch": 29.79050279329609, + "grad_norm": 1.4896053075790405, + "learning_rate": 2.936312849162011e-05, + "loss": 0.119, + "step": 10665 + }, + { + "epoch": 29.793296089385475, + "grad_norm": 0.4400220215320587, + "learning_rate": 2.9362756052141528e-05, + "loss": 0.1514, + "step": 10666 + }, + { + "epoch": 29.79608938547486, + "grad_norm": 1.0473942756652832, + "learning_rate": 2.9362383612662944e-05, + "loss": 0.1703, + "step": 10667 + }, + { + "epoch": 29.798882681564244, + "grad_norm": 0.8684031367301941, + "learning_rate": 2.9362011173184358e-05, + "loss": 0.1221, + "step": 10668 + }, + { + "epoch": 29.80167597765363, + "grad_norm": 1.1273804903030396, + "learning_rate": 2.9361638733705774e-05, + "loss": 0.1244, + "step": 10669 + }, + { + "epoch": 29.804469273743017, + "grad_norm": 0.7158713340759277, + "learning_rate": 2.936126629422719e-05, + "loss": 0.1306, + "step": 10670 + }, + { + "epoch": 29.807262569832403, + "grad_norm": 0.6423994898796082, + "learning_rate": 2.9360893854748604e-05, + "loss": 0.1415, + "step": 10671 + }, + { + "epoch": 29.810055865921786, + "grad_norm": 0.6010428071022034, + "learning_rate": 2.9360521415270017e-05, + "loss": 0.1134, + "step": 10672 + }, + { + "epoch": 29.812849162011172, + "grad_norm": 0.6045321822166443, + "learning_rate": 2.9360148975791433e-05, + "loss": 0.1163, + "step": 10673 + }, + { + "epoch": 29.81564245810056, + "grad_norm": 1.2525767087936401, + "learning_rate": 2.935977653631285e-05, + "loss": 0.1177, + "step": 10674 + }, + { + "epoch": 29.818435754189945, + "grad_norm": 0.7399109601974487, + "learning_rate": 2.9359404096834267e-05, + "loss": 0.1399, + "step": 10675 + }, + { + "epoch": 29.821229050279328, + "grad_norm": 0.652255117893219, + "learning_rate": 2.9359031657355683e-05, + "loss": 0.1416, + "step": 10676 + }, + { + "epoch": 29.824022346368714, + "grad_norm": 0.877556562423706, + "learning_rate": 2.9358659217877093e-05, + "loss": 0.1267, + "step": 10677 + }, + { + "epoch": 29.8268156424581, + "grad_norm": 1.3405183553695679, + "learning_rate": 2.935828677839851e-05, + "loss": 0.1493, + "step": 10678 + }, + { + "epoch": 29.829608938547487, + "grad_norm": 0.5100284218788147, + "learning_rate": 2.9357914338919926e-05, + "loss": 0.1178, + "step": 10679 + }, + { + "epoch": 29.83240223463687, + "grad_norm": 1.0250123739242554, + "learning_rate": 2.9357541899441342e-05, + "loss": 0.1344, + "step": 10680 + }, + { + "epoch": 29.835195530726256, + "grad_norm": 1.6386967897415161, + "learning_rate": 2.9357169459962756e-05, + "loss": 0.1651, + "step": 10681 + }, + { + "epoch": 29.837988826815643, + "grad_norm": 1.969767689704895, + "learning_rate": 2.9356797020484172e-05, + "loss": 0.1824, + "step": 10682 + }, + { + "epoch": 29.84078212290503, + "grad_norm": 0.5189602971076965, + "learning_rate": 2.935642458100559e-05, + "loss": 0.1549, + "step": 10683 + }, + { + "epoch": 29.843575418994412, + "grad_norm": 0.5187779068946838, + "learning_rate": 2.9356052141527002e-05, + "loss": 0.1579, + "step": 10684 + }, + { + "epoch": 29.8463687150838, + "grad_norm": 0.648497998714447, + "learning_rate": 2.935567970204842e-05, + "loss": 0.1368, + "step": 10685 + }, + { + "epoch": 29.849162011173185, + "grad_norm": 0.4926854372024536, + "learning_rate": 2.935530726256983e-05, + "loss": 0.1681, + "step": 10686 + }, + { + "epoch": 29.85195530726257, + "grad_norm": 0.946799099445343, + "learning_rate": 2.9354934823091248e-05, + "loss": 0.1453, + "step": 10687 + }, + { + "epoch": 29.854748603351954, + "grad_norm": 0.6755886673927307, + "learning_rate": 2.9354562383612665e-05, + "loss": 0.1573, + "step": 10688 + }, + { + "epoch": 29.85754189944134, + "grad_norm": 0.535248875617981, + "learning_rate": 2.935418994413408e-05, + "loss": 0.1417, + "step": 10689 + }, + { + "epoch": 29.860335195530727, + "grad_norm": 0.8182392120361328, + "learning_rate": 2.9353817504655494e-05, + "loss": 0.147, + "step": 10690 + }, + { + "epoch": 29.863128491620113, + "grad_norm": 0.5701040625572205, + "learning_rate": 2.9353445065176907e-05, + "loss": 0.1308, + "step": 10691 + }, + { + "epoch": 29.865921787709496, + "grad_norm": 0.48231396079063416, + "learning_rate": 2.9353072625698324e-05, + "loss": 0.1258, + "step": 10692 + }, + { + "epoch": 29.868715083798882, + "grad_norm": 0.7018629312515259, + "learning_rate": 2.935270018621974e-05, + "loss": 0.1409, + "step": 10693 + }, + { + "epoch": 29.87150837988827, + "grad_norm": 0.9043267369270325, + "learning_rate": 2.9352327746741157e-05, + "loss": 0.1325, + "step": 10694 + }, + { + "epoch": 29.874301675977655, + "grad_norm": 0.6016144752502441, + "learning_rate": 2.935195530726257e-05, + "loss": 0.1256, + "step": 10695 + }, + { + "epoch": 29.877094972067038, + "grad_norm": 0.6148431897163391, + "learning_rate": 2.9351582867783987e-05, + "loss": 0.1183, + "step": 10696 + }, + { + "epoch": 29.879888268156424, + "grad_norm": 0.5569833517074585, + "learning_rate": 2.93512104283054e-05, + "loss": 0.1281, + "step": 10697 + }, + { + "epoch": 29.88268156424581, + "grad_norm": 0.4574359655380249, + "learning_rate": 2.9350837988826816e-05, + "loss": 0.1203, + "step": 10698 + }, + { + "epoch": 29.885474860335197, + "grad_norm": 1.0791407823562622, + "learning_rate": 2.935046554934823e-05, + "loss": 0.1365, + "step": 10699 + }, + { + "epoch": 29.88826815642458, + "grad_norm": 0.6333727836608887, + "learning_rate": 2.9350093109869646e-05, + "loss": 0.1213, + "step": 10700 + }, + { + "epoch": 29.891061452513966, + "grad_norm": 0.7145021557807922, + "learning_rate": 2.9349720670391063e-05, + "loss": 0.1275, + "step": 10701 + }, + { + "epoch": 29.893854748603353, + "grad_norm": 1.4067751169204712, + "learning_rate": 2.934934823091248e-05, + "loss": 0.1399, + "step": 10702 + }, + { + "epoch": 29.89664804469274, + "grad_norm": 0.7626563310623169, + "learning_rate": 2.9348975791433896e-05, + "loss": 0.1298, + "step": 10703 + }, + { + "epoch": 29.899441340782122, + "grad_norm": 0.9082857966423035, + "learning_rate": 2.9348603351955305e-05, + "loss": 0.1316, + "step": 10704 + }, + { + "epoch": 29.90223463687151, + "grad_norm": 0.5436176657676697, + "learning_rate": 2.9348230912476722e-05, + "loss": 0.1174, + "step": 10705 + }, + { + "epoch": 29.905027932960895, + "grad_norm": 4.377463340759277, + "learning_rate": 2.934785847299814e-05, + "loss": 0.1561, + "step": 10706 + }, + { + "epoch": 29.907821229050278, + "grad_norm": 1.3648546934127808, + "learning_rate": 2.9347486033519555e-05, + "loss": 0.1851, + "step": 10707 + }, + { + "epoch": 29.910614525139664, + "grad_norm": 0.40779852867126465, + "learning_rate": 2.9347113594040968e-05, + "loss": 0.1398, + "step": 10708 + }, + { + "epoch": 29.91340782122905, + "grad_norm": 0.5987035632133484, + "learning_rate": 2.9346741154562385e-05, + "loss": 0.144, + "step": 10709 + }, + { + "epoch": 29.916201117318437, + "grad_norm": 0.5447646975517273, + "learning_rate": 2.93463687150838e-05, + "loss": 0.1512, + "step": 10710 + }, + { + "epoch": 29.91899441340782, + "grad_norm": 0.9231038093566895, + "learning_rate": 2.9345996275605214e-05, + "loss": 0.1697, + "step": 10711 + }, + { + "epoch": 29.921787709497206, + "grad_norm": 0.7327036261558533, + "learning_rate": 2.934562383612663e-05, + "loss": 0.1467, + "step": 10712 + }, + { + "epoch": 29.924581005586592, + "grad_norm": 0.7058854699134827, + "learning_rate": 2.9345251396648044e-05, + "loss": 0.1394, + "step": 10713 + }, + { + "epoch": 29.92737430167598, + "grad_norm": 1.8419588804244995, + "learning_rate": 2.934487895716946e-05, + "loss": 0.1825, + "step": 10714 + }, + { + "epoch": 29.93016759776536, + "grad_norm": 0.9127386808395386, + "learning_rate": 2.9344506517690877e-05, + "loss": 0.159, + "step": 10715 + }, + { + "epoch": 29.932960893854748, + "grad_norm": 0.5845786333084106, + "learning_rate": 2.9344134078212294e-05, + "loss": 0.1172, + "step": 10716 + }, + { + "epoch": 29.935754189944134, + "grad_norm": 0.5541092753410339, + "learning_rate": 2.9343761638733703e-05, + "loss": 0.1256, + "step": 10717 + }, + { + "epoch": 29.93854748603352, + "grad_norm": 1.048414707183838, + "learning_rate": 2.934338919925512e-05, + "loss": 0.133, + "step": 10718 + }, + { + "epoch": 29.941340782122904, + "grad_norm": 0.476756751537323, + "learning_rate": 2.9343016759776537e-05, + "loss": 0.1241, + "step": 10719 + }, + { + "epoch": 29.94413407821229, + "grad_norm": 0.9811753630638123, + "learning_rate": 2.9342644320297953e-05, + "loss": 0.1108, + "step": 10720 + }, + { + "epoch": 29.946927374301676, + "grad_norm": 1.0004475116729736, + "learning_rate": 2.934227188081937e-05, + "loss": 0.1229, + "step": 10721 + }, + { + "epoch": 29.949720670391063, + "grad_norm": 1.0050222873687744, + "learning_rate": 2.9341899441340783e-05, + "loss": 0.1297, + "step": 10722 + }, + { + "epoch": 29.952513966480446, + "grad_norm": 0.5971828699111938, + "learning_rate": 2.93415270018622e-05, + "loss": 0.1343, + "step": 10723 + }, + { + "epoch": 29.955307262569832, + "grad_norm": 0.693327009677887, + "learning_rate": 2.9341154562383612e-05, + "loss": 0.1162, + "step": 10724 + }, + { + "epoch": 29.95810055865922, + "grad_norm": 1.4522182941436768, + "learning_rate": 2.934078212290503e-05, + "loss": 0.1343, + "step": 10725 + }, + { + "epoch": 29.960893854748605, + "grad_norm": 0.7131670117378235, + "learning_rate": 2.9340409683426442e-05, + "loss": 0.1317, + "step": 10726 + }, + { + "epoch": 29.963687150837988, + "grad_norm": 0.811343789100647, + "learning_rate": 2.934003724394786e-05, + "loss": 0.1305, + "step": 10727 + }, + { + "epoch": 29.966480446927374, + "grad_norm": 1.2809664011001587, + "learning_rate": 2.9339664804469275e-05, + "loss": 0.1214, + "step": 10728 + }, + { + "epoch": 29.96927374301676, + "grad_norm": 0.7985628247261047, + "learning_rate": 2.9339292364990692e-05, + "loss": 0.1181, + "step": 10729 + }, + { + "epoch": 29.972067039106147, + "grad_norm": 1.3230293989181519, + "learning_rate": 2.9338919925512105e-05, + "loss": 0.1289, + "step": 10730 + }, + { + "epoch": 29.97486033519553, + "grad_norm": 1.247899055480957, + "learning_rate": 2.9338547486033518e-05, + "loss": 0.1486, + "step": 10731 + }, + { + "epoch": 29.977653631284916, + "grad_norm": 1.3585870265960693, + "learning_rate": 2.9338175046554935e-05, + "loss": 0.1548, + "step": 10732 + }, + { + "epoch": 29.980446927374302, + "grad_norm": 1.4187811613082886, + "learning_rate": 2.933780260707635e-05, + "loss": 0.1373, + "step": 10733 + }, + { + "epoch": 29.98324022346369, + "grad_norm": 1.6363410949707031, + "learning_rate": 2.9337430167597768e-05, + "loss": 0.1519, + "step": 10734 + }, + { + "epoch": 29.98603351955307, + "grad_norm": 0.8634381294250488, + "learning_rate": 2.933705772811918e-05, + "loss": 0.1205, + "step": 10735 + }, + { + "epoch": 29.988826815642458, + "grad_norm": 0.6560643315315247, + "learning_rate": 2.9336685288640597e-05, + "loss": 0.1302, + "step": 10736 + }, + { + "epoch": 29.991620111731844, + "grad_norm": 0.5352717638015747, + "learning_rate": 2.933631284916201e-05, + "loss": 0.117, + "step": 10737 + }, + { + "epoch": 29.994413407821227, + "grad_norm": 0.7482457160949707, + "learning_rate": 2.9335940409683427e-05, + "loss": 0.1354, + "step": 10738 + }, + { + "epoch": 29.997206703910614, + "grad_norm": 0.6608924865722656, + "learning_rate": 2.933556797020484e-05, + "loss": 0.1147, + "step": 10739 + }, + { + "epoch": 30.0, + "grad_norm": 1.5212706327438354, + "learning_rate": 2.9335195530726257e-05, + "loss": 0.1551, + "step": 10740 + }, + { + "epoch": 30.002793296089386, + "grad_norm": 0.4389514625072479, + "learning_rate": 2.9334823091247673e-05, + "loss": 0.1764, + "step": 10741 + }, + { + "epoch": 30.00558659217877, + "grad_norm": 1.0548597574234009, + "learning_rate": 2.933445065176909e-05, + "loss": 0.1549, + "step": 10742 + }, + { + "epoch": 30.008379888268156, + "grad_norm": 0.47398829460144043, + "learning_rate": 2.9334078212290506e-05, + "loss": 0.1279, + "step": 10743 + }, + { + "epoch": 30.011173184357542, + "grad_norm": 0.45794180035591125, + "learning_rate": 2.9333705772811916e-05, + "loss": 0.1257, + "step": 10744 + }, + { + "epoch": 30.01396648044693, + "grad_norm": 0.555305540561676, + "learning_rate": 2.9333333333333333e-05, + "loss": 0.1367, + "step": 10745 + }, + { + "epoch": 30.01675977653631, + "grad_norm": 0.9073511362075806, + "learning_rate": 2.933296089385475e-05, + "loss": 0.1332, + "step": 10746 + }, + { + "epoch": 30.019553072625698, + "grad_norm": 0.9497601389884949, + "learning_rate": 2.9332588454376166e-05, + "loss": 0.1404, + "step": 10747 + }, + { + "epoch": 30.022346368715084, + "grad_norm": 0.8284384608268738, + "learning_rate": 2.933221601489758e-05, + "loss": 0.1389, + "step": 10748 + }, + { + "epoch": 30.02513966480447, + "grad_norm": 0.8162916898727417, + "learning_rate": 2.9331843575418995e-05, + "loss": 0.1369, + "step": 10749 + }, + { + "epoch": 30.027932960893853, + "grad_norm": 0.8605371713638306, + "learning_rate": 2.9331471135940412e-05, + "loss": 0.1172, + "step": 10750 + }, + { + "epoch": 30.03072625698324, + "grad_norm": 0.9187843799591064, + "learning_rate": 2.9331098696461825e-05, + "loss": 0.1277, + "step": 10751 + }, + { + "epoch": 30.033519553072626, + "grad_norm": 0.4780046045780182, + "learning_rate": 2.933072625698324e-05, + "loss": 0.1127, + "step": 10752 + }, + { + "epoch": 30.036312849162012, + "grad_norm": 0.9003934264183044, + "learning_rate": 2.9330353817504655e-05, + "loss": 0.1317, + "step": 10753 + }, + { + "epoch": 30.039106145251395, + "grad_norm": 0.9905088543891907, + "learning_rate": 2.932998137802607e-05, + "loss": 0.1271, + "step": 10754 + }, + { + "epoch": 30.04189944134078, + "grad_norm": 0.605608344078064, + "learning_rate": 2.9329608938547488e-05, + "loss": 0.122, + "step": 10755 + }, + { + "epoch": 30.044692737430168, + "grad_norm": 0.6809398531913757, + "learning_rate": 2.9329236499068904e-05, + "loss": 0.1288, + "step": 10756 + }, + { + "epoch": 30.047486033519554, + "grad_norm": 0.6633028388023376, + "learning_rate": 2.9328864059590314e-05, + "loss": 0.1107, + "step": 10757 + }, + { + "epoch": 30.050279329608937, + "grad_norm": 0.9642558097839355, + "learning_rate": 2.932849162011173e-05, + "loss": 0.1036, + "step": 10758 + }, + { + "epoch": 30.053072625698324, + "grad_norm": 0.5509170293807983, + "learning_rate": 2.9328119180633147e-05, + "loss": 0.1329, + "step": 10759 + }, + { + "epoch": 30.05586592178771, + "grad_norm": 0.7056469917297363, + "learning_rate": 2.9327746741154564e-05, + "loss": 0.1239, + "step": 10760 + }, + { + "epoch": 30.058659217877096, + "grad_norm": 0.5772311687469482, + "learning_rate": 2.932737430167598e-05, + "loss": 0.1036, + "step": 10761 + }, + { + "epoch": 30.06145251396648, + "grad_norm": 0.545687198638916, + "learning_rate": 2.9327001862197393e-05, + "loss": 0.1136, + "step": 10762 + }, + { + "epoch": 30.064245810055866, + "grad_norm": 1.60112464427948, + "learning_rate": 2.932662942271881e-05, + "loss": 0.1369, + "step": 10763 + }, + { + "epoch": 30.067039106145252, + "grad_norm": 0.9481348395347595, + "learning_rate": 2.9326256983240223e-05, + "loss": 0.1453, + "step": 10764 + }, + { + "epoch": 30.06983240223464, + "grad_norm": 1.3088154792785645, + "learning_rate": 2.932588454376164e-05, + "loss": 0.1371, + "step": 10765 + }, + { + "epoch": 30.07262569832402, + "grad_norm": 0.5230773687362671, + "learning_rate": 2.9325512104283053e-05, + "loss": 0.1644, + "step": 10766 + }, + { + "epoch": 30.075418994413408, + "grad_norm": 0.39163196086883545, + "learning_rate": 2.932513966480447e-05, + "loss": 0.1659, + "step": 10767 + }, + { + "epoch": 30.078212290502794, + "grad_norm": 0.7163559198379517, + "learning_rate": 2.9324767225325886e-05, + "loss": 0.1415, + "step": 10768 + }, + { + "epoch": 30.08100558659218, + "grad_norm": 0.38742905855178833, + "learning_rate": 2.9324394785847302e-05, + "loss": 0.153, + "step": 10769 + }, + { + "epoch": 30.083798882681563, + "grad_norm": 0.4856635630130768, + "learning_rate": 2.932402234636872e-05, + "loss": 0.1369, + "step": 10770 + }, + { + "epoch": 30.08659217877095, + "grad_norm": 0.5509588122367859, + "learning_rate": 2.932364990689013e-05, + "loss": 0.1217, + "step": 10771 + }, + { + "epoch": 30.089385474860336, + "grad_norm": 0.7062911987304688, + "learning_rate": 2.9323277467411545e-05, + "loss": 0.1239, + "step": 10772 + }, + { + "epoch": 30.092178770949722, + "grad_norm": 0.51119464635849, + "learning_rate": 2.932290502793296e-05, + "loss": 0.128, + "step": 10773 + }, + { + "epoch": 30.094972067039105, + "grad_norm": 0.7665387988090515, + "learning_rate": 2.9322532588454378e-05, + "loss": 0.1255, + "step": 10774 + }, + { + "epoch": 30.09776536312849, + "grad_norm": 0.5074460506439209, + "learning_rate": 2.932216014897579e-05, + "loss": 0.1327, + "step": 10775 + }, + { + "epoch": 30.100558659217878, + "grad_norm": 0.4890937805175781, + "learning_rate": 2.9321787709497208e-05, + "loss": 0.1239, + "step": 10776 + }, + { + "epoch": 30.10335195530726, + "grad_norm": 0.9356038570404053, + "learning_rate": 2.932141527001862e-05, + "loss": 0.1065, + "step": 10777 + }, + { + "epoch": 30.106145251396647, + "grad_norm": 0.5013797283172607, + "learning_rate": 2.9321042830540038e-05, + "loss": 0.1368, + "step": 10778 + }, + { + "epoch": 30.108938547486034, + "grad_norm": 0.5724867582321167, + "learning_rate": 2.9320670391061454e-05, + "loss": 0.1232, + "step": 10779 + }, + { + "epoch": 30.11173184357542, + "grad_norm": 0.5355210900306702, + "learning_rate": 2.9320297951582867e-05, + "loss": 0.1335, + "step": 10780 + }, + { + "epoch": 30.114525139664803, + "grad_norm": 1.0843520164489746, + "learning_rate": 2.9319925512104284e-05, + "loss": 0.131, + "step": 10781 + }, + { + "epoch": 30.11731843575419, + "grad_norm": 0.6976938843727112, + "learning_rate": 2.93195530726257e-05, + "loss": 0.1285, + "step": 10782 + }, + { + "epoch": 30.120111731843576, + "grad_norm": 0.8384305238723755, + "learning_rate": 2.9319180633147117e-05, + "loss": 0.1425, + "step": 10783 + }, + { + "epoch": 30.122905027932962, + "grad_norm": 0.45478296279907227, + "learning_rate": 2.9318808193668527e-05, + "loss": 0.1, + "step": 10784 + }, + { + "epoch": 30.125698324022345, + "grad_norm": 0.8733988404273987, + "learning_rate": 2.9318435754189943e-05, + "loss": 0.1356, + "step": 10785 + }, + { + "epoch": 30.12849162011173, + "grad_norm": 1.2705235481262207, + "learning_rate": 2.931806331471136e-05, + "loss": 0.1187, + "step": 10786 + }, + { + "epoch": 30.131284916201118, + "grad_norm": 0.7307395339012146, + "learning_rate": 2.9317690875232776e-05, + "loss": 0.1011, + "step": 10787 + }, + { + "epoch": 30.134078212290504, + "grad_norm": 1.6086333990097046, + "learning_rate": 2.9317318435754193e-05, + "loss": 0.1025, + "step": 10788 + }, + { + "epoch": 30.136871508379887, + "grad_norm": 1.3234703540802002, + "learning_rate": 2.9316945996275606e-05, + "loss": 0.1526, + "step": 10789 + }, + { + "epoch": 30.139664804469273, + "grad_norm": 1.2869313955307007, + "learning_rate": 2.9316573556797022e-05, + "loss": 0.1875, + "step": 10790 + }, + { + "epoch": 30.14245810055866, + "grad_norm": 0.6259729862213135, + "learning_rate": 2.9316201117318436e-05, + "loss": 0.1356, + "step": 10791 + }, + { + "epoch": 30.145251396648046, + "grad_norm": 0.3444909453392029, + "learning_rate": 2.9315828677839852e-05, + "loss": 0.1413, + "step": 10792 + }, + { + "epoch": 30.14804469273743, + "grad_norm": 0.6505109667778015, + "learning_rate": 2.9315456238361265e-05, + "loss": 0.1276, + "step": 10793 + }, + { + "epoch": 30.150837988826815, + "grad_norm": 0.6537352204322815, + "learning_rate": 2.9315083798882682e-05, + "loss": 0.1218, + "step": 10794 + }, + { + "epoch": 30.1536312849162, + "grad_norm": 0.48368868231773376, + "learning_rate": 2.93147113594041e-05, + "loss": 0.1466, + "step": 10795 + }, + { + "epoch": 30.156424581005588, + "grad_norm": 0.3414881229400635, + "learning_rate": 2.9314338919925515e-05, + "loss": 0.1245, + "step": 10796 + }, + { + "epoch": 30.15921787709497, + "grad_norm": 1.1636502742767334, + "learning_rate": 2.9313966480446928e-05, + "loss": 0.1363, + "step": 10797 + }, + { + "epoch": 30.162011173184357, + "grad_norm": 1.1944196224212646, + "learning_rate": 2.931359404096834e-05, + "loss": 0.1337, + "step": 10798 + }, + { + "epoch": 30.164804469273744, + "grad_norm": 0.4433516561985016, + "learning_rate": 2.9313221601489758e-05, + "loss": 0.1268, + "step": 10799 + }, + { + "epoch": 30.16759776536313, + "grad_norm": 0.7922617793083191, + "learning_rate": 2.9312849162011174e-05, + "loss": 0.1109, + "step": 10800 + }, + { + "epoch": 30.170391061452513, + "grad_norm": 0.567241370677948, + "learning_rate": 2.931247672253259e-05, + "loss": 0.1329, + "step": 10801 + }, + { + "epoch": 30.1731843575419, + "grad_norm": 0.5357523560523987, + "learning_rate": 2.9312104283054004e-05, + "loss": 0.1125, + "step": 10802 + }, + { + "epoch": 30.175977653631286, + "grad_norm": 0.5322859883308411, + "learning_rate": 2.931173184357542e-05, + "loss": 0.1236, + "step": 10803 + }, + { + "epoch": 30.178770949720672, + "grad_norm": 0.7781312465667725, + "learning_rate": 2.9311359404096834e-05, + "loss": 0.1133, + "step": 10804 + }, + { + "epoch": 30.181564245810055, + "grad_norm": 0.6291888952255249, + "learning_rate": 2.931098696461825e-05, + "loss": 0.1282, + "step": 10805 + }, + { + "epoch": 30.18435754189944, + "grad_norm": 0.7900036573410034, + "learning_rate": 2.9310614525139667e-05, + "loss": 0.1344, + "step": 10806 + }, + { + "epoch": 30.187150837988828, + "grad_norm": 0.8931881189346313, + "learning_rate": 2.931024208566108e-05, + "loss": 0.1331, + "step": 10807 + }, + { + "epoch": 30.189944134078214, + "grad_norm": 0.9883906245231628, + "learning_rate": 2.9309869646182496e-05, + "loss": 0.1008, + "step": 10808 + }, + { + "epoch": 30.192737430167597, + "grad_norm": 1.036879301071167, + "learning_rate": 2.9309497206703913e-05, + "loss": 0.0953, + "step": 10809 + }, + { + "epoch": 30.195530726256983, + "grad_norm": 0.7139865159988403, + "learning_rate": 2.930912476722533e-05, + "loss": 0.1015, + "step": 10810 + }, + { + "epoch": 30.19832402234637, + "grad_norm": 0.9333999156951904, + "learning_rate": 2.930875232774674e-05, + "loss": 0.1136, + "step": 10811 + }, + { + "epoch": 30.201117318435756, + "grad_norm": 1.1830919981002808, + "learning_rate": 2.9308379888268156e-05, + "loss": 0.1152, + "step": 10812 + }, + { + "epoch": 30.20391061452514, + "grad_norm": 0.5223439335823059, + "learning_rate": 2.9308007448789572e-05, + "loss": 0.1012, + "step": 10813 + }, + { + "epoch": 30.206703910614525, + "grad_norm": 1.049330472946167, + "learning_rate": 2.930763500931099e-05, + "loss": 0.1118, + "step": 10814 + }, + { + "epoch": 30.20949720670391, + "grad_norm": 1.1240369081497192, + "learning_rate": 2.9307262569832405e-05, + "loss": 0.1674, + "step": 10815 + }, + { + "epoch": 30.212290502793294, + "grad_norm": 0.4339815676212311, + "learning_rate": 2.930689013035382e-05, + "loss": 0.1524, + "step": 10816 + }, + { + "epoch": 30.21508379888268, + "grad_norm": 0.40020614862442017, + "learning_rate": 2.930651769087523e-05, + "loss": 0.1319, + "step": 10817 + }, + { + "epoch": 30.217877094972067, + "grad_norm": 0.6314427852630615, + "learning_rate": 2.9306145251396648e-05, + "loss": 0.1282, + "step": 10818 + }, + { + "epoch": 30.220670391061454, + "grad_norm": 0.47901681065559387, + "learning_rate": 2.9305772811918065e-05, + "loss": 0.1415, + "step": 10819 + }, + { + "epoch": 30.223463687150836, + "grad_norm": 0.4218054711818695, + "learning_rate": 2.9305400372439478e-05, + "loss": 0.1241, + "step": 10820 + }, + { + "epoch": 30.226256983240223, + "grad_norm": 0.8817117810249329, + "learning_rate": 2.9305027932960894e-05, + "loss": 0.1281, + "step": 10821 + }, + { + "epoch": 30.22905027932961, + "grad_norm": 0.5895299315452576, + "learning_rate": 2.930465549348231e-05, + "loss": 0.1452, + "step": 10822 + }, + { + "epoch": 30.231843575418996, + "grad_norm": 0.82712721824646, + "learning_rate": 2.9304283054003727e-05, + "loss": 0.1191, + "step": 10823 + }, + { + "epoch": 30.23463687150838, + "grad_norm": 0.5503485798835754, + "learning_rate": 2.930391061452514e-05, + "loss": 0.1269, + "step": 10824 + }, + { + "epoch": 30.237430167597765, + "grad_norm": 0.5506543517112732, + "learning_rate": 2.9303538175046554e-05, + "loss": 0.1333, + "step": 10825 + }, + { + "epoch": 30.24022346368715, + "grad_norm": 0.9568360447883606, + "learning_rate": 2.930316573556797e-05, + "loss": 0.1377, + "step": 10826 + }, + { + "epoch": 30.243016759776538, + "grad_norm": 0.6476728320121765, + "learning_rate": 2.9302793296089387e-05, + "loss": 0.1254, + "step": 10827 + }, + { + "epoch": 30.24581005586592, + "grad_norm": 0.5175853371620178, + "learning_rate": 2.9302420856610803e-05, + "loss": 0.1382, + "step": 10828 + }, + { + "epoch": 30.248603351955307, + "grad_norm": 0.5694970488548279, + "learning_rate": 2.9302048417132217e-05, + "loss": 0.11, + "step": 10829 + }, + { + "epoch": 30.251396648044693, + "grad_norm": 0.4471593201160431, + "learning_rate": 2.9301675977653633e-05, + "loss": 0.1, + "step": 10830 + }, + { + "epoch": 30.25418994413408, + "grad_norm": 0.7166264653205872, + "learning_rate": 2.9301303538175046e-05, + "loss": 0.1085, + "step": 10831 + }, + { + "epoch": 30.256983240223462, + "grad_norm": 0.4068407416343689, + "learning_rate": 2.9300931098696463e-05, + "loss": 0.11, + "step": 10832 + }, + { + "epoch": 30.25977653631285, + "grad_norm": 1.295580506324768, + "learning_rate": 2.9300558659217876e-05, + "loss": 0.1099, + "step": 10833 + }, + { + "epoch": 30.262569832402235, + "grad_norm": 1.1231085062026978, + "learning_rate": 2.9300186219739292e-05, + "loss": 0.1187, + "step": 10834 + }, + { + "epoch": 30.26536312849162, + "grad_norm": 1.0896168947219849, + "learning_rate": 2.929981378026071e-05, + "loss": 0.1419, + "step": 10835 + }, + { + "epoch": 30.268156424581004, + "grad_norm": 0.5710717439651489, + "learning_rate": 2.9299441340782125e-05, + "loss": 0.1045, + "step": 10836 + }, + { + "epoch": 30.27094972067039, + "grad_norm": 1.1251825094223022, + "learning_rate": 2.929906890130354e-05, + "loss": 0.115, + "step": 10837 + }, + { + "epoch": 30.273743016759777, + "grad_norm": 1.9052543640136719, + "learning_rate": 2.9298696461824952e-05, + "loss": 0.1225, + "step": 10838 + }, + { + "epoch": 30.276536312849164, + "grad_norm": 1.8028885126113892, + "learning_rate": 2.929832402234637e-05, + "loss": 0.1231, + "step": 10839 + }, + { + "epoch": 30.279329608938546, + "grad_norm": 1.8405770063400269, + "learning_rate": 2.9297951582867785e-05, + "loss": 0.1442, + "step": 10840 + }, + { + "epoch": 30.282122905027933, + "grad_norm": 1.868221402168274, + "learning_rate": 2.92975791433892e-05, + "loss": 0.1585, + "step": 10841 + }, + { + "epoch": 30.28491620111732, + "grad_norm": 0.4665271043777466, + "learning_rate": 2.9297206703910615e-05, + "loss": 0.132, + "step": 10842 + }, + { + "epoch": 30.287709497206706, + "grad_norm": 1.1202598810195923, + "learning_rate": 2.929683426443203e-05, + "loss": 0.1519, + "step": 10843 + }, + { + "epoch": 30.29050279329609, + "grad_norm": 0.48046159744262695, + "learning_rate": 2.9296461824953444e-05, + "loss": 0.1504, + "step": 10844 + }, + { + "epoch": 30.293296089385475, + "grad_norm": 0.5506300330162048, + "learning_rate": 2.929608938547486e-05, + "loss": 0.1568, + "step": 10845 + }, + { + "epoch": 30.29608938547486, + "grad_norm": 0.548572301864624, + "learning_rate": 2.9295716945996277e-05, + "loss": 0.1454, + "step": 10846 + }, + { + "epoch": 30.298882681564244, + "grad_norm": 0.4262256324291229, + "learning_rate": 2.929534450651769e-05, + "loss": 0.1425, + "step": 10847 + }, + { + "epoch": 30.30167597765363, + "grad_norm": 0.5511898994445801, + "learning_rate": 2.9294972067039107e-05, + "loss": 0.1238, + "step": 10848 + }, + { + "epoch": 30.304469273743017, + "grad_norm": 0.5512363910675049, + "learning_rate": 2.9294599627560524e-05, + "loss": 0.1256, + "step": 10849 + }, + { + "epoch": 30.307262569832403, + "grad_norm": 0.7440720796585083, + "learning_rate": 2.929422718808194e-05, + "loss": 0.135, + "step": 10850 + }, + { + "epoch": 30.310055865921786, + "grad_norm": 0.8135368824005127, + "learning_rate": 2.929385474860335e-05, + "loss": 0.1295, + "step": 10851 + }, + { + "epoch": 30.312849162011172, + "grad_norm": 0.7327372431755066, + "learning_rate": 2.9293482309124766e-05, + "loss": 0.1192, + "step": 10852 + }, + { + "epoch": 30.31564245810056, + "grad_norm": 0.9743462204933167, + "learning_rate": 2.9293109869646183e-05, + "loss": 0.0999, + "step": 10853 + }, + { + "epoch": 30.318435754189945, + "grad_norm": 0.7415388822555542, + "learning_rate": 2.92927374301676e-05, + "loss": 0.1105, + "step": 10854 + }, + { + "epoch": 30.321229050279328, + "grad_norm": 0.5574302077293396, + "learning_rate": 2.9292364990689016e-05, + "loss": 0.1216, + "step": 10855 + }, + { + "epoch": 30.324022346368714, + "grad_norm": 0.4512060880661011, + "learning_rate": 2.929199255121043e-05, + "loss": 0.136, + "step": 10856 + }, + { + "epoch": 30.3268156424581, + "grad_norm": 0.8332117795944214, + "learning_rate": 2.9291620111731842e-05, + "loss": 0.1167, + "step": 10857 + }, + { + "epoch": 30.329608938547487, + "grad_norm": 0.7633121609687805, + "learning_rate": 2.929124767225326e-05, + "loss": 0.1296, + "step": 10858 + }, + { + "epoch": 30.33240223463687, + "grad_norm": 0.6479407548904419, + "learning_rate": 2.9290875232774675e-05, + "loss": 0.1177, + "step": 10859 + }, + { + "epoch": 30.335195530726256, + "grad_norm": 2.0657358169555664, + "learning_rate": 2.929050279329609e-05, + "loss": 0.1366, + "step": 10860 + }, + { + "epoch": 30.337988826815643, + "grad_norm": 0.8548824787139893, + "learning_rate": 2.9290130353817505e-05, + "loss": 0.125, + "step": 10861 + }, + { + "epoch": 30.34078212290503, + "grad_norm": 1.1212801933288574, + "learning_rate": 2.928975791433892e-05, + "loss": 0.1171, + "step": 10862 + }, + { + "epoch": 30.343575418994412, + "grad_norm": 0.8357458710670471, + "learning_rate": 2.9289385474860338e-05, + "loss": 0.1359, + "step": 10863 + }, + { + "epoch": 30.3463687150838, + "grad_norm": 1.5184500217437744, + "learning_rate": 2.928901303538175e-05, + "loss": 0.1162, + "step": 10864 + }, + { + "epoch": 30.349162011173185, + "grad_norm": 1.9766309261322021, + "learning_rate": 2.9288640595903164e-05, + "loss": 0.1175, + "step": 10865 + }, + { + "epoch": 30.35195530726257, + "grad_norm": 0.8563112020492554, + "learning_rate": 2.928826815642458e-05, + "loss": 0.1675, + "step": 10866 + }, + { + "epoch": 30.354748603351954, + "grad_norm": 0.8108881115913391, + "learning_rate": 2.9287895716945997e-05, + "loss": 0.15, + "step": 10867 + }, + { + "epoch": 30.35754189944134, + "grad_norm": 0.41150376200675964, + "learning_rate": 2.9287523277467414e-05, + "loss": 0.1486, + "step": 10868 + }, + { + "epoch": 30.360335195530727, + "grad_norm": 0.5472218990325928, + "learning_rate": 2.9287150837988827e-05, + "loss": 0.1502, + "step": 10869 + }, + { + "epoch": 30.363128491620113, + "grad_norm": 0.8257962465286255, + "learning_rate": 2.9286778398510244e-05, + "loss": 0.1329, + "step": 10870 + }, + { + "epoch": 30.365921787709496, + "grad_norm": 0.7862657308578491, + "learning_rate": 2.9286405959031657e-05, + "loss": 0.1415, + "step": 10871 + }, + { + "epoch": 30.368715083798882, + "grad_norm": 0.5698227286338806, + "learning_rate": 2.9286033519553073e-05, + "loss": 0.1373, + "step": 10872 + }, + { + "epoch": 30.37150837988827, + "grad_norm": 0.5088604688644409, + "learning_rate": 2.928566108007449e-05, + "loss": 0.1395, + "step": 10873 + }, + { + "epoch": 30.374301675977655, + "grad_norm": 0.5495511889457703, + "learning_rate": 2.9285288640595903e-05, + "loss": 0.1571, + "step": 10874 + }, + { + "epoch": 30.377094972067038, + "grad_norm": 1.5615309476852417, + "learning_rate": 2.928491620111732e-05, + "loss": 0.1496, + "step": 10875 + }, + { + "epoch": 30.379888268156424, + "grad_norm": 0.8063206076622009, + "learning_rate": 2.9284543761638736e-05, + "loss": 0.1505, + "step": 10876 + }, + { + "epoch": 30.38268156424581, + "grad_norm": 0.7514229416847229, + "learning_rate": 2.928417132216015e-05, + "loss": 0.1339, + "step": 10877 + }, + { + "epoch": 30.385474860335197, + "grad_norm": 0.6186892986297607, + "learning_rate": 2.9283798882681562e-05, + "loss": 0.1363, + "step": 10878 + }, + { + "epoch": 30.38826815642458, + "grad_norm": 0.7613847851753235, + "learning_rate": 2.928342644320298e-05, + "loss": 0.1356, + "step": 10879 + }, + { + "epoch": 30.391061452513966, + "grad_norm": 0.47592592239379883, + "learning_rate": 2.9283054003724395e-05, + "loss": 0.1004, + "step": 10880 + }, + { + "epoch": 30.393854748603353, + "grad_norm": 0.658026397228241, + "learning_rate": 2.9282681564245812e-05, + "loss": 0.0955, + "step": 10881 + }, + { + "epoch": 30.39664804469274, + "grad_norm": 0.5682522058486938, + "learning_rate": 2.928230912476723e-05, + "loss": 0.1165, + "step": 10882 + }, + { + "epoch": 30.399441340782122, + "grad_norm": 1.0308337211608887, + "learning_rate": 2.928193668528864e-05, + "loss": 0.1262, + "step": 10883 + }, + { + "epoch": 30.40223463687151, + "grad_norm": 0.7502932548522949, + "learning_rate": 2.9281564245810055e-05, + "loss": 0.1308, + "step": 10884 + }, + { + "epoch": 30.405027932960895, + "grad_norm": 0.4897265136241913, + "learning_rate": 2.928119180633147e-05, + "loss": 0.1036, + "step": 10885 + }, + { + "epoch": 30.407821229050278, + "grad_norm": 1.0478448867797852, + "learning_rate": 2.9280819366852888e-05, + "loss": 0.1339, + "step": 10886 + }, + { + "epoch": 30.410614525139664, + "grad_norm": 0.981343150138855, + "learning_rate": 2.92804469273743e-05, + "loss": 0.1049, + "step": 10887 + }, + { + "epoch": 30.41340782122905, + "grad_norm": 0.7250065803527832, + "learning_rate": 2.9280074487895718e-05, + "loss": 0.101, + "step": 10888 + }, + { + "epoch": 30.416201117318437, + "grad_norm": 0.6474355459213257, + "learning_rate": 2.9279702048417134e-05, + "loss": 0.1264, + "step": 10889 + }, + { + "epoch": 30.41899441340782, + "grad_norm": 2.7161409854888916, + "learning_rate": 2.927932960893855e-05, + "loss": 0.1795, + "step": 10890 + }, + { + "epoch": 30.421787709497206, + "grad_norm": 0.4642121195793152, + "learning_rate": 2.9278957169459964e-05, + "loss": 0.1393, + "step": 10891 + }, + { + "epoch": 30.424581005586592, + "grad_norm": 0.5611233115196228, + "learning_rate": 2.9278584729981377e-05, + "loss": 0.1515, + "step": 10892 + }, + { + "epoch": 30.42737430167598, + "grad_norm": 0.49881866574287415, + "learning_rate": 2.9278212290502793e-05, + "loss": 0.1318, + "step": 10893 + }, + { + "epoch": 30.43016759776536, + "grad_norm": 0.6504005789756775, + "learning_rate": 2.927783985102421e-05, + "loss": 0.1403, + "step": 10894 + }, + { + "epoch": 30.432960893854748, + "grad_norm": 0.5921235680580139, + "learning_rate": 2.9277467411545627e-05, + "loss": 0.1587, + "step": 10895 + }, + { + "epoch": 30.435754189944134, + "grad_norm": 0.4185747504234314, + "learning_rate": 2.927709497206704e-05, + "loss": 0.1238, + "step": 10896 + }, + { + "epoch": 30.43854748603352, + "grad_norm": 0.3684191107749939, + "learning_rate": 2.9276722532588453e-05, + "loss": 0.1341, + "step": 10897 + }, + { + "epoch": 30.441340782122904, + "grad_norm": 0.48470941185951233, + "learning_rate": 2.927635009310987e-05, + "loss": 0.1263, + "step": 10898 + }, + { + "epoch": 30.44413407821229, + "grad_norm": 0.5004915595054626, + "learning_rate": 2.9275977653631286e-05, + "loss": 0.1257, + "step": 10899 + }, + { + "epoch": 30.446927374301676, + "grad_norm": 0.5581909418106079, + "learning_rate": 2.9275605214152702e-05, + "loss": 0.1206, + "step": 10900 + }, + { + "epoch": 30.449720670391063, + "grad_norm": 0.4861087501049042, + "learning_rate": 2.9275232774674116e-05, + "loss": 0.1553, + "step": 10901 + }, + { + "epoch": 30.452513966480446, + "grad_norm": 1.4719380140304565, + "learning_rate": 2.9274860335195532e-05, + "loss": 0.1262, + "step": 10902 + }, + { + "epoch": 30.455307262569832, + "grad_norm": 0.5321345925331116, + "learning_rate": 2.927448789571695e-05, + "loss": 0.1234, + "step": 10903 + }, + { + "epoch": 30.45810055865922, + "grad_norm": 0.7850427627563477, + "learning_rate": 2.9274115456238362e-05, + "loss": 0.1134, + "step": 10904 + }, + { + "epoch": 30.460893854748605, + "grad_norm": 0.6084005236625671, + "learning_rate": 2.9273743016759775e-05, + "loss": 0.1241, + "step": 10905 + }, + { + "epoch": 30.463687150837988, + "grad_norm": 1.177004098892212, + "learning_rate": 2.927337057728119e-05, + "loss": 0.115, + "step": 10906 + }, + { + "epoch": 30.466480446927374, + "grad_norm": 0.6097121238708496, + "learning_rate": 2.9272998137802608e-05, + "loss": 0.1209, + "step": 10907 + }, + { + "epoch": 30.46927374301676, + "grad_norm": 1.4867726564407349, + "learning_rate": 2.9272625698324025e-05, + "loss": 0.1211, + "step": 10908 + }, + { + "epoch": 30.472067039106147, + "grad_norm": 0.931998610496521, + "learning_rate": 2.927225325884544e-05, + "loss": 0.1293, + "step": 10909 + }, + { + "epoch": 30.47486033519553, + "grad_norm": 0.7968723773956299, + "learning_rate": 2.9271880819366854e-05, + "loss": 0.1019, + "step": 10910 + }, + { + "epoch": 30.477653631284916, + "grad_norm": 0.959830105304718, + "learning_rate": 2.9271508379888267e-05, + "loss": 0.1245, + "step": 10911 + }, + { + "epoch": 30.480446927374302, + "grad_norm": 0.8602876663208008, + "learning_rate": 2.9271135940409684e-05, + "loss": 0.1384, + "step": 10912 + }, + { + "epoch": 30.48324022346369, + "grad_norm": 1.2473728656768799, + "learning_rate": 2.92707635009311e-05, + "loss": 0.1406, + "step": 10913 + }, + { + "epoch": 30.48603351955307, + "grad_norm": 0.8826327323913574, + "learning_rate": 2.9270391061452514e-05, + "loss": 0.1189, + "step": 10914 + }, + { + "epoch": 30.488826815642458, + "grad_norm": 2.110964059829712, + "learning_rate": 2.927001862197393e-05, + "loss": 0.145, + "step": 10915 + }, + { + "epoch": 30.491620111731844, + "grad_norm": 0.44271162152290344, + "learning_rate": 2.9269646182495347e-05, + "loss": 0.1624, + "step": 10916 + }, + { + "epoch": 30.49441340782123, + "grad_norm": 0.8379058241844177, + "learning_rate": 2.926927374301676e-05, + "loss": 0.1606, + "step": 10917 + }, + { + "epoch": 30.497206703910614, + "grad_norm": 1.020081639289856, + "learning_rate": 2.9268901303538176e-05, + "loss": 0.135, + "step": 10918 + }, + { + "epoch": 30.5, + "grad_norm": 0.6568204164505005, + "learning_rate": 2.926852886405959e-05, + "loss": 0.1451, + "step": 10919 + }, + { + "epoch": 30.502793296089386, + "grad_norm": 0.8407003879547119, + "learning_rate": 2.9268156424581006e-05, + "loss": 0.1416, + "step": 10920 + }, + { + "epoch": 30.505586592178773, + "grad_norm": 0.7281261086463928, + "learning_rate": 2.9267783985102423e-05, + "loss": 0.1302, + "step": 10921 + }, + { + "epoch": 30.508379888268156, + "grad_norm": 0.45415738224983215, + "learning_rate": 2.926741154562384e-05, + "loss": 0.1434, + "step": 10922 + }, + { + "epoch": 30.511173184357542, + "grad_norm": 0.5068362951278687, + "learning_rate": 2.9267039106145252e-05, + "loss": 0.1277, + "step": 10923 + }, + { + "epoch": 30.51396648044693, + "grad_norm": 0.5227986574172974, + "learning_rate": 2.9266666666666665e-05, + "loss": 0.1118, + "step": 10924 + }, + { + "epoch": 30.51675977653631, + "grad_norm": 1.0815397500991821, + "learning_rate": 2.9266294227188082e-05, + "loss": 0.1215, + "step": 10925 + }, + { + "epoch": 30.519553072625698, + "grad_norm": 0.7598541378974915, + "learning_rate": 2.92659217877095e-05, + "loss": 0.1163, + "step": 10926 + }, + { + "epoch": 30.522346368715084, + "grad_norm": 0.7266508340835571, + "learning_rate": 2.9265549348230915e-05, + "loss": 0.1081, + "step": 10927 + }, + { + "epoch": 30.52513966480447, + "grad_norm": 0.9228708148002625, + "learning_rate": 2.9265176908752328e-05, + "loss": 0.1332, + "step": 10928 + }, + { + "epoch": 30.527932960893853, + "grad_norm": 0.8515303730964661, + "learning_rate": 2.9264804469273745e-05, + "loss": 0.116, + "step": 10929 + }, + { + "epoch": 30.53072625698324, + "grad_norm": 0.6465170383453369, + "learning_rate": 2.926443202979516e-05, + "loss": 0.1052, + "step": 10930 + }, + { + "epoch": 30.533519553072626, + "grad_norm": 0.9780614376068115, + "learning_rate": 2.9264059590316574e-05, + "loss": 0.1052, + "step": 10931 + }, + { + "epoch": 30.536312849162012, + "grad_norm": 0.7662989497184753, + "learning_rate": 2.9263687150837988e-05, + "loss": 0.1164, + "step": 10932 + }, + { + "epoch": 30.539106145251395, + "grad_norm": 1.2347103357315063, + "learning_rate": 2.9263314711359404e-05, + "loss": 0.1287, + "step": 10933 + }, + { + "epoch": 30.54189944134078, + "grad_norm": 0.8335418105125427, + "learning_rate": 2.926294227188082e-05, + "loss": 0.1403, + "step": 10934 + }, + { + "epoch": 30.544692737430168, + "grad_norm": 0.8232866525650024, + "learning_rate": 2.9262569832402237e-05, + "loss": 0.1266, + "step": 10935 + }, + { + "epoch": 30.547486033519554, + "grad_norm": 0.8729012608528137, + "learning_rate": 2.926219739292365e-05, + "loss": 0.1161, + "step": 10936 + }, + { + "epoch": 30.550279329608937, + "grad_norm": 1.079271912574768, + "learning_rate": 2.9261824953445063e-05, + "loss": 0.113, + "step": 10937 + }, + { + "epoch": 30.553072625698324, + "grad_norm": 1.9026201963424683, + "learning_rate": 2.926145251396648e-05, + "loss": 0.124, + "step": 10938 + }, + { + "epoch": 30.55586592178771, + "grad_norm": 0.9166541695594788, + "learning_rate": 2.9261080074487897e-05, + "loss": 0.1118, + "step": 10939 + }, + { + "epoch": 30.558659217877096, + "grad_norm": 1.1458464860916138, + "learning_rate": 2.9260707635009313e-05, + "loss": 0.1578, + "step": 10940 + }, + { + "epoch": 30.56145251396648, + "grad_norm": 0.6625373363494873, + "learning_rate": 2.9260335195530726e-05, + "loss": 0.1577, + "step": 10941 + }, + { + "epoch": 30.564245810055866, + "grad_norm": 0.7648371458053589, + "learning_rate": 2.9259962756052143e-05, + "loss": 0.1445, + "step": 10942 + }, + { + "epoch": 30.567039106145252, + "grad_norm": 0.6653331518173218, + "learning_rate": 2.925959031657356e-05, + "loss": 0.153, + "step": 10943 + }, + { + "epoch": 30.56983240223464, + "grad_norm": 0.6977198719978333, + "learning_rate": 2.9259217877094972e-05, + "loss": 0.147, + "step": 10944 + }, + { + "epoch": 30.57262569832402, + "grad_norm": 0.5753094553947449, + "learning_rate": 2.9258845437616386e-05, + "loss": 0.1555, + "step": 10945 + }, + { + "epoch": 30.575418994413408, + "grad_norm": 0.559613823890686, + "learning_rate": 2.9258472998137802e-05, + "loss": 0.1297, + "step": 10946 + }, + { + "epoch": 30.578212290502794, + "grad_norm": 0.7003271579742432, + "learning_rate": 2.925810055865922e-05, + "loss": 0.1544, + "step": 10947 + }, + { + "epoch": 30.58100558659218, + "grad_norm": 0.6713724136352539, + "learning_rate": 2.9257728119180635e-05, + "loss": 0.1216, + "step": 10948 + }, + { + "epoch": 30.583798882681563, + "grad_norm": 0.6262597441673279, + "learning_rate": 2.9257355679702052e-05, + "loss": 0.1365, + "step": 10949 + }, + { + "epoch": 30.58659217877095, + "grad_norm": 0.7206240296363831, + "learning_rate": 2.9256983240223465e-05, + "loss": 0.1564, + "step": 10950 + }, + { + "epoch": 30.589385474860336, + "grad_norm": 0.500984787940979, + "learning_rate": 2.9256610800744878e-05, + "loss": 0.1399, + "step": 10951 + }, + { + "epoch": 30.592178770949722, + "grad_norm": 0.5471373200416565, + "learning_rate": 2.9256238361266295e-05, + "loss": 0.128, + "step": 10952 + }, + { + "epoch": 30.594972067039105, + "grad_norm": 0.454300194978714, + "learning_rate": 2.925586592178771e-05, + "loss": 0.1315, + "step": 10953 + }, + { + "epoch": 30.59776536312849, + "grad_norm": 0.7245542407035828, + "learning_rate": 2.9255493482309124e-05, + "loss": 0.1321, + "step": 10954 + }, + { + "epoch": 30.600558659217878, + "grad_norm": 0.7796682119369507, + "learning_rate": 2.925512104283054e-05, + "loss": 0.1402, + "step": 10955 + }, + { + "epoch": 30.60335195530726, + "grad_norm": 0.5802582502365112, + "learning_rate": 2.9254748603351957e-05, + "loss": 0.1064, + "step": 10956 + }, + { + "epoch": 30.606145251396647, + "grad_norm": 0.9436224699020386, + "learning_rate": 2.925437616387337e-05, + "loss": 0.1271, + "step": 10957 + }, + { + "epoch": 30.608938547486034, + "grad_norm": 2.5055058002471924, + "learning_rate": 2.9254003724394787e-05, + "loss": 0.1235, + "step": 10958 + }, + { + "epoch": 30.61173184357542, + "grad_norm": 2.9140493869781494, + "learning_rate": 2.92536312849162e-05, + "loss": 0.1163, + "step": 10959 + }, + { + "epoch": 30.614525139664803, + "grad_norm": 1.334713101387024, + "learning_rate": 2.9253258845437617e-05, + "loss": 0.1382, + "step": 10960 + }, + { + "epoch": 30.61731843575419, + "grad_norm": 0.7957078218460083, + "learning_rate": 2.9252886405959033e-05, + "loss": 0.1244, + "step": 10961 + }, + { + "epoch": 30.620111731843576, + "grad_norm": 0.7607032060623169, + "learning_rate": 2.925251396648045e-05, + "loss": 0.1289, + "step": 10962 + }, + { + "epoch": 30.622905027932962, + "grad_norm": 0.7584272027015686, + "learning_rate": 2.9252141527001863e-05, + "loss": 0.1349, + "step": 10963 + }, + { + "epoch": 30.625698324022345, + "grad_norm": 0.9344617128372192, + "learning_rate": 2.9251769087523276e-05, + "loss": 0.1387, + "step": 10964 + }, + { + "epoch": 30.62849162011173, + "grad_norm": 1.7150990962982178, + "learning_rate": 2.9251396648044693e-05, + "loss": 0.1616, + "step": 10965 + }, + { + "epoch": 30.631284916201118, + "grad_norm": 0.3860221803188324, + "learning_rate": 2.925102420856611e-05, + "loss": 0.1342, + "step": 10966 + }, + { + "epoch": 30.634078212290504, + "grad_norm": 0.3596689999103546, + "learning_rate": 2.9250651769087526e-05, + "loss": 0.1436, + "step": 10967 + }, + { + "epoch": 30.636871508379887, + "grad_norm": 0.5161229372024536, + "learning_rate": 2.925027932960894e-05, + "loss": 0.1466, + "step": 10968 + }, + { + "epoch": 30.639664804469273, + "grad_norm": 0.4382386803627014, + "learning_rate": 2.9249906890130355e-05, + "loss": 0.142, + "step": 10969 + }, + { + "epoch": 30.64245810055866, + "grad_norm": 0.44098392128944397, + "learning_rate": 2.9249534450651772e-05, + "loss": 0.1146, + "step": 10970 + }, + { + "epoch": 30.645251396648046, + "grad_norm": 1.0223309993743896, + "learning_rate": 2.9249162011173185e-05, + "loss": 0.1711, + "step": 10971 + }, + { + "epoch": 30.64804469273743, + "grad_norm": 0.7126517295837402, + "learning_rate": 2.9248789571694598e-05, + "loss": 0.1475, + "step": 10972 + }, + { + "epoch": 30.650837988826815, + "grad_norm": 0.7285624146461487, + "learning_rate": 2.9248417132216015e-05, + "loss": 0.1466, + "step": 10973 + }, + { + "epoch": 30.6536312849162, + "grad_norm": 1.3476494550704956, + "learning_rate": 2.924804469273743e-05, + "loss": 0.1295, + "step": 10974 + }, + { + "epoch": 30.656424581005588, + "grad_norm": 0.704710066318512, + "learning_rate": 2.9247672253258848e-05, + "loss": 0.1548, + "step": 10975 + }, + { + "epoch": 30.65921787709497, + "grad_norm": 0.4753796458244324, + "learning_rate": 2.9247299813780264e-05, + "loss": 0.1299, + "step": 10976 + }, + { + "epoch": 30.662011173184357, + "grad_norm": 0.7920334935188293, + "learning_rate": 2.9246927374301674e-05, + "loss": 0.1423, + "step": 10977 + }, + { + "epoch": 30.664804469273744, + "grad_norm": 0.6246286034584045, + "learning_rate": 2.924655493482309e-05, + "loss": 0.1306, + "step": 10978 + }, + { + "epoch": 30.66759776536313, + "grad_norm": 0.6063803434371948, + "learning_rate": 2.9246182495344507e-05, + "loss": 0.123, + "step": 10979 + }, + { + "epoch": 30.670391061452513, + "grad_norm": 0.9164617657661438, + "learning_rate": 2.9245810055865924e-05, + "loss": 0.098, + "step": 10980 + }, + { + "epoch": 30.6731843575419, + "grad_norm": 0.56934654712677, + "learning_rate": 2.9245437616387337e-05, + "loss": 0.1186, + "step": 10981 + }, + { + "epoch": 30.675977653631286, + "grad_norm": 0.5561362504959106, + "learning_rate": 2.9245065176908753e-05, + "loss": 0.1182, + "step": 10982 + }, + { + "epoch": 30.678770949720672, + "grad_norm": 0.7261382341384888, + "learning_rate": 2.924469273743017e-05, + "loss": 0.1178, + "step": 10983 + }, + { + "epoch": 30.681564245810055, + "grad_norm": 1.964884877204895, + "learning_rate": 2.9244320297951583e-05, + "loss": 0.1204, + "step": 10984 + }, + { + "epoch": 30.68435754189944, + "grad_norm": 1.7754485607147217, + "learning_rate": 2.9243947858473e-05, + "loss": 0.1399, + "step": 10985 + }, + { + "epoch": 30.687150837988828, + "grad_norm": 0.5915618538856506, + "learning_rate": 2.9243575418994413e-05, + "loss": 0.0916, + "step": 10986 + }, + { + "epoch": 30.689944134078214, + "grad_norm": 0.7070540189743042, + "learning_rate": 2.924320297951583e-05, + "loss": 0.1261, + "step": 10987 + }, + { + "epoch": 30.692737430167597, + "grad_norm": 0.713016927242279, + "learning_rate": 2.9242830540037246e-05, + "loss": 0.1357, + "step": 10988 + }, + { + "epoch": 30.695530726256983, + "grad_norm": 1.3962632417678833, + "learning_rate": 2.9242458100558662e-05, + "loss": 0.1199, + "step": 10989 + }, + { + "epoch": 30.69832402234637, + "grad_norm": 1.799930214881897, + "learning_rate": 2.9242085661080075e-05, + "loss": 0.1706, + "step": 10990 + }, + { + "epoch": 30.701117318435756, + "grad_norm": 0.38034602999687195, + "learning_rate": 2.924171322160149e-05, + "loss": 0.1504, + "step": 10991 + }, + { + "epoch": 30.70391061452514, + "grad_norm": 0.73758465051651, + "learning_rate": 2.9241340782122905e-05, + "loss": 0.1374, + "step": 10992 + }, + { + "epoch": 30.706703910614525, + "grad_norm": 0.5242428779602051, + "learning_rate": 2.924096834264432e-05, + "loss": 0.1519, + "step": 10993 + }, + { + "epoch": 30.70949720670391, + "grad_norm": 0.4497721493244171, + "learning_rate": 2.9240595903165738e-05, + "loss": 0.1399, + "step": 10994 + }, + { + "epoch": 30.712290502793294, + "grad_norm": 0.4681183099746704, + "learning_rate": 2.924022346368715e-05, + "loss": 0.1777, + "step": 10995 + }, + { + "epoch": 30.71508379888268, + "grad_norm": 0.4304003119468689, + "learning_rate": 2.9239851024208568e-05, + "loss": 0.1298, + "step": 10996 + }, + { + "epoch": 30.717877094972067, + "grad_norm": 1.2363581657409668, + "learning_rate": 2.923947858472998e-05, + "loss": 0.1279, + "step": 10997 + }, + { + "epoch": 30.720670391061454, + "grad_norm": 0.6725017428398132, + "learning_rate": 2.9239106145251398e-05, + "loss": 0.1465, + "step": 10998 + }, + { + "epoch": 30.723463687150836, + "grad_norm": 0.5074008703231812, + "learning_rate": 2.923873370577281e-05, + "loss": 0.1194, + "step": 10999 + }, + { + "epoch": 30.726256983240223, + "grad_norm": 1.1014437675476074, + "learning_rate": 2.9238361266294227e-05, + "loss": 0.1493, + "step": 11000 + }, + { + "epoch": 30.726256983240223, + "eval_cer": 0.19232867445691904, + "eval_loss": 0.2757295072078705, + "eval_runtime": 24.4127, + "eval_samples_per_second": 185.887, + "eval_steps_per_second": 0.614, + "eval_wer": 0.34152677299516776, + "step": 11000 + }, + { + "epoch": 30.72905027932961, + "grad_norm": 0.7551240921020508, + "learning_rate": 2.9237988826815644e-05, + "loss": 0.133, + "step": 11001 + }, + { + "epoch": 30.731843575418996, + "grad_norm": 0.48168542981147766, + "learning_rate": 2.923761638733706e-05, + "loss": 0.1304, + "step": 11002 + }, + { + "epoch": 30.73463687150838, + "grad_norm": 0.49907779693603516, + "learning_rate": 2.9237243947858477e-05, + "loss": 0.108, + "step": 11003 + }, + { + "epoch": 30.737430167597765, + "grad_norm": 0.5818973183631897, + "learning_rate": 2.9236871508379887e-05, + "loss": 0.119, + "step": 11004 + }, + { + "epoch": 30.74022346368715, + "grad_norm": 0.5697147846221924, + "learning_rate": 2.9236499068901303e-05, + "loss": 0.1078, + "step": 11005 + }, + { + "epoch": 30.743016759776538, + "grad_norm": 0.5157054662704468, + "learning_rate": 2.923612662942272e-05, + "loss": 0.1083, + "step": 11006 + }, + { + "epoch": 30.74581005586592, + "grad_norm": 0.5101779699325562, + "learning_rate": 2.9235754189944136e-05, + "loss": 0.14, + "step": 11007 + }, + { + "epoch": 30.748603351955307, + "grad_norm": 0.7372326254844666, + "learning_rate": 2.923538175046555e-05, + "loss": 0.1199, + "step": 11008 + }, + { + "epoch": 30.751396648044693, + "grad_norm": 1.4695380926132202, + "learning_rate": 2.9235009310986966e-05, + "loss": 0.1095, + "step": 11009 + }, + { + "epoch": 30.75418994413408, + "grad_norm": 0.7206535935401917, + "learning_rate": 2.9234636871508382e-05, + "loss": 0.1324, + "step": 11010 + }, + { + "epoch": 30.756983240223462, + "grad_norm": 0.7854723930358887, + "learning_rate": 2.9234264432029796e-05, + "loss": 0.1071, + "step": 11011 + }, + { + "epoch": 30.75977653631285, + "grad_norm": 1.5812605619430542, + "learning_rate": 2.9233891992551212e-05, + "loss": 0.1404, + "step": 11012 + }, + { + "epoch": 30.762569832402235, + "grad_norm": 0.809922993183136, + "learning_rate": 2.9233519553072625e-05, + "loss": 0.1441, + "step": 11013 + }, + { + "epoch": 30.76536312849162, + "grad_norm": 1.4736754894256592, + "learning_rate": 2.9233147113594042e-05, + "loss": 0.1298, + "step": 11014 + }, + { + "epoch": 30.768156424581004, + "grad_norm": 1.6582483053207397, + "learning_rate": 2.923277467411546e-05, + "loss": 0.1879, + "step": 11015 + }, + { + "epoch": 30.77094972067039, + "grad_norm": 0.5692858695983887, + "learning_rate": 2.9232402234636875e-05, + "loss": 0.1336, + "step": 11016 + }, + { + "epoch": 30.773743016759777, + "grad_norm": 0.3737396001815796, + "learning_rate": 2.9232029795158285e-05, + "loss": 0.1438, + "step": 11017 + }, + { + "epoch": 30.776536312849164, + "grad_norm": 0.4468024671077728, + "learning_rate": 2.92316573556797e-05, + "loss": 0.1421, + "step": 11018 + }, + { + "epoch": 30.779329608938546, + "grad_norm": 0.5824623107910156, + "learning_rate": 2.9231284916201118e-05, + "loss": 0.1434, + "step": 11019 + }, + { + "epoch": 30.782122905027933, + "grad_norm": 0.8865940570831299, + "learning_rate": 2.9230912476722534e-05, + "loss": 0.1461, + "step": 11020 + }, + { + "epoch": 30.78491620111732, + "grad_norm": 0.5878365635871887, + "learning_rate": 2.923054003724395e-05, + "loss": 0.1537, + "step": 11021 + }, + { + "epoch": 30.787709497206706, + "grad_norm": 0.5971664786338806, + "learning_rate": 2.9230167597765364e-05, + "loss": 0.1229, + "step": 11022 + }, + { + "epoch": 30.79050279329609, + "grad_norm": 0.45406803488731384, + "learning_rate": 2.922979515828678e-05, + "loss": 0.1296, + "step": 11023 + }, + { + "epoch": 30.793296089385475, + "grad_norm": 0.4770466089248657, + "learning_rate": 2.9229422718808194e-05, + "loss": 0.1287, + "step": 11024 + }, + { + "epoch": 30.79608938547486, + "grad_norm": 0.6409130096435547, + "learning_rate": 2.922905027932961e-05, + "loss": 0.1335, + "step": 11025 + }, + { + "epoch": 30.798882681564244, + "grad_norm": 0.5106768608093262, + "learning_rate": 2.9228677839851023e-05, + "loss": 0.1325, + "step": 11026 + }, + { + "epoch": 30.80167597765363, + "grad_norm": 0.7379457950592041, + "learning_rate": 2.922830540037244e-05, + "loss": 0.1345, + "step": 11027 + }, + { + "epoch": 30.804469273743017, + "grad_norm": 0.6335598826408386, + "learning_rate": 2.9227932960893856e-05, + "loss": 0.1213, + "step": 11028 + }, + { + "epoch": 30.807262569832403, + "grad_norm": 1.1691824197769165, + "learning_rate": 2.9227560521415273e-05, + "loss": 0.1238, + "step": 11029 + }, + { + "epoch": 30.810055865921786, + "grad_norm": 0.9525341391563416, + "learning_rate": 2.9227188081936686e-05, + "loss": 0.1161, + "step": 11030 + }, + { + "epoch": 30.812849162011172, + "grad_norm": 0.49627014994621277, + "learning_rate": 2.92268156424581e-05, + "loss": 0.1178, + "step": 11031 + }, + { + "epoch": 30.81564245810056, + "grad_norm": 0.4216959476470947, + "learning_rate": 2.9226443202979516e-05, + "loss": 0.1312, + "step": 11032 + }, + { + "epoch": 30.818435754189945, + "grad_norm": 0.6517633199691772, + "learning_rate": 2.9226070763500932e-05, + "loss": 0.1328, + "step": 11033 + }, + { + "epoch": 30.821229050279328, + "grad_norm": 0.7244470715522766, + "learning_rate": 2.922569832402235e-05, + "loss": 0.1355, + "step": 11034 + }, + { + "epoch": 30.824022346368714, + "grad_norm": 0.8207921385765076, + "learning_rate": 2.9225325884543762e-05, + "loss": 0.1278, + "step": 11035 + }, + { + "epoch": 30.8268156424581, + "grad_norm": 0.9865902662277222, + "learning_rate": 2.922495344506518e-05, + "loss": 0.1422, + "step": 11036 + }, + { + "epoch": 30.829608938547487, + "grad_norm": 1.171858549118042, + "learning_rate": 2.922458100558659e-05, + "loss": 0.1089, + "step": 11037 + }, + { + "epoch": 30.83240223463687, + "grad_norm": 1.5440659523010254, + "learning_rate": 2.9224208566108008e-05, + "loss": 0.1267, + "step": 11038 + }, + { + "epoch": 30.835195530726256, + "grad_norm": 0.8752313852310181, + "learning_rate": 2.922383612662942e-05, + "loss": 0.1422, + "step": 11039 + }, + { + "epoch": 30.837988826815643, + "grad_norm": 0.7936633825302124, + "learning_rate": 2.9223463687150838e-05, + "loss": 0.166, + "step": 11040 + }, + { + "epoch": 30.84078212290503, + "grad_norm": 0.37372681498527527, + "learning_rate": 2.9223091247672254e-05, + "loss": 0.1599, + "step": 11041 + }, + { + "epoch": 30.843575418994412, + "grad_norm": 0.738226592540741, + "learning_rate": 2.922271880819367e-05, + "loss": 0.1423, + "step": 11042 + }, + { + "epoch": 30.8463687150838, + "grad_norm": 1.181384563446045, + "learning_rate": 2.9222346368715087e-05, + "loss": 0.1544, + "step": 11043 + }, + { + "epoch": 30.849162011173185, + "grad_norm": 0.5932697057723999, + "learning_rate": 2.9221973929236497e-05, + "loss": 0.144, + "step": 11044 + }, + { + "epoch": 30.85195530726257, + "grad_norm": 0.653597891330719, + "learning_rate": 2.9221601489757914e-05, + "loss": 0.1525, + "step": 11045 + }, + { + "epoch": 30.854748603351954, + "grad_norm": 0.512799859046936, + "learning_rate": 2.922122905027933e-05, + "loss": 0.1266, + "step": 11046 + }, + { + "epoch": 30.85754189944134, + "grad_norm": 0.6629946231842041, + "learning_rate": 2.9220856610800747e-05, + "loss": 0.1497, + "step": 11047 + }, + { + "epoch": 30.860335195530727, + "grad_norm": 0.4790458381175995, + "learning_rate": 2.922048417132216e-05, + "loss": 0.1507, + "step": 11048 + }, + { + "epoch": 30.863128491620113, + "grad_norm": 0.5879828333854675, + "learning_rate": 2.9220111731843576e-05, + "loss": 0.1232, + "step": 11049 + }, + { + "epoch": 30.865921787709496, + "grad_norm": 0.5144869685173035, + "learning_rate": 2.9219739292364993e-05, + "loss": 0.1534, + "step": 11050 + }, + { + "epoch": 30.868715083798882, + "grad_norm": 0.4670201241970062, + "learning_rate": 2.9219366852886406e-05, + "loss": 0.1301, + "step": 11051 + }, + { + "epoch": 30.87150837988827, + "grad_norm": 0.7957239747047424, + "learning_rate": 2.9218994413407823e-05, + "loss": 0.159, + "step": 11052 + }, + { + "epoch": 30.874301675977655, + "grad_norm": 0.5923554301261902, + "learning_rate": 2.9218621973929236e-05, + "loss": 0.1187, + "step": 11053 + }, + { + "epoch": 30.877094972067038, + "grad_norm": 0.9406750202178955, + "learning_rate": 2.9218249534450652e-05, + "loss": 0.1247, + "step": 11054 + }, + { + "epoch": 30.879888268156424, + "grad_norm": 1.435390591621399, + "learning_rate": 2.921787709497207e-05, + "loss": 0.1232, + "step": 11055 + }, + { + "epoch": 30.88268156424581, + "grad_norm": 0.6745941638946533, + "learning_rate": 2.9217504655493485e-05, + "loss": 0.1532, + "step": 11056 + }, + { + "epoch": 30.885474860335197, + "grad_norm": 0.6642698049545288, + "learning_rate": 2.9217132216014895e-05, + "loss": 0.1169, + "step": 11057 + }, + { + "epoch": 30.88826815642458, + "grad_norm": 0.9181659817695618, + "learning_rate": 2.9216759776536312e-05, + "loss": 0.1154, + "step": 11058 + }, + { + "epoch": 30.891061452513966, + "grad_norm": 0.5119708776473999, + "learning_rate": 2.9216387337057728e-05, + "loss": 0.115, + "step": 11059 + }, + { + "epoch": 30.893854748603353, + "grad_norm": 1.2788416147232056, + "learning_rate": 2.9216014897579145e-05, + "loss": 0.123, + "step": 11060 + }, + { + "epoch": 30.89664804469274, + "grad_norm": 0.7071927785873413, + "learning_rate": 2.921564245810056e-05, + "loss": 0.1242, + "step": 11061 + }, + { + "epoch": 30.899441340782122, + "grad_norm": 1.4204421043395996, + "learning_rate": 2.9215270018621975e-05, + "loss": 0.1395, + "step": 11062 + }, + { + "epoch": 30.90223463687151, + "grad_norm": 0.9198746681213379, + "learning_rate": 2.921489757914339e-05, + "loss": 0.1303, + "step": 11063 + }, + { + "epoch": 30.905027932960895, + "grad_norm": 1.2826063632965088, + "learning_rate": 2.9214525139664804e-05, + "loss": 0.1322, + "step": 11064 + }, + { + "epoch": 30.907821229050278, + "grad_norm": 1.441228985786438, + "learning_rate": 2.921415270018622e-05, + "loss": 0.1834, + "step": 11065 + }, + { + "epoch": 30.910614525139664, + "grad_norm": 0.4177148640155792, + "learning_rate": 2.9213780260707634e-05, + "loss": 0.1471, + "step": 11066 + }, + { + "epoch": 30.91340782122905, + "grad_norm": 0.6040217876434326, + "learning_rate": 2.921340782122905e-05, + "loss": 0.1787, + "step": 11067 + }, + { + "epoch": 30.916201117318437, + "grad_norm": 0.6897390484809875, + "learning_rate": 2.9213035381750467e-05, + "loss": 0.1673, + "step": 11068 + }, + { + "epoch": 30.91899441340782, + "grad_norm": 0.5190335512161255, + "learning_rate": 2.9212662942271883e-05, + "loss": 0.1317, + "step": 11069 + }, + { + "epoch": 30.921787709497206, + "grad_norm": 0.5291154980659485, + "learning_rate": 2.92122905027933e-05, + "loss": 0.1441, + "step": 11070 + }, + { + "epoch": 30.924581005586592, + "grad_norm": 0.6432134509086609, + "learning_rate": 2.921191806331471e-05, + "loss": 0.18, + "step": 11071 + }, + { + "epoch": 30.92737430167598, + "grad_norm": 0.8610014915466309, + "learning_rate": 2.9211545623836126e-05, + "loss": 0.1233, + "step": 11072 + }, + { + "epoch": 30.93016759776536, + "grad_norm": 0.6640138030052185, + "learning_rate": 2.9211173184357543e-05, + "loss": 0.153, + "step": 11073 + }, + { + "epoch": 30.932960893854748, + "grad_norm": 0.5257469415664673, + "learning_rate": 2.921080074487896e-05, + "loss": 0.1431, + "step": 11074 + }, + { + "epoch": 30.935754189944134, + "grad_norm": 0.4968760311603546, + "learning_rate": 2.9210428305400373e-05, + "loss": 0.1473, + "step": 11075 + }, + { + "epoch": 30.93854748603352, + "grad_norm": 0.7321038246154785, + "learning_rate": 2.921005586592179e-05, + "loss": 0.1254, + "step": 11076 + }, + { + "epoch": 30.941340782122904, + "grad_norm": 0.4648311138153076, + "learning_rate": 2.9209683426443202e-05, + "loss": 0.1163, + "step": 11077 + }, + { + "epoch": 30.94413407821229, + "grad_norm": 0.6952062845230103, + "learning_rate": 2.920931098696462e-05, + "loss": 0.1189, + "step": 11078 + }, + { + "epoch": 30.946927374301676, + "grad_norm": 0.8002465963363647, + "learning_rate": 2.9208938547486035e-05, + "loss": 0.1452, + "step": 11079 + }, + { + "epoch": 30.949720670391063, + "grad_norm": 0.6411896347999573, + "learning_rate": 2.920856610800745e-05, + "loss": 0.1306, + "step": 11080 + }, + { + "epoch": 30.952513966480446, + "grad_norm": 0.5562252402305603, + "learning_rate": 2.9208193668528865e-05, + "loss": 0.1314, + "step": 11081 + }, + { + "epoch": 30.955307262569832, + "grad_norm": 3.422788143157959, + "learning_rate": 2.920782122905028e-05, + "loss": 0.1472, + "step": 11082 + }, + { + "epoch": 30.95810055865922, + "grad_norm": 1.2467641830444336, + "learning_rate": 2.9207448789571698e-05, + "loss": 0.1233, + "step": 11083 + }, + { + "epoch": 30.960893854748605, + "grad_norm": 0.8434257507324219, + "learning_rate": 2.9207076350093108e-05, + "loss": 0.1067, + "step": 11084 + }, + { + "epoch": 30.963687150837988, + "grad_norm": 0.7765750885009766, + "learning_rate": 2.9206703910614524e-05, + "loss": 0.1311, + "step": 11085 + }, + { + "epoch": 30.966480446927374, + "grad_norm": 0.6842732429504395, + "learning_rate": 2.920633147113594e-05, + "loss": 0.1168, + "step": 11086 + }, + { + "epoch": 30.96927374301676, + "grad_norm": 0.8778892755508423, + "learning_rate": 2.9205959031657357e-05, + "loss": 0.1145, + "step": 11087 + }, + { + "epoch": 30.972067039106147, + "grad_norm": 0.807911217212677, + "learning_rate": 2.9205586592178774e-05, + "loss": 0.1047, + "step": 11088 + }, + { + "epoch": 30.97486033519553, + "grad_norm": 0.7919840216636658, + "learning_rate": 2.9205214152700187e-05, + "loss": 0.1141, + "step": 11089 + }, + { + "epoch": 30.977653631284916, + "grad_norm": 1.752585530281067, + "learning_rate": 2.92048417132216e-05, + "loss": 0.1283, + "step": 11090 + }, + { + "epoch": 30.980446927374302, + "grad_norm": 0.6209720969200134, + "learning_rate": 2.9204469273743017e-05, + "loss": 0.1368, + "step": 11091 + }, + { + "epoch": 30.98324022346369, + "grad_norm": 0.38330164551734924, + "learning_rate": 2.9204096834264433e-05, + "loss": 0.141, + "step": 11092 + }, + { + "epoch": 30.98603351955307, + "grad_norm": 0.5753977298736572, + "learning_rate": 2.9203724394785846e-05, + "loss": 0.1322, + "step": 11093 + }, + { + "epoch": 30.988826815642458, + "grad_norm": 0.6430714726448059, + "learning_rate": 2.9203351955307263e-05, + "loss": 0.1436, + "step": 11094 + }, + { + "epoch": 30.991620111731844, + "grad_norm": 3.870543956756592, + "learning_rate": 2.920297951582868e-05, + "loss": 0.13, + "step": 11095 + }, + { + "epoch": 30.994413407821227, + "grad_norm": 0.6240790486335754, + "learning_rate": 2.9202607076350096e-05, + "loss": 0.1059, + "step": 11096 + }, + { + "epoch": 30.997206703910614, + "grad_norm": 0.7928051948547363, + "learning_rate": 2.920223463687151e-05, + "loss": 0.1209, + "step": 11097 + }, + { + "epoch": 31.0, + "grad_norm": 0.6602972745895386, + "learning_rate": 2.9201862197392922e-05, + "loss": 0.1375, + "step": 11098 + }, + { + "epoch": 31.002793296089386, + "grad_norm": 0.48760637640953064, + "learning_rate": 2.920148975791434e-05, + "loss": 0.1718, + "step": 11099 + }, + { + "epoch": 31.00558659217877, + "grad_norm": 0.564897358417511, + "learning_rate": 2.9201117318435755e-05, + "loss": 0.1685, + "step": 11100 + }, + { + "epoch": 31.008379888268156, + "grad_norm": 0.3948921859264374, + "learning_rate": 2.9200744878957172e-05, + "loss": 0.1644, + "step": 11101 + }, + { + "epoch": 31.011173184357542, + "grad_norm": 0.551472008228302, + "learning_rate": 2.9200372439478585e-05, + "loss": 0.1389, + "step": 11102 + }, + { + "epoch": 31.01396648044693, + "grad_norm": 0.5005613565444946, + "learning_rate": 2.92e-05, + "loss": 0.1259, + "step": 11103 + }, + { + "epoch": 31.01675977653631, + "grad_norm": 0.5690290331840515, + "learning_rate": 2.9199627560521415e-05, + "loss": 0.1549, + "step": 11104 + }, + { + "epoch": 31.019553072625698, + "grad_norm": 0.44449639320373535, + "learning_rate": 2.919925512104283e-05, + "loss": 0.1288, + "step": 11105 + }, + { + "epoch": 31.022346368715084, + "grad_norm": 0.48968592286109924, + "learning_rate": 2.9198882681564248e-05, + "loss": 0.1319, + "step": 11106 + }, + { + "epoch": 31.02513966480447, + "grad_norm": 0.5675985217094421, + "learning_rate": 2.919851024208566e-05, + "loss": 0.1185, + "step": 11107 + }, + { + "epoch": 31.027932960893853, + "grad_norm": 0.65571528673172, + "learning_rate": 2.9198137802607078e-05, + "loss": 0.1321, + "step": 11108 + }, + { + "epoch": 31.03072625698324, + "grad_norm": 1.7297593355178833, + "learning_rate": 2.9197765363128494e-05, + "loss": 0.1319, + "step": 11109 + }, + { + "epoch": 31.033519553072626, + "grad_norm": 0.5273337364196777, + "learning_rate": 2.9197392923649907e-05, + "loss": 0.1226, + "step": 11110 + }, + { + "epoch": 31.036312849162012, + "grad_norm": 0.7260307669639587, + "learning_rate": 2.919702048417132e-05, + "loss": 0.1249, + "step": 11111 + }, + { + "epoch": 31.039106145251395, + "grad_norm": 0.3700547218322754, + "learning_rate": 2.9196648044692737e-05, + "loss": 0.1046, + "step": 11112 + }, + { + "epoch": 31.04189944134078, + "grad_norm": 0.7575254440307617, + "learning_rate": 2.9196275605214153e-05, + "loss": 0.1071, + "step": 11113 + }, + { + "epoch": 31.044692737430168, + "grad_norm": 0.5272620916366577, + "learning_rate": 2.919590316573557e-05, + "loss": 0.1188, + "step": 11114 + }, + { + "epoch": 31.047486033519554, + "grad_norm": 0.5048580169677734, + "learning_rate": 2.9195530726256987e-05, + "loss": 0.1048, + "step": 11115 + }, + { + "epoch": 31.050279329608937, + "grad_norm": 0.5638339519500732, + "learning_rate": 2.91951582867784e-05, + "loss": 0.1054, + "step": 11116 + }, + { + "epoch": 31.053072625698324, + "grad_norm": 0.9367580413818359, + "learning_rate": 2.9194785847299813e-05, + "loss": 0.1031, + "step": 11117 + }, + { + "epoch": 31.05586592178771, + "grad_norm": 2.018308162689209, + "learning_rate": 2.919441340782123e-05, + "loss": 0.0904, + "step": 11118 + }, + { + "epoch": 31.058659217877096, + "grad_norm": 2.00692081451416, + "learning_rate": 2.9194040968342646e-05, + "loss": 0.1017, + "step": 11119 + }, + { + "epoch": 31.06145251396648, + "grad_norm": 0.7968051433563232, + "learning_rate": 2.919366852886406e-05, + "loss": 0.0971, + "step": 11120 + }, + { + "epoch": 31.064245810055866, + "grad_norm": 0.6614850759506226, + "learning_rate": 2.9193296089385476e-05, + "loss": 0.0995, + "step": 11121 + }, + { + "epoch": 31.067039106145252, + "grad_norm": 1.1388533115386963, + "learning_rate": 2.9192923649906892e-05, + "loss": 0.1318, + "step": 11122 + }, + { + "epoch": 31.06983240223464, + "grad_norm": 0.9293944239616394, + "learning_rate": 2.919255121042831e-05, + "loss": 0.1516, + "step": 11123 + }, + { + "epoch": 31.07262569832402, + "grad_norm": 5.621562480926514, + "learning_rate": 2.9192178770949722e-05, + "loss": 0.1295, + "step": 11124 + }, + { + "epoch": 31.075418994413408, + "grad_norm": 0.469963401556015, + "learning_rate": 2.9191806331471135e-05, + "loss": 0.1345, + "step": 11125 + }, + { + "epoch": 31.078212290502794, + "grad_norm": 0.623753547668457, + "learning_rate": 2.919143389199255e-05, + "loss": 0.1399, + "step": 11126 + }, + { + "epoch": 31.08100558659218, + "grad_norm": 0.4264279007911682, + "learning_rate": 2.9191061452513968e-05, + "loss": 0.1193, + "step": 11127 + }, + { + "epoch": 31.083798882681563, + "grad_norm": 0.4660208523273468, + "learning_rate": 2.9190689013035385e-05, + "loss": 0.156, + "step": 11128 + }, + { + "epoch": 31.08659217877095, + "grad_norm": 0.3440682590007782, + "learning_rate": 2.9190316573556798e-05, + "loss": 0.1188, + "step": 11129 + }, + { + "epoch": 31.089385474860336, + "grad_norm": 0.49728551506996155, + "learning_rate": 2.918994413407821e-05, + "loss": 0.1308, + "step": 11130 + }, + { + "epoch": 31.092178770949722, + "grad_norm": 0.6862742304801941, + "learning_rate": 2.9189571694599627e-05, + "loss": 0.1225, + "step": 11131 + }, + { + "epoch": 31.094972067039105, + "grad_norm": 0.8641822338104248, + "learning_rate": 2.9189199255121044e-05, + "loss": 0.1319, + "step": 11132 + }, + { + "epoch": 31.09776536312849, + "grad_norm": 0.5784920454025269, + "learning_rate": 2.9188826815642457e-05, + "loss": 0.1271, + "step": 11133 + }, + { + "epoch": 31.100558659217878, + "grad_norm": 0.5532668828964233, + "learning_rate": 2.9188454376163874e-05, + "loss": 0.114, + "step": 11134 + }, + { + "epoch": 31.10335195530726, + "grad_norm": 0.40502455830574036, + "learning_rate": 2.918808193668529e-05, + "loss": 0.1232, + "step": 11135 + }, + { + "epoch": 31.106145251396647, + "grad_norm": 0.5328342318534851, + "learning_rate": 2.9187709497206707e-05, + "loss": 0.1074, + "step": 11136 + }, + { + "epoch": 31.108938547486034, + "grad_norm": 1.0358999967575073, + "learning_rate": 2.918733705772812e-05, + "loss": 0.1108, + "step": 11137 + }, + { + "epoch": 31.11173184357542, + "grad_norm": 0.5908676981925964, + "learning_rate": 2.9186964618249533e-05, + "loss": 0.0955, + "step": 11138 + }, + { + "epoch": 31.114525139664803, + "grad_norm": 0.5693071484565735, + "learning_rate": 2.918659217877095e-05, + "loss": 0.1364, + "step": 11139 + }, + { + "epoch": 31.11731843575419, + "grad_norm": 0.6910690069198608, + "learning_rate": 2.9186219739292366e-05, + "loss": 0.1168, + "step": 11140 + }, + { + "epoch": 31.120111731843576, + "grad_norm": 0.5765452980995178, + "learning_rate": 2.9185847299813783e-05, + "loss": 0.121, + "step": 11141 + }, + { + "epoch": 31.122905027932962, + "grad_norm": 1.8958898782730103, + "learning_rate": 2.9185474860335196e-05, + "loss": 0.1336, + "step": 11142 + }, + { + "epoch": 31.125698324022345, + "grad_norm": 0.4404650330543518, + "learning_rate": 2.9185102420856612e-05, + "loss": 0.1038, + "step": 11143 + }, + { + "epoch": 31.12849162011173, + "grad_norm": 0.5306821465492249, + "learning_rate": 2.9184729981378025e-05, + "loss": 0.1015, + "step": 11144 + }, + { + "epoch": 31.131284916201118, + "grad_norm": 0.5296844244003296, + "learning_rate": 2.9184357541899442e-05, + "loss": 0.0995, + "step": 11145 + }, + { + "epoch": 31.134078212290504, + "grad_norm": 1.117905616760254, + "learning_rate": 2.918398510242086e-05, + "loss": 0.1328, + "step": 11146 + }, + { + "epoch": 31.136871508379887, + "grad_norm": 0.6559600234031677, + "learning_rate": 2.918361266294227e-05, + "loss": 0.1057, + "step": 11147 + }, + { + "epoch": 31.139664804469273, + "grad_norm": 7.9397382736206055, + "learning_rate": 2.9183240223463688e-05, + "loss": 0.1511, + "step": 11148 + }, + { + "epoch": 31.14245810055866, + "grad_norm": 0.42574751377105713, + "learning_rate": 2.9182867783985105e-05, + "loss": 0.1414, + "step": 11149 + }, + { + "epoch": 31.145251396648046, + "grad_norm": 0.7254141569137573, + "learning_rate": 2.9182495344506518e-05, + "loss": 0.1467, + "step": 11150 + }, + { + "epoch": 31.14804469273743, + "grad_norm": 0.3913133144378662, + "learning_rate": 2.918212290502793e-05, + "loss": 0.1201, + "step": 11151 + }, + { + "epoch": 31.150837988826815, + "grad_norm": 0.45497387647628784, + "learning_rate": 2.9181750465549348e-05, + "loss": 0.1398, + "step": 11152 + }, + { + "epoch": 31.1536312849162, + "grad_norm": 0.5041510462760925, + "learning_rate": 2.9181378026070764e-05, + "loss": 0.1195, + "step": 11153 + }, + { + "epoch": 31.156424581005588, + "grad_norm": 0.4097708463668823, + "learning_rate": 2.918100558659218e-05, + "loss": 0.1377, + "step": 11154 + }, + { + "epoch": 31.15921787709497, + "grad_norm": 0.595705509185791, + "learning_rate": 2.9180633147113597e-05, + "loss": 0.1361, + "step": 11155 + }, + { + "epoch": 31.162011173184357, + "grad_norm": 0.43529993295669556, + "learning_rate": 2.918026070763501e-05, + "loss": 0.1313, + "step": 11156 + }, + { + "epoch": 31.164804469273744, + "grad_norm": 0.8735937476158142, + "learning_rate": 2.9179888268156423e-05, + "loss": 0.1133, + "step": 11157 + }, + { + "epoch": 31.16759776536313, + "grad_norm": 0.42274972796440125, + "learning_rate": 2.917951582867784e-05, + "loss": 0.1449, + "step": 11158 + }, + { + "epoch": 31.170391061452513, + "grad_norm": 0.7005733847618103, + "learning_rate": 2.9179143389199256e-05, + "loss": 0.118, + "step": 11159 + }, + { + "epoch": 31.1731843575419, + "grad_norm": 0.5046589374542236, + "learning_rate": 2.917877094972067e-05, + "loss": 0.1323, + "step": 11160 + }, + { + "epoch": 31.175977653631286, + "grad_norm": 0.7535504102706909, + "learning_rate": 2.9178398510242086e-05, + "loss": 0.1277, + "step": 11161 + }, + { + "epoch": 31.178770949720672, + "grad_norm": 0.4841661751270294, + "learning_rate": 2.9178026070763503e-05, + "loss": 0.1264, + "step": 11162 + }, + { + "epoch": 31.181564245810055, + "grad_norm": 0.4449787139892578, + "learning_rate": 2.917765363128492e-05, + "loss": 0.1133, + "step": 11163 + }, + { + "epoch": 31.18435754189944, + "grad_norm": 0.4880211651325226, + "learning_rate": 2.9177281191806332e-05, + "loss": 0.1189, + "step": 11164 + }, + { + "epoch": 31.187150837988828, + "grad_norm": 0.7201145887374878, + "learning_rate": 2.9176908752327746e-05, + "loss": 0.1006, + "step": 11165 + }, + { + "epoch": 31.189944134078214, + "grad_norm": 0.4769839644432068, + "learning_rate": 2.9176536312849162e-05, + "loss": 0.0929, + "step": 11166 + }, + { + "epoch": 31.192737430167597, + "grad_norm": 0.3895951211452484, + "learning_rate": 2.917616387337058e-05, + "loss": 0.1173, + "step": 11167 + }, + { + "epoch": 31.195530726256983, + "grad_norm": 0.8985816836357117, + "learning_rate": 2.9175791433891995e-05, + "loss": 0.1124, + "step": 11168 + }, + { + "epoch": 31.19832402234637, + "grad_norm": 2.2142035961151123, + "learning_rate": 2.9175418994413408e-05, + "loss": 0.1088, + "step": 11169 + }, + { + "epoch": 31.201117318435756, + "grad_norm": 0.7613591551780701, + "learning_rate": 2.917504655493482e-05, + "loss": 0.1486, + "step": 11170 + }, + { + "epoch": 31.20391061452514, + "grad_norm": 0.819797694683075, + "learning_rate": 2.9174674115456238e-05, + "loss": 0.1132, + "step": 11171 + }, + { + "epoch": 31.206703910614525, + "grad_norm": 1.3848758935928345, + "learning_rate": 2.9174301675977654e-05, + "loss": 0.1189, + "step": 11172 + }, + { + "epoch": 31.20949720670391, + "grad_norm": 1.2316429615020752, + "learning_rate": 2.917392923649907e-05, + "loss": 0.1534, + "step": 11173 + }, + { + "epoch": 31.212290502793294, + "grad_norm": 0.5841204524040222, + "learning_rate": 2.9173556797020484e-05, + "loss": 0.1506, + "step": 11174 + }, + { + "epoch": 31.21508379888268, + "grad_norm": 0.5192066431045532, + "learning_rate": 2.91731843575419e-05, + "loss": 0.1343, + "step": 11175 + }, + { + "epoch": 31.217877094972067, + "grad_norm": 0.5138654112815857, + "learning_rate": 2.9172811918063317e-05, + "loss": 0.1394, + "step": 11176 + }, + { + "epoch": 31.220670391061454, + "grad_norm": 0.6665380597114563, + "learning_rate": 2.917243947858473e-05, + "loss": 0.1259, + "step": 11177 + }, + { + "epoch": 31.223463687150836, + "grad_norm": 0.8793531060218811, + "learning_rate": 2.9172067039106144e-05, + "loss": 0.1356, + "step": 11178 + }, + { + "epoch": 31.226256983240223, + "grad_norm": 0.5928593277931213, + "learning_rate": 2.917169459962756e-05, + "loss": 0.1397, + "step": 11179 + }, + { + "epoch": 31.22905027932961, + "grad_norm": 0.6511921882629395, + "learning_rate": 2.9171322160148977e-05, + "loss": 0.1375, + "step": 11180 + }, + { + "epoch": 31.231843575418996, + "grad_norm": 1.014257788658142, + "learning_rate": 2.9170949720670393e-05, + "loss": 0.1351, + "step": 11181 + }, + { + "epoch": 31.23463687150838, + "grad_norm": 1.5401153564453125, + "learning_rate": 2.917057728119181e-05, + "loss": 0.1227, + "step": 11182 + }, + { + "epoch": 31.237430167597765, + "grad_norm": 0.49825388193130493, + "learning_rate": 2.9170204841713223e-05, + "loss": 0.1376, + "step": 11183 + }, + { + "epoch": 31.24022346368715, + "grad_norm": 0.4024001955986023, + "learning_rate": 2.9169832402234636e-05, + "loss": 0.1309, + "step": 11184 + }, + { + "epoch": 31.243016759776538, + "grad_norm": 0.647185206413269, + "learning_rate": 2.9169459962756053e-05, + "loss": 0.11, + "step": 11185 + }, + { + "epoch": 31.24581005586592, + "grad_norm": 0.6265434622764587, + "learning_rate": 2.916908752327747e-05, + "loss": 0.1145, + "step": 11186 + }, + { + "epoch": 31.248603351955307, + "grad_norm": 0.7807632088661194, + "learning_rate": 2.9168715083798882e-05, + "loss": 0.1072, + "step": 11187 + }, + { + "epoch": 31.251396648044693, + "grad_norm": 0.7583564519882202, + "learning_rate": 2.91683426443203e-05, + "loss": 0.119, + "step": 11188 + }, + { + "epoch": 31.25418994413408, + "grad_norm": 0.825576901435852, + "learning_rate": 2.9167970204841715e-05, + "loss": 0.1025, + "step": 11189 + }, + { + "epoch": 31.256983240223462, + "grad_norm": 1.4972659349441528, + "learning_rate": 2.916759776536313e-05, + "loss": 0.1006, + "step": 11190 + }, + { + "epoch": 31.25977653631285, + "grad_norm": 1.3257824182510376, + "learning_rate": 2.9167225325884545e-05, + "loss": 0.0941, + "step": 11191 + }, + { + "epoch": 31.262569832402235, + "grad_norm": 0.9562399387359619, + "learning_rate": 2.9166852886405958e-05, + "loss": 0.1111, + "step": 11192 + }, + { + "epoch": 31.26536312849162, + "grad_norm": 1.1841588020324707, + "learning_rate": 2.9166480446927375e-05, + "loss": 0.1265, + "step": 11193 + }, + { + "epoch": 31.268156424581004, + "grad_norm": 0.8276757597923279, + "learning_rate": 2.916610800744879e-05, + "loss": 0.1282, + "step": 11194 + }, + { + "epoch": 31.27094972067039, + "grad_norm": 0.7408550977706909, + "learning_rate": 2.9165735567970208e-05, + "loss": 0.1279, + "step": 11195 + }, + { + "epoch": 31.273743016759777, + "grad_norm": 1.0708444118499756, + "learning_rate": 2.916536312849162e-05, + "loss": 0.1109, + "step": 11196 + }, + { + "epoch": 31.276536312849164, + "grad_norm": 1.0657051801681519, + "learning_rate": 2.9164990689013034e-05, + "loss": 0.1064, + "step": 11197 + }, + { + "epoch": 31.279329608938546, + "grad_norm": 1.9242299795150757, + "learning_rate": 2.916461824953445e-05, + "loss": 0.1467, + "step": 11198 + }, + { + "epoch": 31.282122905027933, + "grad_norm": 0.8867298364639282, + "learning_rate": 2.9164245810055867e-05, + "loss": 0.1448, + "step": 11199 + }, + { + "epoch": 31.28491620111732, + "grad_norm": 0.40576767921447754, + "learning_rate": 2.9163873370577284e-05, + "loss": 0.1278, + "step": 11200 + }, + { + "epoch": 31.287709497206706, + "grad_norm": 0.5285935401916504, + "learning_rate": 2.9163500931098697e-05, + "loss": 0.1458, + "step": 11201 + }, + { + "epoch": 31.29050279329609, + "grad_norm": 0.47125130891799927, + "learning_rate": 2.9163128491620113e-05, + "loss": 0.1238, + "step": 11202 + }, + { + "epoch": 31.293296089385475, + "grad_norm": 0.3635161817073822, + "learning_rate": 2.916275605214153e-05, + "loss": 0.1294, + "step": 11203 + }, + { + "epoch": 31.29608938547486, + "grad_norm": 0.43909817934036255, + "learning_rate": 2.9162383612662943e-05, + "loss": 0.1108, + "step": 11204 + }, + { + "epoch": 31.298882681564244, + "grad_norm": 0.40635430812835693, + "learning_rate": 2.9162011173184356e-05, + "loss": 0.1363, + "step": 11205 + }, + { + "epoch": 31.30167597765363, + "grad_norm": 0.8468971252441406, + "learning_rate": 2.9161638733705773e-05, + "loss": 0.1315, + "step": 11206 + }, + { + "epoch": 31.304469273743017, + "grad_norm": 0.7817533612251282, + "learning_rate": 2.916126629422719e-05, + "loss": 0.1216, + "step": 11207 + }, + { + "epoch": 31.307262569832403, + "grad_norm": 0.7234654426574707, + "learning_rate": 2.9160893854748606e-05, + "loss": 0.1294, + "step": 11208 + }, + { + "epoch": 31.310055865921786, + "grad_norm": 0.9988625645637512, + "learning_rate": 2.9160521415270022e-05, + "loss": 0.1269, + "step": 11209 + }, + { + "epoch": 31.312849162011172, + "grad_norm": 0.5437435507774353, + "learning_rate": 2.9160148975791432e-05, + "loss": 0.1348, + "step": 11210 + }, + { + "epoch": 31.31564245810056, + "grad_norm": 0.3717828094959259, + "learning_rate": 2.915977653631285e-05, + "loss": 0.1267, + "step": 11211 + }, + { + "epoch": 31.318435754189945, + "grad_norm": 0.5756182074546814, + "learning_rate": 2.9159404096834265e-05, + "loss": 0.1421, + "step": 11212 + }, + { + "epoch": 31.321229050279328, + "grad_norm": 0.6181471943855286, + "learning_rate": 2.915903165735568e-05, + "loss": 0.1464, + "step": 11213 + }, + { + "epoch": 31.324022346368714, + "grad_norm": 0.539965033531189, + "learning_rate": 2.9158659217877095e-05, + "loss": 0.109, + "step": 11214 + }, + { + "epoch": 31.3268156424581, + "grad_norm": 0.4433780610561371, + "learning_rate": 2.915828677839851e-05, + "loss": 0.1127, + "step": 11215 + }, + { + "epoch": 31.329608938547487, + "grad_norm": 0.9737433195114136, + "learning_rate": 2.9157914338919928e-05, + "loss": 0.1083, + "step": 11216 + }, + { + "epoch": 31.33240223463687, + "grad_norm": 0.45517122745513916, + "learning_rate": 2.915754189944134e-05, + "loss": 0.0952, + "step": 11217 + }, + { + "epoch": 31.335195530726256, + "grad_norm": 0.4626595079898834, + "learning_rate": 2.9157169459962758e-05, + "loss": 0.1141, + "step": 11218 + }, + { + "epoch": 31.337988826815643, + "grad_norm": 0.803513765335083, + "learning_rate": 2.915679702048417e-05, + "loss": 0.1353, + "step": 11219 + }, + { + "epoch": 31.34078212290503, + "grad_norm": 0.8342534899711609, + "learning_rate": 2.9156424581005587e-05, + "loss": 0.1275, + "step": 11220 + }, + { + "epoch": 31.343575418994412, + "grad_norm": 1.0899252891540527, + "learning_rate": 2.9156052141527004e-05, + "loss": 0.1134, + "step": 11221 + }, + { + "epoch": 31.3463687150838, + "grad_norm": 0.9254817366600037, + "learning_rate": 2.915567970204842e-05, + "loss": 0.1615, + "step": 11222 + }, + { + "epoch": 31.349162011173185, + "grad_norm": 0.9588879942893982, + "learning_rate": 2.9155307262569833e-05, + "loss": 0.125, + "step": 11223 + }, + { + "epoch": 31.35195530726257, + "grad_norm": 0.3830050528049469, + "learning_rate": 2.9154934823091247e-05, + "loss": 0.1473, + "step": 11224 + }, + { + "epoch": 31.354748603351954, + "grad_norm": 0.829472005367279, + "learning_rate": 2.9154562383612663e-05, + "loss": 0.1425, + "step": 11225 + }, + { + "epoch": 31.35754189944134, + "grad_norm": 0.3419938385486603, + "learning_rate": 2.915418994413408e-05, + "loss": 0.1453, + "step": 11226 + }, + { + "epoch": 31.360335195530727, + "grad_norm": 0.6421366930007935, + "learning_rate": 2.9153817504655493e-05, + "loss": 0.1205, + "step": 11227 + }, + { + "epoch": 31.363128491620113, + "grad_norm": 0.5547392964363098, + "learning_rate": 2.915344506517691e-05, + "loss": 0.1512, + "step": 11228 + }, + { + "epoch": 31.365921787709496, + "grad_norm": 0.41139599680900574, + "learning_rate": 2.9153072625698326e-05, + "loss": 0.1407, + "step": 11229 + }, + { + "epoch": 31.368715083798882, + "grad_norm": 0.8481231927871704, + "learning_rate": 2.915270018621974e-05, + "loss": 0.1313, + "step": 11230 + }, + { + "epoch": 31.37150837988827, + "grad_norm": 0.7011939287185669, + "learning_rate": 2.9152327746741156e-05, + "loss": 0.1204, + "step": 11231 + }, + { + "epoch": 31.374301675977655, + "grad_norm": 0.48114824295043945, + "learning_rate": 2.915195530726257e-05, + "loss": 0.1194, + "step": 11232 + }, + { + "epoch": 31.377094972067038, + "grad_norm": 2.0640876293182373, + "learning_rate": 2.9151582867783985e-05, + "loss": 0.1115, + "step": 11233 + }, + { + "epoch": 31.379888268156424, + "grad_norm": 1.034644603729248, + "learning_rate": 2.9151210428305402e-05, + "loss": 0.1284, + "step": 11234 + }, + { + "epoch": 31.38268156424581, + "grad_norm": 0.4535059928894043, + "learning_rate": 2.9150837988826818e-05, + "loss": 0.1453, + "step": 11235 + }, + { + "epoch": 31.385474860335197, + "grad_norm": 0.7382097840309143, + "learning_rate": 2.915046554934823e-05, + "loss": 0.1257, + "step": 11236 + }, + { + "epoch": 31.38826815642458, + "grad_norm": 0.6040434241294861, + "learning_rate": 2.9150093109869645e-05, + "loss": 0.1236, + "step": 11237 + }, + { + "epoch": 31.391061452513966, + "grad_norm": 0.5116978287696838, + "learning_rate": 2.914972067039106e-05, + "loss": 0.1249, + "step": 11238 + }, + { + "epoch": 31.393854748603353, + "grad_norm": 0.5775531530380249, + "learning_rate": 2.9149348230912478e-05, + "loss": 0.1221, + "step": 11239 + }, + { + "epoch": 31.39664804469274, + "grad_norm": 0.6317403316497803, + "learning_rate": 2.9148975791433894e-05, + "loss": 0.1305, + "step": 11240 + }, + { + "epoch": 31.399441340782122, + "grad_norm": 0.3982962965965271, + "learning_rate": 2.9148603351955307e-05, + "loss": 0.0876, + "step": 11241 + }, + { + "epoch": 31.40223463687151, + "grad_norm": 0.5275425910949707, + "learning_rate": 2.9148230912476724e-05, + "loss": 0.1123, + "step": 11242 + }, + { + "epoch": 31.405027932960895, + "grad_norm": 1.3028274774551392, + "learning_rate": 2.914785847299814e-05, + "loss": 0.1165, + "step": 11243 + }, + { + "epoch": 31.407821229050278, + "grad_norm": 0.7831704020500183, + "learning_rate": 2.9147486033519554e-05, + "loss": 0.1053, + "step": 11244 + }, + { + "epoch": 31.410614525139664, + "grad_norm": 1.4969260692596436, + "learning_rate": 2.9147113594040967e-05, + "loss": 0.1049, + "step": 11245 + }, + { + "epoch": 31.41340782122905, + "grad_norm": 0.8118186593055725, + "learning_rate": 2.9146741154562383e-05, + "loss": 0.1124, + "step": 11246 + }, + { + "epoch": 31.416201117318437, + "grad_norm": 0.9867529273033142, + "learning_rate": 2.91463687150838e-05, + "loss": 0.1367, + "step": 11247 + }, + { + "epoch": 31.41899441340782, + "grad_norm": 1.0371507406234741, + "learning_rate": 2.9145996275605216e-05, + "loss": 0.1248, + "step": 11248 + }, + { + "epoch": 31.421787709497206, + "grad_norm": 0.4341827630996704, + "learning_rate": 2.9145623836126633e-05, + "loss": 0.142, + "step": 11249 + }, + { + "epoch": 31.424581005586592, + "grad_norm": 0.3753769099712372, + "learning_rate": 2.9145251396648043e-05, + "loss": 0.1473, + "step": 11250 + }, + { + "epoch": 31.42737430167598, + "grad_norm": 0.6967756748199463, + "learning_rate": 2.914487895716946e-05, + "loss": 0.1387, + "step": 11251 + }, + { + "epoch": 31.43016759776536, + "grad_norm": 0.7745687365531921, + "learning_rate": 2.9144506517690876e-05, + "loss": 0.1659, + "step": 11252 + }, + { + "epoch": 31.432960893854748, + "grad_norm": 0.4845737814903259, + "learning_rate": 2.9144134078212292e-05, + "loss": 0.1507, + "step": 11253 + }, + { + "epoch": 31.435754189944134, + "grad_norm": 0.6170945763587952, + "learning_rate": 2.9143761638733705e-05, + "loss": 0.1205, + "step": 11254 + }, + { + "epoch": 31.43854748603352, + "grad_norm": 0.6779460906982422, + "learning_rate": 2.9143389199255122e-05, + "loss": 0.1302, + "step": 11255 + }, + { + "epoch": 31.441340782122904, + "grad_norm": 0.5153147578239441, + "learning_rate": 2.914301675977654e-05, + "loss": 0.1551, + "step": 11256 + }, + { + "epoch": 31.44413407821229, + "grad_norm": 0.48761433362960815, + "learning_rate": 2.914264432029795e-05, + "loss": 0.1305, + "step": 11257 + }, + { + "epoch": 31.446927374301676, + "grad_norm": 0.658875048160553, + "learning_rate": 2.9142271880819368e-05, + "loss": 0.1302, + "step": 11258 + }, + { + "epoch": 31.449720670391063, + "grad_norm": 0.5726484060287476, + "learning_rate": 2.914189944134078e-05, + "loss": 0.1182, + "step": 11259 + }, + { + "epoch": 31.452513966480446, + "grad_norm": 0.5406266450881958, + "learning_rate": 2.9141527001862198e-05, + "loss": 0.1157, + "step": 11260 + }, + { + "epoch": 31.455307262569832, + "grad_norm": 1.038266897201538, + "learning_rate": 2.9141154562383614e-05, + "loss": 0.1049, + "step": 11261 + }, + { + "epoch": 31.45810055865922, + "grad_norm": 0.6423352956771851, + "learning_rate": 2.914078212290503e-05, + "loss": 0.1081, + "step": 11262 + }, + { + "epoch": 31.460893854748605, + "grad_norm": 0.5203008055686951, + "learning_rate": 2.9140409683426444e-05, + "loss": 0.1129, + "step": 11263 + }, + { + "epoch": 31.463687150837988, + "grad_norm": 0.7936501502990723, + "learning_rate": 2.9140037243947857e-05, + "loss": 0.1217, + "step": 11264 + }, + { + "epoch": 31.466480446927374, + "grad_norm": 0.5317016243934631, + "learning_rate": 2.9139664804469274e-05, + "loss": 0.0974, + "step": 11265 + }, + { + "epoch": 31.46927374301676, + "grad_norm": 0.9439084529876709, + "learning_rate": 2.913929236499069e-05, + "loss": 0.1013, + "step": 11266 + }, + { + "epoch": 31.472067039106147, + "grad_norm": 0.7311908006668091, + "learning_rate": 2.9138919925512107e-05, + "loss": 0.1259, + "step": 11267 + }, + { + "epoch": 31.47486033519553, + "grad_norm": 0.7596652507781982, + "learning_rate": 2.913854748603352e-05, + "loss": 0.1172, + "step": 11268 + }, + { + "epoch": 31.477653631284916, + "grad_norm": 0.6238402724266052, + "learning_rate": 2.9138175046554936e-05, + "loss": 0.1206, + "step": 11269 + }, + { + "epoch": 31.480446927374302, + "grad_norm": 0.5644253492355347, + "learning_rate": 2.913780260707635e-05, + "loss": 0.0959, + "step": 11270 + }, + { + "epoch": 31.48324022346369, + "grad_norm": 1.0567772388458252, + "learning_rate": 2.9137430167597766e-05, + "loss": 0.1248, + "step": 11271 + }, + { + "epoch": 31.48603351955307, + "grad_norm": 0.6794947981834412, + "learning_rate": 2.913705772811918e-05, + "loss": 0.1283, + "step": 11272 + }, + { + "epoch": 31.488826815642458, + "grad_norm": 1.139284610748291, + "learning_rate": 2.9136685288640596e-05, + "loss": 0.1884, + "step": 11273 + }, + { + "epoch": 31.491620111731844, + "grad_norm": 0.6784729361534119, + "learning_rate": 2.9136312849162012e-05, + "loss": 0.151, + "step": 11274 + }, + { + "epoch": 31.49441340782123, + "grad_norm": 0.8385449051856995, + "learning_rate": 2.913594040968343e-05, + "loss": 0.1488, + "step": 11275 + }, + { + "epoch": 31.497206703910614, + "grad_norm": 0.5701112747192383, + "learning_rate": 2.9135567970204845e-05, + "loss": 0.1287, + "step": 11276 + }, + { + "epoch": 31.5, + "grad_norm": 0.5470848083496094, + "learning_rate": 2.9135195530726255e-05, + "loss": 0.1333, + "step": 11277 + }, + { + "epoch": 31.502793296089386, + "grad_norm": 0.4494888484477997, + "learning_rate": 2.9134823091247672e-05, + "loss": 0.1323, + "step": 11278 + }, + { + "epoch": 31.505586592178773, + "grad_norm": 0.45605167746543884, + "learning_rate": 2.9134450651769088e-05, + "loss": 0.1222, + "step": 11279 + }, + { + "epoch": 31.508379888268156, + "grad_norm": 0.5419972538948059, + "learning_rate": 2.9134078212290505e-05, + "loss": 0.1476, + "step": 11280 + }, + { + "epoch": 31.511173184357542, + "grad_norm": 0.48797333240509033, + "learning_rate": 2.9133705772811918e-05, + "loss": 0.1117, + "step": 11281 + }, + { + "epoch": 31.51396648044693, + "grad_norm": 1.2916760444641113, + "learning_rate": 2.9133333333333334e-05, + "loss": 0.1235, + "step": 11282 + }, + { + "epoch": 31.51675977653631, + "grad_norm": 1.0662083625793457, + "learning_rate": 2.913296089385475e-05, + "loss": 0.1166, + "step": 11283 + }, + { + "epoch": 31.519553072625698, + "grad_norm": 0.4281464219093323, + "learning_rate": 2.9132588454376164e-05, + "loss": 0.1133, + "step": 11284 + }, + { + "epoch": 31.522346368715084, + "grad_norm": 1.1821345090866089, + "learning_rate": 2.913221601489758e-05, + "loss": 0.1116, + "step": 11285 + }, + { + "epoch": 31.52513966480447, + "grad_norm": 0.7619142532348633, + "learning_rate": 2.9131843575418994e-05, + "loss": 0.123, + "step": 11286 + }, + { + "epoch": 31.527932960893853, + "grad_norm": 0.6409810185432434, + "learning_rate": 2.913147113594041e-05, + "loss": 0.1149, + "step": 11287 + }, + { + "epoch": 31.53072625698324, + "grad_norm": 0.35231053829193115, + "learning_rate": 2.9131098696461827e-05, + "loss": 0.0922, + "step": 11288 + }, + { + "epoch": 31.533519553072626, + "grad_norm": 2.018141269683838, + "learning_rate": 2.9130726256983243e-05, + "loss": 0.1203, + "step": 11289 + }, + { + "epoch": 31.536312849162012, + "grad_norm": 1.0238337516784668, + "learning_rate": 2.9130353817504653e-05, + "loss": 0.1154, + "step": 11290 + }, + { + "epoch": 31.539106145251395, + "grad_norm": 0.47847360372543335, + "learning_rate": 2.912998137802607e-05, + "loss": 0.0963, + "step": 11291 + }, + { + "epoch": 31.54189944134078, + "grad_norm": 1.0600863695144653, + "learning_rate": 2.9129608938547486e-05, + "loss": 0.1388, + "step": 11292 + }, + { + "epoch": 31.544692737430168, + "grad_norm": 1.2997592687606812, + "learning_rate": 2.9129236499068903e-05, + "loss": 0.1161, + "step": 11293 + }, + { + "epoch": 31.547486033519554, + "grad_norm": 0.9025399684906006, + "learning_rate": 2.912886405959032e-05, + "loss": 0.1135, + "step": 11294 + }, + { + "epoch": 31.550279329608937, + "grad_norm": 0.7423285245895386, + "learning_rate": 2.9128491620111733e-05, + "loss": 0.1355, + "step": 11295 + }, + { + "epoch": 31.553072625698324, + "grad_norm": 0.951409101486206, + "learning_rate": 2.912811918063315e-05, + "loss": 0.1201, + "step": 11296 + }, + { + "epoch": 31.55586592178771, + "grad_norm": 2.026644706726074, + "learning_rate": 2.9127746741154562e-05, + "loss": 0.1169, + "step": 11297 + }, + { + "epoch": 31.558659217877096, + "grad_norm": 2.0168943405151367, + "learning_rate": 2.912737430167598e-05, + "loss": 0.1453, + "step": 11298 + }, + { + "epoch": 31.56145251396648, + "grad_norm": 0.6606597900390625, + "learning_rate": 2.9127001862197392e-05, + "loss": 0.1271, + "step": 11299 + }, + { + "epoch": 31.564245810055866, + "grad_norm": 0.5266720652580261, + "learning_rate": 2.912662942271881e-05, + "loss": 0.1399, + "step": 11300 + }, + { + "epoch": 31.567039106145252, + "grad_norm": 0.6017882227897644, + "learning_rate": 2.9126256983240225e-05, + "loss": 0.1331, + "step": 11301 + }, + { + "epoch": 31.56983240223464, + "grad_norm": 0.431729257106781, + "learning_rate": 2.912588454376164e-05, + "loss": 0.1322, + "step": 11302 + }, + { + "epoch": 31.57262569832402, + "grad_norm": 0.8739426732063293, + "learning_rate": 2.9125512104283058e-05, + "loss": 0.1307, + "step": 11303 + }, + { + "epoch": 31.575418994413408, + "grad_norm": 0.5554105043411255, + "learning_rate": 2.9125139664804468e-05, + "loss": 0.1304, + "step": 11304 + }, + { + "epoch": 31.578212290502794, + "grad_norm": 0.5587800145149231, + "learning_rate": 2.9124767225325884e-05, + "loss": 0.1459, + "step": 11305 + }, + { + "epoch": 31.58100558659218, + "grad_norm": 0.8306577205657959, + "learning_rate": 2.91243947858473e-05, + "loss": 0.1288, + "step": 11306 + }, + { + "epoch": 31.583798882681563, + "grad_norm": 0.5139084458351135, + "learning_rate": 2.9124022346368717e-05, + "loss": 0.1174, + "step": 11307 + }, + { + "epoch": 31.58659217877095, + "grad_norm": 0.6466895937919617, + "learning_rate": 2.912364990689013e-05, + "loss": 0.1422, + "step": 11308 + }, + { + "epoch": 31.589385474860336, + "grad_norm": 0.819442629814148, + "learning_rate": 2.9123277467411547e-05, + "loss": 0.1184, + "step": 11309 + }, + { + "epoch": 31.592178770949722, + "grad_norm": 0.5551138520240784, + "learning_rate": 2.912290502793296e-05, + "loss": 0.1299, + "step": 11310 + }, + { + "epoch": 31.594972067039105, + "grad_norm": 0.8254559636116028, + "learning_rate": 2.9122532588454377e-05, + "loss": 0.1175, + "step": 11311 + }, + { + "epoch": 31.59776536312849, + "grad_norm": 0.8160226941108704, + "learning_rate": 2.9122160148975793e-05, + "loss": 0.1145, + "step": 11312 + }, + { + "epoch": 31.600558659217878, + "grad_norm": 1.3198901414871216, + "learning_rate": 2.9121787709497206e-05, + "loss": 0.1069, + "step": 11313 + }, + { + "epoch": 31.60335195530726, + "grad_norm": 0.5703531503677368, + "learning_rate": 2.9121415270018623e-05, + "loss": 0.1098, + "step": 11314 + }, + { + "epoch": 31.606145251396647, + "grad_norm": 0.5508652329444885, + "learning_rate": 2.912104283054004e-05, + "loss": 0.1323, + "step": 11315 + }, + { + "epoch": 31.608938547486034, + "grad_norm": 1.1219124794006348, + "learning_rate": 2.9120670391061456e-05, + "loss": 0.1297, + "step": 11316 + }, + { + "epoch": 31.61173184357542, + "grad_norm": 0.9740353226661682, + "learning_rate": 2.9120297951582866e-05, + "loss": 0.1077, + "step": 11317 + }, + { + "epoch": 31.614525139664803, + "grad_norm": 0.579616129398346, + "learning_rate": 2.9119925512104282e-05, + "loss": 0.1206, + "step": 11318 + }, + { + "epoch": 31.61731843575419, + "grad_norm": 0.8031436800956726, + "learning_rate": 2.91195530726257e-05, + "loss": 0.1413, + "step": 11319 + }, + { + "epoch": 31.620111731843576, + "grad_norm": 0.8484092950820923, + "learning_rate": 2.9119180633147115e-05, + "loss": 0.1013, + "step": 11320 + }, + { + "epoch": 31.622905027932962, + "grad_norm": 1.3192845582962036, + "learning_rate": 2.911880819366853e-05, + "loss": 0.1326, + "step": 11321 + }, + { + "epoch": 31.625698324022345, + "grad_norm": 1.288801670074463, + "learning_rate": 2.9118435754189945e-05, + "loss": 0.1267, + "step": 11322 + }, + { + "epoch": 31.62849162011173, + "grad_norm": 1.378746747970581, + "learning_rate": 2.911806331471136e-05, + "loss": 0.1577, + "step": 11323 + }, + { + "epoch": 31.631284916201118, + "grad_norm": 0.4791553020477295, + "learning_rate": 2.9117690875232775e-05, + "loss": 0.1479, + "step": 11324 + }, + { + "epoch": 31.634078212290504, + "grad_norm": 0.5797760486602783, + "learning_rate": 2.911731843575419e-05, + "loss": 0.1464, + "step": 11325 + }, + { + "epoch": 31.636871508379887, + "grad_norm": 0.9512999057769775, + "learning_rate": 2.9116945996275604e-05, + "loss": 0.1337, + "step": 11326 + }, + { + "epoch": 31.639664804469273, + "grad_norm": 0.4446868300437927, + "learning_rate": 2.911657355679702e-05, + "loss": 0.146, + "step": 11327 + }, + { + "epoch": 31.64245810055866, + "grad_norm": 0.5506141185760498, + "learning_rate": 2.9116201117318438e-05, + "loss": 0.1453, + "step": 11328 + }, + { + "epoch": 31.645251396648046, + "grad_norm": 0.716937780380249, + "learning_rate": 2.9115828677839854e-05, + "loss": 0.1442, + "step": 11329 + }, + { + "epoch": 31.64804469273743, + "grad_norm": 0.5722537040710449, + "learning_rate": 2.9115456238361264e-05, + "loss": 0.126, + "step": 11330 + }, + { + "epoch": 31.650837988826815, + "grad_norm": 0.655984103679657, + "learning_rate": 2.911508379888268e-05, + "loss": 0.1531, + "step": 11331 + }, + { + "epoch": 31.6536312849162, + "grad_norm": 0.5576936602592468, + "learning_rate": 2.9114711359404097e-05, + "loss": 0.1153, + "step": 11332 + }, + { + "epoch": 31.656424581005588, + "grad_norm": 0.6796316504478455, + "learning_rate": 2.9114338919925513e-05, + "loss": 0.1161, + "step": 11333 + }, + { + "epoch": 31.65921787709497, + "grad_norm": 0.8578693270683289, + "learning_rate": 2.911396648044693e-05, + "loss": 0.1381, + "step": 11334 + }, + { + "epoch": 31.662011173184357, + "grad_norm": 0.668703556060791, + "learning_rate": 2.9113594040968343e-05, + "loss": 0.1402, + "step": 11335 + }, + { + "epoch": 31.664804469273744, + "grad_norm": 0.4319070875644684, + "learning_rate": 2.911322160148976e-05, + "loss": 0.1212, + "step": 11336 + }, + { + "epoch": 31.66759776536313, + "grad_norm": 1.0139859914779663, + "learning_rate": 2.9112849162011173e-05, + "loss": 0.1423, + "step": 11337 + }, + { + "epoch": 31.670391061452513, + "grad_norm": 0.5772187113761902, + "learning_rate": 2.911247672253259e-05, + "loss": 0.115, + "step": 11338 + }, + { + "epoch": 31.6731843575419, + "grad_norm": 0.6372597217559814, + "learning_rate": 2.9112104283054002e-05, + "loss": 0.1077, + "step": 11339 + }, + { + "epoch": 31.675977653631286, + "grad_norm": 0.7503645420074463, + "learning_rate": 2.911173184357542e-05, + "loss": 0.1048, + "step": 11340 + }, + { + "epoch": 31.678770949720672, + "grad_norm": 0.6373572945594788, + "learning_rate": 2.9111359404096836e-05, + "loss": 0.113, + "step": 11341 + }, + { + "epoch": 31.681564245810055, + "grad_norm": 0.6178513765335083, + "learning_rate": 2.9110986964618252e-05, + "loss": 0.1383, + "step": 11342 + }, + { + "epoch": 31.68435754189944, + "grad_norm": 1.8578002452850342, + "learning_rate": 2.911061452513967e-05, + "loss": 0.1214, + "step": 11343 + }, + { + "epoch": 31.687150837988828, + "grad_norm": 3.616570472717285, + "learning_rate": 2.911024208566108e-05, + "loss": 0.1097, + "step": 11344 + }, + { + "epoch": 31.689944134078214, + "grad_norm": 1.5079550743103027, + "learning_rate": 2.9109869646182495e-05, + "loss": 0.1605, + "step": 11345 + }, + { + "epoch": 31.692737430167597, + "grad_norm": 1.7322475910186768, + "learning_rate": 2.910949720670391e-05, + "loss": 0.1301, + "step": 11346 + }, + { + "epoch": 31.695530726256983, + "grad_norm": 0.8724284768104553, + "learning_rate": 2.9109124767225328e-05, + "loss": 0.1295, + "step": 11347 + }, + { + "epoch": 31.69832402234637, + "grad_norm": 4.2533860206604, + "learning_rate": 2.910875232774674e-05, + "loss": 0.1738, + "step": 11348 + }, + { + "epoch": 31.701117318435756, + "grad_norm": 0.42804980278015137, + "learning_rate": 2.9108379888268158e-05, + "loss": 0.1472, + "step": 11349 + }, + { + "epoch": 31.70391061452514, + "grad_norm": 0.635755181312561, + "learning_rate": 2.910800744878957e-05, + "loss": 0.1519, + "step": 11350 + }, + { + "epoch": 31.706703910614525, + "grad_norm": 0.3821738064289093, + "learning_rate": 2.9107635009310987e-05, + "loss": 0.1184, + "step": 11351 + }, + { + "epoch": 31.70949720670391, + "grad_norm": 0.5296958088874817, + "learning_rate": 2.9107262569832404e-05, + "loss": 0.1503, + "step": 11352 + }, + { + "epoch": 31.712290502793294, + "grad_norm": 0.5313336849212646, + "learning_rate": 2.9106890130353817e-05, + "loss": 0.1275, + "step": 11353 + }, + { + "epoch": 31.71508379888268, + "grad_norm": 0.43348440527915955, + "learning_rate": 2.9106517690875234e-05, + "loss": 0.1584, + "step": 11354 + }, + { + "epoch": 31.717877094972067, + "grad_norm": 0.4813261926174164, + "learning_rate": 2.910614525139665e-05, + "loss": 0.1426, + "step": 11355 + }, + { + "epoch": 31.720670391061454, + "grad_norm": 0.6635324358940125, + "learning_rate": 2.9105772811918067e-05, + "loss": 0.1263, + "step": 11356 + }, + { + "epoch": 31.723463687150836, + "grad_norm": 0.3761115074157715, + "learning_rate": 2.9105400372439476e-05, + "loss": 0.1159, + "step": 11357 + }, + { + "epoch": 31.726256983240223, + "grad_norm": 0.49832892417907715, + "learning_rate": 2.9105027932960893e-05, + "loss": 0.1158, + "step": 11358 + }, + { + "epoch": 31.72905027932961, + "grad_norm": 1.3012127876281738, + "learning_rate": 2.910465549348231e-05, + "loss": 0.1133, + "step": 11359 + }, + { + "epoch": 31.731843575418996, + "grad_norm": 0.400896281003952, + "learning_rate": 2.9104283054003726e-05, + "loss": 0.1137, + "step": 11360 + }, + { + "epoch": 31.73463687150838, + "grad_norm": 0.4874321520328522, + "learning_rate": 2.9103910614525143e-05, + "loss": 0.1299, + "step": 11361 + }, + { + "epoch": 31.737430167597765, + "grad_norm": 0.46231335401535034, + "learning_rate": 2.9103538175046556e-05, + "loss": 0.1266, + "step": 11362 + }, + { + "epoch": 31.74022346368715, + "grad_norm": 0.6087988018989563, + "learning_rate": 2.9103165735567972e-05, + "loss": 0.095, + "step": 11363 + }, + { + "epoch": 31.743016759776538, + "grad_norm": 0.6937782168388367, + "learning_rate": 2.9102793296089385e-05, + "loss": 0.1169, + "step": 11364 + }, + { + "epoch": 31.74581005586592, + "grad_norm": 1.2886908054351807, + "learning_rate": 2.9102420856610802e-05, + "loss": 0.1289, + "step": 11365 + }, + { + "epoch": 31.748603351955307, + "grad_norm": 0.6705276966094971, + "learning_rate": 2.9102048417132215e-05, + "loss": 0.1361, + "step": 11366 + }, + { + "epoch": 31.751396648044693, + "grad_norm": 1.0643564462661743, + "learning_rate": 2.910167597765363e-05, + "loss": 0.1284, + "step": 11367 + }, + { + "epoch": 31.75418994413408, + "grad_norm": 0.754845917224884, + "learning_rate": 2.9101303538175048e-05, + "loss": 0.1239, + "step": 11368 + }, + { + "epoch": 31.756983240223462, + "grad_norm": 0.6815353631973267, + "learning_rate": 2.9100931098696465e-05, + "loss": 0.1187, + "step": 11369 + }, + { + "epoch": 31.75977653631285, + "grad_norm": 0.7705675363540649, + "learning_rate": 2.9100558659217878e-05, + "loss": 0.0894, + "step": 11370 + }, + { + "epoch": 31.762569832402235, + "grad_norm": 1.2816483974456787, + "learning_rate": 2.910018621973929e-05, + "loss": 0.1239, + "step": 11371 + }, + { + "epoch": 31.76536312849162, + "grad_norm": 0.7328917980194092, + "learning_rate": 2.9099813780260707e-05, + "loss": 0.1067, + "step": 11372 + }, + { + "epoch": 31.768156424581004, + "grad_norm": 1.3683905601501465, + "learning_rate": 2.9099441340782124e-05, + "loss": 0.1605, + "step": 11373 + }, + { + "epoch": 31.77094972067039, + "grad_norm": 0.49384063482284546, + "learning_rate": 2.909906890130354e-05, + "loss": 0.1395, + "step": 11374 + }, + { + "epoch": 31.773743016759777, + "grad_norm": 0.4456234574317932, + "learning_rate": 2.9098696461824954e-05, + "loss": 0.1459, + "step": 11375 + }, + { + "epoch": 31.776536312849164, + "grad_norm": 0.43068134784698486, + "learning_rate": 2.909832402234637e-05, + "loss": 0.1183, + "step": 11376 + }, + { + "epoch": 31.779329608938546, + "grad_norm": 0.4045012295246124, + "learning_rate": 2.9097951582867783e-05, + "loss": 0.1324, + "step": 11377 + }, + { + "epoch": 31.782122905027933, + "grad_norm": 0.6231347918510437, + "learning_rate": 2.90975791433892e-05, + "loss": 0.1399, + "step": 11378 + }, + { + "epoch": 31.78491620111732, + "grad_norm": 0.4700104892253876, + "learning_rate": 2.9097206703910616e-05, + "loss": 0.1197, + "step": 11379 + }, + { + "epoch": 31.787709497206706, + "grad_norm": 0.42046988010406494, + "learning_rate": 2.909683426443203e-05, + "loss": 0.1417, + "step": 11380 + }, + { + "epoch": 31.79050279329609, + "grad_norm": 0.48085761070251465, + "learning_rate": 2.9096461824953446e-05, + "loss": 0.1247, + "step": 11381 + }, + { + "epoch": 31.793296089385475, + "grad_norm": 0.4583817422389984, + "learning_rate": 2.9096089385474863e-05, + "loss": 0.1304, + "step": 11382 + }, + { + "epoch": 31.79608938547486, + "grad_norm": 0.7636438012123108, + "learning_rate": 2.909571694599628e-05, + "loss": 0.131, + "step": 11383 + }, + { + "epoch": 31.798882681564244, + "grad_norm": 0.5904665589332581, + "learning_rate": 2.909534450651769e-05, + "loss": 0.1299, + "step": 11384 + }, + { + "epoch": 31.80167597765363, + "grad_norm": 0.8593294024467468, + "learning_rate": 2.9094972067039106e-05, + "loss": 0.1244, + "step": 11385 + }, + { + "epoch": 31.804469273743017, + "grad_norm": 0.44437646865844727, + "learning_rate": 2.9094599627560522e-05, + "loss": 0.1014, + "step": 11386 + }, + { + "epoch": 31.807262569832403, + "grad_norm": 0.7184320092201233, + "learning_rate": 2.909422718808194e-05, + "loss": 0.134, + "step": 11387 + }, + { + "epoch": 31.810055865921786, + "grad_norm": 0.9309052228927612, + "learning_rate": 2.9093854748603355e-05, + "loss": 0.0979, + "step": 11388 + }, + { + "epoch": 31.812849162011172, + "grad_norm": 0.8557402491569519, + "learning_rate": 2.9093482309124768e-05, + "loss": 0.1381, + "step": 11389 + }, + { + "epoch": 31.81564245810056, + "grad_norm": 0.752841591835022, + "learning_rate": 2.909310986964618e-05, + "loss": 0.1332, + "step": 11390 + }, + { + "epoch": 31.818435754189945, + "grad_norm": 0.5481062531471252, + "learning_rate": 2.9092737430167598e-05, + "loss": 0.134, + "step": 11391 + }, + { + "epoch": 31.821229050279328, + "grad_norm": 0.8775721788406372, + "learning_rate": 2.9092364990689014e-05, + "loss": 0.0876, + "step": 11392 + }, + { + "epoch": 31.824022346368714, + "grad_norm": 0.4577224850654602, + "learning_rate": 2.9091992551210428e-05, + "loss": 0.1238, + "step": 11393 + }, + { + "epoch": 31.8268156424581, + "grad_norm": 0.6334572434425354, + "learning_rate": 2.9091620111731844e-05, + "loss": 0.1254, + "step": 11394 + }, + { + "epoch": 31.829608938547487, + "grad_norm": 0.5112488269805908, + "learning_rate": 2.909124767225326e-05, + "loss": 0.0942, + "step": 11395 + }, + { + "epoch": 31.83240223463687, + "grad_norm": 1.0063005685806274, + "learning_rate": 2.9090875232774677e-05, + "loss": 0.1067, + "step": 11396 + }, + { + "epoch": 31.835195530726256, + "grad_norm": 1.1976594924926758, + "learning_rate": 2.909050279329609e-05, + "loss": 0.1074, + "step": 11397 + }, + { + "epoch": 31.837988826815643, + "grad_norm": 0.9952809810638428, + "learning_rate": 2.9090130353817504e-05, + "loss": 0.1789, + "step": 11398 + }, + { + "epoch": 31.84078212290503, + "grad_norm": 0.5269639492034912, + "learning_rate": 2.908975791433892e-05, + "loss": 0.1709, + "step": 11399 + }, + { + "epoch": 31.843575418994412, + "grad_norm": 0.4594257175922394, + "learning_rate": 2.9089385474860337e-05, + "loss": 0.1371, + "step": 11400 + }, + { + "epoch": 31.8463687150838, + "grad_norm": 0.8154005408287048, + "learning_rate": 2.9089013035381753e-05, + "loss": 0.1513, + "step": 11401 + }, + { + "epoch": 31.849162011173185, + "grad_norm": 0.9869506359100342, + "learning_rate": 2.9088640595903166e-05, + "loss": 0.145, + "step": 11402 + }, + { + "epoch": 31.85195530726257, + "grad_norm": 0.4339911639690399, + "learning_rate": 2.9088268156424583e-05, + "loss": 0.1329, + "step": 11403 + }, + { + "epoch": 31.854748603351954, + "grad_norm": 0.548836350440979, + "learning_rate": 2.9087895716945996e-05, + "loss": 0.1549, + "step": 11404 + }, + { + "epoch": 31.85754189944134, + "grad_norm": 0.5507139563560486, + "learning_rate": 2.9087523277467412e-05, + "loss": 0.1312, + "step": 11405 + }, + { + "epoch": 31.860335195530727, + "grad_norm": 0.7226369380950928, + "learning_rate": 2.908715083798883e-05, + "loss": 0.1337, + "step": 11406 + }, + { + "epoch": 31.863128491620113, + "grad_norm": 0.4804980158805847, + "learning_rate": 2.9086778398510242e-05, + "loss": 0.1112, + "step": 11407 + }, + { + "epoch": 31.865921787709496, + "grad_norm": 1.5079970359802246, + "learning_rate": 2.908640595903166e-05, + "loss": 0.1352, + "step": 11408 + }, + { + "epoch": 31.868715083798882, + "grad_norm": 0.6937280893325806, + "learning_rate": 2.9086033519553075e-05, + "loss": 0.124, + "step": 11409 + }, + { + "epoch": 31.87150837988827, + "grad_norm": 0.5681544542312622, + "learning_rate": 2.908566108007449e-05, + "loss": 0.1043, + "step": 11410 + }, + { + "epoch": 31.874301675977655, + "grad_norm": 0.7296879887580872, + "learning_rate": 2.90852886405959e-05, + "loss": 0.1232, + "step": 11411 + }, + { + "epoch": 31.877094972067038, + "grad_norm": 0.5165708661079407, + "learning_rate": 2.9084916201117318e-05, + "loss": 0.1118, + "step": 11412 + }, + { + "epoch": 31.879888268156424, + "grad_norm": 0.5993463397026062, + "learning_rate": 2.9084543761638735e-05, + "loss": 0.1131, + "step": 11413 + }, + { + "epoch": 31.88268156424581, + "grad_norm": 0.8188669681549072, + "learning_rate": 2.908417132216015e-05, + "loss": 0.1115, + "step": 11414 + }, + { + "epoch": 31.885474860335197, + "grad_norm": 0.6825377345085144, + "learning_rate": 2.9083798882681564e-05, + "loss": 0.1307, + "step": 11415 + }, + { + "epoch": 31.88826815642458, + "grad_norm": 1.595306634902954, + "learning_rate": 2.908342644320298e-05, + "loss": 0.1188, + "step": 11416 + }, + { + "epoch": 31.891061452513966, + "grad_norm": 0.6538015604019165, + "learning_rate": 2.9083054003724394e-05, + "loss": 0.1098, + "step": 11417 + }, + { + "epoch": 31.893854748603353, + "grad_norm": 0.768258273601532, + "learning_rate": 2.908268156424581e-05, + "loss": 0.1359, + "step": 11418 + }, + { + "epoch": 31.89664804469274, + "grad_norm": 0.7873837351799011, + "learning_rate": 2.9082309124767227e-05, + "loss": 0.1183, + "step": 11419 + }, + { + "epoch": 31.899441340782122, + "grad_norm": 0.8126099109649658, + "learning_rate": 2.908193668528864e-05, + "loss": 0.1182, + "step": 11420 + }, + { + "epoch": 31.90223463687151, + "grad_norm": 1.7466037273406982, + "learning_rate": 2.9081564245810057e-05, + "loss": 0.0973, + "step": 11421 + }, + { + "epoch": 31.905027932960895, + "grad_norm": 1.1594281196594238, + "learning_rate": 2.9081191806331473e-05, + "loss": 0.1379, + "step": 11422 + }, + { + "epoch": 31.907821229050278, + "grad_norm": 1.8684005737304688, + "learning_rate": 2.908081936685289e-05, + "loss": 0.1461, + "step": 11423 + }, + { + "epoch": 31.910614525139664, + "grad_norm": 0.5899036526679993, + "learning_rate": 2.90804469273743e-05, + "loss": 0.1512, + "step": 11424 + }, + { + "epoch": 31.91340782122905, + "grad_norm": 0.4627734422683716, + "learning_rate": 2.9080074487895716e-05, + "loss": 0.1421, + "step": 11425 + }, + { + "epoch": 31.916201117318437, + "grad_norm": 0.6056705713272095, + "learning_rate": 2.9079702048417133e-05, + "loss": 0.1349, + "step": 11426 + }, + { + "epoch": 31.91899441340782, + "grad_norm": 0.6041312217712402, + "learning_rate": 2.907932960893855e-05, + "loss": 0.1351, + "step": 11427 + }, + { + "epoch": 31.921787709497206, + "grad_norm": 0.6099762320518494, + "learning_rate": 2.9078957169459966e-05, + "loss": 0.1291, + "step": 11428 + }, + { + "epoch": 31.924581005586592, + "grad_norm": 0.7641177177429199, + "learning_rate": 2.907858472998138e-05, + "loss": 0.1493, + "step": 11429 + }, + { + "epoch": 31.92737430167598, + "grad_norm": 0.6777753233909607, + "learning_rate": 2.9078212290502792e-05, + "loss": 0.1509, + "step": 11430 + }, + { + "epoch": 31.93016759776536, + "grad_norm": 0.42045456171035767, + "learning_rate": 2.907783985102421e-05, + "loss": 0.119, + "step": 11431 + }, + { + "epoch": 31.932960893854748, + "grad_norm": 1.2904971837997437, + "learning_rate": 2.9077467411545625e-05, + "loss": 0.1346, + "step": 11432 + }, + { + "epoch": 31.935754189944134, + "grad_norm": 0.4699160158634186, + "learning_rate": 2.9077094972067038e-05, + "loss": 0.1392, + "step": 11433 + }, + { + "epoch": 31.93854748603352, + "grad_norm": 0.47551435232162476, + "learning_rate": 2.9076722532588455e-05, + "loss": 0.1418, + "step": 11434 + }, + { + "epoch": 31.941340782122904, + "grad_norm": 0.5792334675788879, + "learning_rate": 2.907635009310987e-05, + "loss": 0.1282, + "step": 11435 + }, + { + "epoch": 31.94413407821229, + "grad_norm": 0.7025512456893921, + "learning_rate": 2.9075977653631288e-05, + "loss": 0.1113, + "step": 11436 + }, + { + "epoch": 31.946927374301676, + "grad_norm": 0.4939000904560089, + "learning_rate": 2.90756052141527e-05, + "loss": 0.1259, + "step": 11437 + }, + { + "epoch": 31.949720670391063, + "grad_norm": 0.9503028988838196, + "learning_rate": 2.9075232774674114e-05, + "loss": 0.12, + "step": 11438 + }, + { + "epoch": 31.952513966480446, + "grad_norm": 0.5583747029304504, + "learning_rate": 2.907486033519553e-05, + "loss": 0.1121, + "step": 11439 + }, + { + "epoch": 31.955307262569832, + "grad_norm": 0.7051616311073303, + "learning_rate": 2.9074487895716947e-05, + "loss": 0.1285, + "step": 11440 + }, + { + "epoch": 31.95810055865922, + "grad_norm": 0.6712101697921753, + "learning_rate": 2.9074115456238364e-05, + "loss": 0.1237, + "step": 11441 + }, + { + "epoch": 31.960893854748605, + "grad_norm": 0.5988596677780151, + "learning_rate": 2.9073743016759777e-05, + "loss": 0.1144, + "step": 11442 + }, + { + "epoch": 31.963687150837988, + "grad_norm": 1.1294323205947876, + "learning_rate": 2.9073370577281193e-05, + "loss": 0.1185, + "step": 11443 + }, + { + "epoch": 31.966480446927374, + "grad_norm": 0.6354877352714539, + "learning_rate": 2.9072998137802607e-05, + "loss": 0.0849, + "step": 11444 + }, + { + "epoch": 31.96927374301676, + "grad_norm": 1.2091822624206543, + "learning_rate": 2.9072625698324023e-05, + "loss": 0.1126, + "step": 11445 + }, + { + "epoch": 31.972067039106147, + "grad_norm": 2.2007827758789062, + "learning_rate": 2.907225325884544e-05, + "loss": 0.1505, + "step": 11446 + }, + { + "epoch": 31.97486033519553, + "grad_norm": 0.884936511516571, + "learning_rate": 2.9071880819366853e-05, + "loss": 0.0961, + "step": 11447 + }, + { + "epoch": 31.977653631284916, + "grad_norm": 1.7838548421859741, + "learning_rate": 2.907150837988827e-05, + "loss": 0.1533, + "step": 11448 + }, + { + "epoch": 31.980446927374302, + "grad_norm": 1.1893609762191772, + "learning_rate": 2.9071135940409686e-05, + "loss": 0.1543, + "step": 11449 + }, + { + "epoch": 31.98324022346369, + "grad_norm": 0.6737279295921326, + "learning_rate": 2.90707635009311e-05, + "loss": 0.1571, + "step": 11450 + }, + { + "epoch": 31.98603351955307, + "grad_norm": 0.8352808356285095, + "learning_rate": 2.9070391061452512e-05, + "loss": 0.1408, + "step": 11451 + }, + { + "epoch": 31.988826815642458, + "grad_norm": 0.6774897575378418, + "learning_rate": 2.907001862197393e-05, + "loss": 0.1469, + "step": 11452 + }, + { + "epoch": 31.991620111731844, + "grad_norm": 0.6046210527420044, + "learning_rate": 2.9069646182495345e-05, + "loss": 0.1199, + "step": 11453 + }, + { + "epoch": 31.994413407821227, + "grad_norm": 0.4696289002895355, + "learning_rate": 2.9069273743016762e-05, + "loss": 0.1168, + "step": 11454 + }, + { + "epoch": 31.997206703910614, + "grad_norm": 0.7318704128265381, + "learning_rate": 2.9068901303538178e-05, + "loss": 0.1161, + "step": 11455 + }, + { + "epoch": 32.0, + "grad_norm": 1.1841037273406982, + "learning_rate": 2.906852886405959e-05, + "loss": 0.1618, + "step": 11456 + }, + { + "epoch": 32.002793296089386, + "grad_norm": 0.5360762476921082, + "learning_rate": 2.9068156424581005e-05, + "loss": 0.1351, + "step": 11457 + }, + { + "epoch": 32.00558659217877, + "grad_norm": 0.5302725434303284, + "learning_rate": 2.906778398510242e-05, + "loss": 0.1532, + "step": 11458 + }, + { + "epoch": 32.00837988826816, + "grad_norm": 0.5450730323791504, + "learning_rate": 2.9067411545623838e-05, + "loss": 0.1369, + "step": 11459 + }, + { + "epoch": 32.01117318435754, + "grad_norm": 1.0436546802520752, + "learning_rate": 2.906703910614525e-05, + "loss": 0.1298, + "step": 11460 + }, + { + "epoch": 32.013966480446925, + "grad_norm": 0.48300543427467346, + "learning_rate": 2.9066666666666667e-05, + "loss": 0.1137, + "step": 11461 + }, + { + "epoch": 32.01675977653631, + "grad_norm": 0.6465678215026855, + "learning_rate": 2.9066294227188084e-05, + "loss": 0.1338, + "step": 11462 + }, + { + "epoch": 32.0195530726257, + "grad_norm": 0.5027114152908325, + "learning_rate": 2.90659217877095e-05, + "loss": 0.1312, + "step": 11463 + }, + { + "epoch": 32.022346368715084, + "grad_norm": 0.7549874782562256, + "learning_rate": 2.9065549348230914e-05, + "loss": 0.1187, + "step": 11464 + }, + { + "epoch": 32.02513966480447, + "grad_norm": 1.094934105873108, + "learning_rate": 2.9065176908752327e-05, + "loss": 0.1041, + "step": 11465 + }, + { + "epoch": 32.02793296089386, + "grad_norm": 0.7447637915611267, + "learning_rate": 2.9064804469273743e-05, + "loss": 0.1416, + "step": 11466 + }, + { + "epoch": 32.03072625698324, + "grad_norm": 0.6258590221405029, + "learning_rate": 2.906443202979516e-05, + "loss": 0.139, + "step": 11467 + }, + { + "epoch": 32.03351955307262, + "grad_norm": 0.6215617060661316, + "learning_rate": 2.9064059590316576e-05, + "loss": 0.1362, + "step": 11468 + }, + { + "epoch": 32.03631284916201, + "grad_norm": 0.6026117205619812, + "learning_rate": 2.906368715083799e-05, + "loss": 0.1125, + "step": 11469 + }, + { + "epoch": 32.039106145251395, + "grad_norm": 0.3941272497177124, + "learning_rate": 2.9063314711359403e-05, + "loss": 0.1157, + "step": 11470 + }, + { + "epoch": 32.04189944134078, + "grad_norm": 0.5017328262329102, + "learning_rate": 2.906294227188082e-05, + "loss": 0.0998, + "step": 11471 + }, + { + "epoch": 32.04469273743017, + "grad_norm": 0.7573413252830505, + "learning_rate": 2.9062569832402236e-05, + "loss": 0.1039, + "step": 11472 + }, + { + "epoch": 32.047486033519554, + "grad_norm": 0.6481353640556335, + "learning_rate": 2.9062197392923652e-05, + "loss": 0.0928, + "step": 11473 + }, + { + "epoch": 32.05027932960894, + "grad_norm": 0.5883134603500366, + "learning_rate": 2.9061824953445065e-05, + "loss": 0.1217, + "step": 11474 + }, + { + "epoch": 32.05307262569833, + "grad_norm": 0.7470808625221252, + "learning_rate": 2.9061452513966482e-05, + "loss": 0.1091, + "step": 11475 + }, + { + "epoch": 32.055865921787706, + "grad_norm": 1.240668773651123, + "learning_rate": 2.90610800744879e-05, + "loss": 0.1234, + "step": 11476 + }, + { + "epoch": 32.05865921787709, + "grad_norm": 0.9502052068710327, + "learning_rate": 2.906070763500931e-05, + "loss": 0.1235, + "step": 11477 + }, + { + "epoch": 32.06145251396648, + "grad_norm": 0.6283586025238037, + "learning_rate": 2.9060335195530725e-05, + "loss": 0.0876, + "step": 11478 + }, + { + "epoch": 32.064245810055866, + "grad_norm": 0.8814488053321838, + "learning_rate": 2.905996275605214e-05, + "loss": 0.114, + "step": 11479 + }, + { + "epoch": 32.06703910614525, + "grad_norm": 0.9993333220481873, + "learning_rate": 2.9059590316573558e-05, + "loss": 0.1041, + "step": 11480 + }, + { + "epoch": 32.06983240223464, + "grad_norm": 2.0455000400543213, + "learning_rate": 2.9059217877094974e-05, + "loss": 0.1553, + "step": 11481 + }, + { + "epoch": 32.072625698324025, + "grad_norm": 0.7848247289657593, + "learning_rate": 2.905884543761639e-05, + "loss": 0.1523, + "step": 11482 + }, + { + "epoch": 32.07541899441341, + "grad_norm": 0.5151913166046143, + "learning_rate": 2.9058472998137804e-05, + "loss": 0.1443, + "step": 11483 + }, + { + "epoch": 32.07821229050279, + "grad_norm": 0.499904602766037, + "learning_rate": 2.9058100558659217e-05, + "loss": 0.1311, + "step": 11484 + }, + { + "epoch": 32.08100558659218, + "grad_norm": 0.7043700218200684, + "learning_rate": 2.9057728119180634e-05, + "loss": 0.1546, + "step": 11485 + }, + { + "epoch": 32.08379888268156, + "grad_norm": 0.9507973194122314, + "learning_rate": 2.905735567970205e-05, + "loss": 0.148, + "step": 11486 + }, + { + "epoch": 32.08659217877095, + "grad_norm": 0.6600408554077148, + "learning_rate": 2.9056983240223463e-05, + "loss": 0.1356, + "step": 11487 + }, + { + "epoch": 32.089385474860336, + "grad_norm": 0.6454641222953796, + "learning_rate": 2.905661080074488e-05, + "loss": 0.1316, + "step": 11488 + }, + { + "epoch": 32.09217877094972, + "grad_norm": 0.4962007403373718, + "learning_rate": 2.9056238361266296e-05, + "loss": 0.1278, + "step": 11489 + }, + { + "epoch": 32.09497206703911, + "grad_norm": 0.6393634676933289, + "learning_rate": 2.905586592178771e-05, + "loss": 0.1182, + "step": 11490 + }, + { + "epoch": 32.097765363128495, + "grad_norm": 0.6664623022079468, + "learning_rate": 2.9055493482309126e-05, + "loss": 0.1327, + "step": 11491 + }, + { + "epoch": 32.100558659217874, + "grad_norm": 1.17244553565979, + "learning_rate": 2.905512104283054e-05, + "loss": 0.1145, + "step": 11492 + }, + { + "epoch": 32.10335195530726, + "grad_norm": 0.8781715035438538, + "learning_rate": 2.9054748603351956e-05, + "loss": 0.1097, + "step": 11493 + }, + { + "epoch": 32.10614525139665, + "grad_norm": 0.5913236141204834, + "learning_rate": 2.9054376163873372e-05, + "loss": 0.1224, + "step": 11494 + }, + { + "epoch": 32.108938547486034, + "grad_norm": 0.7578981518745422, + "learning_rate": 2.905400372439479e-05, + "loss": 0.1127, + "step": 11495 + }, + { + "epoch": 32.11173184357542, + "grad_norm": 0.4904521405696869, + "learning_rate": 2.9053631284916202e-05, + "loss": 0.092, + "step": 11496 + }, + { + "epoch": 32.114525139664806, + "grad_norm": 0.5301882028579712, + "learning_rate": 2.9053258845437615e-05, + "loss": 0.1105, + "step": 11497 + }, + { + "epoch": 32.11731843575419, + "grad_norm": 0.7162693738937378, + "learning_rate": 2.9052886405959032e-05, + "loss": 0.121, + "step": 11498 + }, + { + "epoch": 32.12011173184357, + "grad_norm": 0.6231658458709717, + "learning_rate": 2.9052513966480448e-05, + "loss": 0.1166, + "step": 11499 + }, + { + "epoch": 32.12290502793296, + "grad_norm": 0.5544078946113586, + "learning_rate": 2.9052141527001865e-05, + "loss": 0.1101, + "step": 11500 + }, + { + "epoch": 32.125698324022345, + "grad_norm": 0.6637757420539856, + "learning_rate": 2.9051769087523278e-05, + "loss": 0.1119, + "step": 11501 + }, + { + "epoch": 32.12849162011173, + "grad_norm": 0.8283183574676514, + "learning_rate": 2.9051396648044694e-05, + "loss": 0.1131, + "step": 11502 + }, + { + "epoch": 32.13128491620112, + "grad_norm": 0.7258477210998535, + "learning_rate": 2.9051024208566108e-05, + "loss": 0.102, + "step": 11503 + }, + { + "epoch": 32.134078212290504, + "grad_norm": 1.5331097841262817, + "learning_rate": 2.9050651769087524e-05, + "loss": 0.1332, + "step": 11504 + }, + { + "epoch": 32.13687150837989, + "grad_norm": 1.2706594467163086, + "learning_rate": 2.9050279329608937e-05, + "loss": 0.1283, + "step": 11505 + }, + { + "epoch": 32.13966480446928, + "grad_norm": 1.9998676776885986, + "learning_rate": 2.9049906890130354e-05, + "loss": 0.1186, + "step": 11506 + }, + { + "epoch": 32.142458100558656, + "grad_norm": 0.47011858224868774, + "learning_rate": 2.904953445065177e-05, + "loss": 0.1563, + "step": 11507 + }, + { + "epoch": 32.14525139664804, + "grad_norm": 0.48870590329170227, + "learning_rate": 2.9049162011173187e-05, + "loss": 0.0999, + "step": 11508 + }, + { + "epoch": 32.14804469273743, + "grad_norm": 0.5872325301170349, + "learning_rate": 2.90487895716946e-05, + "loss": 0.1305, + "step": 11509 + }, + { + "epoch": 32.150837988826815, + "grad_norm": 0.814748227596283, + "learning_rate": 2.9048417132216013e-05, + "loss": 0.1239, + "step": 11510 + }, + { + "epoch": 32.1536312849162, + "grad_norm": 0.45315638184547424, + "learning_rate": 2.904804469273743e-05, + "loss": 0.1634, + "step": 11511 + }, + { + "epoch": 32.15642458100559, + "grad_norm": 0.6266511082649231, + "learning_rate": 2.9047672253258846e-05, + "loss": 0.1362, + "step": 11512 + }, + { + "epoch": 32.159217877094974, + "grad_norm": 0.4814375042915344, + "learning_rate": 2.9047299813780263e-05, + "loss": 0.126, + "step": 11513 + }, + { + "epoch": 32.16201117318436, + "grad_norm": 0.462228387594223, + "learning_rate": 2.9046927374301676e-05, + "loss": 0.1196, + "step": 11514 + }, + { + "epoch": 32.16480446927374, + "grad_norm": 0.6652405261993408, + "learning_rate": 2.9046554934823092e-05, + "loss": 0.1146, + "step": 11515 + }, + { + "epoch": 32.167597765363126, + "grad_norm": 0.542783796787262, + "learning_rate": 2.904618249534451e-05, + "loss": 0.1481, + "step": 11516 + }, + { + "epoch": 32.17039106145251, + "grad_norm": 0.5119122862815857, + "learning_rate": 2.9045810055865922e-05, + "loss": 0.1364, + "step": 11517 + }, + { + "epoch": 32.1731843575419, + "grad_norm": 0.6590242385864258, + "learning_rate": 2.9045437616387335e-05, + "loss": 0.1433, + "step": 11518 + }, + { + "epoch": 32.175977653631286, + "grad_norm": 0.4834156334400177, + "learning_rate": 2.9045065176908752e-05, + "loss": 0.1097, + "step": 11519 + }, + { + "epoch": 32.17877094972067, + "grad_norm": 3.051438331604004, + "learning_rate": 2.904469273743017e-05, + "loss": 0.1071, + "step": 11520 + }, + { + "epoch": 32.18156424581006, + "grad_norm": 0.49365168809890747, + "learning_rate": 2.9044320297951585e-05, + "loss": 0.1199, + "step": 11521 + }, + { + "epoch": 32.184357541899445, + "grad_norm": 1.0759117603302002, + "learning_rate": 2.9043947858473e-05, + "loss": 0.1047, + "step": 11522 + }, + { + "epoch": 32.187150837988824, + "grad_norm": 0.7220920324325562, + "learning_rate": 2.904357541899441e-05, + "loss": 0.1265, + "step": 11523 + }, + { + "epoch": 32.18994413407821, + "grad_norm": 0.5836476683616638, + "learning_rate": 2.9043202979515828e-05, + "loss": 0.1174, + "step": 11524 + }, + { + "epoch": 32.1927374301676, + "grad_norm": 0.8359343409538269, + "learning_rate": 2.9042830540037244e-05, + "loss": 0.0822, + "step": 11525 + }, + { + "epoch": 32.19553072625698, + "grad_norm": 1.0536203384399414, + "learning_rate": 2.904245810055866e-05, + "loss": 0.1018, + "step": 11526 + }, + { + "epoch": 32.19832402234637, + "grad_norm": 0.8754586577415466, + "learning_rate": 2.9042085661080074e-05, + "loss": 0.1337, + "step": 11527 + }, + { + "epoch": 32.201117318435756, + "grad_norm": 0.5908018946647644, + "learning_rate": 2.904171322160149e-05, + "loss": 0.0934, + "step": 11528 + }, + { + "epoch": 32.20391061452514, + "grad_norm": 0.8667889833450317, + "learning_rate": 2.9041340782122907e-05, + "loss": 0.0969, + "step": 11529 + }, + { + "epoch": 32.20670391061452, + "grad_norm": 1.5182418823242188, + "learning_rate": 2.904096834264432e-05, + "loss": 0.1187, + "step": 11530 + }, + { + "epoch": 32.20949720670391, + "grad_norm": 1.8611503839492798, + "learning_rate": 2.9040595903165737e-05, + "loss": 0.1413, + "step": 11531 + }, + { + "epoch": 32.212290502793294, + "grad_norm": 0.5644127726554871, + "learning_rate": 2.904022346368715e-05, + "loss": 0.1515, + "step": 11532 + }, + { + "epoch": 32.21508379888268, + "grad_norm": 0.5445064902305603, + "learning_rate": 2.9039851024208566e-05, + "loss": 0.1306, + "step": 11533 + }, + { + "epoch": 32.21787709497207, + "grad_norm": 0.8680050373077393, + "learning_rate": 2.9039478584729983e-05, + "loss": 0.1247, + "step": 11534 + }, + { + "epoch": 32.220670391061454, + "grad_norm": 0.41354113817214966, + "learning_rate": 2.90391061452514e-05, + "loss": 0.1414, + "step": 11535 + }, + { + "epoch": 32.22346368715084, + "grad_norm": 0.46267974376678467, + "learning_rate": 2.9038733705772813e-05, + "loss": 0.1309, + "step": 11536 + }, + { + "epoch": 32.226256983240226, + "grad_norm": 0.39061981439590454, + "learning_rate": 2.9038361266294226e-05, + "loss": 0.1293, + "step": 11537 + }, + { + "epoch": 32.229050279329606, + "grad_norm": 0.6516837477684021, + "learning_rate": 2.9037988826815642e-05, + "loss": 0.1667, + "step": 11538 + }, + { + "epoch": 32.23184357541899, + "grad_norm": 0.5648525357246399, + "learning_rate": 2.903761638733706e-05, + "loss": 0.1391, + "step": 11539 + }, + { + "epoch": 32.23463687150838, + "grad_norm": 0.5568559169769287, + "learning_rate": 2.9037243947858475e-05, + "loss": 0.1156, + "step": 11540 + }, + { + "epoch": 32.237430167597765, + "grad_norm": 1.0369322299957275, + "learning_rate": 2.903687150837989e-05, + "loss": 0.13, + "step": 11541 + }, + { + "epoch": 32.24022346368715, + "grad_norm": 0.7856486439704895, + "learning_rate": 2.9036499068901305e-05, + "loss": 0.1229, + "step": 11542 + }, + { + "epoch": 32.24301675977654, + "grad_norm": 0.6375740766525269, + "learning_rate": 2.9036126629422718e-05, + "loss": 0.123, + "step": 11543 + }, + { + "epoch": 32.245810055865924, + "grad_norm": 1.2076802253723145, + "learning_rate": 2.9035754189944135e-05, + "loss": 0.0996, + "step": 11544 + }, + { + "epoch": 32.24860335195531, + "grad_norm": 0.8328583836555481, + "learning_rate": 2.9035381750465548e-05, + "loss": 0.1269, + "step": 11545 + }, + { + "epoch": 32.25139664804469, + "grad_norm": 0.6343109011650085, + "learning_rate": 2.9035009310986964e-05, + "loss": 0.1176, + "step": 11546 + }, + { + "epoch": 32.254189944134076, + "grad_norm": 0.7076234817504883, + "learning_rate": 2.903463687150838e-05, + "loss": 0.0945, + "step": 11547 + }, + { + "epoch": 32.25698324022346, + "grad_norm": 0.4342292845249176, + "learning_rate": 2.9034264432029797e-05, + "loss": 0.0963, + "step": 11548 + }, + { + "epoch": 32.25977653631285, + "grad_norm": 0.4531785249710083, + "learning_rate": 2.9033891992551214e-05, + "loss": 0.0942, + "step": 11549 + }, + { + "epoch": 32.262569832402235, + "grad_norm": 0.6963884234428406, + "learning_rate": 2.9033519553072624e-05, + "loss": 0.1152, + "step": 11550 + }, + { + "epoch": 32.26536312849162, + "grad_norm": 0.649150013923645, + "learning_rate": 2.903314711359404e-05, + "loss": 0.1042, + "step": 11551 + }, + { + "epoch": 32.26815642458101, + "grad_norm": 0.7063801884651184, + "learning_rate": 2.9032774674115457e-05, + "loss": 0.1058, + "step": 11552 + }, + { + "epoch": 32.270949720670394, + "grad_norm": 0.7553907632827759, + "learning_rate": 2.9032402234636873e-05, + "loss": 0.1053, + "step": 11553 + }, + { + "epoch": 32.273743016759774, + "grad_norm": 0.49191009998321533, + "learning_rate": 2.9032029795158287e-05, + "loss": 0.0892, + "step": 11554 + }, + { + "epoch": 32.27653631284916, + "grad_norm": NaN, + "learning_rate": 2.9032029795158287e-05, + "loss": 0.1301, + "step": 11555 + }, + { + "epoch": 32.279329608938546, + "grad_norm": 0.9798621535301208, + "learning_rate": 2.9031657355679703e-05, + "loss": 0.1272, + "step": 11556 + }, + { + "epoch": 32.28212290502793, + "grad_norm": 0.600005030632019, + "learning_rate": 2.903128491620112e-05, + "loss": 0.151, + "step": 11557 + }, + { + "epoch": 32.28491620111732, + "grad_norm": 0.6046677231788635, + "learning_rate": 2.9030912476722533e-05, + "loss": 0.1756, + "step": 11558 + }, + { + "epoch": 32.287709497206706, + "grad_norm": 0.5978448987007141, + "learning_rate": 2.903054003724395e-05, + "loss": 0.1262, + "step": 11559 + }, + { + "epoch": 32.29050279329609, + "grad_norm": 0.4611630141735077, + "learning_rate": 2.9030167597765362e-05, + "loss": 0.1249, + "step": 11560 + }, + { + "epoch": 32.29329608938548, + "grad_norm": 1.9989442825317383, + "learning_rate": 2.902979515828678e-05, + "loss": 0.1575, + "step": 11561 + }, + { + "epoch": 32.29608938547486, + "grad_norm": 0.4649551510810852, + "learning_rate": 2.9029422718808196e-05, + "loss": 0.1349, + "step": 11562 + }, + { + "epoch": 32.298882681564244, + "grad_norm": 0.42807361483573914, + "learning_rate": 2.9029050279329612e-05, + "loss": 0.1286, + "step": 11563 + }, + { + "epoch": 32.30167597765363, + "grad_norm": 1.1387466192245483, + "learning_rate": 2.9028677839851022e-05, + "loss": 0.1283, + "step": 11564 + }, + { + "epoch": 32.30446927374302, + "grad_norm": 0.44623246788978577, + "learning_rate": 2.902830540037244e-05, + "loss": 0.1232, + "step": 11565 + }, + { + "epoch": 32.3072625698324, + "grad_norm": 0.5277854204177856, + "learning_rate": 2.9027932960893855e-05, + "loss": 0.1259, + "step": 11566 + }, + { + "epoch": 32.31005586592179, + "grad_norm": 0.4881921708583832, + "learning_rate": 2.902756052141527e-05, + "loss": 0.1182, + "step": 11567 + }, + { + "epoch": 32.312849162011176, + "grad_norm": 0.7005869746208191, + "learning_rate": 2.9027188081936688e-05, + "loss": 0.1229, + "step": 11568 + }, + { + "epoch": 32.315642458100555, + "grad_norm": 0.8172481656074524, + "learning_rate": 2.90268156424581e-05, + "loss": 0.1154, + "step": 11569 + }, + { + "epoch": 32.31843575418994, + "grad_norm": 1.0141992568969727, + "learning_rate": 2.9026443202979518e-05, + "loss": 0.1172, + "step": 11570 + }, + { + "epoch": 32.32122905027933, + "grad_norm": 0.5154245495796204, + "learning_rate": 2.902607076350093e-05, + "loss": 0.1138, + "step": 11571 + }, + { + "epoch": 32.324022346368714, + "grad_norm": 0.8285523056983948, + "learning_rate": 2.9025698324022347e-05, + "loss": 0.1053, + "step": 11572 + }, + { + "epoch": 32.3268156424581, + "grad_norm": 0.6182783842086792, + "learning_rate": 2.902532588454376e-05, + "loss": 0.1387, + "step": 11573 + }, + { + "epoch": 32.32960893854749, + "grad_norm": 0.712374746799469, + "learning_rate": 2.9024953445065177e-05, + "loss": 0.1086, + "step": 11574 + }, + { + "epoch": 32.332402234636874, + "grad_norm": 0.7277215123176575, + "learning_rate": 2.9024581005586594e-05, + "loss": 0.1002, + "step": 11575 + }, + { + "epoch": 32.33519553072626, + "grad_norm": 0.7789361476898193, + "learning_rate": 2.902420856610801e-05, + "loss": 0.1107, + "step": 11576 + }, + { + "epoch": 32.33798882681564, + "grad_norm": 1.255947470664978, + "learning_rate": 2.9023836126629427e-05, + "loss": 0.1104, + "step": 11577 + }, + { + "epoch": 32.340782122905026, + "grad_norm": 11.37190055847168, + "learning_rate": 2.9023463687150836e-05, + "loss": 0.1075, + "step": 11578 + }, + { + "epoch": 32.34357541899441, + "grad_norm": 1.2183692455291748, + "learning_rate": 2.9023091247672253e-05, + "loss": 0.1174, + "step": 11579 + }, + { + "epoch": 32.3463687150838, + "grad_norm": 1.4370073080062866, + "learning_rate": 2.902271880819367e-05, + "loss": 0.1352, + "step": 11580 + }, + { + "epoch": 32.349162011173185, + "grad_norm": 1.2790968418121338, + "learning_rate": 2.9022346368715086e-05, + "loss": 0.1393, + "step": 11581 + }, + { + "epoch": 32.35195530726257, + "grad_norm": 0.39996299147605896, + "learning_rate": 2.90219739292365e-05, + "loss": 0.1346, + "step": 11582 + }, + { + "epoch": 32.35474860335196, + "grad_norm": 0.4168819785118103, + "learning_rate": 2.9021601489757916e-05, + "loss": 0.1228, + "step": 11583 + }, + { + "epoch": 32.357541899441344, + "grad_norm": 0.3590599000453949, + "learning_rate": 2.902122905027933e-05, + "loss": 0.1549, + "step": 11584 + }, + { + "epoch": 32.36033519553072, + "grad_norm": 0.5941213965415955, + "learning_rate": 2.9020856610800745e-05, + "loss": 0.15, + "step": 11585 + }, + { + "epoch": 32.36312849162011, + "grad_norm": 0.3619312644004822, + "learning_rate": 2.9020484171322162e-05, + "loss": 0.1316, + "step": 11586 + }, + { + "epoch": 32.365921787709496, + "grad_norm": 0.385081946849823, + "learning_rate": 2.9020111731843575e-05, + "loss": 0.1198, + "step": 11587 + }, + { + "epoch": 32.36871508379888, + "grad_norm": 0.3606245815753937, + "learning_rate": 2.901973929236499e-05, + "loss": 0.1148, + "step": 11588 + }, + { + "epoch": 32.37150837988827, + "grad_norm": 0.5366318225860596, + "learning_rate": 2.9019366852886408e-05, + "loss": 0.116, + "step": 11589 + }, + { + "epoch": 32.374301675977655, + "grad_norm": 0.648208737373352, + "learning_rate": 2.9018994413407825e-05, + "loss": 0.1163, + "step": 11590 + }, + { + "epoch": 32.37709497206704, + "grad_norm": 0.5565825700759888, + "learning_rate": 2.9018621973929234e-05, + "loss": 0.1118, + "step": 11591 + }, + { + "epoch": 32.37988826815643, + "grad_norm": 0.4164585471153259, + "learning_rate": 2.901824953445065e-05, + "loss": 0.1134, + "step": 11592 + }, + { + "epoch": 32.38268156424581, + "grad_norm": 0.6844955682754517, + "learning_rate": 2.9017877094972067e-05, + "loss": 0.139, + "step": 11593 + }, + { + "epoch": 32.385474860335194, + "grad_norm": 0.8008732199668884, + "learning_rate": 2.9017504655493484e-05, + "loss": 0.1185, + "step": 11594 + }, + { + "epoch": 32.38826815642458, + "grad_norm": 0.6207792162895203, + "learning_rate": 2.90171322160149e-05, + "loss": 0.1179, + "step": 11595 + }, + { + "epoch": 32.391061452513966, + "grad_norm": 0.764324963092804, + "learning_rate": 2.9016759776536314e-05, + "loss": 0.1133, + "step": 11596 + }, + { + "epoch": 32.39385474860335, + "grad_norm": 0.8764557838439941, + "learning_rate": 2.901638733705773e-05, + "loss": 0.1203, + "step": 11597 + }, + { + "epoch": 32.39664804469274, + "grad_norm": 0.5222392082214355, + "learning_rate": 2.9016014897579143e-05, + "loss": 0.0957, + "step": 11598 + }, + { + "epoch": 32.399441340782126, + "grad_norm": 0.6611840724945068, + "learning_rate": 2.901564245810056e-05, + "loss": 0.1186, + "step": 11599 + }, + { + "epoch": 32.40223463687151, + "grad_norm": 0.6834790110588074, + "learning_rate": 2.9015270018621973e-05, + "loss": 0.1083, + "step": 11600 + }, + { + "epoch": 32.40502793296089, + "grad_norm": 0.6019606590270996, + "learning_rate": 2.901489757914339e-05, + "loss": 0.0986, + "step": 11601 + }, + { + "epoch": 32.40782122905028, + "grad_norm": 1.1443127393722534, + "learning_rate": 2.9014525139664806e-05, + "loss": 0.0867, + "step": 11602 + }, + { + "epoch": 32.410614525139664, + "grad_norm": 0.6570228338241577, + "learning_rate": 2.9014152700186223e-05, + "loss": 0.1238, + "step": 11603 + }, + { + "epoch": 32.41340782122905, + "grad_norm": 0.9677271246910095, + "learning_rate": 2.9013780260707636e-05, + "loss": 0.0986, + "step": 11604 + }, + { + "epoch": 32.41620111731844, + "grad_norm": 0.5858938694000244, + "learning_rate": 2.901340782122905e-05, + "loss": 0.1131, + "step": 11605 + }, + { + "epoch": 32.41899441340782, + "grad_norm": 1.9492919445037842, + "learning_rate": 2.9013035381750465e-05, + "loss": 0.1607, + "step": 11606 + }, + { + "epoch": 32.42178770949721, + "grad_norm": 0.8044222593307495, + "learning_rate": 2.9012662942271882e-05, + "loss": 0.1552, + "step": 11607 + }, + { + "epoch": 32.42458100558659, + "grad_norm": 0.6043156385421753, + "learning_rate": 2.90122905027933e-05, + "loss": 0.1315, + "step": 11608 + }, + { + "epoch": 32.427374301675975, + "grad_norm": 0.4839250147342682, + "learning_rate": 2.901191806331471e-05, + "loss": 0.1507, + "step": 11609 + }, + { + "epoch": 32.43016759776536, + "grad_norm": 0.6688704490661621, + "learning_rate": 2.9011545623836128e-05, + "loss": 0.1317, + "step": 11610 + }, + { + "epoch": 32.43296089385475, + "grad_norm": 0.5286478400230408, + "learning_rate": 2.901117318435754e-05, + "loss": 0.118, + "step": 11611 + }, + { + "epoch": 32.435754189944134, + "grad_norm": 0.7872605919837952, + "learning_rate": 2.9010800744878958e-05, + "loss": 0.1275, + "step": 11612 + }, + { + "epoch": 32.43854748603352, + "grad_norm": 1.0222618579864502, + "learning_rate": 2.901042830540037e-05, + "loss": 0.1169, + "step": 11613 + }, + { + "epoch": 32.44134078212291, + "grad_norm": 0.48011314868927, + "learning_rate": 2.9010055865921788e-05, + "loss": 0.1181, + "step": 11614 + }, + { + "epoch": 32.444134078212294, + "grad_norm": 0.461617112159729, + "learning_rate": 2.9009683426443204e-05, + "loss": 0.1357, + "step": 11615 + }, + { + "epoch": 32.44692737430167, + "grad_norm": 0.5446233153343201, + "learning_rate": 2.900931098696462e-05, + "loss": 0.1228, + "step": 11616 + }, + { + "epoch": 32.44972067039106, + "grad_norm": 0.5562061667442322, + "learning_rate": 2.9008938547486037e-05, + "loss": 0.1163, + "step": 11617 + }, + { + "epoch": 32.452513966480446, + "grad_norm": 0.5054976940155029, + "learning_rate": 2.9008566108007447e-05, + "loss": 0.1235, + "step": 11618 + }, + { + "epoch": 32.45530726256983, + "grad_norm": 1.00011146068573, + "learning_rate": 2.9008193668528863e-05, + "loss": 0.1362, + "step": 11619 + }, + { + "epoch": 32.45810055865922, + "grad_norm": 0.8260816335678101, + "learning_rate": 2.900782122905028e-05, + "loss": 0.1149, + "step": 11620 + }, + { + "epoch": 32.460893854748605, + "grad_norm": 1.1240321397781372, + "learning_rate": 2.9007448789571697e-05, + "loss": 0.1099, + "step": 11621 + }, + { + "epoch": 32.46368715083799, + "grad_norm": 0.4680924713611603, + "learning_rate": 2.900707635009311e-05, + "loss": 0.09, + "step": 11622 + }, + { + "epoch": 32.46648044692738, + "grad_norm": 0.6165708899497986, + "learning_rate": 2.9006703910614526e-05, + "loss": 0.0973, + "step": 11623 + }, + { + "epoch": 32.46927374301676, + "grad_norm": 0.5037407875061035, + "learning_rate": 2.900633147113594e-05, + "loss": 0.1104, + "step": 11624 + }, + { + "epoch": 32.47206703910614, + "grad_norm": 0.6603978872299194, + "learning_rate": 2.9005959031657356e-05, + "loss": 0.1059, + "step": 11625 + }, + { + "epoch": 32.47486033519553, + "grad_norm": 0.9982938170433044, + "learning_rate": 2.9005586592178772e-05, + "loss": 0.1203, + "step": 11626 + }, + { + "epoch": 32.477653631284916, + "grad_norm": 0.7667892575263977, + "learning_rate": 2.9005214152700186e-05, + "loss": 0.1125, + "step": 11627 + }, + { + "epoch": 32.4804469273743, + "grad_norm": 0.9383038282394409, + "learning_rate": 2.9004841713221602e-05, + "loss": 0.0985, + "step": 11628 + }, + { + "epoch": 32.48324022346369, + "grad_norm": 1.0465164184570312, + "learning_rate": 2.900446927374302e-05, + "loss": 0.1202, + "step": 11629 + }, + { + "epoch": 32.486033519553075, + "grad_norm": 1.4612559080123901, + "learning_rate": 2.9004096834264435e-05, + "loss": 0.101, + "step": 11630 + }, + { + "epoch": 32.48882681564246, + "grad_norm": 0.8116002678871155, + "learning_rate": 2.9003724394785845e-05, + "loss": 0.1258, + "step": 11631 + }, + { + "epoch": 32.49162011173184, + "grad_norm": 0.510444700717926, + "learning_rate": 2.900335195530726e-05, + "loss": 0.1424, + "step": 11632 + }, + { + "epoch": 32.49441340782123, + "grad_norm": 0.41458451747894287, + "learning_rate": 2.9002979515828678e-05, + "loss": 0.1182, + "step": 11633 + }, + { + "epoch": 32.497206703910614, + "grad_norm": 0.7024139165878296, + "learning_rate": 2.9002607076350095e-05, + "loss": 0.1358, + "step": 11634 + }, + { + "epoch": 32.5, + "grad_norm": 0.4586012065410614, + "learning_rate": 2.900223463687151e-05, + "loss": 0.1305, + "step": 11635 + }, + { + "epoch": 32.502793296089386, + "grad_norm": 0.833459198474884, + "learning_rate": 2.9001862197392924e-05, + "loss": 0.1277, + "step": 11636 + }, + { + "epoch": 32.50558659217877, + "grad_norm": 1.0530120134353638, + "learning_rate": 2.900148975791434e-05, + "loss": 0.1291, + "step": 11637 + }, + { + "epoch": 32.50837988826816, + "grad_norm": 0.4031282663345337, + "learning_rate": 2.9001117318435754e-05, + "loss": 0.1233, + "step": 11638 + }, + { + "epoch": 32.51117318435754, + "grad_norm": 0.5475565791130066, + "learning_rate": 2.900074487895717e-05, + "loss": 0.1355, + "step": 11639 + }, + { + "epoch": 32.513966480446925, + "grad_norm": 0.5864057540893555, + "learning_rate": 2.9000372439478584e-05, + "loss": 0.1114, + "step": 11640 + }, + { + "epoch": 32.51675977653631, + "grad_norm": 0.6513293981552124, + "learning_rate": 2.9e-05, + "loss": 0.1255, + "step": 11641 + }, + { + "epoch": 32.5195530726257, + "grad_norm": 1.3443348407745361, + "learning_rate": 2.8999627560521417e-05, + "loss": 0.121, + "step": 11642 + }, + { + "epoch": 32.522346368715084, + "grad_norm": 0.5730998516082764, + "learning_rate": 2.8999255121042833e-05, + "loss": 0.1105, + "step": 11643 + }, + { + "epoch": 32.52513966480447, + "grad_norm": 0.43546152114868164, + "learning_rate": 2.8998882681564246e-05, + "loss": 0.0927, + "step": 11644 + }, + { + "epoch": 32.52793296089386, + "grad_norm": 0.7054458260536194, + "learning_rate": 2.899851024208566e-05, + "loss": 0.1185, + "step": 11645 + }, + { + "epoch": 32.53072625698324, + "grad_norm": 0.5634428262710571, + "learning_rate": 2.8998137802607076e-05, + "loss": 0.1053, + "step": 11646 + }, + { + "epoch": 32.53351955307262, + "grad_norm": 0.371175080537796, + "learning_rate": 2.8997765363128493e-05, + "loss": 0.1174, + "step": 11647 + }, + { + "epoch": 32.53631284916201, + "grad_norm": 2.5914647579193115, + "learning_rate": 2.899739292364991e-05, + "loss": 0.1128, + "step": 11648 + }, + { + "epoch": 32.539106145251395, + "grad_norm": 1.147901177406311, + "learning_rate": 2.8997020484171322e-05, + "loss": 0.1069, + "step": 11649 + }, + { + "epoch": 32.54189944134078, + "grad_norm": 0.5383266806602478, + "learning_rate": 2.899664804469274e-05, + "loss": 0.1068, + "step": 11650 + }, + { + "epoch": 32.54469273743017, + "grad_norm": 1.0204461812973022, + "learning_rate": 2.8996275605214152e-05, + "loss": 0.1089, + "step": 11651 + }, + { + "epoch": 32.547486033519554, + "grad_norm": 0.6548675894737244, + "learning_rate": 2.899590316573557e-05, + "loss": 0.102, + "step": 11652 + }, + { + "epoch": 32.55027932960894, + "grad_norm": 0.5740475654602051, + "learning_rate": 2.8995530726256985e-05, + "loss": 0.1078, + "step": 11653 + }, + { + "epoch": 32.55307262569833, + "grad_norm": 0.6245483756065369, + "learning_rate": 2.8995158286778398e-05, + "loss": 0.1219, + "step": 11654 + }, + { + "epoch": 32.555865921787706, + "grad_norm": 5.039485931396484, + "learning_rate": 2.8994785847299815e-05, + "loss": 0.1154, + "step": 11655 + }, + { + "epoch": 32.55865921787709, + "grad_norm": 1.1401432752609253, + "learning_rate": 2.899441340782123e-05, + "loss": 0.1847, + "step": 11656 + }, + { + "epoch": 32.56145251396648, + "grad_norm": 0.48428383469581604, + "learning_rate": 2.8994040968342648e-05, + "loss": 0.1431, + "step": 11657 + }, + { + "epoch": 32.564245810055866, + "grad_norm": 0.6597126722335815, + "learning_rate": 2.8993668528864058e-05, + "loss": 0.142, + "step": 11658 + }, + { + "epoch": 32.56703910614525, + "grad_norm": 0.4747900664806366, + "learning_rate": 2.8993296089385474e-05, + "loss": 0.1532, + "step": 11659 + }, + { + "epoch": 32.56983240223464, + "grad_norm": 0.41927337646484375, + "learning_rate": 2.899292364990689e-05, + "loss": 0.1254, + "step": 11660 + }, + { + "epoch": 32.572625698324025, + "grad_norm": 0.7458277344703674, + "learning_rate": 2.8992551210428307e-05, + "loss": 0.1233, + "step": 11661 + }, + { + "epoch": 32.57541899441341, + "grad_norm": 0.7658627033233643, + "learning_rate": 2.8992178770949724e-05, + "loss": 0.1317, + "step": 11662 + }, + { + "epoch": 32.57821229050279, + "grad_norm": 0.49048006534576416, + "learning_rate": 2.8991806331471137e-05, + "loss": 0.1278, + "step": 11663 + }, + { + "epoch": 32.58100558659218, + "grad_norm": 1.0322487354278564, + "learning_rate": 2.899143389199255e-05, + "loss": 0.1254, + "step": 11664 + }, + { + "epoch": 32.58379888268156, + "grad_norm": 0.46417757868766785, + "learning_rate": 2.8991061452513967e-05, + "loss": 0.1082, + "step": 11665 + }, + { + "epoch": 32.58659217877095, + "grad_norm": 0.7952748537063599, + "learning_rate": 2.8990689013035383e-05, + "loss": 0.115, + "step": 11666 + }, + { + "epoch": 32.589385474860336, + "grad_norm": 0.5298309326171875, + "learning_rate": 2.8990316573556796e-05, + "loss": 0.1238, + "step": 11667 + }, + { + "epoch": 32.59217877094972, + "grad_norm": 0.48072707653045654, + "learning_rate": 2.8989944134078213e-05, + "loss": 0.1137, + "step": 11668 + }, + { + "epoch": 32.59497206703911, + "grad_norm": 0.48389825224876404, + "learning_rate": 2.898957169459963e-05, + "loss": 0.1131, + "step": 11669 + }, + { + "epoch": 32.59776536312849, + "grad_norm": 0.43596351146698, + "learning_rate": 2.8989199255121046e-05, + "loss": 0.1119, + "step": 11670 + }, + { + "epoch": 32.600558659217874, + "grad_norm": 0.6929168105125427, + "learning_rate": 2.898882681564246e-05, + "loss": 0.1004, + "step": 11671 + }, + { + "epoch": 32.60335195530726, + "grad_norm": 0.7404049038887024, + "learning_rate": 2.8988454376163872e-05, + "loss": 0.1125, + "step": 11672 + }, + { + "epoch": 32.60614525139665, + "grad_norm": 0.7057824730873108, + "learning_rate": 2.898808193668529e-05, + "loss": 0.1038, + "step": 11673 + }, + { + "epoch": 32.608938547486034, + "grad_norm": 0.5819087624549866, + "learning_rate": 2.8987709497206705e-05, + "loss": 0.1181, + "step": 11674 + }, + { + "epoch": 32.61173184357542, + "grad_norm": 1.0849462747573853, + "learning_rate": 2.8987337057728122e-05, + "loss": 0.1105, + "step": 11675 + }, + { + "epoch": 32.614525139664806, + "grad_norm": 0.5873315930366516, + "learning_rate": 2.8986964618249535e-05, + "loss": 0.1189, + "step": 11676 + }, + { + "epoch": 32.61731843575419, + "grad_norm": 0.7975081205368042, + "learning_rate": 2.898659217877095e-05, + "loss": 0.1416, + "step": 11677 + }, + { + "epoch": 32.62011173184357, + "grad_norm": 1.0046249628067017, + "learning_rate": 2.8986219739292365e-05, + "loss": 0.1067, + "step": 11678 + }, + { + "epoch": 32.62290502793296, + "grad_norm": 0.8293424248695374, + "learning_rate": 2.898584729981378e-05, + "loss": 0.11, + "step": 11679 + }, + { + "epoch": 32.625698324022345, + "grad_norm": 6.157870292663574, + "learning_rate": 2.8985474860335198e-05, + "loss": 0.1019, + "step": 11680 + }, + { + "epoch": 32.62849162011173, + "grad_norm": 1.7385804653167725, + "learning_rate": 2.898510242085661e-05, + "loss": 0.1518, + "step": 11681 + }, + { + "epoch": 32.63128491620112, + "grad_norm": 0.6534480452537537, + "learning_rate": 2.8984729981378027e-05, + "loss": 0.1436, + "step": 11682 + }, + { + "epoch": 32.634078212290504, + "grad_norm": 0.5643952488899231, + "learning_rate": 2.8984357541899444e-05, + "loss": 0.1525, + "step": 11683 + }, + { + "epoch": 32.63687150837989, + "grad_norm": 0.3524082601070404, + "learning_rate": 2.8983985102420857e-05, + "loss": 0.1455, + "step": 11684 + }, + { + "epoch": 32.63966480446928, + "grad_norm": 0.5116269588470459, + "learning_rate": 2.898361266294227e-05, + "loss": 0.1255, + "step": 11685 + }, + { + "epoch": 32.642458100558656, + "grad_norm": 0.5811818242073059, + "learning_rate": 2.8983240223463687e-05, + "loss": 0.1375, + "step": 11686 + }, + { + "epoch": 32.64525139664804, + "grad_norm": 0.44462546706199646, + "learning_rate": 2.8982867783985103e-05, + "loss": 0.1295, + "step": 11687 + }, + { + "epoch": 32.64804469273743, + "grad_norm": 0.8963475823402405, + "learning_rate": 2.898249534450652e-05, + "loss": 0.1171, + "step": 11688 + }, + { + "epoch": 32.650837988826815, + "grad_norm": 0.5777872800827026, + "learning_rate": 2.8982122905027936e-05, + "loss": 0.1278, + "step": 11689 + }, + { + "epoch": 32.6536312849162, + "grad_norm": 0.5025116205215454, + "learning_rate": 2.898175046554935e-05, + "loss": 0.1257, + "step": 11690 + }, + { + "epoch": 32.65642458100559, + "grad_norm": 0.7681517004966736, + "learning_rate": 2.8981378026070763e-05, + "loss": 0.1323, + "step": 11691 + }, + { + "epoch": 32.659217877094974, + "grad_norm": 0.5635277628898621, + "learning_rate": 2.898100558659218e-05, + "loss": 0.116, + "step": 11692 + }, + { + "epoch": 32.66201117318436, + "grad_norm": 0.4179731607437134, + "learning_rate": 2.8980633147113596e-05, + "loss": 0.119, + "step": 11693 + }, + { + "epoch": 32.66480446927374, + "grad_norm": 1.1320351362228394, + "learning_rate": 2.898026070763501e-05, + "loss": 0.1206, + "step": 11694 + }, + { + "epoch": 32.667597765363126, + "grad_norm": 1.1458282470703125, + "learning_rate": 2.8979888268156425e-05, + "loss": 0.1137, + "step": 11695 + }, + { + "epoch": 32.67039106145251, + "grad_norm": 0.9579174518585205, + "learning_rate": 2.8979515828677842e-05, + "loss": 0.1028, + "step": 11696 + }, + { + "epoch": 32.6731843575419, + "grad_norm": 0.7252718806266785, + "learning_rate": 2.897914338919926e-05, + "loss": 0.1173, + "step": 11697 + }, + { + "epoch": 32.675977653631286, + "grad_norm": 0.6374531984329224, + "learning_rate": 2.897877094972067e-05, + "loss": 0.1347, + "step": 11698 + }, + { + "epoch": 32.67877094972067, + "grad_norm": 0.5857641696929932, + "learning_rate": 2.8978398510242085e-05, + "loss": 0.0873, + "step": 11699 + }, + { + "epoch": 32.68156424581006, + "grad_norm": 0.7718000411987305, + "learning_rate": 2.89780260707635e-05, + "loss": 0.1292, + "step": 11700 + }, + { + "epoch": 32.684357541899445, + "grad_norm": 0.7356138825416565, + "learning_rate": 2.8977653631284918e-05, + "loss": 0.1384, + "step": 11701 + }, + { + "epoch": 32.687150837988824, + "grad_norm": 0.7168378829956055, + "learning_rate": 2.8977281191806334e-05, + "loss": 0.0961, + "step": 11702 + }, + { + "epoch": 32.68994413407821, + "grad_norm": 0.6431041359901428, + "learning_rate": 2.8976908752327747e-05, + "loss": 0.1108, + "step": 11703 + }, + { + "epoch": 32.6927374301676, + "grad_norm": 1.300775408744812, + "learning_rate": 2.897653631284916e-05, + "loss": 0.1151, + "step": 11704 + }, + { + "epoch": 32.69553072625698, + "grad_norm": 0.5599760413169861, + "learning_rate": 2.8976163873370577e-05, + "loss": 0.0971, + "step": 11705 + }, + { + "epoch": 32.69832402234637, + "grad_norm": 0.9854787588119507, + "learning_rate": 2.8975791433891994e-05, + "loss": 0.1502, + "step": 11706 + }, + { + "epoch": 32.701117318435756, + "grad_norm": 1.1462829113006592, + "learning_rate": 2.8975418994413407e-05, + "loss": 0.1366, + "step": 11707 + }, + { + "epoch": 32.70391061452514, + "grad_norm": 0.5143190026283264, + "learning_rate": 2.8975046554934823e-05, + "loss": 0.1449, + "step": 11708 + }, + { + "epoch": 32.70670391061452, + "grad_norm": 0.7899280190467834, + "learning_rate": 2.897467411545624e-05, + "loss": 0.141, + "step": 11709 + }, + { + "epoch": 32.70949720670391, + "grad_norm": 0.38011038303375244, + "learning_rate": 2.8974301675977656e-05, + "loss": 0.1468, + "step": 11710 + }, + { + "epoch": 32.712290502793294, + "grad_norm": 0.7771890163421631, + "learning_rate": 2.897392923649907e-05, + "loss": 0.1398, + "step": 11711 + }, + { + "epoch": 32.71508379888268, + "grad_norm": 0.5740870833396912, + "learning_rate": 2.8973556797020483e-05, + "loss": 0.1488, + "step": 11712 + }, + { + "epoch": 32.71787709497207, + "grad_norm": 0.5477670431137085, + "learning_rate": 2.89731843575419e-05, + "loss": 0.1635, + "step": 11713 + }, + { + "epoch": 32.720670391061454, + "grad_norm": 0.6732445359230042, + "learning_rate": 2.8972811918063316e-05, + "loss": 0.1248, + "step": 11714 + }, + { + "epoch": 32.72346368715084, + "grad_norm": 0.4767455756664276, + "learning_rate": 2.8972439478584732e-05, + "loss": 0.1243, + "step": 11715 + }, + { + "epoch": 32.726256983240226, + "grad_norm": 0.6214262247085571, + "learning_rate": 2.8972067039106145e-05, + "loss": 0.1301, + "step": 11716 + }, + { + "epoch": 32.729050279329606, + "grad_norm": 0.6384440064430237, + "learning_rate": 2.8971694599627562e-05, + "loss": 0.1217, + "step": 11717 + }, + { + "epoch": 32.73184357541899, + "grad_norm": 0.47134125232696533, + "learning_rate": 2.8971322160148975e-05, + "loss": 0.1037, + "step": 11718 + }, + { + "epoch": 32.73463687150838, + "grad_norm": 0.4598710238933563, + "learning_rate": 2.897094972067039e-05, + "loss": 0.1173, + "step": 11719 + }, + { + "epoch": 32.737430167597765, + "grad_norm": 0.724292516708374, + "learning_rate": 2.8970577281191808e-05, + "loss": 0.1051, + "step": 11720 + }, + { + "epoch": 32.74022346368715, + "grad_norm": 0.7679964900016785, + "learning_rate": 2.897020484171322e-05, + "loss": 0.1123, + "step": 11721 + }, + { + "epoch": 32.74301675977654, + "grad_norm": 0.7937564253807068, + "learning_rate": 2.8969832402234638e-05, + "loss": 0.1174, + "step": 11722 + }, + { + "epoch": 32.745810055865924, + "grad_norm": 0.54521244764328, + "learning_rate": 2.8969459962756054e-05, + "loss": 0.1409, + "step": 11723 + }, + { + "epoch": 32.74860335195531, + "grad_norm": 1.0139082670211792, + "learning_rate": 2.8969087523277468e-05, + "loss": 0.1344, + "step": 11724 + }, + { + "epoch": 32.75139664804469, + "grad_norm": 0.6386202573776245, + "learning_rate": 2.896871508379888e-05, + "loss": 0.1328, + "step": 11725 + }, + { + "epoch": 32.754189944134076, + "grad_norm": 0.7377815842628479, + "learning_rate": 2.8968342644320297e-05, + "loss": 0.1057, + "step": 11726 + }, + { + "epoch": 32.75698324022346, + "grad_norm": 0.7540372610092163, + "learning_rate": 2.8967970204841714e-05, + "loss": 0.1167, + "step": 11727 + }, + { + "epoch": 32.75977653631285, + "grad_norm": 0.6726033091545105, + "learning_rate": 2.896759776536313e-05, + "loss": 0.1162, + "step": 11728 + }, + { + "epoch": 32.762569832402235, + "grad_norm": 0.6294278502464294, + "learning_rate": 2.8967225325884547e-05, + "loss": 0.1077, + "step": 11729 + }, + { + "epoch": 32.76536312849162, + "grad_norm": 0.828324556350708, + "learning_rate": 2.896685288640596e-05, + "loss": 0.1105, + "step": 11730 + }, + { + "epoch": 32.76815642458101, + "grad_norm": 1.8297585248947144, + "learning_rate": 2.8966480446927373e-05, + "loss": 0.1337, + "step": 11731 + }, + { + "epoch": 32.770949720670394, + "grad_norm": 0.4164298176765442, + "learning_rate": 2.896610800744879e-05, + "loss": 0.1653, + "step": 11732 + }, + { + "epoch": 32.773743016759774, + "grad_norm": 1.048378586769104, + "learning_rate": 2.8965735567970206e-05, + "loss": 0.1383, + "step": 11733 + }, + { + "epoch": 32.77653631284916, + "grad_norm": 0.8624338507652283, + "learning_rate": 2.896536312849162e-05, + "loss": 0.1363, + "step": 11734 + }, + { + "epoch": 32.779329608938546, + "grad_norm": 0.5666893124580383, + "learning_rate": 2.8964990689013036e-05, + "loss": 0.1278, + "step": 11735 + }, + { + "epoch": 32.78212290502793, + "grad_norm": 0.48236891627311707, + "learning_rate": 2.8964618249534452e-05, + "loss": 0.1324, + "step": 11736 + }, + { + "epoch": 32.78491620111732, + "grad_norm": 0.5194763541221619, + "learning_rate": 2.896424581005587e-05, + "loss": 0.1317, + "step": 11737 + }, + { + "epoch": 32.787709497206706, + "grad_norm": 0.5682879686355591, + "learning_rate": 2.8963873370577282e-05, + "loss": 0.1329, + "step": 11738 + }, + { + "epoch": 32.79050279329609, + "grad_norm": 0.7940146327018738, + "learning_rate": 2.8963500931098695e-05, + "loss": 0.1101, + "step": 11739 + }, + { + "epoch": 32.79329608938548, + "grad_norm": 2.9935081005096436, + "learning_rate": 2.8963128491620112e-05, + "loss": 0.1321, + "step": 11740 + }, + { + "epoch": 32.79608938547486, + "grad_norm": 0.7734065055847168, + "learning_rate": 2.896275605214153e-05, + "loss": 0.1148, + "step": 11741 + }, + { + "epoch": 32.798882681564244, + "grad_norm": 0.5733015537261963, + "learning_rate": 2.8962383612662945e-05, + "loss": 0.12, + "step": 11742 + }, + { + "epoch": 32.80167597765363, + "grad_norm": 0.5576422214508057, + "learning_rate": 2.8962011173184358e-05, + "loss": 0.123, + "step": 11743 + }, + { + "epoch": 32.80446927374302, + "grad_norm": 0.5820799469947815, + "learning_rate": 2.896163873370577e-05, + "loss": 0.114, + "step": 11744 + }, + { + "epoch": 32.8072625698324, + "grad_norm": 0.527786374092102, + "learning_rate": 2.8961266294227188e-05, + "loss": 0.1149, + "step": 11745 + }, + { + "epoch": 32.81005586592179, + "grad_norm": 1.2941192388534546, + "learning_rate": 2.8960893854748604e-05, + "loss": 0.1361, + "step": 11746 + }, + { + "epoch": 32.812849162011176, + "grad_norm": 1.5482568740844727, + "learning_rate": 2.896052141527002e-05, + "loss": 0.1068, + "step": 11747 + }, + { + "epoch": 32.815642458100555, + "grad_norm": 0.5295989513397217, + "learning_rate": 2.8960148975791434e-05, + "loss": 0.1117, + "step": 11748 + }, + { + "epoch": 32.81843575418994, + "grad_norm": 0.505101203918457, + "learning_rate": 2.895977653631285e-05, + "loss": 0.0921, + "step": 11749 + }, + { + "epoch": 32.82122905027933, + "grad_norm": 0.6305166482925415, + "learning_rate": 2.8959404096834267e-05, + "loss": 0.0996, + "step": 11750 + }, + { + "epoch": 32.824022346368714, + "grad_norm": 0.7805349230766296, + "learning_rate": 2.895903165735568e-05, + "loss": 0.118, + "step": 11751 + }, + { + "epoch": 32.8268156424581, + "grad_norm": 1.5062118768692017, + "learning_rate": 2.8958659217877093e-05, + "loss": 0.1094, + "step": 11752 + }, + { + "epoch": 32.82960893854749, + "grad_norm": 1.0550320148468018, + "learning_rate": 2.895828677839851e-05, + "loss": 0.1005, + "step": 11753 + }, + { + "epoch": 32.832402234636874, + "grad_norm": 0.6459513306617737, + "learning_rate": 2.8957914338919926e-05, + "loss": 0.1079, + "step": 11754 + }, + { + "epoch": 32.83519553072626, + "grad_norm": 1.244171142578125, + "learning_rate": 2.8957541899441343e-05, + "loss": 0.1188, + "step": 11755 + }, + { + "epoch": 32.83798882681564, + "grad_norm": 0.8736401200294495, + "learning_rate": 2.895716945996276e-05, + "loss": 0.1267, + "step": 11756 + }, + { + "epoch": 32.840782122905026, + "grad_norm": 0.4675360321998596, + "learning_rate": 2.8956797020484173e-05, + "loss": 0.1386, + "step": 11757 + }, + { + "epoch": 32.84357541899441, + "grad_norm": 0.47127360105514526, + "learning_rate": 2.8956424581005586e-05, + "loss": 0.1232, + "step": 11758 + }, + { + "epoch": 32.8463687150838, + "grad_norm": 0.604851484298706, + "learning_rate": 2.8956052141527002e-05, + "loss": 0.1334, + "step": 11759 + }, + { + "epoch": 32.849162011173185, + "grad_norm": 0.5738930106163025, + "learning_rate": 2.895567970204842e-05, + "loss": 0.1299, + "step": 11760 + }, + { + "epoch": 32.85195530726257, + "grad_norm": 0.6112295985221863, + "learning_rate": 2.8955307262569832e-05, + "loss": 0.1388, + "step": 11761 + }, + { + "epoch": 32.85474860335196, + "grad_norm": 0.4301196038722992, + "learning_rate": 2.895493482309125e-05, + "loss": 0.119, + "step": 11762 + }, + { + "epoch": 32.857541899441344, + "grad_norm": 0.7580046057701111, + "learning_rate": 2.8954562383612665e-05, + "loss": 0.1211, + "step": 11763 + }, + { + "epoch": 32.86033519553072, + "grad_norm": 0.6082758903503418, + "learning_rate": 2.8954189944134078e-05, + "loss": 0.1247, + "step": 11764 + }, + { + "epoch": 32.86312849162011, + "grad_norm": 0.5993214845657349, + "learning_rate": 2.8953817504655495e-05, + "loss": 0.1329, + "step": 11765 + }, + { + "epoch": 32.865921787709496, + "grad_norm": 0.7101327776908875, + "learning_rate": 2.8953445065176908e-05, + "loss": 0.1402, + "step": 11766 + }, + { + "epoch": 32.86871508379888, + "grad_norm": 1.5042301416397095, + "learning_rate": 2.8953072625698324e-05, + "loss": 0.1187, + "step": 11767 + }, + { + "epoch": 32.87150837988827, + "grad_norm": 0.7729624509811401, + "learning_rate": 2.895270018621974e-05, + "loss": 0.1303, + "step": 11768 + }, + { + "epoch": 32.874301675977655, + "grad_norm": 0.44607242941856384, + "learning_rate": 2.8952327746741157e-05, + "loss": 0.1002, + "step": 11769 + }, + { + "epoch": 32.87709497206704, + "grad_norm": 0.6521061658859253, + "learning_rate": 2.895195530726257e-05, + "loss": 0.1075, + "step": 11770 + }, + { + "epoch": 32.87988826815643, + "grad_norm": 0.6207712888717651, + "learning_rate": 2.8951582867783984e-05, + "loss": 0.122, + "step": 11771 + }, + { + "epoch": 32.88268156424581, + "grad_norm": 0.6382371187210083, + "learning_rate": 2.89512104283054e-05, + "loss": 0.1243, + "step": 11772 + }, + { + "epoch": 32.885474860335194, + "grad_norm": 0.5781164169311523, + "learning_rate": 2.8950837988826817e-05, + "loss": 0.1085, + "step": 11773 + }, + { + "epoch": 32.88826815642458, + "grad_norm": 0.6507511734962463, + "learning_rate": 2.8950465549348233e-05, + "loss": 0.0846, + "step": 11774 + }, + { + "epoch": 32.891061452513966, + "grad_norm": 1.047013759613037, + "learning_rate": 2.8950093109869647e-05, + "loss": 0.1263, + "step": 11775 + }, + { + "epoch": 32.89385474860335, + "grad_norm": 1.9171062707901, + "learning_rate": 2.8949720670391063e-05, + "loss": 0.1123, + "step": 11776 + }, + { + "epoch": 32.89664804469274, + "grad_norm": 0.7856870293617249, + "learning_rate": 2.894934823091248e-05, + "loss": 0.1111, + "step": 11777 + }, + { + "epoch": 32.899441340782126, + "grad_norm": 0.9122694134712219, + "learning_rate": 2.8948975791433893e-05, + "loss": 0.1202, + "step": 11778 + }, + { + "epoch": 32.90223463687151, + "grad_norm": 0.6787843108177185, + "learning_rate": 2.8948603351955306e-05, + "loss": 0.0929, + "step": 11779 + }, + { + "epoch": 32.90502793296089, + "grad_norm": 1.060819387435913, + "learning_rate": 2.8948230912476722e-05, + "loss": 0.0949, + "step": 11780 + }, + { + "epoch": 32.90782122905028, + "grad_norm": 2.5148229598999023, + "learning_rate": 2.894785847299814e-05, + "loss": 0.1623, + "step": 11781 + }, + { + "epoch": 32.910614525139664, + "grad_norm": 0.5718089938163757, + "learning_rate": 2.8947486033519555e-05, + "loss": 0.1562, + "step": 11782 + }, + { + "epoch": 32.91340782122905, + "grad_norm": 0.4739844799041748, + "learning_rate": 2.8947113594040972e-05, + "loss": 0.1652, + "step": 11783 + }, + { + "epoch": 32.91620111731844, + "grad_norm": 0.7173149585723877, + "learning_rate": 2.8946741154562382e-05, + "loss": 0.1439, + "step": 11784 + }, + { + "epoch": 32.91899441340782, + "grad_norm": 0.6075518131256104, + "learning_rate": 2.89463687150838e-05, + "loss": 0.1343, + "step": 11785 + }, + { + "epoch": 32.92178770949721, + "grad_norm": 0.3609885275363922, + "learning_rate": 2.8945996275605215e-05, + "loss": 0.1435, + "step": 11786 + }, + { + "epoch": 32.92458100558659, + "grad_norm": 0.42353710532188416, + "learning_rate": 2.894562383612663e-05, + "loss": 0.1265, + "step": 11787 + }, + { + "epoch": 32.927374301675975, + "grad_norm": 0.7388096451759338, + "learning_rate": 2.8945251396648045e-05, + "loss": 0.1432, + "step": 11788 + }, + { + "epoch": 32.93016759776536, + "grad_norm": 0.5735319256782532, + "learning_rate": 2.894487895716946e-05, + "loss": 0.1337, + "step": 11789 + }, + { + "epoch": 32.93296089385475, + "grad_norm": 0.7574446201324463, + "learning_rate": 2.8944506517690878e-05, + "loss": 0.137, + "step": 11790 + }, + { + "epoch": 32.935754189944134, + "grad_norm": 0.677788257598877, + "learning_rate": 2.894413407821229e-05, + "loss": 0.1383, + "step": 11791 + }, + { + "epoch": 32.93854748603352, + "grad_norm": 0.774433434009552, + "learning_rate": 2.8943761638733707e-05, + "loss": 0.1319, + "step": 11792 + }, + { + "epoch": 32.94134078212291, + "grad_norm": 0.5948057174682617, + "learning_rate": 2.894338919925512e-05, + "loss": 0.1143, + "step": 11793 + }, + { + "epoch": 32.944134078212294, + "grad_norm": 0.6802560687065125, + "learning_rate": 2.8943016759776537e-05, + "loss": 0.1095, + "step": 11794 + }, + { + "epoch": 32.94692737430167, + "grad_norm": 0.711823582649231, + "learning_rate": 2.8942644320297954e-05, + "loss": 0.1148, + "step": 11795 + }, + { + "epoch": 32.94972067039106, + "grad_norm": 2.0489375591278076, + "learning_rate": 2.894227188081937e-05, + "loss": 0.1055, + "step": 11796 + }, + { + "epoch": 32.952513966480446, + "grad_norm": 0.4744565486907959, + "learning_rate": 2.8941899441340783e-05, + "loss": 0.0972, + "step": 11797 + }, + { + "epoch": 32.95530726256983, + "grad_norm": 0.6116234064102173, + "learning_rate": 2.8941527001862196e-05, + "loss": 0.1089, + "step": 11798 + }, + { + "epoch": 32.95810055865922, + "grad_norm": 0.605070948600769, + "learning_rate": 2.8941154562383613e-05, + "loss": 0.112, + "step": 11799 + }, + { + "epoch": 32.960893854748605, + "grad_norm": 1.7834542989730835, + "learning_rate": 2.894078212290503e-05, + "loss": 0.1047, + "step": 11800 + }, + { + "epoch": 32.96368715083799, + "grad_norm": 0.47254642844200134, + "learning_rate": 2.8940409683426443e-05, + "loss": 0.088, + "step": 11801 + }, + { + "epoch": 32.96648044692738, + "grad_norm": 1.2655344009399414, + "learning_rate": 2.894003724394786e-05, + "loss": 0.0983, + "step": 11802 + }, + { + "epoch": 32.96927374301676, + "grad_norm": 1.083546757698059, + "learning_rate": 2.8939664804469276e-05, + "loss": 0.0942, + "step": 11803 + }, + { + "epoch": 32.97206703910614, + "grad_norm": 0.847129225730896, + "learning_rate": 2.893929236499069e-05, + "loss": 0.1228, + "step": 11804 + }, + { + "epoch": 32.97486033519553, + "grad_norm": 2.415562629699707, + "learning_rate": 2.8938919925512105e-05, + "loss": 0.124, + "step": 11805 + }, + { + "epoch": 32.977653631284916, + "grad_norm": 1.2506825923919678, + "learning_rate": 2.893854748603352e-05, + "loss": 0.1419, + "step": 11806 + }, + { + "epoch": 32.9804469273743, + "grad_norm": 0.38439974188804626, + "learning_rate": 2.8938175046554935e-05, + "loss": 0.1339, + "step": 11807 + }, + { + "epoch": 32.98324022346369, + "grad_norm": 0.476211816072464, + "learning_rate": 2.893780260707635e-05, + "loss": 0.1323, + "step": 11808 + }, + { + "epoch": 32.986033519553075, + "grad_norm": 0.47028812766075134, + "learning_rate": 2.8937430167597768e-05, + "loss": 0.1077, + "step": 11809 + }, + { + "epoch": 32.98882681564246, + "grad_norm": 0.39275336265563965, + "learning_rate": 2.893705772811918e-05, + "loss": 0.121, + "step": 11810 + }, + { + "epoch": 32.99162011173184, + "grad_norm": 1.1896722316741943, + "learning_rate": 2.8936685288640594e-05, + "loss": 0.0952, + "step": 11811 + }, + { + "epoch": 32.99441340782123, + "grad_norm": 1.1106173992156982, + "learning_rate": 2.893631284916201e-05, + "loss": 0.1082, + "step": 11812 + }, + { + "epoch": 32.997206703910614, + "grad_norm": 0.6822313666343689, + "learning_rate": 2.8935940409683427e-05, + "loss": 0.0975, + "step": 11813 + }, + { + "epoch": 33.0, + "grad_norm": 2.41569447517395, + "learning_rate": 2.8935567970204844e-05, + "loss": 0.1174, + "step": 11814 + }, + { + "epoch": 33.002793296089386, + "grad_norm": 0.43861421942710876, + "learning_rate": 2.8935195530726257e-05, + "loss": 0.1402, + "step": 11815 + }, + { + "epoch": 33.00558659217877, + "grad_norm": 0.4086211621761322, + "learning_rate": 2.8934823091247674e-05, + "loss": 0.1356, + "step": 11816 + }, + { + "epoch": 33.00837988826816, + "grad_norm": 0.5466194152832031, + "learning_rate": 2.893445065176909e-05, + "loss": 0.1431, + "step": 11817 + }, + { + "epoch": 33.01117318435754, + "grad_norm": 0.49200335144996643, + "learning_rate": 2.8934078212290503e-05, + "loss": 0.1219, + "step": 11818 + }, + { + "epoch": 33.013966480446925, + "grad_norm": 0.6775206923484802, + "learning_rate": 2.8933705772811916e-05, + "loss": 0.1262, + "step": 11819 + }, + { + "epoch": 33.01675977653631, + "grad_norm": 0.5161769390106201, + "learning_rate": 2.8933333333333333e-05, + "loss": 0.1554, + "step": 11820 + }, + { + "epoch": 33.0195530726257, + "grad_norm": 0.43343257904052734, + "learning_rate": 2.893296089385475e-05, + "loss": 0.1369, + "step": 11821 + }, + { + "epoch": 33.022346368715084, + "grad_norm": 0.5806891918182373, + "learning_rate": 2.8932588454376166e-05, + "loss": 0.1105, + "step": 11822 + }, + { + "epoch": 33.02513966480447, + "grad_norm": 0.3567426800727844, + "learning_rate": 2.8932216014897583e-05, + "loss": 0.1195, + "step": 11823 + }, + { + "epoch": 33.02793296089386, + "grad_norm": 0.507453441619873, + "learning_rate": 2.8931843575418992e-05, + "loss": 0.1031, + "step": 11824 + }, + { + "epoch": 33.03072625698324, + "grad_norm": 0.7409469485282898, + "learning_rate": 2.893147113594041e-05, + "loss": 0.1142, + "step": 11825 + }, + { + "epoch": 33.03351955307262, + "grad_norm": 0.6951045393943787, + "learning_rate": 2.8931098696461825e-05, + "loss": 0.1248, + "step": 11826 + }, + { + "epoch": 33.03631284916201, + "grad_norm": 0.44877034425735474, + "learning_rate": 2.8930726256983242e-05, + "loss": 0.0974, + "step": 11827 + }, + { + "epoch": 33.039106145251395, + "grad_norm": 0.7529577612876892, + "learning_rate": 2.8930353817504655e-05, + "loss": 0.1166, + "step": 11828 + }, + { + "epoch": 33.04189944134078, + "grad_norm": 0.7848125100135803, + "learning_rate": 2.892998137802607e-05, + "loss": 0.0936, + "step": 11829 + }, + { + "epoch": 33.04469273743017, + "grad_norm": 0.5167027115821838, + "learning_rate": 2.8929608938547488e-05, + "loss": 0.0875, + "step": 11830 + }, + { + "epoch": 33.047486033519554, + "grad_norm": 0.7369900345802307, + "learning_rate": 2.89292364990689e-05, + "loss": 0.1082, + "step": 11831 + }, + { + "epoch": 33.05027932960894, + "grad_norm": 1.0744457244873047, + "learning_rate": 2.8928864059590318e-05, + "loss": 0.1024, + "step": 11832 + }, + { + "epoch": 33.05307262569833, + "grad_norm": 0.9741693735122681, + "learning_rate": 2.892849162011173e-05, + "loss": 0.1025, + "step": 11833 + }, + { + "epoch": 33.055865921787706, + "grad_norm": 0.5866530537605286, + "learning_rate": 2.8928119180633148e-05, + "loss": 0.1032, + "step": 11834 + }, + { + "epoch": 33.05865921787709, + "grad_norm": 0.737775981426239, + "learning_rate": 2.8927746741154564e-05, + "loss": 0.1009, + "step": 11835 + }, + { + "epoch": 33.06145251396648, + "grad_norm": 0.709430456161499, + "learning_rate": 2.892737430167598e-05, + "loss": 0.0928, + "step": 11836 + }, + { + "epoch": 33.064245810055866, + "grad_norm": 1.0749318599700928, + "learning_rate": 2.8927001862197394e-05, + "loss": 0.0871, + "step": 11837 + }, + { + "epoch": 33.06703910614525, + "grad_norm": 1.9895274639129639, + "learning_rate": 2.8926629422718807e-05, + "loss": 0.105, + "step": 11838 + }, + { + "epoch": 33.06983240223464, + "grad_norm": 1.1601412296295166, + "learning_rate": 2.8926256983240223e-05, + "loss": 0.1454, + "step": 11839 + }, + { + "epoch": 33.072625698324025, + "grad_norm": 0.5644034147262573, + "learning_rate": 2.892588454376164e-05, + "loss": 0.1359, + "step": 11840 + }, + { + "epoch": 33.07541899441341, + "grad_norm": 0.5380356907844543, + "learning_rate": 2.8925512104283057e-05, + "loss": 0.1279, + "step": 11841 + }, + { + "epoch": 33.07821229050279, + "grad_norm": 0.5799257755279541, + "learning_rate": 2.892513966480447e-05, + "loss": 0.1361, + "step": 11842 + }, + { + "epoch": 33.08100558659218, + "grad_norm": 0.4598289132118225, + "learning_rate": 2.8924767225325886e-05, + "loss": 0.129, + "step": 11843 + }, + { + "epoch": 33.08379888268156, + "grad_norm": 1.2657113075256348, + "learning_rate": 2.89243947858473e-05, + "loss": 0.1194, + "step": 11844 + }, + { + "epoch": 33.08659217877095, + "grad_norm": 0.4033646583557129, + "learning_rate": 2.8924022346368716e-05, + "loss": 0.1178, + "step": 11845 + }, + { + "epoch": 33.089385474860336, + "grad_norm": 0.6537470817565918, + "learning_rate": 2.892364990689013e-05, + "loss": 0.1112, + "step": 11846 + }, + { + "epoch": 33.09217877094972, + "grad_norm": 0.47220510244369507, + "learning_rate": 2.8923277467411546e-05, + "loss": 0.1175, + "step": 11847 + }, + { + "epoch": 33.09497206703911, + "grad_norm": 0.4278092086315155, + "learning_rate": 2.8922905027932962e-05, + "loss": 0.1093, + "step": 11848 + }, + { + "epoch": 33.097765363128495, + "grad_norm": 0.6259655356407166, + "learning_rate": 2.892253258845438e-05, + "loss": 0.1162, + "step": 11849 + }, + { + "epoch": 33.100558659217874, + "grad_norm": 0.44524043798446655, + "learning_rate": 2.8922160148975795e-05, + "loss": 0.1213, + "step": 11850 + }, + { + "epoch": 33.10335195530726, + "grad_norm": 0.6989511847496033, + "learning_rate": 2.8921787709497205e-05, + "loss": 0.11, + "step": 11851 + }, + { + "epoch": 33.10614525139665, + "grad_norm": 0.6646243333816528, + "learning_rate": 2.892141527001862e-05, + "loss": 0.1045, + "step": 11852 + }, + { + "epoch": 33.108938547486034, + "grad_norm": 0.602091372013092, + "learning_rate": 2.8921042830540038e-05, + "loss": 0.1054, + "step": 11853 + }, + { + "epoch": 33.11173184357542, + "grad_norm": 0.3871191143989563, + "learning_rate": 2.8920670391061455e-05, + "loss": 0.1086, + "step": 11854 + }, + { + "epoch": 33.114525139664806, + "grad_norm": 0.5027832984924316, + "learning_rate": 2.8920297951582868e-05, + "loss": 0.1134, + "step": 11855 + }, + { + "epoch": 33.11731843575419, + "grad_norm": 0.9379854798316956, + "learning_rate": 2.8919925512104284e-05, + "loss": 0.1042, + "step": 11856 + }, + { + "epoch": 33.12011173184357, + "grad_norm": 1.2426095008850098, + "learning_rate": 2.89195530726257e-05, + "loss": 0.1051, + "step": 11857 + }, + { + "epoch": 33.12290502793296, + "grad_norm": 0.5156109929084778, + "learning_rate": 2.8919180633147114e-05, + "loss": 0.1026, + "step": 11858 + }, + { + "epoch": 33.125698324022345, + "grad_norm": 0.5144765377044678, + "learning_rate": 2.891880819366853e-05, + "loss": 0.0818, + "step": 11859 + }, + { + "epoch": 33.12849162011173, + "grad_norm": 1.227052927017212, + "learning_rate": 2.8918435754189944e-05, + "loss": 0.0948, + "step": 11860 + }, + { + "epoch": 33.13128491620112, + "grad_norm": 2.1337954998016357, + "learning_rate": 2.891806331471136e-05, + "loss": 0.0811, + "step": 11861 + }, + { + "epoch": 33.134078212290504, + "grad_norm": 0.6854702234268188, + "learning_rate": 2.8917690875232777e-05, + "loss": 0.0877, + "step": 11862 + }, + { + "epoch": 33.13687150837989, + "grad_norm": 1.6550198793411255, + "learning_rate": 2.8917318435754193e-05, + "loss": 0.1005, + "step": 11863 + }, + { + "epoch": 33.13966480446928, + "grad_norm": 1.427643895149231, + "learning_rate": 2.8916945996275603e-05, + "loss": 0.1484, + "step": 11864 + }, + { + "epoch": 33.142458100558656, + "grad_norm": 0.4611775577068329, + "learning_rate": 2.891657355679702e-05, + "loss": 0.1325, + "step": 11865 + }, + { + "epoch": 33.14525139664804, + "grad_norm": 0.5122498273849487, + "learning_rate": 2.8916201117318436e-05, + "loss": 0.1417, + "step": 11866 + }, + { + "epoch": 33.14804469273743, + "grad_norm": 0.4476025700569153, + "learning_rate": 2.8915828677839853e-05, + "loss": 0.1393, + "step": 11867 + }, + { + "epoch": 33.150837988826815, + "grad_norm": 0.6849700212478638, + "learning_rate": 2.891545623836127e-05, + "loss": 0.1416, + "step": 11868 + }, + { + "epoch": 33.1536312849162, + "grad_norm": 0.8973786234855652, + "learning_rate": 2.8915083798882682e-05, + "loss": 0.1142, + "step": 11869 + }, + { + "epoch": 33.15642458100559, + "grad_norm": 0.5232882499694824, + "learning_rate": 2.89147113594041e-05, + "loss": 0.111, + "step": 11870 + }, + { + "epoch": 33.159217877094974, + "grad_norm": 0.47905585169792175, + "learning_rate": 2.8914338919925512e-05, + "loss": 0.1224, + "step": 11871 + }, + { + "epoch": 33.16201117318436, + "grad_norm": 0.42836809158325195, + "learning_rate": 2.891396648044693e-05, + "loss": 0.1211, + "step": 11872 + }, + { + "epoch": 33.16480446927374, + "grad_norm": 0.6051400303840637, + "learning_rate": 2.891359404096834e-05, + "loss": 0.1004, + "step": 11873 + }, + { + "epoch": 33.167597765363126, + "grad_norm": 0.7157866358757019, + "learning_rate": 2.8913221601489758e-05, + "loss": 0.105, + "step": 11874 + }, + { + "epoch": 33.17039106145251, + "grad_norm": 0.6608445644378662, + "learning_rate": 2.8912849162011175e-05, + "loss": 0.1001, + "step": 11875 + }, + { + "epoch": 33.1731843575419, + "grad_norm": 0.7600589990615845, + "learning_rate": 2.891247672253259e-05, + "loss": 0.1344, + "step": 11876 + }, + { + "epoch": 33.175977653631286, + "grad_norm": 0.6446343660354614, + "learning_rate": 2.8912104283054008e-05, + "loss": 0.0991, + "step": 11877 + }, + { + "epoch": 33.17877094972067, + "grad_norm": 0.5405754446983337, + "learning_rate": 2.8911731843575418e-05, + "loss": 0.1062, + "step": 11878 + }, + { + "epoch": 33.18156424581006, + "grad_norm": 0.5944010615348816, + "learning_rate": 2.8911359404096834e-05, + "loss": 0.1011, + "step": 11879 + }, + { + "epoch": 33.184357541899445, + "grad_norm": 0.947073757648468, + "learning_rate": 2.891098696461825e-05, + "loss": 0.0918, + "step": 11880 + }, + { + "epoch": 33.187150837988824, + "grad_norm": 0.562556266784668, + "learning_rate": 2.8910614525139667e-05, + "loss": 0.1301, + "step": 11881 + }, + { + "epoch": 33.18994413407821, + "grad_norm": 0.5718497633934021, + "learning_rate": 2.891024208566108e-05, + "loss": 0.087, + "step": 11882 + }, + { + "epoch": 33.1927374301676, + "grad_norm": 0.9126362800598145, + "learning_rate": 2.8909869646182497e-05, + "loss": 0.1128, + "step": 11883 + }, + { + "epoch": 33.19553072625698, + "grad_norm": 0.6396532654762268, + "learning_rate": 2.890949720670391e-05, + "loss": 0.1008, + "step": 11884 + }, + { + "epoch": 33.19832402234637, + "grad_norm": 0.708503007888794, + "learning_rate": 2.8909124767225327e-05, + "loss": 0.1129, + "step": 11885 + }, + { + "epoch": 33.201117318435756, + "grad_norm": 1.704394817352295, + "learning_rate": 2.8908752327746743e-05, + "loss": 0.1052, + "step": 11886 + }, + { + "epoch": 33.20391061452514, + "grad_norm": 0.7068504095077515, + "learning_rate": 2.8908379888268156e-05, + "loss": 0.1195, + "step": 11887 + }, + { + "epoch": 33.20670391061452, + "grad_norm": 0.6166456937789917, + "learning_rate": 2.8908007448789573e-05, + "loss": 0.0999, + "step": 11888 + }, + { + "epoch": 33.20949720670391, + "grad_norm": 0.920451283454895, + "learning_rate": 2.890763500931099e-05, + "loss": 0.134, + "step": 11889 + }, + { + "epoch": 33.212290502793294, + "grad_norm": 0.33773255348205566, + "learning_rate": 2.8907262569832406e-05, + "loss": 0.1403, + "step": 11890 + }, + { + "epoch": 33.21508379888268, + "grad_norm": 0.6431445479393005, + "learning_rate": 2.8906890130353816e-05, + "loss": 0.1251, + "step": 11891 + }, + { + "epoch": 33.21787709497207, + "grad_norm": 0.42108678817749023, + "learning_rate": 2.8906517690875232e-05, + "loss": 0.1024, + "step": 11892 + }, + { + "epoch": 33.220670391061454, + "grad_norm": 0.45759743452072144, + "learning_rate": 2.890614525139665e-05, + "loss": 0.1098, + "step": 11893 + }, + { + "epoch": 33.22346368715084, + "grad_norm": 0.6496647000312805, + "learning_rate": 2.8905772811918065e-05, + "loss": 0.149, + "step": 11894 + }, + { + "epoch": 33.226256983240226, + "grad_norm": 0.46367108821868896, + "learning_rate": 2.8905400372439478e-05, + "loss": 0.139, + "step": 11895 + }, + { + "epoch": 33.229050279329606, + "grad_norm": 0.49851760268211365, + "learning_rate": 2.8905027932960895e-05, + "loss": 0.1289, + "step": 11896 + }, + { + "epoch": 33.23184357541899, + "grad_norm": 1.427432656288147, + "learning_rate": 2.890465549348231e-05, + "loss": 0.1103, + "step": 11897 + }, + { + "epoch": 33.23463687150838, + "grad_norm": 0.6100799441337585, + "learning_rate": 2.8904283054003725e-05, + "loss": 0.109, + "step": 11898 + }, + { + "epoch": 33.237430167597765, + "grad_norm": 0.6036418080329895, + "learning_rate": 2.890391061452514e-05, + "loss": 0.1401, + "step": 11899 + }, + { + "epoch": 33.24022346368715, + "grad_norm": 0.8606651425361633, + "learning_rate": 2.8903538175046554e-05, + "loss": 0.1167, + "step": 11900 + }, + { + "epoch": 33.24301675977654, + "grad_norm": 0.9987406134605408, + "learning_rate": 2.890316573556797e-05, + "loss": 0.1128, + "step": 11901 + }, + { + "epoch": 33.245810055865924, + "grad_norm": 1.2171399593353271, + "learning_rate": 2.8902793296089387e-05, + "loss": 0.0955, + "step": 11902 + }, + { + "epoch": 33.24860335195531, + "grad_norm": 2.0975472927093506, + "learning_rate": 2.8902420856610804e-05, + "loss": 0.1177, + "step": 11903 + }, + { + "epoch": 33.25139664804469, + "grad_norm": 0.5347232818603516, + "learning_rate": 2.8902048417132214e-05, + "loss": 0.1108, + "step": 11904 + }, + { + "epoch": 33.254189944134076, + "grad_norm": 0.6976109147071838, + "learning_rate": 2.890167597765363e-05, + "loss": 0.1123, + "step": 11905 + }, + { + "epoch": 33.25698324022346, + "grad_norm": 0.894544243812561, + "learning_rate": 2.8901303538175047e-05, + "loss": 0.0957, + "step": 11906 + }, + { + "epoch": 33.25977653631285, + "grad_norm": 0.6270454525947571, + "learning_rate": 2.8900931098696463e-05, + "loss": 0.099, + "step": 11907 + }, + { + "epoch": 33.262569832402235, + "grad_norm": 0.5920355319976807, + "learning_rate": 2.890055865921788e-05, + "loss": 0.1012, + "step": 11908 + }, + { + "epoch": 33.26536312849162, + "grad_norm": 0.5719515681266785, + "learning_rate": 2.8900186219739293e-05, + "loss": 0.0902, + "step": 11909 + }, + { + "epoch": 33.26815642458101, + "grad_norm": 0.7015680074691772, + "learning_rate": 2.889981378026071e-05, + "loss": 0.105, + "step": 11910 + }, + { + "epoch": 33.270949720670394, + "grad_norm": 0.8339635133743286, + "learning_rate": 2.8899441340782123e-05, + "loss": 0.0945, + "step": 11911 + }, + { + "epoch": 33.273743016759774, + "grad_norm": 0.8978230953216553, + "learning_rate": 2.889906890130354e-05, + "loss": 0.1017, + "step": 11912 + }, + { + "epoch": 33.27653631284916, + "grad_norm": 0.6773324012756348, + "learning_rate": 2.8898696461824952e-05, + "loss": 0.0966, + "step": 11913 + }, + { + "epoch": 33.279329608938546, + "grad_norm": 1.1373486518859863, + "learning_rate": 2.889832402234637e-05, + "loss": 0.1272, + "step": 11914 + }, + { + "epoch": 33.28212290502793, + "grad_norm": 0.5292335748672485, + "learning_rate": 2.8897951582867785e-05, + "loss": 0.131, + "step": 11915 + }, + { + "epoch": 33.28491620111732, + "grad_norm": 0.35299041867256165, + "learning_rate": 2.8897579143389202e-05, + "loss": 0.1207, + "step": 11916 + }, + { + "epoch": 33.287709497206706, + "grad_norm": 0.6361880302429199, + "learning_rate": 2.889720670391062e-05, + "loss": 0.1179, + "step": 11917 + }, + { + "epoch": 33.29050279329609, + "grad_norm": 0.6454640626907349, + "learning_rate": 2.8896834264432028e-05, + "loss": 0.1368, + "step": 11918 + }, + { + "epoch": 33.29329608938548, + "grad_norm": 0.9586564302444458, + "learning_rate": 2.8896461824953445e-05, + "loss": 0.1432, + "step": 11919 + }, + { + "epoch": 33.29608938547486, + "grad_norm": 0.5333968997001648, + "learning_rate": 2.889608938547486e-05, + "loss": 0.1307, + "step": 11920 + }, + { + "epoch": 33.298882681564244, + "grad_norm": 0.4975660443305969, + "learning_rate": 2.8895716945996278e-05, + "loss": 0.1384, + "step": 11921 + }, + { + "epoch": 33.30167597765363, + "grad_norm": 0.4461301565170288, + "learning_rate": 2.889534450651769e-05, + "loss": 0.1017, + "step": 11922 + }, + { + "epoch": 33.30446927374302, + "grad_norm": 0.5027980208396912, + "learning_rate": 2.8894972067039107e-05, + "loss": 0.1334, + "step": 11923 + }, + { + "epoch": 33.3072625698324, + "grad_norm": 0.5641487240791321, + "learning_rate": 2.889459962756052e-05, + "loss": 0.123, + "step": 11924 + }, + { + "epoch": 33.31005586592179, + "grad_norm": 0.5850476622581482, + "learning_rate": 2.8894227188081937e-05, + "loss": 0.1067, + "step": 11925 + }, + { + "epoch": 33.312849162011176, + "grad_norm": 0.601762592792511, + "learning_rate": 2.8893854748603354e-05, + "loss": 0.1291, + "step": 11926 + }, + { + "epoch": 33.315642458100555, + "grad_norm": 0.9589539766311646, + "learning_rate": 2.8893482309124767e-05, + "loss": 0.1209, + "step": 11927 + }, + { + "epoch": 33.31843575418994, + "grad_norm": 0.6934160590171814, + "learning_rate": 2.8893109869646183e-05, + "loss": 0.0997, + "step": 11928 + }, + { + "epoch": 33.32122905027933, + "grad_norm": 0.6341589093208313, + "learning_rate": 2.88927374301676e-05, + "loss": 0.1013, + "step": 11929 + }, + { + "epoch": 33.324022346368714, + "grad_norm": 0.6728426218032837, + "learning_rate": 2.8892364990689016e-05, + "loss": 0.0918, + "step": 11930 + }, + { + "epoch": 33.3268156424581, + "grad_norm": 0.7008306384086609, + "learning_rate": 2.8891992551210426e-05, + "loss": 0.1072, + "step": 11931 + }, + { + "epoch": 33.32960893854749, + "grad_norm": 0.5848775506019592, + "learning_rate": 2.8891620111731843e-05, + "loss": 0.1067, + "step": 11932 + }, + { + "epoch": 33.332402234636874, + "grad_norm": 1.3687366247177124, + "learning_rate": 2.889124767225326e-05, + "loss": 0.0975, + "step": 11933 + }, + { + "epoch": 33.33519553072626, + "grad_norm": 0.8178009390830994, + "learning_rate": 2.8890875232774676e-05, + "loss": 0.1049, + "step": 11934 + }, + { + "epoch": 33.33798882681564, + "grad_norm": 0.790607213973999, + "learning_rate": 2.8890502793296092e-05, + "loss": 0.0938, + "step": 11935 + }, + { + "epoch": 33.340782122905026, + "grad_norm": 2.32475209236145, + "learning_rate": 2.8890130353817505e-05, + "loss": 0.0878, + "step": 11936 + }, + { + "epoch": 33.34357541899441, + "grad_norm": 0.8475244045257568, + "learning_rate": 2.888975791433892e-05, + "loss": 0.1096, + "step": 11937 + }, + { + "epoch": 33.3463687150838, + "grad_norm": 0.9567390084266663, + "learning_rate": 2.8889385474860335e-05, + "loss": 0.1117, + "step": 11938 + }, + { + "epoch": 33.349162011173185, + "grad_norm": 1.0737760066986084, + "learning_rate": 2.888901303538175e-05, + "loss": 0.1288, + "step": 11939 + }, + { + "epoch": 33.35195530726257, + "grad_norm": 0.5669879913330078, + "learning_rate": 2.8888640595903165e-05, + "loss": 0.1389, + "step": 11940 + }, + { + "epoch": 33.35474860335196, + "grad_norm": 1.0674482583999634, + "learning_rate": 2.888826815642458e-05, + "loss": 0.1384, + "step": 11941 + }, + { + "epoch": 33.357541899441344, + "grad_norm": 0.651211142539978, + "learning_rate": 2.8887895716945998e-05, + "loss": 0.1301, + "step": 11942 + }, + { + "epoch": 33.36033519553072, + "grad_norm": 0.8863028287887573, + "learning_rate": 2.8887523277467414e-05, + "loss": 0.1356, + "step": 11943 + }, + { + "epoch": 33.36312849162011, + "grad_norm": 0.47466549277305603, + "learning_rate": 2.8887150837988828e-05, + "loss": 0.1171, + "step": 11944 + }, + { + "epoch": 33.365921787709496, + "grad_norm": 0.6014111638069153, + "learning_rate": 2.888677839851024e-05, + "loss": 0.148, + "step": 11945 + }, + { + "epoch": 33.36871508379888, + "grad_norm": 0.6862956285476685, + "learning_rate": 2.8886405959031657e-05, + "loss": 0.1402, + "step": 11946 + }, + { + "epoch": 33.37150837988827, + "grad_norm": 0.5442752838134766, + "learning_rate": 2.8886033519553074e-05, + "loss": 0.118, + "step": 11947 + }, + { + "epoch": 33.374301675977655, + "grad_norm": 0.7002395987510681, + "learning_rate": 2.888566108007449e-05, + "loss": 0.1287, + "step": 11948 + }, + { + "epoch": 33.37709497206704, + "grad_norm": 0.7353047728538513, + "learning_rate": 2.8885288640595903e-05, + "loss": 0.1001, + "step": 11949 + }, + { + "epoch": 33.37988826815643, + "grad_norm": 1.0501681566238403, + "learning_rate": 2.888491620111732e-05, + "loss": 0.1044, + "step": 11950 + }, + { + "epoch": 33.38268156424581, + "grad_norm": 1.3877969980239868, + "learning_rate": 2.8884543761638733e-05, + "loss": 0.1304, + "step": 11951 + }, + { + "epoch": 33.385474860335194, + "grad_norm": 0.5502230525016785, + "learning_rate": 2.888417132216015e-05, + "loss": 0.104, + "step": 11952 + }, + { + "epoch": 33.38826815642458, + "grad_norm": 0.8231713771820068, + "learning_rate": 2.8883798882681566e-05, + "loss": 0.1246, + "step": 11953 + }, + { + "epoch": 33.391061452513966, + "grad_norm": 1.0367318391799927, + "learning_rate": 2.888342644320298e-05, + "loss": 0.1085, + "step": 11954 + }, + { + "epoch": 33.39385474860335, + "grad_norm": 0.8085458874702454, + "learning_rate": 2.8883054003724396e-05, + "loss": 0.1189, + "step": 11955 + }, + { + "epoch": 33.39664804469274, + "grad_norm": 2.0722734928131104, + "learning_rate": 2.8882681564245812e-05, + "loss": 0.1196, + "step": 11956 + }, + { + "epoch": 33.399441340782126, + "grad_norm": 0.7507209777832031, + "learning_rate": 2.8882309124767226e-05, + "loss": 0.1094, + "step": 11957 + }, + { + "epoch": 33.40223463687151, + "grad_norm": 0.6941487789154053, + "learning_rate": 2.888193668528864e-05, + "loss": 0.0928, + "step": 11958 + }, + { + "epoch": 33.40502793296089, + "grad_norm": 1.6512529850006104, + "learning_rate": 2.8881564245810055e-05, + "loss": 0.1008, + "step": 11959 + }, + { + "epoch": 33.40782122905028, + "grad_norm": 0.9180213212966919, + "learning_rate": 2.8881191806331472e-05, + "loss": 0.1187, + "step": 11960 + }, + { + "epoch": 33.410614525139664, + "grad_norm": 0.9321079850196838, + "learning_rate": 2.888081936685289e-05, + "loss": 0.0934, + "step": 11961 + }, + { + "epoch": 33.41340782122905, + "grad_norm": 1.0341862440109253, + "learning_rate": 2.8880446927374305e-05, + "loss": 0.1102, + "step": 11962 + }, + { + "epoch": 33.41620111731844, + "grad_norm": 0.855726420879364, + "learning_rate": 2.8880074487895718e-05, + "loss": 0.1244, + "step": 11963 + }, + { + "epoch": 33.41899441340782, + "grad_norm": 1.1319292783737183, + "learning_rate": 2.887970204841713e-05, + "loss": 0.1439, + "step": 11964 + }, + { + "epoch": 33.42178770949721, + "grad_norm": 0.5387114882469177, + "learning_rate": 2.8879329608938548e-05, + "loss": 0.1498, + "step": 11965 + }, + { + "epoch": 33.42458100558659, + "grad_norm": 0.4311525821685791, + "learning_rate": 2.8878957169459964e-05, + "loss": 0.1423, + "step": 11966 + }, + { + "epoch": 33.427374301675975, + "grad_norm": 0.5798940062522888, + "learning_rate": 2.8878584729981377e-05, + "loss": 0.1405, + "step": 11967 + }, + { + "epoch": 33.43016759776536, + "grad_norm": 0.4414791762828827, + "learning_rate": 2.8878212290502794e-05, + "loss": 0.1279, + "step": 11968 + }, + { + "epoch": 33.43296089385475, + "grad_norm": 0.346048504114151, + "learning_rate": 2.887783985102421e-05, + "loss": 0.1219, + "step": 11969 + }, + { + "epoch": 33.435754189944134, + "grad_norm": 0.4155499339103699, + "learning_rate": 2.8877467411545627e-05, + "loss": 0.1196, + "step": 11970 + }, + { + "epoch": 33.43854748603352, + "grad_norm": 0.5778445601463318, + "learning_rate": 2.887709497206704e-05, + "loss": 0.1401, + "step": 11971 + }, + { + "epoch": 33.44134078212291, + "grad_norm": 0.48537784814834595, + "learning_rate": 2.8876722532588453e-05, + "loss": 0.1058, + "step": 11972 + }, + { + "epoch": 33.444134078212294, + "grad_norm": 0.578041672706604, + "learning_rate": 2.887635009310987e-05, + "loss": 0.1302, + "step": 11973 + }, + { + "epoch": 33.44692737430167, + "grad_norm": 0.7752585411071777, + "learning_rate": 2.8875977653631286e-05, + "loss": 0.1287, + "step": 11974 + }, + { + "epoch": 33.44972067039106, + "grad_norm": 0.8605770468711853, + "learning_rate": 2.8875605214152703e-05, + "loss": 0.1285, + "step": 11975 + }, + { + "epoch": 33.452513966480446, + "grad_norm": 0.6376892328262329, + "learning_rate": 2.8875232774674116e-05, + "loss": 0.1084, + "step": 11976 + }, + { + "epoch": 33.45530726256983, + "grad_norm": 1.0281418561935425, + "learning_rate": 2.887486033519553e-05, + "loss": 0.1265, + "step": 11977 + }, + { + "epoch": 33.45810055865922, + "grad_norm": 0.7573197484016418, + "learning_rate": 2.8874487895716946e-05, + "loss": 0.0992, + "step": 11978 + }, + { + "epoch": 33.460893854748605, + "grad_norm": 0.5508536100387573, + "learning_rate": 2.8874115456238362e-05, + "loss": 0.0987, + "step": 11979 + }, + { + "epoch": 33.46368715083799, + "grad_norm": 1.257166862487793, + "learning_rate": 2.887374301675978e-05, + "loss": 0.0962, + "step": 11980 + }, + { + "epoch": 33.46648044692738, + "grad_norm": 0.9305388927459717, + "learning_rate": 2.8873370577281192e-05, + "loss": 0.1015, + "step": 11981 + }, + { + "epoch": 33.46927374301676, + "grad_norm": 0.6357961297035217, + "learning_rate": 2.887299813780261e-05, + "loss": 0.1065, + "step": 11982 + }, + { + "epoch": 33.47206703910614, + "grad_norm": 0.7340711355209351, + "learning_rate": 2.8872625698324025e-05, + "loss": 0.117, + "step": 11983 + }, + { + "epoch": 33.47486033519553, + "grad_norm": 0.9506404399871826, + "learning_rate": 2.8872253258845438e-05, + "loss": 0.134, + "step": 11984 + }, + { + "epoch": 33.477653631284916, + "grad_norm": 1.1830238103866577, + "learning_rate": 2.887188081936685e-05, + "loss": 0.1239, + "step": 11985 + }, + { + "epoch": 33.4804469273743, + "grad_norm": 0.5905396342277527, + "learning_rate": 2.8871508379888268e-05, + "loss": 0.1056, + "step": 11986 + }, + { + "epoch": 33.48324022346369, + "grad_norm": 0.8434512615203857, + "learning_rate": 2.8871135940409684e-05, + "loss": 0.0916, + "step": 11987 + }, + { + "epoch": 33.486033519553075, + "grad_norm": 1.0223803520202637, + "learning_rate": 2.88707635009311e-05, + "loss": 0.1066, + "step": 11988 + }, + { + "epoch": 33.48882681564246, + "grad_norm": 1.4853408336639404, + "learning_rate": 2.8870391061452514e-05, + "loss": 0.1315, + "step": 11989 + }, + { + "epoch": 33.49162011173184, + "grad_norm": 0.5566100478172302, + "learning_rate": 2.887001862197393e-05, + "loss": 0.1423, + "step": 11990 + }, + { + "epoch": 33.49441340782123, + "grad_norm": 1.9042680263519287, + "learning_rate": 2.8869646182495344e-05, + "loss": 0.1257, + "step": 11991 + }, + { + "epoch": 33.497206703910614, + "grad_norm": 0.38484513759613037, + "learning_rate": 2.886927374301676e-05, + "loss": 0.125, + "step": 11992 + }, + { + "epoch": 33.5, + "grad_norm": 0.7436890006065369, + "learning_rate": 2.8868901303538177e-05, + "loss": 0.1299, + "step": 11993 + }, + { + "epoch": 33.502793296089386, + "grad_norm": 0.48428410291671753, + "learning_rate": 2.886852886405959e-05, + "loss": 0.1132, + "step": 11994 + }, + { + "epoch": 33.50558659217877, + "grad_norm": 0.42450639605522156, + "learning_rate": 2.8868156424581006e-05, + "loss": 0.1323, + "step": 11995 + }, + { + "epoch": 33.50837988826816, + "grad_norm": 0.5892863869667053, + "learning_rate": 2.8867783985102423e-05, + "loss": 0.1186, + "step": 11996 + }, + { + "epoch": 33.51117318435754, + "grad_norm": 0.4602404236793518, + "learning_rate": 2.8867411545623836e-05, + "loss": 0.1272, + "step": 11997 + }, + { + "epoch": 33.513966480446925, + "grad_norm": 0.44554421305656433, + "learning_rate": 2.886703910614525e-05, + "loss": 0.1183, + "step": 11998 + }, + { + "epoch": 33.51675977653631, + "grad_norm": 0.42025020718574524, + "learning_rate": 2.8866666666666666e-05, + "loss": 0.1193, + "step": 11999 + }, + { + "epoch": 33.5195530726257, + "grad_norm": 0.6173800230026245, + "learning_rate": 2.8866294227188082e-05, + "loss": 0.1267, + "step": 12000 + }, + { + "epoch": 33.5195530726257, + "eval_cer": 0.1923832280449085, + "eval_loss": 0.28262877464294434, + "eval_runtime": 23.4257, + "eval_samples_per_second": 193.719, + "eval_steps_per_second": 0.64, + "eval_wer": 0.3481746319935197, + "step": 12000 + }, + { + "epoch": 33.522346368715084, + "grad_norm": 0.8347731232643127, + "learning_rate": 2.88659217877095e-05, + "loss": 0.1046, + "step": 12001 + }, + { + "epoch": 33.52513966480447, + "grad_norm": 0.38569268584251404, + "learning_rate": 2.8865549348230915e-05, + "loss": 0.0924, + "step": 12002 + }, + { + "epoch": 33.52793296089386, + "grad_norm": 0.42692431807518005, + "learning_rate": 2.886517690875233e-05, + "loss": 0.0877, + "step": 12003 + }, + { + "epoch": 33.53072625698324, + "grad_norm": 0.5952602624893188, + "learning_rate": 2.8864804469273742e-05, + "loss": 0.1013, + "step": 12004 + }, + { + "epoch": 33.53351955307262, + "grad_norm": 0.5392391681671143, + "learning_rate": 2.8864432029795158e-05, + "loss": 0.0955, + "step": 12005 + }, + { + "epoch": 33.53631284916201, + "grad_norm": 0.8842974305152893, + "learning_rate": 2.8864059590316575e-05, + "loss": 0.1253, + "step": 12006 + }, + { + "epoch": 33.539106145251395, + "grad_norm": 0.49084413051605225, + "learning_rate": 2.8863687150837988e-05, + "loss": 0.1076, + "step": 12007 + }, + { + "epoch": 33.54189944134078, + "grad_norm": 0.8543521165847778, + "learning_rate": 2.8863314711359405e-05, + "loss": 0.0931, + "step": 12008 + }, + { + "epoch": 33.54469273743017, + "grad_norm": 0.983264684677124, + "learning_rate": 2.886294227188082e-05, + "loss": 0.1119, + "step": 12009 + }, + { + "epoch": 33.547486033519554, + "grad_norm": 0.6532970666885376, + "learning_rate": 2.8862569832402238e-05, + "loss": 0.0882, + "step": 12010 + }, + { + "epoch": 33.55027932960894, + "grad_norm": 0.6392433047294617, + "learning_rate": 2.886219739292365e-05, + "loss": 0.1025, + "step": 12011 + }, + { + "epoch": 33.55307262569833, + "grad_norm": 0.7274671196937561, + "learning_rate": 2.8861824953445064e-05, + "loss": 0.1044, + "step": 12012 + }, + { + "epoch": 33.555865921787706, + "grad_norm": 1.9564698934555054, + "learning_rate": 2.886145251396648e-05, + "loss": 0.1096, + "step": 12013 + }, + { + "epoch": 33.55865921787709, + "grad_norm": 1.633194088935852, + "learning_rate": 2.8861080074487897e-05, + "loss": 0.1276, + "step": 12014 + }, + { + "epoch": 33.56145251396648, + "grad_norm": 0.5619674324989319, + "learning_rate": 2.8860707635009313e-05, + "loss": 0.1559, + "step": 12015 + }, + { + "epoch": 33.564245810055866, + "grad_norm": 0.9252403378486633, + "learning_rate": 2.8860335195530727e-05, + "loss": 0.1464, + "step": 12016 + }, + { + "epoch": 33.56703910614525, + "grad_norm": 0.6081932783126831, + "learning_rate": 2.885996275605214e-05, + "loss": 0.1378, + "step": 12017 + }, + { + "epoch": 33.56983240223464, + "grad_norm": 0.6333334445953369, + "learning_rate": 2.8859590316573556e-05, + "loss": 0.1116, + "step": 12018 + }, + { + "epoch": 33.572625698324025, + "grad_norm": 0.9584927558898926, + "learning_rate": 2.8859217877094973e-05, + "loss": 0.1252, + "step": 12019 + }, + { + "epoch": 33.57541899441341, + "grad_norm": 0.4221067726612091, + "learning_rate": 2.885884543761639e-05, + "loss": 0.1321, + "step": 12020 + }, + { + "epoch": 33.57821229050279, + "grad_norm": 0.5741831064224243, + "learning_rate": 2.8858472998137803e-05, + "loss": 0.1128, + "step": 12021 + }, + { + "epoch": 33.58100558659218, + "grad_norm": 0.7997621893882751, + "learning_rate": 2.885810055865922e-05, + "loss": 0.1106, + "step": 12022 + }, + { + "epoch": 33.58379888268156, + "grad_norm": 0.7686396837234497, + "learning_rate": 2.8857728119180636e-05, + "loss": 0.1029, + "step": 12023 + }, + { + "epoch": 33.58659217877095, + "grad_norm": 1.809342384338379, + "learning_rate": 2.885735567970205e-05, + "loss": 0.1169, + "step": 12024 + }, + { + "epoch": 33.589385474860336, + "grad_norm": 0.49419406056404114, + "learning_rate": 2.8856983240223462e-05, + "loss": 0.1051, + "step": 12025 + }, + { + "epoch": 33.59217877094972, + "grad_norm": 0.5091065764427185, + "learning_rate": 2.885661080074488e-05, + "loss": 0.1103, + "step": 12026 + }, + { + "epoch": 33.59497206703911, + "grad_norm": 0.6425865292549133, + "learning_rate": 2.8856238361266295e-05, + "loss": 0.1225, + "step": 12027 + }, + { + "epoch": 33.59776536312849, + "grad_norm": 0.7058922648429871, + "learning_rate": 2.885586592178771e-05, + "loss": 0.1198, + "step": 12028 + }, + { + "epoch": 33.600558659217874, + "grad_norm": 0.5020976662635803, + "learning_rate": 2.8855493482309128e-05, + "loss": 0.126, + "step": 12029 + }, + { + "epoch": 33.60335195530726, + "grad_norm": 0.6380429267883301, + "learning_rate": 2.885512104283054e-05, + "loss": 0.1033, + "step": 12030 + }, + { + "epoch": 33.60614525139665, + "grad_norm": 0.5840025544166565, + "learning_rate": 2.8854748603351954e-05, + "loss": 0.1319, + "step": 12031 + }, + { + "epoch": 33.608938547486034, + "grad_norm": 0.574286937713623, + "learning_rate": 2.885437616387337e-05, + "loss": 0.0987, + "step": 12032 + }, + { + "epoch": 33.61173184357542, + "grad_norm": 0.6874840259552002, + "learning_rate": 2.8854003724394787e-05, + "loss": 0.0968, + "step": 12033 + }, + { + "epoch": 33.614525139664806, + "grad_norm": 1.861464500427246, + "learning_rate": 2.88536312849162e-05, + "loss": 0.091, + "step": 12034 + }, + { + "epoch": 33.61731843575419, + "grad_norm": 0.9345552325248718, + "learning_rate": 2.8853258845437617e-05, + "loss": 0.11, + "step": 12035 + }, + { + "epoch": 33.62011173184357, + "grad_norm": 1.5989056825637817, + "learning_rate": 2.8852886405959034e-05, + "loss": 0.0977, + "step": 12036 + }, + { + "epoch": 33.62290502793296, + "grad_norm": 0.750248372554779, + "learning_rate": 2.8852513966480447e-05, + "loss": 0.0958, + "step": 12037 + }, + { + "epoch": 33.625698324022345, + "grad_norm": 1.3077244758605957, + "learning_rate": 2.8852141527001863e-05, + "loss": 0.1142, + "step": 12038 + }, + { + "epoch": 33.62849162011173, + "grad_norm": 1.0738492012023926, + "learning_rate": 2.8851769087523276e-05, + "loss": 0.1322, + "step": 12039 + }, + { + "epoch": 33.63128491620112, + "grad_norm": 0.3760049641132355, + "learning_rate": 2.8851396648044693e-05, + "loss": 0.1415, + "step": 12040 + }, + { + "epoch": 33.634078212290504, + "grad_norm": 0.5350551605224609, + "learning_rate": 2.885102420856611e-05, + "loss": 0.1538, + "step": 12041 + }, + { + "epoch": 33.63687150837989, + "grad_norm": 0.40974342823028564, + "learning_rate": 2.8850651769087526e-05, + "loss": 0.1261, + "step": 12042 + }, + { + "epoch": 33.63966480446928, + "grad_norm": 0.8398497700691223, + "learning_rate": 2.885027932960894e-05, + "loss": 0.1448, + "step": 12043 + }, + { + "epoch": 33.642458100558656, + "grad_norm": 0.4438527524471283, + "learning_rate": 2.8849906890130352e-05, + "loss": 0.1429, + "step": 12044 + }, + { + "epoch": 33.64525139664804, + "grad_norm": 0.5960460901260376, + "learning_rate": 2.884953445065177e-05, + "loss": 0.1256, + "step": 12045 + }, + { + "epoch": 33.64804469273743, + "grad_norm": 0.737170934677124, + "learning_rate": 2.8849162011173185e-05, + "loss": 0.1243, + "step": 12046 + }, + { + "epoch": 33.650837988826815, + "grad_norm": 0.40275463461875916, + "learning_rate": 2.8848789571694602e-05, + "loss": 0.1292, + "step": 12047 + }, + { + "epoch": 33.6536312849162, + "grad_norm": 0.40619465708732605, + "learning_rate": 2.8848417132216015e-05, + "loss": 0.1051, + "step": 12048 + }, + { + "epoch": 33.65642458100559, + "grad_norm": 0.6816853284835815, + "learning_rate": 2.884804469273743e-05, + "loss": 0.1557, + "step": 12049 + }, + { + "epoch": 33.659217877094974, + "grad_norm": 0.4252112805843353, + "learning_rate": 2.8847672253258848e-05, + "loss": 0.1321, + "step": 12050 + }, + { + "epoch": 33.66201117318436, + "grad_norm": 0.8524900078773499, + "learning_rate": 2.884729981378026e-05, + "loss": 0.1068, + "step": 12051 + }, + { + "epoch": 33.66480446927374, + "grad_norm": 0.5277634859085083, + "learning_rate": 2.8846927374301674e-05, + "loss": 0.1043, + "step": 12052 + }, + { + "epoch": 33.667597765363126, + "grad_norm": 0.47487518191337585, + "learning_rate": 2.884655493482309e-05, + "loss": 0.1159, + "step": 12053 + }, + { + "epoch": 33.67039106145251, + "grad_norm": 0.7155695557594299, + "learning_rate": 2.8846182495344508e-05, + "loss": 0.0993, + "step": 12054 + }, + { + "epoch": 33.6731843575419, + "grad_norm": 0.6022620797157288, + "learning_rate": 2.8845810055865924e-05, + "loss": 0.1147, + "step": 12055 + }, + { + "epoch": 33.675977653631286, + "grad_norm": 0.407840371131897, + "learning_rate": 2.884543761638734e-05, + "loss": 0.102, + "step": 12056 + }, + { + "epoch": 33.67877094972067, + "grad_norm": 0.6576537489891052, + "learning_rate": 2.884506517690875e-05, + "loss": 0.1172, + "step": 12057 + }, + { + "epoch": 33.68156424581006, + "grad_norm": 0.6776052713394165, + "learning_rate": 2.8844692737430167e-05, + "loss": 0.0949, + "step": 12058 + }, + { + "epoch": 33.684357541899445, + "grad_norm": 0.9328562617301941, + "learning_rate": 2.8844320297951583e-05, + "loss": 0.1043, + "step": 12059 + }, + { + "epoch": 33.687150837988824, + "grad_norm": 0.5954485535621643, + "learning_rate": 2.8843947858473e-05, + "loss": 0.1068, + "step": 12060 + }, + { + "epoch": 33.68994413407821, + "grad_norm": 0.8498448133468628, + "learning_rate": 2.8843575418994413e-05, + "loss": 0.0952, + "step": 12061 + }, + { + "epoch": 33.6927374301676, + "grad_norm": 1.9723049402236938, + "learning_rate": 2.884320297951583e-05, + "loss": 0.1075, + "step": 12062 + }, + { + "epoch": 33.69553072625698, + "grad_norm": 0.5684854984283447, + "learning_rate": 2.8842830540037246e-05, + "loss": 0.0882, + "step": 12063 + }, + { + "epoch": 33.69832402234637, + "grad_norm": 1.3210735321044922, + "learning_rate": 2.884245810055866e-05, + "loss": 0.1256, + "step": 12064 + }, + { + "epoch": 33.701117318435756, + "grad_norm": 0.6330487132072449, + "learning_rate": 2.8842085661080076e-05, + "loss": 0.1418, + "step": 12065 + }, + { + "epoch": 33.70391061452514, + "grad_norm": 0.5472257137298584, + "learning_rate": 2.884171322160149e-05, + "loss": 0.1178, + "step": 12066 + }, + { + "epoch": 33.70670391061452, + "grad_norm": 0.4341663718223572, + "learning_rate": 2.8841340782122906e-05, + "loss": 0.1419, + "step": 12067 + }, + { + "epoch": 33.70949720670391, + "grad_norm": 1.6247992515563965, + "learning_rate": 2.8840968342644322e-05, + "loss": 0.1458, + "step": 12068 + }, + { + "epoch": 33.712290502793294, + "grad_norm": 0.6505889296531677, + "learning_rate": 2.884059590316574e-05, + "loss": 0.1371, + "step": 12069 + }, + { + "epoch": 33.71508379888268, + "grad_norm": 1.4585626125335693, + "learning_rate": 2.8840223463687152e-05, + "loss": 0.1376, + "step": 12070 + }, + { + "epoch": 33.71787709497207, + "grad_norm": 0.4142107367515564, + "learning_rate": 2.8839851024208565e-05, + "loss": 0.115, + "step": 12071 + }, + { + "epoch": 33.720670391061454, + "grad_norm": 1.304884672164917, + "learning_rate": 2.883947858472998e-05, + "loss": 0.1367, + "step": 12072 + }, + { + "epoch": 33.72346368715084, + "grad_norm": 0.5887250304222107, + "learning_rate": 2.8839106145251398e-05, + "loss": 0.1152, + "step": 12073 + }, + { + "epoch": 33.726256983240226, + "grad_norm": 1.820916771888733, + "learning_rate": 2.8838733705772815e-05, + "loss": 0.1383, + "step": 12074 + }, + { + "epoch": 33.729050279329606, + "grad_norm": 0.9225091338157654, + "learning_rate": 2.8838361266294228e-05, + "loss": 0.1328, + "step": 12075 + }, + { + "epoch": 33.73184357541899, + "grad_norm": 0.48643213510513306, + "learning_rate": 2.8837988826815644e-05, + "loss": 0.103, + "step": 12076 + }, + { + "epoch": 33.73463687150838, + "grad_norm": 0.6281489729881287, + "learning_rate": 2.8837616387337057e-05, + "loss": 0.1419, + "step": 12077 + }, + { + "epoch": 33.737430167597765, + "grad_norm": 0.4533924162387848, + "learning_rate": 2.8837243947858474e-05, + "loss": 0.1098, + "step": 12078 + }, + { + "epoch": 33.74022346368715, + "grad_norm": 0.5373750329017639, + "learning_rate": 2.8836871508379887e-05, + "loss": 0.091, + "step": 12079 + }, + { + "epoch": 33.74301675977654, + "grad_norm": 0.5219509601593018, + "learning_rate": 2.8836499068901304e-05, + "loss": 0.0969, + "step": 12080 + }, + { + "epoch": 33.745810055865924, + "grad_norm": 0.45361557602882385, + "learning_rate": 2.883612662942272e-05, + "loss": 0.1056, + "step": 12081 + }, + { + "epoch": 33.74860335195531, + "grad_norm": 0.5206710696220398, + "learning_rate": 2.8835754189944137e-05, + "loss": 0.1026, + "step": 12082 + }, + { + "epoch": 33.75139664804469, + "grad_norm": 0.4368840157985687, + "learning_rate": 2.883538175046555e-05, + "loss": 0.1061, + "step": 12083 + }, + { + "epoch": 33.754189944134076, + "grad_norm": 0.5068817138671875, + "learning_rate": 2.8835009310986963e-05, + "loss": 0.0774, + "step": 12084 + }, + { + "epoch": 33.75698324022346, + "grad_norm": 1.0821444988250732, + "learning_rate": 2.883463687150838e-05, + "loss": 0.1047, + "step": 12085 + }, + { + "epoch": 33.75977653631285, + "grad_norm": 0.6247003078460693, + "learning_rate": 2.8834264432029796e-05, + "loss": 0.0977, + "step": 12086 + }, + { + "epoch": 33.762569832402235, + "grad_norm": 0.979183554649353, + "learning_rate": 2.8833891992551213e-05, + "loss": 0.1136, + "step": 12087 + }, + { + "epoch": 33.76536312849162, + "grad_norm": 1.4834308624267578, + "learning_rate": 2.8833519553072626e-05, + "loss": 0.1454, + "step": 12088 + }, + { + "epoch": 33.76815642458101, + "grad_norm": 1.4447088241577148, + "learning_rate": 2.8833147113594042e-05, + "loss": 0.1559, + "step": 12089 + }, + { + "epoch": 33.770949720670394, + "grad_norm": 0.48456767201423645, + "learning_rate": 2.883277467411546e-05, + "loss": 0.1301, + "step": 12090 + }, + { + "epoch": 33.773743016759774, + "grad_norm": 0.9529103636741638, + "learning_rate": 2.8832402234636872e-05, + "loss": 0.1362, + "step": 12091 + }, + { + "epoch": 33.77653631284916, + "grad_norm": 0.8141891360282898, + "learning_rate": 2.8832029795158285e-05, + "loss": 0.1581, + "step": 12092 + }, + { + "epoch": 33.779329608938546, + "grad_norm": 0.5127532482147217, + "learning_rate": 2.88316573556797e-05, + "loss": 0.1222, + "step": 12093 + }, + { + "epoch": 33.78212290502793, + "grad_norm": 0.5042409896850586, + "learning_rate": 2.8831284916201118e-05, + "loss": 0.1391, + "step": 12094 + }, + { + "epoch": 33.78491620111732, + "grad_norm": 0.4732040762901306, + "learning_rate": 2.8830912476722535e-05, + "loss": 0.1197, + "step": 12095 + }, + { + "epoch": 33.787709497206706, + "grad_norm": 0.40316933393478394, + "learning_rate": 2.883054003724395e-05, + "loss": 0.136, + "step": 12096 + }, + { + "epoch": 33.79050279329609, + "grad_norm": 0.5223598480224609, + "learning_rate": 2.883016759776536e-05, + "loss": 0.1144, + "step": 12097 + }, + { + "epoch": 33.79329608938548, + "grad_norm": 0.8223846554756165, + "learning_rate": 2.8829795158286778e-05, + "loss": 0.1176, + "step": 12098 + }, + { + "epoch": 33.79608938547486, + "grad_norm": 0.6258419752120972, + "learning_rate": 2.8829422718808194e-05, + "loss": 0.1275, + "step": 12099 + }, + { + "epoch": 33.798882681564244, + "grad_norm": 1.1711677312850952, + "learning_rate": 2.882905027932961e-05, + "loss": 0.1273, + "step": 12100 + }, + { + "epoch": 33.80167597765363, + "grad_norm": 0.43600520491600037, + "learning_rate": 2.8828677839851024e-05, + "loss": 0.0977, + "step": 12101 + }, + { + "epoch": 33.80446927374302, + "grad_norm": 0.7021886110305786, + "learning_rate": 2.882830540037244e-05, + "loss": 0.1044, + "step": 12102 + }, + { + "epoch": 33.8072625698324, + "grad_norm": 0.6917007565498352, + "learning_rate": 2.8827932960893857e-05, + "loss": 0.1004, + "step": 12103 + }, + { + "epoch": 33.81005586592179, + "grad_norm": 0.9114732146263123, + "learning_rate": 2.882756052141527e-05, + "loss": 0.1197, + "step": 12104 + }, + { + "epoch": 33.812849162011176, + "grad_norm": 0.6026027798652649, + "learning_rate": 2.8827188081936686e-05, + "loss": 0.1101, + "step": 12105 + }, + { + "epoch": 33.815642458100555, + "grad_norm": 0.4195392429828644, + "learning_rate": 2.88268156424581e-05, + "loss": 0.0849, + "step": 12106 + }, + { + "epoch": 33.81843575418994, + "grad_norm": 1.3626302480697632, + "learning_rate": 2.8826443202979516e-05, + "loss": 0.0973, + "step": 12107 + }, + { + "epoch": 33.82122905027933, + "grad_norm": 1.1943187713623047, + "learning_rate": 2.8826070763500933e-05, + "loss": 0.1325, + "step": 12108 + }, + { + "epoch": 33.824022346368714, + "grad_norm": 0.6423885822296143, + "learning_rate": 2.882569832402235e-05, + "loss": 0.1161, + "step": 12109 + }, + { + "epoch": 33.8268156424581, + "grad_norm": 1.3500010967254639, + "learning_rate": 2.8825325884543762e-05, + "loss": 0.1112, + "step": 12110 + }, + { + "epoch": 33.82960893854749, + "grad_norm": 0.8457313776016235, + "learning_rate": 2.8824953445065176e-05, + "loss": 0.1122, + "step": 12111 + }, + { + "epoch": 33.832402234636874, + "grad_norm": 0.8273063898086548, + "learning_rate": 2.8824581005586592e-05, + "loss": 0.1134, + "step": 12112 + }, + { + "epoch": 33.83519553072626, + "grad_norm": 0.7616357207298279, + "learning_rate": 2.882420856610801e-05, + "loss": 0.1119, + "step": 12113 + }, + { + "epoch": 33.83798882681564, + "grad_norm": 2.1788299083709717, + "learning_rate": 2.8823836126629425e-05, + "loss": 0.1441, + "step": 12114 + }, + { + "epoch": 33.840782122905026, + "grad_norm": 0.4490053653717041, + "learning_rate": 2.8823463687150838e-05, + "loss": 0.16, + "step": 12115 + }, + { + "epoch": 33.84357541899441, + "grad_norm": 0.5000353455543518, + "learning_rate": 2.8823091247672255e-05, + "loss": 0.1415, + "step": 12116 + }, + { + "epoch": 33.8463687150838, + "grad_norm": 0.44445353746414185, + "learning_rate": 2.8822718808193668e-05, + "loss": 0.1416, + "step": 12117 + }, + { + "epoch": 33.849162011173185, + "grad_norm": 0.430544376373291, + "learning_rate": 2.8822346368715084e-05, + "loss": 0.1225, + "step": 12118 + }, + { + "epoch": 33.85195530726257, + "grad_norm": 0.48886945843696594, + "learning_rate": 2.8821973929236498e-05, + "loss": 0.121, + "step": 12119 + }, + { + "epoch": 33.85474860335196, + "grad_norm": 0.7372334599494934, + "learning_rate": 2.8821601489757914e-05, + "loss": 0.1078, + "step": 12120 + }, + { + "epoch": 33.857541899441344, + "grad_norm": 0.6973428726196289, + "learning_rate": 2.882122905027933e-05, + "loss": 0.1407, + "step": 12121 + }, + { + "epoch": 33.86033519553072, + "grad_norm": 0.6394141316413879, + "learning_rate": 2.8820856610800747e-05, + "loss": 0.1264, + "step": 12122 + }, + { + "epoch": 33.86312849162011, + "grad_norm": 0.659797728061676, + "learning_rate": 2.8820484171322164e-05, + "loss": 0.1283, + "step": 12123 + }, + { + "epoch": 33.865921787709496, + "grad_norm": 0.5654914975166321, + "learning_rate": 2.8820111731843574e-05, + "loss": 0.1145, + "step": 12124 + }, + { + "epoch": 33.86871508379888, + "grad_norm": 0.42376047372817993, + "learning_rate": 2.881973929236499e-05, + "loss": 0.1059, + "step": 12125 + }, + { + "epoch": 33.87150837988827, + "grad_norm": 0.5230462551116943, + "learning_rate": 2.8819366852886407e-05, + "loss": 0.1096, + "step": 12126 + }, + { + "epoch": 33.874301675977655, + "grad_norm": 0.605443000793457, + "learning_rate": 2.8818994413407823e-05, + "loss": 0.1056, + "step": 12127 + }, + { + "epoch": 33.87709497206704, + "grad_norm": 0.817352831363678, + "learning_rate": 2.8818621973929236e-05, + "loss": 0.1229, + "step": 12128 + }, + { + "epoch": 33.87988826815643, + "grad_norm": 2.2278077602386475, + "learning_rate": 2.8818249534450653e-05, + "loss": 0.108, + "step": 12129 + }, + { + "epoch": 33.88268156424581, + "grad_norm": 0.5802817940711975, + "learning_rate": 2.881787709497207e-05, + "loss": 0.124, + "step": 12130 + }, + { + "epoch": 33.885474860335194, + "grad_norm": 0.853345513343811, + "learning_rate": 2.8817504655493483e-05, + "loss": 0.095, + "step": 12131 + }, + { + "epoch": 33.88826815642458, + "grad_norm": 0.5885377526283264, + "learning_rate": 2.88171322160149e-05, + "loss": 0.1064, + "step": 12132 + }, + { + "epoch": 33.891061452513966, + "grad_norm": 0.5610578060150146, + "learning_rate": 2.8816759776536312e-05, + "loss": 0.1021, + "step": 12133 + }, + { + "epoch": 33.89385474860335, + "grad_norm": 0.5034704804420471, + "learning_rate": 2.881638733705773e-05, + "loss": 0.109, + "step": 12134 + }, + { + "epoch": 33.89664804469274, + "grad_norm": 0.5927660465240479, + "learning_rate": 2.8816014897579145e-05, + "loss": 0.1278, + "step": 12135 + }, + { + "epoch": 33.899441340782126, + "grad_norm": 0.7882961630821228, + "learning_rate": 2.8815642458100562e-05, + "loss": 0.0971, + "step": 12136 + }, + { + "epoch": 33.90223463687151, + "grad_norm": 0.8573343753814697, + "learning_rate": 2.881527001862197e-05, + "loss": 0.1085, + "step": 12137 + }, + { + "epoch": 33.90502793296089, + "grad_norm": 1.0902026891708374, + "learning_rate": 2.8814897579143388e-05, + "loss": 0.107, + "step": 12138 + }, + { + "epoch": 33.90782122905028, + "grad_norm": 2.159317970275879, + "learning_rate": 2.8814525139664805e-05, + "loss": 0.1392, + "step": 12139 + }, + { + "epoch": 33.910614525139664, + "grad_norm": 0.39848846197128296, + "learning_rate": 2.881415270018622e-05, + "loss": 0.1417, + "step": 12140 + }, + { + "epoch": 33.91340782122905, + "grad_norm": 0.5939761400222778, + "learning_rate": 2.8813780260707638e-05, + "loss": 0.1474, + "step": 12141 + }, + { + "epoch": 33.91620111731844, + "grad_norm": 0.4674624502658844, + "learning_rate": 2.881340782122905e-05, + "loss": 0.1246, + "step": 12142 + }, + { + "epoch": 33.91899441340782, + "grad_norm": 0.7521407008171082, + "learning_rate": 2.8813035381750467e-05, + "loss": 0.1548, + "step": 12143 + }, + { + "epoch": 33.92178770949721, + "grad_norm": 0.6610081195831299, + "learning_rate": 2.881266294227188e-05, + "loss": 0.1495, + "step": 12144 + }, + { + "epoch": 33.92458100558659, + "grad_norm": 0.5000593662261963, + "learning_rate": 2.8812290502793297e-05, + "loss": 0.1076, + "step": 12145 + }, + { + "epoch": 33.927374301675975, + "grad_norm": 0.5141152143478394, + "learning_rate": 2.881191806331471e-05, + "loss": 0.1248, + "step": 12146 + }, + { + "epoch": 33.93016759776536, + "grad_norm": 1.2036556005477905, + "learning_rate": 2.8811545623836127e-05, + "loss": 0.1272, + "step": 12147 + }, + { + "epoch": 33.93296089385475, + "grad_norm": 1.0991889238357544, + "learning_rate": 2.8811173184357543e-05, + "loss": 0.123, + "step": 12148 + }, + { + "epoch": 33.935754189944134, + "grad_norm": 0.8348732590675354, + "learning_rate": 2.881080074487896e-05, + "loss": 0.139, + "step": 12149 + }, + { + "epoch": 33.93854748603352, + "grad_norm": 0.7661930918693542, + "learning_rate": 2.8810428305400376e-05, + "loss": 0.1397, + "step": 12150 + }, + { + "epoch": 33.94134078212291, + "grad_norm": 0.5570211410522461, + "learning_rate": 2.8810055865921786e-05, + "loss": 0.0898, + "step": 12151 + }, + { + "epoch": 33.944134078212294, + "grad_norm": 0.39525118470191956, + "learning_rate": 2.8809683426443203e-05, + "loss": 0.1068, + "step": 12152 + }, + { + "epoch": 33.94692737430167, + "grad_norm": 0.4747087359428406, + "learning_rate": 2.880931098696462e-05, + "loss": 0.1231, + "step": 12153 + }, + { + "epoch": 33.94972067039106, + "grad_norm": 0.49552568793296814, + "learning_rate": 2.8808938547486036e-05, + "loss": 0.1001, + "step": 12154 + }, + { + "epoch": 33.952513966480446, + "grad_norm": 0.7116245031356812, + "learning_rate": 2.880856610800745e-05, + "loss": 0.1044, + "step": 12155 + }, + { + "epoch": 33.95530726256983, + "grad_norm": 1.059133768081665, + "learning_rate": 2.8808193668528865e-05, + "loss": 0.1012, + "step": 12156 + }, + { + "epoch": 33.95810055865922, + "grad_norm": 0.6314469575881958, + "learning_rate": 2.880782122905028e-05, + "loss": 0.1081, + "step": 12157 + }, + { + "epoch": 33.960893854748605, + "grad_norm": 0.729124903678894, + "learning_rate": 2.8807448789571695e-05, + "loss": 0.106, + "step": 12158 + }, + { + "epoch": 33.96368715083799, + "grad_norm": 1.275092363357544, + "learning_rate": 2.880707635009311e-05, + "loss": 0.0997, + "step": 12159 + }, + { + "epoch": 33.96648044692738, + "grad_norm": 1.5366345643997192, + "learning_rate": 2.8806703910614525e-05, + "loss": 0.1173, + "step": 12160 + }, + { + "epoch": 33.96927374301676, + "grad_norm": 0.6829842329025269, + "learning_rate": 2.880633147113594e-05, + "loss": 0.0999, + "step": 12161 + }, + { + "epoch": 33.97206703910614, + "grad_norm": 1.4889684915542603, + "learning_rate": 2.8805959031657358e-05, + "loss": 0.1104, + "step": 12162 + }, + { + "epoch": 33.97486033519553, + "grad_norm": 0.9617162942886353, + "learning_rate": 2.8805586592178774e-05, + "loss": 0.1123, + "step": 12163 + }, + { + "epoch": 33.977653631284916, + "grad_norm": 1.9676289558410645, + "learning_rate": 2.8805214152700184e-05, + "loss": 0.1485, + "step": 12164 + }, + { + "epoch": 33.9804469273743, + "grad_norm": 0.5549558401107788, + "learning_rate": 2.88048417132216e-05, + "loss": 0.1287, + "step": 12165 + }, + { + "epoch": 33.98324022346369, + "grad_norm": 0.533507764339447, + "learning_rate": 2.8804469273743017e-05, + "loss": 0.144, + "step": 12166 + }, + { + "epoch": 33.986033519553075, + "grad_norm": 0.43817782402038574, + "learning_rate": 2.8804096834264434e-05, + "loss": 0.114, + "step": 12167 + }, + { + "epoch": 33.98882681564246, + "grad_norm": 0.8531481027603149, + "learning_rate": 2.880372439478585e-05, + "loss": 0.1375, + "step": 12168 + }, + { + "epoch": 33.99162011173184, + "grad_norm": 0.9275565147399902, + "learning_rate": 2.8803351955307263e-05, + "loss": 0.1256, + "step": 12169 + }, + { + "epoch": 33.99441340782123, + "grad_norm": 1.0315141677856445, + "learning_rate": 2.880297951582868e-05, + "loss": 0.1094, + "step": 12170 + }, + { + "epoch": 33.997206703910614, + "grad_norm": 1.327575922012329, + "learning_rate": 2.8802607076350093e-05, + "loss": 0.1106, + "step": 12171 + }, + { + "epoch": 34.0, + "grad_norm": 0.7876173853874207, + "learning_rate": 2.880223463687151e-05, + "loss": 0.1079, + "step": 12172 + }, + { + "epoch": 34.002793296089386, + "grad_norm": 0.36049985885620117, + "learning_rate": 2.8801862197392923e-05, + "loss": 0.1309, + "step": 12173 + }, + { + "epoch": 34.00558659217877, + "grad_norm": 0.4788808822631836, + "learning_rate": 2.880148975791434e-05, + "loss": 0.1479, + "step": 12174 + }, + { + "epoch": 34.00837988826816, + "grad_norm": 0.48985806107521057, + "learning_rate": 2.8801117318435756e-05, + "loss": 0.1435, + "step": 12175 + }, + { + "epoch": 34.01117318435754, + "grad_norm": 0.7011278867721558, + "learning_rate": 2.8800744878957172e-05, + "loss": 0.1603, + "step": 12176 + }, + { + "epoch": 34.013966480446925, + "grad_norm": 0.5953546762466431, + "learning_rate": 2.8800372439478586e-05, + "loss": 0.1462, + "step": 12177 + }, + { + "epoch": 34.01675977653631, + "grad_norm": 1.1388782262802124, + "learning_rate": 2.88e-05, + "loss": 0.1301, + "step": 12178 + }, + { + "epoch": 34.0195530726257, + "grad_norm": 0.6492198705673218, + "learning_rate": 2.8799627560521415e-05, + "loss": 0.1264, + "step": 12179 + }, + { + "epoch": 34.022346368715084, + "grad_norm": 0.672677755355835, + "learning_rate": 2.8799255121042832e-05, + "loss": 0.1021, + "step": 12180 + }, + { + "epoch": 34.02513966480447, + "grad_norm": 1.3372321128845215, + "learning_rate": 2.8798882681564248e-05, + "loss": 0.1022, + "step": 12181 + }, + { + "epoch": 34.02793296089386, + "grad_norm": 0.5798516869544983, + "learning_rate": 2.879851024208566e-05, + "loss": 0.1083, + "step": 12182 + }, + { + "epoch": 34.03072625698324, + "grad_norm": 0.5085784792900085, + "learning_rate": 2.8798137802607078e-05, + "loss": 0.1074, + "step": 12183 + }, + { + "epoch": 34.03351955307262, + "grad_norm": 1.3595062494277954, + "learning_rate": 2.879776536312849e-05, + "loss": 0.113, + "step": 12184 + }, + { + "epoch": 34.03631284916201, + "grad_norm": 0.6317221522331238, + "learning_rate": 2.8797392923649908e-05, + "loss": 0.0996, + "step": 12185 + }, + { + "epoch": 34.039106145251395, + "grad_norm": 0.8338974714279175, + "learning_rate": 2.879702048417132e-05, + "loss": 0.0832, + "step": 12186 + }, + { + "epoch": 34.04189944134078, + "grad_norm": 1.3796659708023071, + "learning_rate": 2.8796648044692737e-05, + "loss": 0.0808, + "step": 12187 + }, + { + "epoch": 34.04469273743017, + "grad_norm": 0.62143474817276, + "learning_rate": 2.8796275605214154e-05, + "loss": 0.0851, + "step": 12188 + }, + { + "epoch": 34.047486033519554, + "grad_norm": 0.542172908782959, + "learning_rate": 2.879590316573557e-05, + "loss": 0.1024, + "step": 12189 + }, + { + "epoch": 34.05027932960894, + "grad_norm": 6.469529151916504, + "learning_rate": 2.8795530726256987e-05, + "loss": 0.0923, + "step": 12190 + }, + { + "epoch": 34.05307262569833, + "grad_norm": 0.6742998957633972, + "learning_rate": 2.8795158286778397e-05, + "loss": 0.0839, + "step": 12191 + }, + { + "epoch": 34.055865921787706, + "grad_norm": 0.6223399043083191, + "learning_rate": 2.8794785847299813e-05, + "loss": 0.0839, + "step": 12192 + }, + { + "epoch": 34.05865921787709, + "grad_norm": 0.8658015131950378, + "learning_rate": 2.879441340782123e-05, + "loss": 0.1061, + "step": 12193 + }, + { + "epoch": 34.06145251396648, + "grad_norm": 0.5656163692474365, + "learning_rate": 2.8794040968342646e-05, + "loss": 0.1161, + "step": 12194 + }, + { + "epoch": 34.064245810055866, + "grad_norm": 1.068800687789917, + "learning_rate": 2.879366852886406e-05, + "loss": 0.1006, + "step": 12195 + }, + { + "epoch": 34.06703910614525, + "grad_norm": 0.6262120008468628, + "learning_rate": 2.8793296089385476e-05, + "loss": 0.0902, + "step": 12196 + }, + { + "epoch": 34.06983240223464, + "grad_norm": 1.0646013021469116, + "learning_rate": 2.879292364990689e-05, + "loss": 0.1395, + "step": 12197 + }, + { + "epoch": 34.072625698324025, + "grad_norm": 0.47066718339920044, + "learning_rate": 2.8792551210428306e-05, + "loss": 0.1227, + "step": 12198 + }, + { + "epoch": 34.07541899441341, + "grad_norm": 0.9049997329711914, + "learning_rate": 2.8792178770949722e-05, + "loss": 0.131, + "step": 12199 + }, + { + "epoch": 34.07821229050279, + "grad_norm": 1.037468671798706, + "learning_rate": 2.8791806331471135e-05, + "loss": 0.1174, + "step": 12200 + }, + { + "epoch": 34.08100558659218, + "grad_norm": 0.8058502078056335, + "learning_rate": 2.8791433891992552e-05, + "loss": 0.1058, + "step": 12201 + }, + { + "epoch": 34.08379888268156, + "grad_norm": 0.5095674395561218, + "learning_rate": 2.879106145251397e-05, + "loss": 0.1201, + "step": 12202 + }, + { + "epoch": 34.08659217877095, + "grad_norm": 0.6402010321617126, + "learning_rate": 2.8790689013035385e-05, + "loss": 0.132, + "step": 12203 + }, + { + "epoch": 34.089385474860336, + "grad_norm": 0.4469653069972992, + "learning_rate": 2.8790316573556795e-05, + "loss": 0.1152, + "step": 12204 + }, + { + "epoch": 34.09217877094972, + "grad_norm": 0.47073206305503845, + "learning_rate": 2.878994413407821e-05, + "loss": 0.1033, + "step": 12205 + }, + { + "epoch": 34.09497206703911, + "grad_norm": 0.5954446196556091, + "learning_rate": 2.8789571694599628e-05, + "loss": 0.1157, + "step": 12206 + }, + { + "epoch": 34.097765363128495, + "grad_norm": 0.5807205438613892, + "learning_rate": 2.8789199255121044e-05, + "loss": 0.1021, + "step": 12207 + }, + { + "epoch": 34.100558659217874, + "grad_norm": 0.5283413529396057, + "learning_rate": 2.878882681564246e-05, + "loss": 0.1039, + "step": 12208 + }, + { + "epoch": 34.10335195530726, + "grad_norm": 0.3920169174671173, + "learning_rate": 2.8788454376163874e-05, + "loss": 0.1157, + "step": 12209 + }, + { + "epoch": 34.10614525139665, + "grad_norm": 0.6349837779998779, + "learning_rate": 2.878808193668529e-05, + "loss": 0.1001, + "step": 12210 + }, + { + "epoch": 34.108938547486034, + "grad_norm": 0.6248700618743896, + "learning_rate": 2.8787709497206704e-05, + "loss": 0.1059, + "step": 12211 + }, + { + "epoch": 34.11173184357542, + "grad_norm": 0.6229177713394165, + "learning_rate": 2.878733705772812e-05, + "loss": 0.0988, + "step": 12212 + }, + { + "epoch": 34.114525139664806, + "grad_norm": 0.7188684344291687, + "learning_rate": 2.8786964618249533e-05, + "loss": 0.0845, + "step": 12213 + }, + { + "epoch": 34.11731843575419, + "grad_norm": 2.283607244491577, + "learning_rate": 2.878659217877095e-05, + "loss": 0.1237, + "step": 12214 + }, + { + "epoch": 34.12011173184357, + "grad_norm": 0.9892615079879761, + "learning_rate": 2.8786219739292366e-05, + "loss": 0.1057, + "step": 12215 + }, + { + "epoch": 34.12290502793296, + "grad_norm": 0.6058937907218933, + "learning_rate": 2.8785847299813783e-05, + "loss": 0.0906, + "step": 12216 + }, + { + "epoch": 34.125698324022345, + "grad_norm": 0.4354842007160187, + "learning_rate": 2.8785474860335196e-05, + "loss": 0.0846, + "step": 12217 + }, + { + "epoch": 34.12849162011173, + "grad_norm": 0.7500052452087402, + "learning_rate": 2.878510242085661e-05, + "loss": 0.1052, + "step": 12218 + }, + { + "epoch": 34.13128491620112, + "grad_norm": 1.1640642881393433, + "learning_rate": 2.8784729981378026e-05, + "loss": 0.0921, + "step": 12219 + }, + { + "epoch": 34.134078212290504, + "grad_norm": 0.7271528244018555, + "learning_rate": 2.8784357541899442e-05, + "loss": 0.0895, + "step": 12220 + }, + { + "epoch": 34.13687150837989, + "grad_norm": 0.8843027353286743, + "learning_rate": 2.878398510242086e-05, + "loss": 0.0993, + "step": 12221 + }, + { + "epoch": 34.13966480446928, + "grad_norm": 0.9962180852890015, + "learning_rate": 2.8783612662942272e-05, + "loss": 0.1303, + "step": 12222 + }, + { + "epoch": 34.142458100558656, + "grad_norm": 0.5881813168525696, + "learning_rate": 2.878324022346369e-05, + "loss": 0.1286, + "step": 12223 + }, + { + "epoch": 34.14525139664804, + "grad_norm": 0.4112342894077301, + "learning_rate": 2.8782867783985102e-05, + "loss": 0.1375, + "step": 12224 + }, + { + "epoch": 34.14804469273743, + "grad_norm": 0.44017621874809265, + "learning_rate": 2.8782495344506518e-05, + "loss": 0.127, + "step": 12225 + }, + { + "epoch": 34.150837988826815, + "grad_norm": 0.4743161201477051, + "learning_rate": 2.8782122905027935e-05, + "loss": 0.1161, + "step": 12226 + }, + { + "epoch": 34.1536312849162, + "grad_norm": 0.7615073323249817, + "learning_rate": 2.8781750465549348e-05, + "loss": 0.1321, + "step": 12227 + }, + { + "epoch": 34.15642458100559, + "grad_norm": 0.5019351243972778, + "learning_rate": 2.8781378026070764e-05, + "loss": 0.1118, + "step": 12228 + }, + { + "epoch": 34.159217877094974, + "grad_norm": 0.5159565806388855, + "learning_rate": 2.878100558659218e-05, + "loss": 0.1426, + "step": 12229 + }, + { + "epoch": 34.16201117318436, + "grad_norm": 0.5144844651222229, + "learning_rate": 2.8780633147113598e-05, + "loss": 0.1117, + "step": 12230 + }, + { + "epoch": 34.16480446927374, + "grad_norm": 0.6701743006706238, + "learning_rate": 2.8780260707635007e-05, + "loss": 0.1153, + "step": 12231 + }, + { + "epoch": 34.167597765363126, + "grad_norm": 0.6476370692253113, + "learning_rate": 2.8779888268156424e-05, + "loss": 0.1017, + "step": 12232 + }, + { + "epoch": 34.17039106145251, + "grad_norm": 0.6108120083808899, + "learning_rate": 2.877951582867784e-05, + "loss": 0.0913, + "step": 12233 + }, + { + "epoch": 34.1731843575419, + "grad_norm": 0.7233288884162903, + "learning_rate": 2.8779143389199257e-05, + "loss": 0.1037, + "step": 12234 + }, + { + "epoch": 34.175977653631286, + "grad_norm": 0.6594313383102417, + "learning_rate": 2.8778770949720673e-05, + "loss": 0.1155, + "step": 12235 + }, + { + "epoch": 34.17877094972067, + "grad_norm": 0.49725350737571716, + "learning_rate": 2.8778398510242087e-05, + "loss": 0.0917, + "step": 12236 + }, + { + "epoch": 34.18156424581006, + "grad_norm": 1.0644292831420898, + "learning_rate": 2.87780260707635e-05, + "loss": 0.1032, + "step": 12237 + }, + { + "epoch": 34.184357541899445, + "grad_norm": 1.991175651550293, + "learning_rate": 2.8777653631284916e-05, + "loss": 0.1009, + "step": 12238 + }, + { + "epoch": 34.187150837988824, + "grad_norm": 0.854800283908844, + "learning_rate": 2.8777281191806333e-05, + "loss": 0.103, + "step": 12239 + }, + { + "epoch": 34.18994413407821, + "grad_norm": 1.5807245969772339, + "learning_rate": 2.8776908752327746e-05, + "loss": 0.1055, + "step": 12240 + }, + { + "epoch": 34.1927374301676, + "grad_norm": 0.7441497445106506, + "learning_rate": 2.8776536312849163e-05, + "loss": 0.1089, + "step": 12241 + }, + { + "epoch": 34.19553072625698, + "grad_norm": 0.6154922842979431, + "learning_rate": 2.877616387337058e-05, + "loss": 0.1024, + "step": 12242 + }, + { + "epoch": 34.19832402234637, + "grad_norm": 2.3905434608459473, + "learning_rate": 2.8775791433891996e-05, + "loss": 0.1096, + "step": 12243 + }, + { + "epoch": 34.201117318435756, + "grad_norm": 3.142505168914795, + "learning_rate": 2.877541899441341e-05, + "loss": 0.0928, + "step": 12244 + }, + { + "epoch": 34.20391061452514, + "grad_norm": 0.7217163443565369, + "learning_rate": 2.8775046554934822e-05, + "loss": 0.0995, + "step": 12245 + }, + { + "epoch": 34.20670391061452, + "grad_norm": 1.5915473699569702, + "learning_rate": 2.877467411545624e-05, + "loss": 0.0938, + "step": 12246 + }, + { + "epoch": 34.20949720670391, + "grad_norm": 3.3858914375305176, + "learning_rate": 2.8774301675977655e-05, + "loss": 0.1291, + "step": 12247 + }, + { + "epoch": 34.212290502793294, + "grad_norm": 0.7657068371772766, + "learning_rate": 2.877392923649907e-05, + "loss": 0.1544, + "step": 12248 + }, + { + "epoch": 34.21508379888268, + "grad_norm": 0.6812044382095337, + "learning_rate": 2.8773556797020485e-05, + "loss": 0.1267, + "step": 12249 + }, + { + "epoch": 34.21787709497207, + "grad_norm": 2.562007188796997, + "learning_rate": 2.87731843575419e-05, + "loss": 0.1497, + "step": 12250 + }, + { + "epoch": 34.220670391061454, + "grad_norm": 0.679125189781189, + "learning_rate": 2.8772811918063314e-05, + "loss": 0.1414, + "step": 12251 + }, + { + "epoch": 34.22346368715084, + "grad_norm": 0.6269919872283936, + "learning_rate": 2.877243947858473e-05, + "loss": 0.1283, + "step": 12252 + }, + { + "epoch": 34.226256983240226, + "grad_norm": 0.9191961884498596, + "learning_rate": 2.8772067039106147e-05, + "loss": 0.1022, + "step": 12253 + }, + { + "epoch": 34.229050279329606, + "grad_norm": 0.6242311000823975, + "learning_rate": 2.877169459962756e-05, + "loss": 0.1041, + "step": 12254 + }, + { + "epoch": 34.23184357541899, + "grad_norm": 0.4041227698326111, + "learning_rate": 2.8771322160148977e-05, + "loss": 0.1283, + "step": 12255 + }, + { + "epoch": 34.23463687150838, + "grad_norm": 0.5380523204803467, + "learning_rate": 2.8770949720670394e-05, + "loss": 0.1158, + "step": 12256 + }, + { + "epoch": 34.237430167597765, + "grad_norm": 0.6204397678375244, + "learning_rate": 2.8770577281191807e-05, + "loss": 0.1308, + "step": 12257 + }, + { + "epoch": 34.24022346368715, + "grad_norm": 0.7925792932510376, + "learning_rate": 2.877020484171322e-05, + "loss": 0.1146, + "step": 12258 + }, + { + "epoch": 34.24301675977654, + "grad_norm": 0.5241414904594421, + "learning_rate": 2.8769832402234636e-05, + "loss": 0.1023, + "step": 12259 + }, + { + "epoch": 34.245810055865924, + "grad_norm": 0.7504146695137024, + "learning_rate": 2.8769459962756053e-05, + "loss": 0.1096, + "step": 12260 + }, + { + "epoch": 34.24860335195531, + "grad_norm": 0.4702037274837494, + "learning_rate": 2.876908752327747e-05, + "loss": 0.0953, + "step": 12261 + }, + { + "epoch": 34.25139664804469, + "grad_norm": 0.8735169768333435, + "learning_rate": 2.8768715083798886e-05, + "loss": 0.118, + "step": 12262 + }, + { + "epoch": 34.254189944134076, + "grad_norm": 0.6174253821372986, + "learning_rate": 2.87683426443203e-05, + "loss": 0.0858, + "step": 12263 + }, + { + "epoch": 34.25698324022346, + "grad_norm": 0.7632341980934143, + "learning_rate": 2.8767970204841712e-05, + "loss": 0.0964, + "step": 12264 + }, + { + "epoch": 34.25977653631285, + "grad_norm": 0.9266559481620789, + "learning_rate": 2.876759776536313e-05, + "loss": 0.1007, + "step": 12265 + }, + { + "epoch": 34.262569832402235, + "grad_norm": 0.6284510493278503, + "learning_rate": 2.8767225325884545e-05, + "loss": 0.1053, + "step": 12266 + }, + { + "epoch": 34.26536312849162, + "grad_norm": 5.757258415222168, + "learning_rate": 2.876685288640596e-05, + "loss": 0.085, + "step": 12267 + }, + { + "epoch": 34.26815642458101, + "grad_norm": 0.8791931867599487, + "learning_rate": 2.8766480446927375e-05, + "loss": 0.0919, + "step": 12268 + }, + { + "epoch": 34.270949720670394, + "grad_norm": 0.8109664916992188, + "learning_rate": 2.876610800744879e-05, + "loss": 0.0952, + "step": 12269 + }, + { + "epoch": 34.273743016759774, + "grad_norm": 0.6538370251655579, + "learning_rate": 2.8765735567970208e-05, + "loss": 0.0843, + "step": 12270 + }, + { + "epoch": 34.27653631284916, + "grad_norm": 0.7202762365341187, + "learning_rate": 2.876536312849162e-05, + "loss": 0.0961, + "step": 12271 + }, + { + "epoch": 34.279329608938546, + "grad_norm": 1.4893349409103394, + "learning_rate": 2.8764990689013034e-05, + "loss": 0.1094, + "step": 12272 + }, + { + "epoch": 34.28212290502793, + "grad_norm": 0.5460959672927856, + "learning_rate": 2.876461824953445e-05, + "loss": 0.15, + "step": 12273 + }, + { + "epoch": 34.28491620111732, + "grad_norm": 0.38679274916648865, + "learning_rate": 2.8764245810055868e-05, + "loss": 0.1522, + "step": 12274 + }, + { + "epoch": 34.287709497206706, + "grad_norm": 0.5955228209495544, + "learning_rate": 2.8763873370577284e-05, + "loss": 0.1302, + "step": 12275 + }, + { + "epoch": 34.29050279329609, + "grad_norm": 1.9894957542419434, + "learning_rate": 2.8763500931098697e-05, + "loss": 0.1381, + "step": 12276 + }, + { + "epoch": 34.29329608938548, + "grad_norm": 0.5826268792152405, + "learning_rate": 2.876312849162011e-05, + "loss": 0.1297, + "step": 12277 + }, + { + "epoch": 34.29608938547486, + "grad_norm": 0.4306769073009491, + "learning_rate": 2.8762756052141527e-05, + "loss": 0.1248, + "step": 12278 + }, + { + "epoch": 34.298882681564244, + "grad_norm": 0.4558275043964386, + "learning_rate": 2.8762383612662943e-05, + "loss": 0.1158, + "step": 12279 + }, + { + "epoch": 34.30167597765363, + "grad_norm": 0.47459742426872253, + "learning_rate": 2.8762011173184357e-05, + "loss": 0.1163, + "step": 12280 + }, + { + "epoch": 34.30446927374302, + "grad_norm": 0.603302001953125, + "learning_rate": 2.8761638733705773e-05, + "loss": 0.1084, + "step": 12281 + }, + { + "epoch": 34.3072625698324, + "grad_norm": 0.6173675656318665, + "learning_rate": 2.876126629422719e-05, + "loss": 0.1173, + "step": 12282 + }, + { + "epoch": 34.31005586592179, + "grad_norm": 0.44608834385871887, + "learning_rate": 2.8760893854748606e-05, + "loss": 0.0983, + "step": 12283 + }, + { + "epoch": 34.312849162011176, + "grad_norm": 0.8414434194564819, + "learning_rate": 2.876052141527002e-05, + "loss": 0.1204, + "step": 12284 + }, + { + "epoch": 34.315642458100555, + "grad_norm": 0.6112593412399292, + "learning_rate": 2.8760148975791432e-05, + "loss": 0.1075, + "step": 12285 + }, + { + "epoch": 34.31843575418994, + "grad_norm": 0.5288552045822144, + "learning_rate": 2.875977653631285e-05, + "loss": 0.1087, + "step": 12286 + }, + { + "epoch": 34.32122905027933, + "grad_norm": 0.5367385149002075, + "learning_rate": 2.8759404096834266e-05, + "loss": 0.0998, + "step": 12287 + }, + { + "epoch": 34.324022346368714, + "grad_norm": 0.5575321912765503, + "learning_rate": 2.8759031657355682e-05, + "loss": 0.1043, + "step": 12288 + }, + { + "epoch": 34.3268156424581, + "grad_norm": 0.9009414911270142, + "learning_rate": 2.8758659217877095e-05, + "loss": 0.098, + "step": 12289 + }, + { + "epoch": 34.32960893854749, + "grad_norm": 0.6777462959289551, + "learning_rate": 2.8758286778398512e-05, + "loss": 0.1003, + "step": 12290 + }, + { + "epoch": 34.332402234636874, + "grad_norm": 0.5753942728042603, + "learning_rate": 2.8757914338919925e-05, + "loss": 0.1066, + "step": 12291 + }, + { + "epoch": 34.33519553072626, + "grad_norm": 0.5781465172767639, + "learning_rate": 2.875754189944134e-05, + "loss": 0.0876, + "step": 12292 + }, + { + "epoch": 34.33798882681564, + "grad_norm": 0.775444507598877, + "learning_rate": 2.8757169459962758e-05, + "loss": 0.1026, + "step": 12293 + }, + { + "epoch": 34.340782122905026, + "grad_norm": 0.7145737409591675, + "learning_rate": 2.875679702048417e-05, + "loss": 0.0834, + "step": 12294 + }, + { + "epoch": 34.34357541899441, + "grad_norm": 0.48206111788749695, + "learning_rate": 2.8756424581005588e-05, + "loss": 0.0843, + "step": 12295 + }, + { + "epoch": 34.3463687150838, + "grad_norm": 0.8930062055587769, + "learning_rate": 2.8756052141527004e-05, + "loss": 0.0867, + "step": 12296 + }, + { + "epoch": 34.349162011173185, + "grad_norm": 2.2014198303222656, + "learning_rate": 2.8755679702048417e-05, + "loss": 0.1215, + "step": 12297 + }, + { + "epoch": 34.35195530726257, + "grad_norm": 0.5093565583229065, + "learning_rate": 2.875530726256983e-05, + "loss": 0.1397, + "step": 12298 + }, + { + "epoch": 34.35474860335196, + "grad_norm": 0.5810315608978271, + "learning_rate": 2.8754934823091247e-05, + "loss": 0.1337, + "step": 12299 + }, + { + "epoch": 34.357541899441344, + "grad_norm": 0.7663044929504395, + "learning_rate": 2.8754562383612664e-05, + "loss": 0.1254, + "step": 12300 + }, + { + "epoch": 34.36033519553072, + "grad_norm": 0.8294664025306702, + "learning_rate": 2.875418994413408e-05, + "loss": 0.12, + "step": 12301 + }, + { + "epoch": 34.36312849162011, + "grad_norm": 0.5388156175613403, + "learning_rate": 2.8753817504655497e-05, + "loss": 0.1153, + "step": 12302 + }, + { + "epoch": 34.365921787709496, + "grad_norm": 0.6688851118087769, + "learning_rate": 2.875344506517691e-05, + "loss": 0.0966, + "step": 12303 + }, + { + "epoch": 34.36871508379888, + "grad_norm": 0.4699972867965698, + "learning_rate": 2.8753072625698323e-05, + "loss": 0.1198, + "step": 12304 + }, + { + "epoch": 34.37150837988827, + "grad_norm": 0.5268261432647705, + "learning_rate": 2.875270018621974e-05, + "loss": 0.1216, + "step": 12305 + }, + { + "epoch": 34.374301675977655, + "grad_norm": 0.47153565287590027, + "learning_rate": 2.8752327746741156e-05, + "loss": 0.1087, + "step": 12306 + }, + { + "epoch": 34.37709497206704, + "grad_norm": 0.47924378514289856, + "learning_rate": 2.875195530726257e-05, + "loss": 0.1113, + "step": 12307 + }, + { + "epoch": 34.37988826815643, + "grad_norm": 1.1344318389892578, + "learning_rate": 2.8751582867783986e-05, + "loss": 0.1417, + "step": 12308 + }, + { + "epoch": 34.38268156424581, + "grad_norm": 0.6694077253341675, + "learning_rate": 2.8751210428305402e-05, + "loss": 0.1046, + "step": 12309 + }, + { + "epoch": 34.385474860335194, + "grad_norm": 0.44611963629722595, + "learning_rate": 2.875083798882682e-05, + "loss": 0.1019, + "step": 12310 + }, + { + "epoch": 34.38826815642458, + "grad_norm": 0.6837864518165588, + "learning_rate": 2.8750465549348232e-05, + "loss": 0.1263, + "step": 12311 + }, + { + "epoch": 34.391061452513966, + "grad_norm": 0.7126997113227844, + "learning_rate": 2.8750093109869645e-05, + "loss": 0.0953, + "step": 12312 + }, + { + "epoch": 34.39385474860335, + "grad_norm": 0.48583850264549255, + "learning_rate": 2.874972067039106e-05, + "loss": 0.0806, + "step": 12313 + }, + { + "epoch": 34.39664804469274, + "grad_norm": 1.6838271617889404, + "learning_rate": 2.8749348230912478e-05, + "loss": 0.1049, + "step": 12314 + }, + { + "epoch": 34.399441340782126, + "grad_norm": 1.2483173608779907, + "learning_rate": 2.8748975791433895e-05, + "loss": 0.0975, + "step": 12315 + }, + { + "epoch": 34.40223463687151, + "grad_norm": 0.7981637716293335, + "learning_rate": 2.8748603351955308e-05, + "loss": 0.0853, + "step": 12316 + }, + { + "epoch": 34.40502793296089, + "grad_norm": 0.4957773983478546, + "learning_rate": 2.874823091247672e-05, + "loss": 0.0803, + "step": 12317 + }, + { + "epoch": 34.40782122905028, + "grad_norm": 1.9487028121948242, + "learning_rate": 2.8747858472998137e-05, + "loss": 0.0888, + "step": 12318 + }, + { + "epoch": 34.410614525139664, + "grad_norm": 1.2367689609527588, + "learning_rate": 2.8747486033519554e-05, + "loss": 0.0905, + "step": 12319 + }, + { + "epoch": 34.41340782122905, + "grad_norm": 0.7017862200737, + "learning_rate": 2.874711359404097e-05, + "loss": 0.0929, + "step": 12320 + }, + { + "epoch": 34.41620111731844, + "grad_norm": 3.7991549968719482, + "learning_rate": 2.8746741154562384e-05, + "loss": 0.076, + "step": 12321 + }, + { + "epoch": 34.41899441340782, + "grad_norm": 1.0378814935684204, + "learning_rate": 2.87463687150838e-05, + "loss": 0.1278, + "step": 12322 + }, + { + "epoch": 34.42178770949721, + "grad_norm": 0.5218268632888794, + "learning_rate": 2.8745996275605217e-05, + "loss": 0.1553, + "step": 12323 + }, + { + "epoch": 34.42458100558659, + "grad_norm": 0.4978460371494293, + "learning_rate": 2.874562383612663e-05, + "loss": 0.1362, + "step": 12324 + }, + { + "epoch": 34.427374301675975, + "grad_norm": 0.6050964593887329, + "learning_rate": 2.8745251396648043e-05, + "loss": 0.1247, + "step": 12325 + }, + { + "epoch": 34.43016759776536, + "grad_norm": 0.47307318449020386, + "learning_rate": 2.874487895716946e-05, + "loss": 0.1244, + "step": 12326 + }, + { + "epoch": 34.43296089385475, + "grad_norm": 0.4680768549442291, + "learning_rate": 2.8744506517690876e-05, + "loss": 0.1394, + "step": 12327 + }, + { + "epoch": 34.435754189944134, + "grad_norm": 0.5703617930412292, + "learning_rate": 2.8744134078212293e-05, + "loss": 0.1307, + "step": 12328 + }, + { + "epoch": 34.43854748603352, + "grad_norm": 0.48793813586235046, + "learning_rate": 2.874376163873371e-05, + "loss": 0.1292, + "step": 12329 + }, + { + "epoch": 34.44134078212291, + "grad_norm": 0.9896544218063354, + "learning_rate": 2.8743389199255122e-05, + "loss": 0.1257, + "step": 12330 + }, + { + "epoch": 34.444134078212294, + "grad_norm": 0.620688796043396, + "learning_rate": 2.8743016759776535e-05, + "loss": 0.0907, + "step": 12331 + }, + { + "epoch": 34.44692737430167, + "grad_norm": 0.5756577849388123, + "learning_rate": 2.8742644320297952e-05, + "loss": 0.1038, + "step": 12332 + }, + { + "epoch": 34.44972067039106, + "grad_norm": 0.5356044769287109, + "learning_rate": 2.874227188081937e-05, + "loss": 0.1231, + "step": 12333 + }, + { + "epoch": 34.452513966480446, + "grad_norm": 0.5273341536521912, + "learning_rate": 2.8741899441340782e-05, + "loss": 0.1153, + "step": 12334 + }, + { + "epoch": 34.45530726256983, + "grad_norm": 1.1868181228637695, + "learning_rate": 2.8741527001862198e-05, + "loss": 0.1073, + "step": 12335 + }, + { + "epoch": 34.45810055865922, + "grad_norm": 0.6669272184371948, + "learning_rate": 2.8741154562383615e-05, + "loss": 0.1126, + "step": 12336 + }, + { + "epoch": 34.460893854748605, + "grad_norm": 0.634226381778717, + "learning_rate": 2.8740782122905028e-05, + "loss": 0.091, + "step": 12337 + }, + { + "epoch": 34.46368715083799, + "grad_norm": 0.379172682762146, + "learning_rate": 2.8740409683426444e-05, + "loss": 0.0875, + "step": 12338 + }, + { + "epoch": 34.46648044692738, + "grad_norm": 0.6124927997589111, + "learning_rate": 2.8740037243947858e-05, + "loss": 0.1026, + "step": 12339 + }, + { + "epoch": 34.46927374301676, + "grad_norm": 0.802702009677887, + "learning_rate": 2.8739664804469274e-05, + "loss": 0.1054, + "step": 12340 + }, + { + "epoch": 34.47206703910614, + "grad_norm": 0.6446529030799866, + "learning_rate": 2.873929236499069e-05, + "loss": 0.0909, + "step": 12341 + }, + { + "epoch": 34.47486033519553, + "grad_norm": 0.8751084208488464, + "learning_rate": 2.8738919925512107e-05, + "loss": 0.0862, + "step": 12342 + }, + { + "epoch": 34.477653631284916, + "grad_norm": 1.0554019212722778, + "learning_rate": 2.873854748603352e-05, + "loss": 0.1112, + "step": 12343 + }, + { + "epoch": 34.4804469273743, + "grad_norm": 0.678268551826477, + "learning_rate": 2.8738175046554934e-05, + "loss": 0.0831, + "step": 12344 + }, + { + "epoch": 34.48324022346369, + "grad_norm": 0.6145634651184082, + "learning_rate": 2.873780260707635e-05, + "loss": 0.0924, + "step": 12345 + }, + { + "epoch": 34.486033519553075, + "grad_norm": 0.9800083637237549, + "learning_rate": 2.8737430167597767e-05, + "loss": 0.0876, + "step": 12346 + }, + { + "epoch": 34.48882681564246, + "grad_norm": 1.4549272060394287, + "learning_rate": 2.8737057728119183e-05, + "loss": 0.1242, + "step": 12347 + }, + { + "epoch": 34.49162011173184, + "grad_norm": 0.3711203336715698, + "learning_rate": 2.8736685288640596e-05, + "loss": 0.1347, + "step": 12348 + }, + { + "epoch": 34.49441340782123, + "grad_norm": 1.6309760808944702, + "learning_rate": 2.8736312849162013e-05, + "loss": 0.1348, + "step": 12349 + }, + { + "epoch": 34.497206703910614, + "grad_norm": 0.5617785453796387, + "learning_rate": 2.8735940409683426e-05, + "loss": 0.1079, + "step": 12350 + }, + { + "epoch": 34.5, + "grad_norm": 0.4274821877479553, + "learning_rate": 2.8735567970204842e-05, + "loss": 0.1334, + "step": 12351 + }, + { + "epoch": 34.502793296089386, + "grad_norm": 5.287921905517578, + "learning_rate": 2.8735195530726256e-05, + "loss": 0.1338, + "step": 12352 + }, + { + "epoch": 34.50558659217877, + "grad_norm": 0.5273119211196899, + "learning_rate": 2.8734823091247672e-05, + "loss": 0.1245, + "step": 12353 + }, + { + "epoch": 34.50837988826816, + "grad_norm": 1.058430790901184, + "learning_rate": 2.873445065176909e-05, + "loss": 0.1221, + "step": 12354 + }, + { + "epoch": 34.51117318435754, + "grad_norm": 0.7142751216888428, + "learning_rate": 2.8734078212290505e-05, + "loss": 0.1088, + "step": 12355 + }, + { + "epoch": 34.513966480446925, + "grad_norm": 1.0941978693008423, + "learning_rate": 2.8733705772811922e-05, + "loss": 0.1065, + "step": 12356 + }, + { + "epoch": 34.51675977653631, + "grad_norm": 0.5854668617248535, + "learning_rate": 2.873333333333333e-05, + "loss": 0.1278, + "step": 12357 + }, + { + "epoch": 34.5195530726257, + "grad_norm": 0.6241967678070068, + "learning_rate": 2.8732960893854748e-05, + "loss": 0.1139, + "step": 12358 + }, + { + "epoch": 34.522346368715084, + "grad_norm": 0.5524423122406006, + "learning_rate": 2.8732588454376165e-05, + "loss": 0.1117, + "step": 12359 + }, + { + "epoch": 34.52513966480447, + "grad_norm": 0.6931923627853394, + "learning_rate": 2.873221601489758e-05, + "loss": 0.1201, + "step": 12360 + }, + { + "epoch": 34.52793296089386, + "grad_norm": 1.6398842334747314, + "learning_rate": 2.8731843575418994e-05, + "loss": 0.1069, + "step": 12361 + }, + { + "epoch": 34.53072625698324, + "grad_norm": 0.5154238343238831, + "learning_rate": 2.873147113594041e-05, + "loss": 0.087, + "step": 12362 + }, + { + "epoch": 34.53351955307262, + "grad_norm": 1.4101554155349731, + "learning_rate": 2.8731098696461827e-05, + "loss": 0.0948, + "step": 12363 + }, + { + "epoch": 34.53631284916201, + "grad_norm": 0.5297505855560303, + "learning_rate": 2.873072625698324e-05, + "loss": 0.1035, + "step": 12364 + }, + { + "epoch": 34.539106145251395, + "grad_norm": 1.1963757276535034, + "learning_rate": 2.8730353817504657e-05, + "loss": 0.0961, + "step": 12365 + }, + { + "epoch": 34.54189944134078, + "grad_norm": 0.6012243628501892, + "learning_rate": 2.872998137802607e-05, + "loss": 0.1026, + "step": 12366 + }, + { + "epoch": 34.54469273743017, + "grad_norm": 1.0139576196670532, + "learning_rate": 2.8729608938547487e-05, + "loss": 0.1111, + "step": 12367 + }, + { + "epoch": 34.547486033519554, + "grad_norm": 0.5612104535102844, + "learning_rate": 2.8729236499068903e-05, + "loss": 0.0831, + "step": 12368 + }, + { + "epoch": 34.55027932960894, + "grad_norm": 1.7200548648834229, + "learning_rate": 2.872886405959032e-05, + "loss": 0.0947, + "step": 12369 + }, + { + "epoch": 34.55307262569833, + "grad_norm": 0.727523922920227, + "learning_rate": 2.872849162011173e-05, + "loss": 0.108, + "step": 12370 + }, + { + "epoch": 34.555865921787706, + "grad_norm": 0.8822000026702881, + "learning_rate": 2.8728119180633146e-05, + "loss": 0.1208, + "step": 12371 + }, + { + "epoch": 34.55865921787709, + "grad_norm": 1.0216048955917358, + "learning_rate": 2.8727746741154563e-05, + "loss": 0.1289, + "step": 12372 + }, + { + "epoch": 34.56145251396648, + "grad_norm": 0.3905128836631775, + "learning_rate": 2.872737430167598e-05, + "loss": 0.1288, + "step": 12373 + }, + { + "epoch": 34.564245810055866, + "grad_norm": 0.4059048295021057, + "learning_rate": 2.8727001862197392e-05, + "loss": 0.1298, + "step": 12374 + }, + { + "epoch": 34.56703910614525, + "grad_norm": 0.35746389627456665, + "learning_rate": 2.872662942271881e-05, + "loss": 0.1257, + "step": 12375 + }, + { + "epoch": 34.56983240223464, + "grad_norm": 0.49064043164253235, + "learning_rate": 2.8726256983240225e-05, + "loss": 0.1195, + "step": 12376 + }, + { + "epoch": 34.572625698324025, + "grad_norm": 1.5336828231811523, + "learning_rate": 2.872588454376164e-05, + "loss": 0.137, + "step": 12377 + }, + { + "epoch": 34.57541899441341, + "grad_norm": 0.5549370646476746, + "learning_rate": 2.8725512104283055e-05, + "loss": 0.1119, + "step": 12378 + }, + { + "epoch": 34.57821229050279, + "grad_norm": 0.9855967164039612, + "learning_rate": 2.8725139664804468e-05, + "loss": 0.1223, + "step": 12379 + }, + { + "epoch": 34.58100558659218, + "grad_norm": 0.5538395643234253, + "learning_rate": 2.8724767225325885e-05, + "loss": 0.1304, + "step": 12380 + }, + { + "epoch": 34.58379888268156, + "grad_norm": 0.8644446730613708, + "learning_rate": 2.87243947858473e-05, + "loss": 0.1072, + "step": 12381 + }, + { + "epoch": 34.58659217877095, + "grad_norm": 0.5990815758705139, + "learning_rate": 2.8724022346368718e-05, + "loss": 0.1158, + "step": 12382 + }, + { + "epoch": 34.589385474860336, + "grad_norm": 0.40621986985206604, + "learning_rate": 2.872364990689013e-05, + "loss": 0.1099, + "step": 12383 + }, + { + "epoch": 34.59217877094972, + "grad_norm": 0.3072119951248169, + "learning_rate": 2.8723277467411544e-05, + "loss": 0.0877, + "step": 12384 + }, + { + "epoch": 34.59497206703911, + "grad_norm": 0.8639836311340332, + "learning_rate": 2.872290502793296e-05, + "loss": 0.1034, + "step": 12385 + }, + { + "epoch": 34.59776536312849, + "grad_norm": 0.5294275283813477, + "learning_rate": 2.8722532588454377e-05, + "loss": 0.1117, + "step": 12386 + }, + { + "epoch": 34.600558659217874, + "grad_norm": 0.7586761116981506, + "learning_rate": 2.8722160148975794e-05, + "loss": 0.1042, + "step": 12387 + }, + { + "epoch": 34.60335195530726, + "grad_norm": 0.7790870070457458, + "learning_rate": 2.8721787709497207e-05, + "loss": 0.1058, + "step": 12388 + }, + { + "epoch": 34.60614525139665, + "grad_norm": 0.5332234501838684, + "learning_rate": 2.8721415270018623e-05, + "loss": 0.0899, + "step": 12389 + }, + { + "epoch": 34.608938547486034, + "grad_norm": 0.9129524230957031, + "learning_rate": 2.8721042830540037e-05, + "loss": 0.0881, + "step": 12390 + }, + { + "epoch": 34.61173184357542, + "grad_norm": 0.561486005783081, + "learning_rate": 2.8720670391061453e-05, + "loss": 0.0927, + "step": 12391 + }, + { + "epoch": 34.614525139664806, + "grad_norm": 1.0032401084899902, + "learning_rate": 2.8720297951582866e-05, + "loss": 0.0799, + "step": 12392 + }, + { + "epoch": 34.61731843575419, + "grad_norm": 0.5267693400382996, + "learning_rate": 2.8719925512104283e-05, + "loss": 0.0854, + "step": 12393 + }, + { + "epoch": 34.62011173184357, + "grad_norm": 0.560483992099762, + "learning_rate": 2.87195530726257e-05, + "loss": 0.0773, + "step": 12394 + }, + { + "epoch": 34.62290502793296, + "grad_norm": 1.2406946420669556, + "learning_rate": 2.8719180633147116e-05, + "loss": 0.0766, + "step": 12395 + }, + { + "epoch": 34.625698324022345, + "grad_norm": 0.6614974141120911, + "learning_rate": 2.8718808193668532e-05, + "loss": 0.0917, + "step": 12396 + }, + { + "epoch": 34.62849162011173, + "grad_norm": 4.341324806213379, + "learning_rate": 2.8718435754189942e-05, + "loss": 0.1091, + "step": 12397 + }, + { + "epoch": 34.63128491620112, + "grad_norm": 0.7191504240036011, + "learning_rate": 2.871806331471136e-05, + "loss": 0.1258, + "step": 12398 + }, + { + "epoch": 34.634078212290504, + "grad_norm": 0.5114927887916565, + "learning_rate": 2.8717690875232775e-05, + "loss": 0.1184, + "step": 12399 + }, + { + "epoch": 34.63687150837989, + "grad_norm": 0.4808073937892914, + "learning_rate": 2.8717318435754192e-05, + "loss": 0.1352, + "step": 12400 + }, + { + "epoch": 34.63966480446928, + "grad_norm": 0.7167580127716064, + "learning_rate": 2.8716945996275605e-05, + "loss": 0.1271, + "step": 12401 + }, + { + "epoch": 34.642458100558656, + "grad_norm": 0.6867895126342773, + "learning_rate": 2.871657355679702e-05, + "loss": 0.1286, + "step": 12402 + }, + { + "epoch": 34.64525139664804, + "grad_norm": 0.524726390838623, + "learning_rate": 2.8716201117318438e-05, + "loss": 0.1272, + "step": 12403 + }, + { + "epoch": 34.64804469273743, + "grad_norm": 0.6452059149742126, + "learning_rate": 2.871582867783985e-05, + "loss": 0.1203, + "step": 12404 + }, + { + "epoch": 34.650837988826815, + "grad_norm": 1.4572561979293823, + "learning_rate": 2.8715456238361268e-05, + "loss": 0.12, + "step": 12405 + }, + { + "epoch": 34.6536312849162, + "grad_norm": 0.5361381769180298, + "learning_rate": 2.871508379888268e-05, + "loss": 0.1113, + "step": 12406 + }, + { + "epoch": 34.65642458100559, + "grad_norm": 0.7374684810638428, + "learning_rate": 2.8714711359404097e-05, + "loss": 0.1093, + "step": 12407 + }, + { + "epoch": 34.659217877094974, + "grad_norm": 0.4823113679885864, + "learning_rate": 2.8714338919925514e-05, + "loss": 0.1278, + "step": 12408 + }, + { + "epoch": 34.66201117318436, + "grad_norm": 0.7109275460243225, + "learning_rate": 2.871396648044693e-05, + "loss": 0.1037, + "step": 12409 + }, + { + "epoch": 34.66480446927374, + "grad_norm": 0.533820390701294, + "learning_rate": 2.871359404096834e-05, + "loss": 0.1026, + "step": 12410 + }, + { + "epoch": 34.667597765363126, + "grad_norm": 1.0012500286102295, + "learning_rate": 2.8713221601489757e-05, + "loss": 0.1057, + "step": 12411 + }, + { + "epoch": 34.67039106145251, + "grad_norm": 1.044945240020752, + "learning_rate": 2.8712849162011173e-05, + "loss": 0.1249, + "step": 12412 + }, + { + "epoch": 34.6731843575419, + "grad_norm": 6.968775272369385, + "learning_rate": 2.871247672253259e-05, + "loss": 0.1063, + "step": 12413 + }, + { + "epoch": 34.675977653631286, + "grad_norm": 0.6296975612640381, + "learning_rate": 2.8712104283054006e-05, + "loss": 0.1031, + "step": 12414 + }, + { + "epoch": 34.67877094972067, + "grad_norm": 0.7429031729698181, + "learning_rate": 2.871173184357542e-05, + "loss": 0.1084, + "step": 12415 + }, + { + "epoch": 34.68156424581006, + "grad_norm": 0.7510053515434265, + "learning_rate": 2.8711359404096836e-05, + "loss": 0.1203, + "step": 12416 + }, + { + "epoch": 34.684357541899445, + "grad_norm": 0.7362044453620911, + "learning_rate": 2.871098696461825e-05, + "loss": 0.1122, + "step": 12417 + }, + { + "epoch": 34.687150837988824, + "grad_norm": 0.5768510699272156, + "learning_rate": 2.8710614525139666e-05, + "loss": 0.0893, + "step": 12418 + }, + { + "epoch": 34.68994413407821, + "grad_norm": 2.2443835735321045, + "learning_rate": 2.871024208566108e-05, + "loss": 0.1028, + "step": 12419 + }, + { + "epoch": 34.6927374301676, + "grad_norm": 0.7875776886940002, + "learning_rate": 2.8709869646182495e-05, + "loss": 0.0899, + "step": 12420 + }, + { + "epoch": 34.69553072625698, + "grad_norm": 2.544219493865967, + "learning_rate": 2.8709497206703912e-05, + "loss": 0.1445, + "step": 12421 + }, + { + "epoch": 34.69832402234637, + "grad_norm": 0.9281243681907654, + "learning_rate": 2.870912476722533e-05, + "loss": 0.1358, + "step": 12422 + }, + { + "epoch": 34.701117318435756, + "grad_norm": 0.5264190435409546, + "learning_rate": 2.8708752327746745e-05, + "loss": 0.1321, + "step": 12423 + }, + { + "epoch": 34.70391061452514, + "grad_norm": 0.5811026692390442, + "learning_rate": 2.8708379888268155e-05, + "loss": 0.1343, + "step": 12424 + }, + { + "epoch": 34.70670391061452, + "grad_norm": 0.45538726449012756, + "learning_rate": 2.870800744878957e-05, + "loss": 0.1378, + "step": 12425 + }, + { + "epoch": 34.70949720670391, + "grad_norm": 0.4086102843284607, + "learning_rate": 2.8707635009310988e-05, + "loss": 0.1223, + "step": 12426 + }, + { + "epoch": 34.712290502793294, + "grad_norm": 0.5298711061477661, + "learning_rate": 2.8707262569832404e-05, + "loss": 0.1265, + "step": 12427 + }, + { + "epoch": 34.71508379888268, + "grad_norm": 0.5780687928199768, + "learning_rate": 2.8706890130353817e-05, + "loss": 0.131, + "step": 12428 + }, + { + "epoch": 34.71787709497207, + "grad_norm": 1.481704592704773, + "learning_rate": 2.8706517690875234e-05, + "loss": 0.1326, + "step": 12429 + }, + { + "epoch": 34.720670391061454, + "grad_norm": 0.816493809223175, + "learning_rate": 2.8706145251396647e-05, + "loss": 0.1264, + "step": 12430 + }, + { + "epoch": 34.72346368715084, + "grad_norm": 0.555926501750946, + "learning_rate": 2.8705772811918064e-05, + "loss": 0.1164, + "step": 12431 + }, + { + "epoch": 34.726256983240226, + "grad_norm": 0.6214694380760193, + "learning_rate": 2.870540037243948e-05, + "loss": 0.1492, + "step": 12432 + }, + { + "epoch": 34.729050279329606, + "grad_norm": 0.5696894526481628, + "learning_rate": 2.8705027932960893e-05, + "loss": 0.1389, + "step": 12433 + }, + { + "epoch": 34.73184357541899, + "grad_norm": 0.5665318965911865, + "learning_rate": 2.870465549348231e-05, + "loss": 0.1148, + "step": 12434 + }, + { + "epoch": 34.73463687150838, + "grad_norm": 0.8854054808616638, + "learning_rate": 2.8704283054003726e-05, + "loss": 0.1066, + "step": 12435 + }, + { + "epoch": 34.737430167597765, + "grad_norm": 0.6257461309432983, + "learning_rate": 2.8703910614525143e-05, + "loss": 0.1283, + "step": 12436 + }, + { + "epoch": 34.74022346368715, + "grad_norm": 0.5824881792068481, + "learning_rate": 2.8703538175046553e-05, + "loss": 0.0927, + "step": 12437 + }, + { + "epoch": 34.74301675977654, + "grad_norm": 1.0441633462905884, + "learning_rate": 2.870316573556797e-05, + "loss": 0.0998, + "step": 12438 + }, + { + "epoch": 34.745810055865924, + "grad_norm": 0.5299841165542603, + "learning_rate": 2.8702793296089386e-05, + "loss": 0.1016, + "step": 12439 + }, + { + "epoch": 34.74860335195531, + "grad_norm": 1.1361135244369507, + "learning_rate": 2.8702420856610802e-05, + "loss": 0.1022, + "step": 12440 + }, + { + "epoch": 34.75139664804469, + "grad_norm": 0.5761517882347107, + "learning_rate": 2.870204841713222e-05, + "loss": 0.0967, + "step": 12441 + }, + { + "epoch": 34.754189944134076, + "grad_norm": 0.6202878355979919, + "learning_rate": 2.8701675977653632e-05, + "loss": 0.0973, + "step": 12442 + }, + { + "epoch": 34.75698324022346, + "grad_norm": 0.6161074042320251, + "learning_rate": 2.870130353817505e-05, + "loss": 0.1124, + "step": 12443 + }, + { + "epoch": 34.75977653631285, + "grad_norm": 0.7700483202934265, + "learning_rate": 2.8700931098696462e-05, + "loss": 0.0951, + "step": 12444 + }, + { + "epoch": 34.762569832402235, + "grad_norm": 0.8365578651428223, + "learning_rate": 2.8700558659217878e-05, + "loss": 0.0982, + "step": 12445 + }, + { + "epoch": 34.76536312849162, + "grad_norm": 1.1674836874008179, + "learning_rate": 2.870018621973929e-05, + "loss": 0.115, + "step": 12446 + }, + { + "epoch": 34.76815642458101, + "grad_norm": 0.801524817943573, + "learning_rate": 2.8699813780260708e-05, + "loss": 0.1095, + "step": 12447 + }, + { + "epoch": 34.770949720670394, + "grad_norm": 0.6177513003349304, + "learning_rate": 2.8699441340782124e-05, + "loss": 0.1573, + "step": 12448 + }, + { + "epoch": 34.773743016759774, + "grad_norm": 0.3991692364215851, + "learning_rate": 2.869906890130354e-05, + "loss": 0.1492, + "step": 12449 + }, + { + "epoch": 34.77653631284916, + "grad_norm": 0.4978353977203369, + "learning_rate": 2.8698696461824954e-05, + "loss": 0.1049, + "step": 12450 + }, + { + "epoch": 34.779329608938546, + "grad_norm": 0.410351425409317, + "learning_rate": 2.8698324022346367e-05, + "loss": 0.1151, + "step": 12451 + }, + { + "epoch": 34.78212290502793, + "grad_norm": 0.6458420753479004, + "learning_rate": 2.8697951582867784e-05, + "loss": 0.112, + "step": 12452 + }, + { + "epoch": 34.78491620111732, + "grad_norm": 0.5826671719551086, + "learning_rate": 2.86975791433892e-05, + "loss": 0.1133, + "step": 12453 + }, + { + "epoch": 34.787709497206706, + "grad_norm": 0.6655648946762085, + "learning_rate": 2.8697206703910617e-05, + "loss": 0.1161, + "step": 12454 + }, + { + "epoch": 34.79050279329609, + "grad_norm": 1.1789259910583496, + "learning_rate": 2.869683426443203e-05, + "loss": 0.1002, + "step": 12455 + }, + { + "epoch": 34.79329608938548, + "grad_norm": 1.095789909362793, + "learning_rate": 2.8696461824953447e-05, + "loss": 0.1105, + "step": 12456 + }, + { + "epoch": 34.79608938547486, + "grad_norm": 0.701210081577301, + "learning_rate": 2.869608938547486e-05, + "loss": 0.126, + "step": 12457 + }, + { + "epoch": 34.798882681564244, + "grad_norm": 0.5247523784637451, + "learning_rate": 2.8695716945996276e-05, + "loss": 0.1239, + "step": 12458 + }, + { + "epoch": 34.80167597765363, + "grad_norm": 1.5028588771820068, + "learning_rate": 2.8695344506517693e-05, + "loss": 0.1181, + "step": 12459 + }, + { + "epoch": 34.80446927374302, + "grad_norm": 0.4735599160194397, + "learning_rate": 2.8694972067039106e-05, + "loss": 0.0982, + "step": 12460 + }, + { + "epoch": 34.8072625698324, + "grad_norm": 0.5018202066421509, + "learning_rate": 2.8694599627560522e-05, + "loss": 0.0999, + "step": 12461 + }, + { + "epoch": 34.81005586592179, + "grad_norm": 1.9164659976959229, + "learning_rate": 2.869422718808194e-05, + "loss": 0.1083, + "step": 12462 + }, + { + "epoch": 34.812849162011176, + "grad_norm": 0.64449143409729, + "learning_rate": 2.8693854748603356e-05, + "loss": 0.1042, + "step": 12463 + }, + { + "epoch": 34.815642458100555, + "grad_norm": 0.8072342872619629, + "learning_rate": 2.8693482309124765e-05, + "loss": 0.0973, + "step": 12464 + }, + { + "epoch": 34.81843575418994, + "grad_norm": 1.0580687522888184, + "learning_rate": 2.8693109869646182e-05, + "loss": 0.0935, + "step": 12465 + }, + { + "epoch": 34.82122905027933, + "grad_norm": 0.8958181738853455, + "learning_rate": 2.86927374301676e-05, + "loss": 0.0925, + "step": 12466 + }, + { + "epoch": 34.824022346368714, + "grad_norm": 0.6203571557998657, + "learning_rate": 2.8692364990689015e-05, + "loss": 0.0859, + "step": 12467 + }, + { + "epoch": 34.8268156424581, + "grad_norm": 0.5666225552558899, + "learning_rate": 2.8691992551210428e-05, + "loss": 0.1126, + "step": 12468 + }, + { + "epoch": 34.82960893854749, + "grad_norm": 0.770258367061615, + "learning_rate": 2.8691620111731845e-05, + "loss": 0.1055, + "step": 12469 + }, + { + "epoch": 34.832402234636874, + "grad_norm": 0.8863216638565063, + "learning_rate": 2.8691247672253258e-05, + "loss": 0.0927, + "step": 12470 + }, + { + "epoch": 34.83519553072626, + "grad_norm": 0.8668456077575684, + "learning_rate": 2.8690875232774674e-05, + "loss": 0.1258, + "step": 12471 + }, + { + "epoch": 34.83798882681564, + "grad_norm": 2.1504433155059814, + "learning_rate": 2.869050279329609e-05, + "loss": 0.1101, + "step": 12472 + }, + { + "epoch": 34.840782122905026, + "grad_norm": 0.41167721152305603, + "learning_rate": 2.8690130353817504e-05, + "loss": 0.1563, + "step": 12473 + }, + { + "epoch": 34.84357541899441, + "grad_norm": 0.6343806982040405, + "learning_rate": 2.868975791433892e-05, + "loss": 0.1217, + "step": 12474 + }, + { + "epoch": 34.8463687150838, + "grad_norm": 0.6648430824279785, + "learning_rate": 2.8689385474860337e-05, + "loss": 0.1342, + "step": 12475 + }, + { + "epoch": 34.849162011173185, + "grad_norm": 0.4553503692150116, + "learning_rate": 2.8689013035381754e-05, + "loss": 0.1285, + "step": 12476 + }, + { + "epoch": 34.85195530726257, + "grad_norm": 0.6306644082069397, + "learning_rate": 2.8688640595903163e-05, + "loss": 0.1421, + "step": 12477 + }, + { + "epoch": 34.85474860335196, + "grad_norm": 0.5963183045387268, + "learning_rate": 2.868826815642458e-05, + "loss": 0.1228, + "step": 12478 + }, + { + "epoch": 34.857541899441344, + "grad_norm": 0.769469678401947, + "learning_rate": 2.8687895716945996e-05, + "loss": 0.1228, + "step": 12479 + }, + { + "epoch": 34.86033519553072, + "grad_norm": 0.5665842294692993, + "learning_rate": 2.8687523277467413e-05, + "loss": 0.1153, + "step": 12480 + }, + { + "epoch": 34.86312849162011, + "grad_norm": 0.6557990312576294, + "learning_rate": 2.868715083798883e-05, + "loss": 0.1316, + "step": 12481 + }, + { + "epoch": 34.865921787709496, + "grad_norm": 1.0329121351242065, + "learning_rate": 2.8686778398510243e-05, + "loss": 0.1155, + "step": 12482 + }, + { + "epoch": 34.86871508379888, + "grad_norm": 1.0834009647369385, + "learning_rate": 2.868640595903166e-05, + "loss": 0.1295, + "step": 12483 + }, + { + "epoch": 34.87150837988827, + "grad_norm": 0.7937959432601929, + "learning_rate": 2.8686033519553072e-05, + "loss": 0.1196, + "step": 12484 + }, + { + "epoch": 34.874301675977655, + "grad_norm": 0.5832961201667786, + "learning_rate": 2.868566108007449e-05, + "loss": 0.1051, + "step": 12485 + }, + { + "epoch": 34.87709497206704, + "grad_norm": 0.40986838936805725, + "learning_rate": 2.8685288640595902e-05, + "loss": 0.0991, + "step": 12486 + }, + { + "epoch": 34.87988826815643, + "grad_norm": 0.4193927049636841, + "learning_rate": 2.868491620111732e-05, + "loss": 0.0998, + "step": 12487 + }, + { + "epoch": 34.88268156424581, + "grad_norm": 0.7037988305091858, + "learning_rate": 2.8684543761638735e-05, + "loss": 0.1292, + "step": 12488 + }, + { + "epoch": 34.885474860335194, + "grad_norm": 0.45512303709983826, + "learning_rate": 2.868417132216015e-05, + "loss": 0.1048, + "step": 12489 + }, + { + "epoch": 34.88826815642458, + "grad_norm": 2.2604198455810547, + "learning_rate": 2.8683798882681565e-05, + "loss": 0.0785, + "step": 12490 + }, + { + "epoch": 34.891061452513966, + "grad_norm": 0.6250013709068298, + "learning_rate": 2.8683426443202978e-05, + "loss": 0.1265, + "step": 12491 + }, + { + "epoch": 34.89385474860335, + "grad_norm": 1.1534614562988281, + "learning_rate": 2.8683054003724394e-05, + "loss": 0.1045, + "step": 12492 + }, + { + "epoch": 34.89664804469274, + "grad_norm": 0.7276747226715088, + "learning_rate": 2.868268156424581e-05, + "loss": 0.0994, + "step": 12493 + }, + { + "epoch": 34.899441340782126, + "grad_norm": 0.7123043537139893, + "learning_rate": 2.8682309124767227e-05, + "loss": 0.1074, + "step": 12494 + }, + { + "epoch": 34.90223463687151, + "grad_norm": 0.913272500038147, + "learning_rate": 2.868193668528864e-05, + "loss": 0.0901, + "step": 12495 + }, + { + "epoch": 34.90502793296089, + "grad_norm": 0.7023056149482727, + "learning_rate": 2.8681564245810057e-05, + "loss": 0.1087, + "step": 12496 + }, + { + "epoch": 34.90782122905028, + "grad_norm": 1.0054320096969604, + "learning_rate": 2.868119180633147e-05, + "loss": 0.1389, + "step": 12497 + }, + { + "epoch": 34.910614525139664, + "grad_norm": 0.40588346123695374, + "learning_rate": 2.8680819366852887e-05, + "loss": 0.1344, + "step": 12498 + }, + { + "epoch": 34.91340782122905, + "grad_norm": 0.4460827708244324, + "learning_rate": 2.8680446927374303e-05, + "loss": 0.1207, + "step": 12499 + }, + { + "epoch": 34.91620111731844, + "grad_norm": 0.5608615875244141, + "learning_rate": 2.8680074487895717e-05, + "loss": 0.1219, + "step": 12500 + }, + { + "epoch": 34.91899441340782, + "grad_norm": 1.165662169456482, + "learning_rate": 2.8679702048417133e-05, + "loss": 0.1387, + "step": 12501 + }, + { + "epoch": 34.92178770949721, + "grad_norm": 0.5131365656852722, + "learning_rate": 2.867932960893855e-05, + "loss": 0.1267, + "step": 12502 + }, + { + "epoch": 34.92458100558659, + "grad_norm": 0.7842077612876892, + "learning_rate": 2.8678957169459966e-05, + "loss": 0.1436, + "step": 12503 + }, + { + "epoch": 34.927374301675975, + "grad_norm": 0.6228480339050293, + "learning_rate": 2.8678584729981376e-05, + "loss": 0.1083, + "step": 12504 + }, + { + "epoch": 34.93016759776536, + "grad_norm": 0.9665493965148926, + "learning_rate": 2.8678212290502792e-05, + "loss": 0.1164, + "step": 12505 + }, + { + "epoch": 34.93296089385475, + "grad_norm": 0.6308093070983887, + "learning_rate": 2.867783985102421e-05, + "loss": 0.1282, + "step": 12506 + }, + { + "epoch": 34.935754189944134, + "grad_norm": 0.4621846079826355, + "learning_rate": 2.8677467411545626e-05, + "loss": 0.0999, + "step": 12507 + }, + { + "epoch": 34.93854748603352, + "grad_norm": 0.9819085597991943, + "learning_rate": 2.8677094972067042e-05, + "loss": 0.1064, + "step": 12508 + }, + { + "epoch": 34.94134078212291, + "grad_norm": 0.6597883105278015, + "learning_rate": 2.8676722532588455e-05, + "loss": 0.0988, + "step": 12509 + }, + { + "epoch": 34.944134078212294, + "grad_norm": 0.5075153112411499, + "learning_rate": 2.867635009310987e-05, + "loss": 0.1214, + "step": 12510 + }, + { + "epoch": 34.94692737430167, + "grad_norm": 0.496648907661438, + "learning_rate": 2.8675977653631285e-05, + "loss": 0.0959, + "step": 12511 + }, + { + "epoch": 34.94972067039106, + "grad_norm": 0.7537051439285278, + "learning_rate": 2.86756052141527e-05, + "loss": 0.1116, + "step": 12512 + }, + { + "epoch": 34.952513966480446, + "grad_norm": 0.653933048248291, + "learning_rate": 2.8675232774674115e-05, + "loss": 0.1129, + "step": 12513 + }, + { + "epoch": 34.95530726256983, + "grad_norm": 0.925032913684845, + "learning_rate": 2.867486033519553e-05, + "loss": 0.1159, + "step": 12514 + }, + { + "epoch": 34.95810055865922, + "grad_norm": 0.6277142763137817, + "learning_rate": 2.8674487895716948e-05, + "loss": 0.0874, + "step": 12515 + }, + { + "epoch": 34.960893854748605, + "grad_norm": 0.6835078001022339, + "learning_rate": 2.8674115456238364e-05, + "loss": 0.1026, + "step": 12516 + }, + { + "epoch": 34.96368715083799, + "grad_norm": 0.4379429817199707, + "learning_rate": 2.8673743016759777e-05, + "loss": 0.0964, + "step": 12517 + }, + { + "epoch": 34.96648044692738, + "grad_norm": 0.9991907477378845, + "learning_rate": 2.867337057728119e-05, + "loss": 0.0875, + "step": 12518 + }, + { + "epoch": 34.96927374301676, + "grad_norm": 0.565986692905426, + "learning_rate": 2.8672998137802607e-05, + "loss": 0.0955, + "step": 12519 + }, + { + "epoch": 34.97206703910614, + "grad_norm": 0.6547340154647827, + "learning_rate": 2.8672625698324024e-05, + "loss": 0.0947, + "step": 12520 + }, + { + "epoch": 34.97486033519553, + "grad_norm": 1.15326988697052, + "learning_rate": 2.867225325884544e-05, + "loss": 0.1199, + "step": 12521 + }, + { + "epoch": 34.977653631284916, + "grad_norm": 1.332696795463562, + "learning_rate": 2.8671880819366853e-05, + "loss": 0.131, + "step": 12522 + }, + { + "epoch": 34.9804469273743, + "grad_norm": 0.3758484125137329, + "learning_rate": 2.867150837988827e-05, + "loss": 0.1286, + "step": 12523 + }, + { + "epoch": 34.98324022346369, + "grad_norm": 0.49063053727149963, + "learning_rate": 2.8671135940409683e-05, + "loss": 0.1237, + "step": 12524 + }, + { + "epoch": 34.986033519553075, + "grad_norm": 0.7230969071388245, + "learning_rate": 2.86707635009311e-05, + "loss": 0.1344, + "step": 12525 + }, + { + "epoch": 34.98882681564246, + "grad_norm": 0.8582870364189148, + "learning_rate": 2.8670391061452516e-05, + "loss": 0.1073, + "step": 12526 + }, + { + "epoch": 34.99162011173184, + "grad_norm": 0.7988040447235107, + "learning_rate": 2.867001862197393e-05, + "loss": 0.1048, + "step": 12527 + }, + { + "epoch": 34.99441340782123, + "grad_norm": 1.032778024673462, + "learning_rate": 2.8669646182495346e-05, + "loss": 0.1036, + "step": 12528 + }, + { + "epoch": 34.997206703910614, + "grad_norm": 0.6394571661949158, + "learning_rate": 2.8669273743016762e-05, + "loss": 0.1023, + "step": 12529 + }, + { + "epoch": 35.0, + "grad_norm": 0.9822726845741272, + "learning_rate": 2.8668901303538175e-05, + "loss": 0.1721, + "step": 12530 + }, + { + "epoch": 35.002793296089386, + "grad_norm": 0.45644325017929077, + "learning_rate": 2.866852886405959e-05, + "loss": 0.123, + "step": 12531 + }, + { + "epoch": 35.00558659217877, + "grad_norm": 0.4861440062522888, + "learning_rate": 2.8668156424581005e-05, + "loss": 0.1186, + "step": 12532 + }, + { + "epoch": 35.00837988826816, + "grad_norm": 0.5961526036262512, + "learning_rate": 2.866778398510242e-05, + "loss": 0.1179, + "step": 12533 + }, + { + "epoch": 35.01117318435754, + "grad_norm": 0.32558730244636536, + "learning_rate": 2.8667411545623838e-05, + "loss": 0.099, + "step": 12534 + }, + { + "epoch": 35.013966480446925, + "grad_norm": 0.5708363056182861, + "learning_rate": 2.8667039106145255e-05, + "loss": 0.1374, + "step": 12535 + }, + { + "epoch": 35.01675977653631, + "grad_norm": 1.3657076358795166, + "learning_rate": 2.8666666666666668e-05, + "loss": 0.1197, + "step": 12536 + }, + { + "epoch": 35.0195530726257, + "grad_norm": 0.5179590582847595, + "learning_rate": 2.866629422718808e-05, + "loss": 0.1192, + "step": 12537 + }, + { + "epoch": 35.022346368715084, + "grad_norm": 0.4806543290615082, + "learning_rate": 2.8665921787709497e-05, + "loss": 0.1041, + "step": 12538 + }, + { + "epoch": 35.02513966480447, + "grad_norm": 0.5884738564491272, + "learning_rate": 2.8665549348230914e-05, + "loss": 0.1182, + "step": 12539 + }, + { + "epoch": 35.02793296089386, + "grad_norm": 0.5764692425727844, + "learning_rate": 2.8665176908752327e-05, + "loss": 0.1031, + "step": 12540 + }, + { + "epoch": 35.03072625698324, + "grad_norm": 0.4058631360530853, + "learning_rate": 2.8664804469273744e-05, + "loss": 0.0968, + "step": 12541 + }, + { + "epoch": 35.03351955307262, + "grad_norm": 1.0484695434570312, + "learning_rate": 2.866443202979516e-05, + "loss": 0.0772, + "step": 12542 + }, + { + "epoch": 35.03631284916201, + "grad_norm": 0.6000149846076965, + "learning_rate": 2.8664059590316577e-05, + "loss": 0.0966, + "step": 12543 + }, + { + "epoch": 35.039106145251395, + "grad_norm": 0.5086411237716675, + "learning_rate": 2.866368715083799e-05, + "loss": 0.1269, + "step": 12544 + }, + { + "epoch": 35.04189944134078, + "grad_norm": 0.6798780560493469, + "learning_rate": 2.8663314711359403e-05, + "loss": 0.0929, + "step": 12545 + }, + { + "epoch": 35.04469273743017, + "grad_norm": 1.4328126907348633, + "learning_rate": 2.866294227188082e-05, + "loss": 0.0952, + "step": 12546 + }, + { + "epoch": 35.047486033519554, + "grad_norm": 0.5250682234764099, + "learning_rate": 2.8662569832402236e-05, + "loss": 0.0823, + "step": 12547 + }, + { + "epoch": 35.05027932960894, + "grad_norm": 0.6736853122711182, + "learning_rate": 2.8662197392923653e-05, + "loss": 0.0848, + "step": 12548 + }, + { + "epoch": 35.05307262569833, + "grad_norm": 0.9679290652275085, + "learning_rate": 2.8661824953445066e-05, + "loss": 0.0964, + "step": 12549 + }, + { + "epoch": 35.055865921787706, + "grad_norm": 0.7215505242347717, + "learning_rate": 2.866145251396648e-05, + "loss": 0.1068, + "step": 12550 + }, + { + "epoch": 35.05865921787709, + "grad_norm": 0.6311858892440796, + "learning_rate": 2.8661080074487895e-05, + "loss": 0.1027, + "step": 12551 + }, + { + "epoch": 35.06145251396648, + "grad_norm": 0.6829830408096313, + "learning_rate": 2.8660707635009312e-05, + "loss": 0.0793, + "step": 12552 + }, + { + "epoch": 35.064245810055866, + "grad_norm": 0.7093613147735596, + "learning_rate": 2.866033519553073e-05, + "loss": 0.086, + "step": 12553 + }, + { + "epoch": 35.06703910614525, + "grad_norm": 0.8886780738830566, + "learning_rate": 2.865996275605214e-05, + "loss": 0.0838, + "step": 12554 + }, + { + "epoch": 35.06983240223464, + "grad_norm": 1.0668368339538574, + "learning_rate": 2.8659590316573558e-05, + "loss": 0.1073, + "step": 12555 + }, + { + "epoch": 35.072625698324025, + "grad_norm": 0.3707360327243805, + "learning_rate": 2.8659217877094975e-05, + "loss": 0.1175, + "step": 12556 + }, + { + "epoch": 35.07541899441341, + "grad_norm": 0.9310334920883179, + "learning_rate": 2.8658845437616388e-05, + "loss": 0.1511, + "step": 12557 + }, + { + "epoch": 35.07821229050279, + "grad_norm": 0.6924193501472473, + "learning_rate": 2.86584729981378e-05, + "loss": 0.1294, + "step": 12558 + }, + { + "epoch": 35.08100558659218, + "grad_norm": 0.5017798542976379, + "learning_rate": 2.8658100558659218e-05, + "loss": 0.1178, + "step": 12559 + }, + { + "epoch": 35.08379888268156, + "grad_norm": 0.5217490792274475, + "learning_rate": 2.8657728119180634e-05, + "loss": 0.1121, + "step": 12560 + }, + { + "epoch": 35.08659217877095, + "grad_norm": 0.5221053957939148, + "learning_rate": 2.865735567970205e-05, + "loss": 0.0985, + "step": 12561 + }, + { + "epoch": 35.089385474860336, + "grad_norm": 0.9596811532974243, + "learning_rate": 2.8656983240223464e-05, + "loss": 0.1185, + "step": 12562 + }, + { + "epoch": 35.09217877094972, + "grad_norm": 0.6762769222259521, + "learning_rate": 2.865661080074488e-05, + "loss": 0.1059, + "step": 12563 + }, + { + "epoch": 35.09497206703911, + "grad_norm": 0.6051248908042908, + "learning_rate": 2.8656238361266293e-05, + "loss": 0.1273, + "step": 12564 + }, + { + "epoch": 35.097765363128495, + "grad_norm": 0.5756537914276123, + "learning_rate": 2.865586592178771e-05, + "loss": 0.1057, + "step": 12565 + }, + { + "epoch": 35.100558659217874, + "grad_norm": 0.504115641117096, + "learning_rate": 2.8655493482309127e-05, + "loss": 0.0976, + "step": 12566 + }, + { + "epoch": 35.10335195530726, + "grad_norm": 0.623311460018158, + "learning_rate": 2.865512104283054e-05, + "loss": 0.0984, + "step": 12567 + }, + { + "epoch": 35.10614525139665, + "grad_norm": 1.0195515155792236, + "learning_rate": 2.8654748603351956e-05, + "loss": 0.0958, + "step": 12568 + }, + { + "epoch": 35.108938547486034, + "grad_norm": 0.4917545020580292, + "learning_rate": 2.8654376163873373e-05, + "loss": 0.0875, + "step": 12569 + }, + { + "epoch": 35.11173184357542, + "grad_norm": 0.6525407433509827, + "learning_rate": 2.8654003724394786e-05, + "loss": 0.0731, + "step": 12570 + }, + { + "epoch": 35.114525139664806, + "grad_norm": 1.4901931285858154, + "learning_rate": 2.86536312849162e-05, + "loss": 0.0863, + "step": 12571 + }, + { + "epoch": 35.11731843575419, + "grad_norm": 0.802433431148529, + "learning_rate": 2.8653258845437616e-05, + "loss": 0.1093, + "step": 12572 + }, + { + "epoch": 35.12011173184357, + "grad_norm": 0.8636652231216431, + "learning_rate": 2.8652886405959032e-05, + "loss": 0.0951, + "step": 12573 + }, + { + "epoch": 35.12290502793296, + "grad_norm": 0.8101543188095093, + "learning_rate": 2.865251396648045e-05, + "loss": 0.0941, + "step": 12574 + }, + { + "epoch": 35.125698324022345, + "grad_norm": 0.5626256465911865, + "learning_rate": 2.8652141527001865e-05, + "loss": 0.1064, + "step": 12575 + }, + { + "epoch": 35.12849162011173, + "grad_norm": 0.605478823184967, + "learning_rate": 2.865176908752328e-05, + "loss": 0.0902, + "step": 12576 + }, + { + "epoch": 35.13128491620112, + "grad_norm": 0.619085431098938, + "learning_rate": 2.865139664804469e-05, + "loss": 0.0906, + "step": 12577 + }, + { + "epoch": 35.134078212290504, + "grad_norm": 1.1203491687774658, + "learning_rate": 2.8651024208566108e-05, + "loss": 0.1034, + "step": 12578 + }, + { + "epoch": 35.13687150837989, + "grad_norm": 1.252799153327942, + "learning_rate": 2.8650651769087525e-05, + "loss": 0.1087, + "step": 12579 + }, + { + "epoch": 35.13966480446928, + "grad_norm": 0.9799112677574158, + "learning_rate": 2.8650279329608938e-05, + "loss": 0.1106, + "step": 12580 + }, + { + "epoch": 35.142458100558656, + "grad_norm": 0.5049086809158325, + "learning_rate": 2.8649906890130354e-05, + "loss": 0.1216, + "step": 12581 + }, + { + "epoch": 35.14525139664804, + "grad_norm": 0.6378149390220642, + "learning_rate": 2.864953445065177e-05, + "loss": 0.1291, + "step": 12582 + }, + { + "epoch": 35.14804469273743, + "grad_norm": 0.6774163246154785, + "learning_rate": 2.8649162011173187e-05, + "loss": 0.1517, + "step": 12583 + }, + { + "epoch": 35.150837988826815, + "grad_norm": 0.9902382493019104, + "learning_rate": 2.86487895716946e-05, + "loss": 0.1414, + "step": 12584 + }, + { + "epoch": 35.1536312849162, + "grad_norm": 0.44367364048957825, + "learning_rate": 2.8648417132216014e-05, + "loss": 0.1167, + "step": 12585 + }, + { + "epoch": 35.15642458100559, + "grad_norm": 1.0062994956970215, + "learning_rate": 2.864804469273743e-05, + "loss": 0.1201, + "step": 12586 + }, + { + "epoch": 35.159217877094974, + "grad_norm": 0.3981975018978119, + "learning_rate": 2.8647672253258847e-05, + "loss": 0.0975, + "step": 12587 + }, + { + "epoch": 35.16201117318436, + "grad_norm": 0.6742017865180969, + "learning_rate": 2.8647299813780263e-05, + "loss": 0.1143, + "step": 12588 + }, + { + "epoch": 35.16480446927374, + "grad_norm": 0.5275003910064697, + "learning_rate": 2.8646927374301676e-05, + "loss": 0.1105, + "step": 12589 + }, + { + "epoch": 35.167597765363126, + "grad_norm": 0.4953584372997284, + "learning_rate": 2.864655493482309e-05, + "loss": 0.1115, + "step": 12590 + }, + { + "epoch": 35.17039106145251, + "grad_norm": 0.8853933811187744, + "learning_rate": 2.8646182495344506e-05, + "loss": 0.1162, + "step": 12591 + }, + { + "epoch": 35.1731843575419, + "grad_norm": 0.5595427751541138, + "learning_rate": 2.8645810055865923e-05, + "loss": 0.1113, + "step": 12592 + }, + { + "epoch": 35.175977653631286, + "grad_norm": 0.3762868344783783, + "learning_rate": 2.864543761638734e-05, + "loss": 0.1035, + "step": 12593 + }, + { + "epoch": 35.17877094972067, + "grad_norm": 3.172611951828003, + "learning_rate": 2.8645065176908752e-05, + "loss": 0.1019, + "step": 12594 + }, + { + "epoch": 35.18156424581006, + "grad_norm": 0.559041440486908, + "learning_rate": 2.864469273743017e-05, + "loss": 0.0948, + "step": 12595 + }, + { + "epoch": 35.184357541899445, + "grad_norm": 0.5812348127365112, + "learning_rate": 2.8644320297951585e-05, + "loss": 0.0828, + "step": 12596 + }, + { + "epoch": 35.187150837988824, + "grad_norm": 0.661872386932373, + "learning_rate": 2.8643947858473e-05, + "loss": 0.1172, + "step": 12597 + }, + { + "epoch": 35.18994413407821, + "grad_norm": 0.4361197352409363, + "learning_rate": 2.864357541899441e-05, + "loss": 0.0647, + "step": 12598 + }, + { + "epoch": 35.1927374301676, + "grad_norm": 1.3499464988708496, + "learning_rate": 2.8643202979515828e-05, + "loss": 0.0934, + "step": 12599 + }, + { + "epoch": 35.19553072625698, + "grad_norm": 1.068103551864624, + "learning_rate": 2.8642830540037245e-05, + "loss": 0.0918, + "step": 12600 + }, + { + "epoch": 35.19832402234637, + "grad_norm": 0.955237627029419, + "learning_rate": 2.864245810055866e-05, + "loss": 0.095, + "step": 12601 + }, + { + "epoch": 35.201117318435756, + "grad_norm": 0.7821932435035706, + "learning_rate": 2.8642085661080078e-05, + "loss": 0.0799, + "step": 12602 + }, + { + "epoch": 35.20391061452514, + "grad_norm": 0.6201761364936829, + "learning_rate": 2.864171322160149e-05, + "loss": 0.1024, + "step": 12603 + }, + { + "epoch": 35.20670391061452, + "grad_norm": 0.9335343837738037, + "learning_rate": 2.8641340782122904e-05, + "loss": 0.0887, + "step": 12604 + }, + { + "epoch": 35.20949720670391, + "grad_norm": 1.9845904111862183, + "learning_rate": 2.864096834264432e-05, + "loss": 0.1113, + "step": 12605 + }, + { + "epoch": 35.212290502793294, + "grad_norm": 0.6334784626960754, + "learning_rate": 2.8640595903165737e-05, + "loss": 0.145, + "step": 12606 + }, + { + "epoch": 35.21508379888268, + "grad_norm": 0.3701227009296417, + "learning_rate": 2.864022346368715e-05, + "loss": 0.1112, + "step": 12607 + }, + { + "epoch": 35.21787709497207, + "grad_norm": 0.4628690481185913, + "learning_rate": 2.8639851024208567e-05, + "loss": 0.1334, + "step": 12608 + }, + { + "epoch": 35.220670391061454, + "grad_norm": 0.6724932789802551, + "learning_rate": 2.8639478584729983e-05, + "loss": 0.1493, + "step": 12609 + }, + { + "epoch": 35.22346368715084, + "grad_norm": 2.1429378986358643, + "learning_rate": 2.8639106145251397e-05, + "loss": 0.1461, + "step": 12610 + }, + { + "epoch": 35.226256983240226, + "grad_norm": 1.004496693611145, + "learning_rate": 2.8638733705772813e-05, + "loss": 0.1129, + "step": 12611 + }, + { + "epoch": 35.229050279329606, + "grad_norm": 0.607039749622345, + "learning_rate": 2.8638361266294226e-05, + "loss": 0.1275, + "step": 12612 + }, + { + "epoch": 35.23184357541899, + "grad_norm": 0.4788110852241516, + "learning_rate": 2.8637988826815643e-05, + "loss": 0.141, + "step": 12613 + }, + { + "epoch": 35.23463687150838, + "grad_norm": 0.5960618853569031, + "learning_rate": 2.863761638733706e-05, + "loss": 0.0896, + "step": 12614 + }, + { + "epoch": 35.237430167597765, + "grad_norm": 0.6030815839767456, + "learning_rate": 2.8637243947858476e-05, + "loss": 0.1012, + "step": 12615 + }, + { + "epoch": 35.24022346368715, + "grad_norm": 0.5562729835510254, + "learning_rate": 2.863687150837989e-05, + "loss": 0.0861, + "step": 12616 + }, + { + "epoch": 35.24301675977654, + "grad_norm": 1.2412105798721313, + "learning_rate": 2.8636499068901302e-05, + "loss": 0.1068, + "step": 12617 + }, + { + "epoch": 35.245810055865924, + "grad_norm": 0.7399153709411621, + "learning_rate": 2.863612662942272e-05, + "loss": 0.105, + "step": 12618 + }, + { + "epoch": 35.24860335195531, + "grad_norm": 1.4547446966171265, + "learning_rate": 2.8635754189944135e-05, + "loss": 0.1031, + "step": 12619 + }, + { + "epoch": 35.25139664804469, + "grad_norm": 0.599181592464447, + "learning_rate": 2.8635381750465552e-05, + "loss": 0.093, + "step": 12620 + }, + { + "epoch": 35.254189944134076, + "grad_norm": 0.6936540007591248, + "learning_rate": 2.8635009310986965e-05, + "loss": 0.0871, + "step": 12621 + }, + { + "epoch": 35.25698324022346, + "grad_norm": 0.7121127247810364, + "learning_rate": 2.863463687150838e-05, + "loss": 0.0983, + "step": 12622 + }, + { + "epoch": 35.25977653631285, + "grad_norm": 1.1034637689590454, + "learning_rate": 2.8634264432029798e-05, + "loss": 0.0878, + "step": 12623 + }, + { + "epoch": 35.262569832402235, + "grad_norm": 0.68548184633255, + "learning_rate": 2.863389199255121e-05, + "loss": 0.0992, + "step": 12624 + }, + { + "epoch": 35.26536312849162, + "grad_norm": 0.5855973958969116, + "learning_rate": 2.8633519553072624e-05, + "loss": 0.0897, + "step": 12625 + }, + { + "epoch": 35.26815642458101, + "grad_norm": 0.5973667502403259, + "learning_rate": 2.863314711359404e-05, + "loss": 0.0901, + "step": 12626 + }, + { + "epoch": 35.270949720670394, + "grad_norm": 0.6468331813812256, + "learning_rate": 2.8632774674115457e-05, + "loss": 0.106, + "step": 12627 + }, + { + "epoch": 35.273743016759774, + "grad_norm": 1.0311312675476074, + "learning_rate": 2.8632402234636874e-05, + "loss": 0.1139, + "step": 12628 + }, + { + "epoch": 35.27653631284916, + "grad_norm": 8.182437896728516, + "learning_rate": 2.863202979515829e-05, + "loss": 0.1228, + "step": 12629 + }, + { + "epoch": 35.279329608938546, + "grad_norm": 1.4373189210891724, + "learning_rate": 2.86316573556797e-05, + "loss": 0.1059, + "step": 12630 + }, + { + "epoch": 35.28212290502793, + "grad_norm": 0.8917000889778137, + "learning_rate": 2.8631284916201117e-05, + "loss": 0.1499, + "step": 12631 + }, + { + "epoch": 35.28491620111732, + "grad_norm": 1.8162322044372559, + "learning_rate": 2.8630912476722533e-05, + "loss": 0.1199, + "step": 12632 + }, + { + "epoch": 35.287709497206706, + "grad_norm": 0.5816372632980347, + "learning_rate": 2.863054003724395e-05, + "loss": 0.1314, + "step": 12633 + }, + { + "epoch": 35.29050279329609, + "grad_norm": 0.5117472410202026, + "learning_rate": 2.8630167597765363e-05, + "loss": 0.1463, + "step": 12634 + }, + { + "epoch": 35.29329608938548, + "grad_norm": 0.4789213240146637, + "learning_rate": 2.862979515828678e-05, + "loss": 0.1221, + "step": 12635 + }, + { + "epoch": 35.29608938547486, + "grad_norm": 0.7155479788780212, + "learning_rate": 2.8629422718808196e-05, + "loss": 0.1151, + "step": 12636 + }, + { + "epoch": 35.298882681564244, + "grad_norm": 0.7502048015594482, + "learning_rate": 2.862905027932961e-05, + "loss": 0.1242, + "step": 12637 + }, + { + "epoch": 35.30167597765363, + "grad_norm": 0.44711077213287354, + "learning_rate": 2.8628677839851026e-05, + "loss": 0.1085, + "step": 12638 + }, + { + "epoch": 35.30446927374302, + "grad_norm": 0.545798659324646, + "learning_rate": 2.862830540037244e-05, + "loss": 0.1221, + "step": 12639 + }, + { + "epoch": 35.3072625698324, + "grad_norm": 0.5038735270500183, + "learning_rate": 2.8627932960893855e-05, + "loss": 0.1181, + "step": 12640 + }, + { + "epoch": 35.31005586592179, + "grad_norm": 1.8946629762649536, + "learning_rate": 2.8627560521415272e-05, + "loss": 0.1036, + "step": 12641 + }, + { + "epoch": 35.312849162011176, + "grad_norm": 0.535828709602356, + "learning_rate": 2.862718808193669e-05, + "loss": 0.1076, + "step": 12642 + }, + { + "epoch": 35.315642458100555, + "grad_norm": 1.0477522611618042, + "learning_rate": 2.86268156424581e-05, + "loss": 0.0852, + "step": 12643 + }, + { + "epoch": 35.31843575418994, + "grad_norm": 0.5067760944366455, + "learning_rate": 2.8626443202979515e-05, + "loss": 0.1195, + "step": 12644 + }, + { + "epoch": 35.32122905027933, + "grad_norm": 0.7768335342407227, + "learning_rate": 2.862607076350093e-05, + "loss": 0.1012, + "step": 12645 + }, + { + "epoch": 35.324022346368714, + "grad_norm": 0.4936356246471405, + "learning_rate": 2.8625698324022348e-05, + "loss": 0.0858, + "step": 12646 + }, + { + "epoch": 35.3268156424581, + "grad_norm": 0.6493908762931824, + "learning_rate": 2.8625325884543764e-05, + "loss": 0.1024, + "step": 12647 + }, + { + "epoch": 35.32960893854749, + "grad_norm": 1.0465030670166016, + "learning_rate": 2.8624953445065177e-05, + "loss": 0.0931, + "step": 12648 + }, + { + "epoch": 35.332402234636874, + "grad_norm": 1.0032364130020142, + "learning_rate": 2.8624581005586594e-05, + "loss": 0.1072, + "step": 12649 + }, + { + "epoch": 35.33519553072626, + "grad_norm": 0.520735502243042, + "learning_rate": 2.8624208566108007e-05, + "loss": 0.0874, + "step": 12650 + }, + { + "epoch": 35.33798882681564, + "grad_norm": 0.9527572989463806, + "learning_rate": 2.8623836126629424e-05, + "loss": 0.0852, + "step": 12651 + }, + { + "epoch": 35.340782122905026, + "grad_norm": 0.8548133969306946, + "learning_rate": 2.8623463687150837e-05, + "loss": 0.0822, + "step": 12652 + }, + { + "epoch": 35.34357541899441, + "grad_norm": 0.6642070412635803, + "learning_rate": 2.8623091247672253e-05, + "loss": 0.0842, + "step": 12653 + }, + { + "epoch": 35.3463687150838, + "grad_norm": 0.663454532623291, + "learning_rate": 2.862271880819367e-05, + "loss": 0.0904, + "step": 12654 + }, + { + "epoch": 35.349162011173185, + "grad_norm": 2.7487425804138184, + "learning_rate": 2.8622346368715086e-05, + "loss": 0.141, + "step": 12655 + }, + { + "epoch": 35.35195530726257, + "grad_norm": 0.6100386381149292, + "learning_rate": 2.86219739292365e-05, + "loss": 0.1307, + "step": 12656 + }, + { + "epoch": 35.35474860335196, + "grad_norm": 0.5931417942047119, + "learning_rate": 2.8621601489757913e-05, + "loss": 0.137, + "step": 12657 + }, + { + "epoch": 35.357541899441344, + "grad_norm": 0.41294941306114197, + "learning_rate": 2.862122905027933e-05, + "loss": 0.1122, + "step": 12658 + }, + { + "epoch": 35.36033519553072, + "grad_norm": 0.5691683888435364, + "learning_rate": 2.8620856610800746e-05, + "loss": 0.1226, + "step": 12659 + }, + { + "epoch": 35.36312849162011, + "grad_norm": 0.4043266773223877, + "learning_rate": 2.8620484171322162e-05, + "loss": 0.1147, + "step": 12660 + }, + { + "epoch": 35.365921787709496, + "grad_norm": 0.8069313168525696, + "learning_rate": 2.8620111731843575e-05, + "loss": 0.1213, + "step": 12661 + }, + { + "epoch": 35.36871508379888, + "grad_norm": 0.5642617344856262, + "learning_rate": 2.8619739292364992e-05, + "loss": 0.1314, + "step": 12662 + }, + { + "epoch": 35.37150837988827, + "grad_norm": 3.5193331241607666, + "learning_rate": 2.861936685288641e-05, + "loss": 0.1105, + "step": 12663 + }, + { + "epoch": 35.374301675977655, + "grad_norm": 0.8121011257171631, + "learning_rate": 2.861899441340782e-05, + "loss": 0.0901, + "step": 12664 + }, + { + "epoch": 35.37709497206704, + "grad_norm": 0.5110476613044739, + "learning_rate": 2.8618621973929235e-05, + "loss": 0.1068, + "step": 12665 + }, + { + "epoch": 35.37988826815643, + "grad_norm": 0.4374030530452728, + "learning_rate": 2.861824953445065e-05, + "loss": 0.0986, + "step": 12666 + }, + { + "epoch": 35.38268156424581, + "grad_norm": 0.6057685613632202, + "learning_rate": 2.8617877094972068e-05, + "loss": 0.1087, + "step": 12667 + }, + { + "epoch": 35.385474860335194, + "grad_norm": 1.0314626693725586, + "learning_rate": 2.8617504655493484e-05, + "loss": 0.1164, + "step": 12668 + }, + { + "epoch": 35.38826815642458, + "grad_norm": 0.601758599281311, + "learning_rate": 2.86171322160149e-05, + "loss": 0.0878, + "step": 12669 + }, + { + "epoch": 35.391061452513966, + "grad_norm": 0.46907806396484375, + "learning_rate": 2.861675977653631e-05, + "loss": 0.0856, + "step": 12670 + }, + { + "epoch": 35.39385474860335, + "grad_norm": 0.5828276872634888, + "learning_rate": 2.8616387337057727e-05, + "loss": 0.1002, + "step": 12671 + }, + { + "epoch": 35.39664804469274, + "grad_norm": 0.5761587619781494, + "learning_rate": 2.8616014897579144e-05, + "loss": 0.0907, + "step": 12672 + }, + { + "epoch": 35.399441340782126, + "grad_norm": 0.875533401966095, + "learning_rate": 2.861564245810056e-05, + "loss": 0.0872, + "step": 12673 + }, + { + "epoch": 35.40223463687151, + "grad_norm": 0.9589071273803711, + "learning_rate": 2.8615270018621973e-05, + "loss": 0.0798, + "step": 12674 + }, + { + "epoch": 35.40502793296089, + "grad_norm": 0.4409010708332062, + "learning_rate": 2.861489757914339e-05, + "loss": 0.0872, + "step": 12675 + }, + { + "epoch": 35.40782122905028, + "grad_norm": 0.5351981520652771, + "learning_rate": 2.8614525139664807e-05, + "loss": 0.0909, + "step": 12676 + }, + { + "epoch": 35.410614525139664, + "grad_norm": 0.7981750965118408, + "learning_rate": 2.861415270018622e-05, + "loss": 0.084, + "step": 12677 + }, + { + "epoch": 35.41340782122905, + "grad_norm": 0.8208580017089844, + "learning_rate": 2.8613780260707636e-05, + "loss": 0.0851, + "step": 12678 + }, + { + "epoch": 35.41620111731844, + "grad_norm": 1.3830796480178833, + "learning_rate": 2.861340782122905e-05, + "loss": 0.0878, + "step": 12679 + }, + { + "epoch": 35.41899441340782, + "grad_norm": 2.5081405639648438, + "learning_rate": 2.8613035381750466e-05, + "loss": 0.111, + "step": 12680 + }, + { + "epoch": 35.42178770949721, + "grad_norm": 0.40884295105934143, + "learning_rate": 2.8612662942271882e-05, + "loss": 0.1269, + "step": 12681 + }, + { + "epoch": 35.42458100558659, + "grad_norm": 0.4836191236972809, + "learning_rate": 2.86122905027933e-05, + "loss": 0.1189, + "step": 12682 + }, + { + "epoch": 35.427374301675975, + "grad_norm": 0.404023677110672, + "learning_rate": 2.8611918063314712e-05, + "loss": 0.1219, + "step": 12683 + }, + { + "epoch": 35.43016759776536, + "grad_norm": 0.3910065293312073, + "learning_rate": 2.8611545623836125e-05, + "loss": 0.1124, + "step": 12684 + }, + { + "epoch": 35.43296089385475, + "grad_norm": 0.42828068137168884, + "learning_rate": 2.8611173184357542e-05, + "loss": 0.1067, + "step": 12685 + }, + { + "epoch": 35.435754189944134, + "grad_norm": 0.4050985276699066, + "learning_rate": 2.861080074487896e-05, + "loss": 0.1284, + "step": 12686 + }, + { + "epoch": 35.43854748603352, + "grad_norm": 0.6300724148750305, + "learning_rate": 2.8610428305400375e-05, + "loss": 0.1364, + "step": 12687 + }, + { + "epoch": 35.44134078212291, + "grad_norm": 0.5587626099586487, + "learning_rate": 2.8610055865921788e-05, + "loss": 0.1039, + "step": 12688 + }, + { + "epoch": 35.444134078212294, + "grad_norm": 0.5257675051689148, + "learning_rate": 2.8609683426443205e-05, + "loss": 0.0975, + "step": 12689 + }, + { + "epoch": 35.44692737430167, + "grad_norm": 0.6634262800216675, + "learning_rate": 2.8609310986964618e-05, + "loss": 0.1285, + "step": 12690 + }, + { + "epoch": 35.44972067039106, + "grad_norm": 0.5079955458641052, + "learning_rate": 2.8608938547486034e-05, + "loss": 0.1349, + "step": 12691 + }, + { + "epoch": 35.452513966480446, + "grad_norm": 1.3186984062194824, + "learning_rate": 2.8608566108007447e-05, + "loss": 0.109, + "step": 12692 + }, + { + "epoch": 35.45530726256983, + "grad_norm": 1.2905592918395996, + "learning_rate": 2.8608193668528864e-05, + "loss": 0.0937, + "step": 12693 + }, + { + "epoch": 35.45810055865922, + "grad_norm": 0.6717334985733032, + "learning_rate": 2.860782122905028e-05, + "loss": 0.0935, + "step": 12694 + }, + { + "epoch": 35.460893854748605, + "grad_norm": 0.7968651056289673, + "learning_rate": 2.8607448789571697e-05, + "loss": 0.1165, + "step": 12695 + }, + { + "epoch": 35.46368715083799, + "grad_norm": 0.9434257745742798, + "learning_rate": 2.8607076350093114e-05, + "loss": 0.0944, + "step": 12696 + }, + { + "epoch": 35.46648044692738, + "grad_norm": 1.0646514892578125, + "learning_rate": 2.8606703910614523e-05, + "loss": 0.0966, + "step": 12697 + }, + { + "epoch": 35.46927374301676, + "grad_norm": 0.5008102059364319, + "learning_rate": 2.860633147113594e-05, + "loss": 0.0803, + "step": 12698 + }, + { + "epoch": 35.47206703910614, + "grad_norm": 0.45859450101852417, + "learning_rate": 2.8605959031657356e-05, + "loss": 0.0919, + "step": 12699 + }, + { + "epoch": 35.47486033519553, + "grad_norm": 0.5608917474746704, + "learning_rate": 2.8605586592178773e-05, + "loss": 0.0919, + "step": 12700 + }, + { + "epoch": 35.477653631284916, + "grad_norm": 0.9109189510345459, + "learning_rate": 2.8605214152700186e-05, + "loss": 0.094, + "step": 12701 + }, + { + "epoch": 35.4804469273743, + "grad_norm": 0.639362633228302, + "learning_rate": 2.8604841713221603e-05, + "loss": 0.0954, + "step": 12702 + }, + { + "epoch": 35.48324022346369, + "grad_norm": 1.2785358428955078, + "learning_rate": 2.860446927374302e-05, + "loss": 0.0971, + "step": 12703 + }, + { + "epoch": 35.486033519553075, + "grad_norm": 1.8585957288742065, + "learning_rate": 2.8604096834264432e-05, + "loss": 0.0942, + "step": 12704 + }, + { + "epoch": 35.48882681564246, + "grad_norm": 1.3540620803833008, + "learning_rate": 2.860372439478585e-05, + "loss": 0.1328, + "step": 12705 + }, + { + "epoch": 35.49162011173184, + "grad_norm": 0.5276389122009277, + "learning_rate": 2.8603351955307262e-05, + "loss": 0.15, + "step": 12706 + }, + { + "epoch": 35.49441340782123, + "grad_norm": 2.168287754058838, + "learning_rate": 2.860297951582868e-05, + "loss": 0.134, + "step": 12707 + }, + { + "epoch": 35.497206703910614, + "grad_norm": 0.7912912368774414, + "learning_rate": 2.8602607076350095e-05, + "loss": 0.11, + "step": 12708 + }, + { + "epoch": 35.5, + "grad_norm": 0.38878828287124634, + "learning_rate": 2.860223463687151e-05, + "loss": 0.122, + "step": 12709 + }, + { + "epoch": 35.502793296089386, + "grad_norm": 0.6501900553703308, + "learning_rate": 2.860186219739292e-05, + "loss": 0.1339, + "step": 12710 + }, + { + "epoch": 35.50558659217877, + "grad_norm": 0.6427205801010132, + "learning_rate": 2.8601489757914338e-05, + "loss": 0.1041, + "step": 12711 + }, + { + "epoch": 35.50837988826816, + "grad_norm": 0.9122547507286072, + "learning_rate": 2.8601117318435754e-05, + "loss": 0.136, + "step": 12712 + }, + { + "epoch": 35.51117318435754, + "grad_norm": 0.6551293134689331, + "learning_rate": 2.860074487895717e-05, + "loss": 0.1169, + "step": 12713 + }, + { + "epoch": 35.513966480446925, + "grad_norm": 0.499549925327301, + "learning_rate": 2.8600372439478587e-05, + "loss": 0.121, + "step": 12714 + }, + { + "epoch": 35.51675977653631, + "grad_norm": 0.46624842286109924, + "learning_rate": 2.86e-05, + "loss": 0.0982, + "step": 12715 + }, + { + "epoch": 35.5195530726257, + "grad_norm": 0.5469024777412415, + "learning_rate": 2.8599627560521417e-05, + "loss": 0.1156, + "step": 12716 + }, + { + "epoch": 35.522346368715084, + "grad_norm": 1.1890872716903687, + "learning_rate": 2.859925512104283e-05, + "loss": 0.1064, + "step": 12717 + }, + { + "epoch": 35.52513966480447, + "grad_norm": 0.43963345885276794, + "learning_rate": 2.8598882681564247e-05, + "loss": 0.1002, + "step": 12718 + }, + { + "epoch": 35.52793296089386, + "grad_norm": 0.474274605512619, + "learning_rate": 2.859851024208566e-05, + "loss": 0.0868, + "step": 12719 + }, + { + "epoch": 35.53072625698324, + "grad_norm": 0.6140235066413879, + "learning_rate": 2.8598137802607077e-05, + "loss": 0.0986, + "step": 12720 + }, + { + "epoch": 35.53351955307262, + "grad_norm": 0.64467853307724, + "learning_rate": 2.8597765363128493e-05, + "loss": 0.0983, + "step": 12721 + }, + { + "epoch": 35.53631284916201, + "grad_norm": 0.48273253440856934, + "learning_rate": 2.859739292364991e-05, + "loss": 0.1024, + "step": 12722 + }, + { + "epoch": 35.539106145251395, + "grad_norm": 0.9401225447654724, + "learning_rate": 2.8597020484171326e-05, + "loss": 0.0912, + "step": 12723 + }, + { + "epoch": 35.54189944134078, + "grad_norm": 1.5433109998703003, + "learning_rate": 2.8596648044692736e-05, + "loss": 0.096, + "step": 12724 + }, + { + "epoch": 35.54469273743017, + "grad_norm": 0.5807570815086365, + "learning_rate": 2.8596275605214152e-05, + "loss": 0.0816, + "step": 12725 + }, + { + "epoch": 35.547486033519554, + "grad_norm": 0.5925974249839783, + "learning_rate": 2.859590316573557e-05, + "loss": 0.0853, + "step": 12726 + }, + { + "epoch": 35.55027932960894, + "grad_norm": 2.099522829055786, + "learning_rate": 2.8595530726256985e-05, + "loss": 0.0816, + "step": 12727 + }, + { + "epoch": 35.55307262569833, + "grad_norm": 0.7767788767814636, + "learning_rate": 2.85951582867784e-05, + "loss": 0.082, + "step": 12728 + }, + { + "epoch": 35.555865921787706, + "grad_norm": 0.8666935563087463, + "learning_rate": 2.8594785847299815e-05, + "loss": 0.0947, + "step": 12729 + }, + { + "epoch": 35.55865921787709, + "grad_norm": 1.1199257373809814, + "learning_rate": 2.859441340782123e-05, + "loss": 0.1323, + "step": 12730 + }, + { + "epoch": 35.56145251396648, + "grad_norm": 0.3649541139602661, + "learning_rate": 2.8594040968342645e-05, + "loss": 0.1353, + "step": 12731 + }, + { + "epoch": 35.564245810055866, + "grad_norm": 0.5578553676605225, + "learning_rate": 2.859366852886406e-05, + "loss": 0.1206, + "step": 12732 + }, + { + "epoch": 35.56703910614525, + "grad_norm": 0.4872959852218628, + "learning_rate": 2.8593296089385475e-05, + "loss": 0.1314, + "step": 12733 + }, + { + "epoch": 35.56983240223464, + "grad_norm": 0.5048848986625671, + "learning_rate": 2.859292364990689e-05, + "loss": 0.1158, + "step": 12734 + }, + { + "epoch": 35.572625698324025, + "grad_norm": 0.6401301622390747, + "learning_rate": 2.8592551210428308e-05, + "loss": 0.1222, + "step": 12735 + }, + { + "epoch": 35.57541899441341, + "grad_norm": 0.6182692050933838, + "learning_rate": 2.8592178770949724e-05, + "loss": 0.1127, + "step": 12736 + }, + { + "epoch": 35.57821229050279, + "grad_norm": 0.919465184211731, + "learning_rate": 2.8591806331471134e-05, + "loss": 0.1239, + "step": 12737 + }, + { + "epoch": 35.58100558659218, + "grad_norm": 0.47898581624031067, + "learning_rate": 2.859143389199255e-05, + "loss": 0.1043, + "step": 12738 + }, + { + "epoch": 35.58379888268156, + "grad_norm": 0.3674750328063965, + "learning_rate": 2.8591061452513967e-05, + "loss": 0.0844, + "step": 12739 + }, + { + "epoch": 35.58659217877095, + "grad_norm": 0.6570559740066528, + "learning_rate": 2.8590689013035384e-05, + "loss": 0.1273, + "step": 12740 + }, + { + "epoch": 35.589385474860336, + "grad_norm": 0.7647868394851685, + "learning_rate": 2.85903165735568e-05, + "loss": 0.1102, + "step": 12741 + }, + { + "epoch": 35.59217877094972, + "grad_norm": 0.6488932371139526, + "learning_rate": 2.8589944134078213e-05, + "loss": 0.1117, + "step": 12742 + }, + { + "epoch": 35.59497206703911, + "grad_norm": 0.7221664190292358, + "learning_rate": 2.858957169459963e-05, + "loss": 0.1031, + "step": 12743 + }, + { + "epoch": 35.59776536312849, + "grad_norm": 0.6849223375320435, + "learning_rate": 2.8589199255121043e-05, + "loss": 0.1132, + "step": 12744 + }, + { + "epoch": 35.600558659217874, + "grad_norm": 0.42869308590888977, + "learning_rate": 2.858882681564246e-05, + "loss": 0.0884, + "step": 12745 + }, + { + "epoch": 35.60335195530726, + "grad_norm": 0.6075124740600586, + "learning_rate": 2.8588454376163873e-05, + "loss": 0.0796, + "step": 12746 + }, + { + "epoch": 35.60614525139665, + "grad_norm": 0.7200224995613098, + "learning_rate": 2.858808193668529e-05, + "loss": 0.1041, + "step": 12747 + }, + { + "epoch": 35.608938547486034, + "grad_norm": 1.1722959280014038, + "learning_rate": 2.8587709497206706e-05, + "loss": 0.0915, + "step": 12748 + }, + { + "epoch": 35.61173184357542, + "grad_norm": 1.6205488443374634, + "learning_rate": 2.8587337057728122e-05, + "loss": 0.1027, + "step": 12749 + }, + { + "epoch": 35.614525139664806, + "grad_norm": 1.1824771165847778, + "learning_rate": 2.8586964618249535e-05, + "loss": 0.093, + "step": 12750 + }, + { + "epoch": 35.61731843575419, + "grad_norm": 2.148683786392212, + "learning_rate": 2.858659217877095e-05, + "loss": 0.0875, + "step": 12751 + }, + { + "epoch": 35.62011173184357, + "grad_norm": 0.6612057089805603, + "learning_rate": 2.8586219739292365e-05, + "loss": 0.0938, + "step": 12752 + }, + { + "epoch": 35.62290502793296, + "grad_norm": 0.8633443713188171, + "learning_rate": 2.858584729981378e-05, + "loss": 0.0784, + "step": 12753 + }, + { + "epoch": 35.625698324022345, + "grad_norm": 0.9909027814865112, + "learning_rate": 2.8585474860335198e-05, + "loss": 0.121, + "step": 12754 + }, + { + "epoch": 35.62849162011173, + "grad_norm": 1.6135245561599731, + "learning_rate": 2.858510242085661e-05, + "loss": 0.1391, + "step": 12755 + }, + { + "epoch": 35.63128491620112, + "grad_norm": 0.501308023929596, + "learning_rate": 2.8584729981378028e-05, + "loss": 0.128, + "step": 12756 + }, + { + "epoch": 35.634078212290504, + "grad_norm": 0.4696389138698578, + "learning_rate": 2.858435754189944e-05, + "loss": 0.1287, + "step": 12757 + }, + { + "epoch": 35.63687150837989, + "grad_norm": 0.38172051310539246, + "learning_rate": 2.8583985102420857e-05, + "loss": 0.1371, + "step": 12758 + }, + { + "epoch": 35.63966480446928, + "grad_norm": 0.5264004468917847, + "learning_rate": 2.858361266294227e-05, + "loss": 0.1166, + "step": 12759 + }, + { + "epoch": 35.642458100558656, + "grad_norm": 0.9633996486663818, + "learning_rate": 2.8583240223463687e-05, + "loss": 0.1231, + "step": 12760 + }, + { + "epoch": 35.64525139664804, + "grad_norm": 0.5824756026268005, + "learning_rate": 2.8582867783985104e-05, + "loss": 0.1343, + "step": 12761 + }, + { + "epoch": 35.64804469273743, + "grad_norm": 0.37935250997543335, + "learning_rate": 2.858249534450652e-05, + "loss": 0.1015, + "step": 12762 + }, + { + "epoch": 35.650837988826815, + "grad_norm": 0.8413099646568298, + "learning_rate": 2.8582122905027937e-05, + "loss": 0.1124, + "step": 12763 + }, + { + "epoch": 35.6536312849162, + "grad_norm": 0.5643170475959778, + "learning_rate": 2.8581750465549346e-05, + "loss": 0.1118, + "step": 12764 + }, + { + "epoch": 35.65642458100559, + "grad_norm": 0.4746699035167694, + "learning_rate": 2.8581378026070763e-05, + "loss": 0.1057, + "step": 12765 + }, + { + "epoch": 35.659217877094974, + "grad_norm": 0.5255001187324524, + "learning_rate": 2.858100558659218e-05, + "loss": 0.0925, + "step": 12766 + }, + { + "epoch": 35.66201117318436, + "grad_norm": 0.535237729549408, + "learning_rate": 2.8580633147113596e-05, + "loss": 0.1243, + "step": 12767 + }, + { + "epoch": 35.66480446927374, + "grad_norm": 0.8225684762001038, + "learning_rate": 2.858026070763501e-05, + "loss": 0.0882, + "step": 12768 + }, + { + "epoch": 35.667597765363126, + "grad_norm": 0.626577615737915, + "learning_rate": 2.8579888268156426e-05, + "loss": 0.1124, + "step": 12769 + }, + { + "epoch": 35.67039106145251, + "grad_norm": 0.8938798308372498, + "learning_rate": 2.857951582867784e-05, + "loss": 0.0913, + "step": 12770 + }, + { + "epoch": 35.6731843575419, + "grad_norm": 0.9357428550720215, + "learning_rate": 2.8579143389199255e-05, + "loss": 0.0958, + "step": 12771 + }, + { + "epoch": 35.675977653631286, + "grad_norm": 0.6729956269264221, + "learning_rate": 2.8578770949720672e-05, + "loss": 0.0878, + "step": 12772 + }, + { + "epoch": 35.67877094972067, + "grad_norm": 0.7665488719940186, + "learning_rate": 2.8578398510242085e-05, + "loss": 0.0859, + "step": 12773 + }, + { + "epoch": 35.68156424581006, + "grad_norm": 0.5552424788475037, + "learning_rate": 2.85780260707635e-05, + "loss": 0.0756, + "step": 12774 + }, + { + "epoch": 35.684357541899445, + "grad_norm": 0.7426563501358032, + "learning_rate": 2.8577653631284918e-05, + "loss": 0.1021, + "step": 12775 + }, + { + "epoch": 35.687150837988824, + "grad_norm": 0.735910177230835, + "learning_rate": 2.8577281191806335e-05, + "loss": 0.0715, + "step": 12776 + }, + { + "epoch": 35.68994413407821, + "grad_norm": 0.6882702112197876, + "learning_rate": 2.8576908752327744e-05, + "loss": 0.0919, + "step": 12777 + }, + { + "epoch": 35.6927374301676, + "grad_norm": 0.8869224190711975, + "learning_rate": 2.857653631284916e-05, + "loss": 0.0763, + "step": 12778 + }, + { + "epoch": 35.69553072625698, + "grad_norm": 0.7544671297073364, + "learning_rate": 2.8576163873370578e-05, + "loss": 0.101, + "step": 12779 + }, + { + "epoch": 35.69832402234637, + "grad_norm": 1.2525566816329956, + "learning_rate": 2.8575791433891994e-05, + "loss": 0.0973, + "step": 12780 + }, + { + "epoch": 35.701117318435756, + "grad_norm": 0.5222823023796082, + "learning_rate": 2.857541899441341e-05, + "loss": 0.1402, + "step": 12781 + }, + { + "epoch": 35.70391061452514, + "grad_norm": 0.46009156107902527, + "learning_rate": 2.8575046554934824e-05, + "loss": 0.1331, + "step": 12782 + }, + { + "epoch": 35.70670391061452, + "grad_norm": 0.3739299774169922, + "learning_rate": 2.8574674115456237e-05, + "loss": 0.1211, + "step": 12783 + }, + { + "epoch": 35.70949720670391, + "grad_norm": 1.033735990524292, + "learning_rate": 2.8574301675977653e-05, + "loss": 0.1095, + "step": 12784 + }, + { + "epoch": 35.712290502793294, + "grad_norm": 0.5259512066841125, + "learning_rate": 2.857392923649907e-05, + "loss": 0.1233, + "step": 12785 + }, + { + "epoch": 35.71508379888268, + "grad_norm": 0.6302036046981812, + "learning_rate": 2.8573556797020483e-05, + "loss": 0.0901, + "step": 12786 + }, + { + "epoch": 35.71787709497207, + "grad_norm": 0.5368133783340454, + "learning_rate": 2.85731843575419e-05, + "loss": 0.1276, + "step": 12787 + }, + { + "epoch": 35.720670391061454, + "grad_norm": 0.5369441509246826, + "learning_rate": 2.8572811918063316e-05, + "loss": 0.1167, + "step": 12788 + }, + { + "epoch": 35.72346368715084, + "grad_norm": 1.3022338151931763, + "learning_rate": 2.8572439478584733e-05, + "loss": 0.1221, + "step": 12789 + }, + { + "epoch": 35.726256983240226, + "grad_norm": 0.4968683421611786, + "learning_rate": 2.8572067039106146e-05, + "loss": 0.115, + "step": 12790 + }, + { + "epoch": 35.729050279329606, + "grad_norm": 0.7155823707580566, + "learning_rate": 2.857169459962756e-05, + "loss": 0.1163, + "step": 12791 + }, + { + "epoch": 35.73184357541899, + "grad_norm": 1.0936763286590576, + "learning_rate": 2.8571322160148976e-05, + "loss": 0.1257, + "step": 12792 + }, + { + "epoch": 35.73463687150838, + "grad_norm": 0.47358986735343933, + "learning_rate": 2.8570949720670392e-05, + "loss": 0.1103, + "step": 12793 + }, + { + "epoch": 35.737430167597765, + "grad_norm": 0.7499448657035828, + "learning_rate": 2.857057728119181e-05, + "loss": 0.0953, + "step": 12794 + }, + { + "epoch": 35.74022346368715, + "grad_norm": 1.7325637340545654, + "learning_rate": 2.8570204841713222e-05, + "loss": 0.1029, + "step": 12795 + }, + { + "epoch": 35.74301675977654, + "grad_norm": 0.5147425532341003, + "learning_rate": 2.856983240223464e-05, + "loss": 0.1181, + "step": 12796 + }, + { + "epoch": 35.745810055865924, + "grad_norm": 0.8368345499038696, + "learning_rate": 2.856945996275605e-05, + "loss": 0.1023, + "step": 12797 + }, + { + "epoch": 35.74860335195531, + "grad_norm": 0.5198988318443298, + "learning_rate": 2.8569087523277468e-05, + "loss": 0.1018, + "step": 12798 + }, + { + "epoch": 35.75139664804469, + "grad_norm": 0.7435005307197571, + "learning_rate": 2.8568715083798885e-05, + "loss": 0.0956, + "step": 12799 + }, + { + "epoch": 35.754189944134076, + "grad_norm": 0.5774767994880676, + "learning_rate": 2.8568342644320298e-05, + "loss": 0.0815, + "step": 12800 + }, + { + "epoch": 35.75698324022346, + "grad_norm": 1.2038090229034424, + "learning_rate": 2.8567970204841714e-05, + "loss": 0.1053, + "step": 12801 + }, + { + "epoch": 35.75977653631285, + "grad_norm": 1.3762905597686768, + "learning_rate": 2.856759776536313e-05, + "loss": 0.0894, + "step": 12802 + }, + { + "epoch": 35.762569832402235, + "grad_norm": 2.9490225315093994, + "learning_rate": 2.8567225325884544e-05, + "loss": 0.1053, + "step": 12803 + }, + { + "epoch": 35.76536312849162, + "grad_norm": 0.6014654636383057, + "learning_rate": 2.8566852886405957e-05, + "loss": 0.0962, + "step": 12804 + }, + { + "epoch": 35.76815642458101, + "grad_norm": 0.9202593564987183, + "learning_rate": 2.8566480446927374e-05, + "loss": 0.0947, + "step": 12805 + }, + { + "epoch": 35.770949720670394, + "grad_norm": 0.7506694793701172, + "learning_rate": 2.856610800744879e-05, + "loss": 0.1463, + "step": 12806 + }, + { + "epoch": 35.773743016759774, + "grad_norm": 0.4424915313720703, + "learning_rate": 2.8565735567970207e-05, + "loss": 0.1518, + "step": 12807 + }, + { + "epoch": 35.77653631284916, + "grad_norm": 0.6723545789718628, + "learning_rate": 2.8565363128491623e-05, + "loss": 0.109, + "step": 12808 + }, + { + "epoch": 35.779329608938546, + "grad_norm": 0.553681492805481, + "learning_rate": 2.8564990689013036e-05, + "loss": 0.1676, + "step": 12809 + }, + { + "epoch": 35.78212290502793, + "grad_norm": 1.8954178094863892, + "learning_rate": 2.856461824953445e-05, + "loss": 0.1213, + "step": 12810 + }, + { + "epoch": 35.78491620111732, + "grad_norm": 0.6780610084533691, + "learning_rate": 2.8564245810055866e-05, + "loss": 0.136, + "step": 12811 + }, + { + "epoch": 35.787709497206706, + "grad_norm": 0.48458331823349, + "learning_rate": 2.8563873370577283e-05, + "loss": 0.1088, + "step": 12812 + }, + { + "epoch": 35.79050279329609, + "grad_norm": 0.4563122093677521, + "learning_rate": 2.8563500931098696e-05, + "loss": 0.1221, + "step": 12813 + }, + { + "epoch": 35.79329608938548, + "grad_norm": 0.3453752398490906, + "learning_rate": 2.8563128491620112e-05, + "loss": 0.0956, + "step": 12814 + }, + { + "epoch": 35.79608938547486, + "grad_norm": 0.731473982334137, + "learning_rate": 2.856275605214153e-05, + "loss": 0.1235, + "step": 12815 + }, + { + "epoch": 35.798882681564244, + "grad_norm": 1.261697769165039, + "learning_rate": 2.8562383612662945e-05, + "loss": 0.1098, + "step": 12816 + }, + { + "epoch": 35.80167597765363, + "grad_norm": 0.6914653778076172, + "learning_rate": 2.856201117318436e-05, + "loss": 0.1088, + "step": 12817 + }, + { + "epoch": 35.80446927374302, + "grad_norm": 0.5401456356048584, + "learning_rate": 2.856163873370577e-05, + "loss": 0.0947, + "step": 12818 + }, + { + "epoch": 35.8072625698324, + "grad_norm": 0.883546769618988, + "learning_rate": 2.8561266294227188e-05, + "loss": 0.0928, + "step": 12819 + }, + { + "epoch": 35.81005586592179, + "grad_norm": 0.5521100759506226, + "learning_rate": 2.8560893854748605e-05, + "loss": 0.1003, + "step": 12820 + }, + { + "epoch": 35.812849162011176, + "grad_norm": 0.6921665668487549, + "learning_rate": 2.856052141527002e-05, + "loss": 0.1081, + "step": 12821 + }, + { + "epoch": 35.815642458100555, + "grad_norm": 1.00566565990448, + "learning_rate": 2.8560148975791434e-05, + "loss": 0.1197, + "step": 12822 + }, + { + "epoch": 35.81843575418994, + "grad_norm": 0.7344270944595337, + "learning_rate": 2.8559776536312848e-05, + "loss": 0.1183, + "step": 12823 + }, + { + "epoch": 35.82122905027933, + "grad_norm": 0.8061746954917908, + "learning_rate": 2.8559404096834264e-05, + "loss": 0.09, + "step": 12824 + }, + { + "epoch": 35.824022346368714, + "grad_norm": 0.9014660716056824, + "learning_rate": 2.855903165735568e-05, + "loss": 0.0837, + "step": 12825 + }, + { + "epoch": 35.8268156424581, + "grad_norm": 0.8619147539138794, + "learning_rate": 2.8558659217877097e-05, + "loss": 0.1044, + "step": 12826 + }, + { + "epoch": 35.82960893854749, + "grad_norm": 1.5132430791854858, + "learning_rate": 2.855828677839851e-05, + "loss": 0.0857, + "step": 12827 + }, + { + "epoch": 35.832402234636874, + "grad_norm": 0.6638464331626892, + "learning_rate": 2.8557914338919927e-05, + "loss": 0.0956, + "step": 12828 + }, + { + "epoch": 35.83519553072626, + "grad_norm": 0.861052393913269, + "learning_rate": 2.8557541899441343e-05, + "loss": 0.1007, + "step": 12829 + }, + { + "epoch": 35.83798882681564, + "grad_norm": 1.1928355693817139, + "learning_rate": 2.8557169459962756e-05, + "loss": 0.108, + "step": 12830 + }, + { + "epoch": 35.840782122905026, + "grad_norm": 0.977036714553833, + "learning_rate": 2.855679702048417e-05, + "loss": 0.1762, + "step": 12831 + }, + { + "epoch": 35.84357541899441, + "grad_norm": 0.6050860285758972, + "learning_rate": 2.8556424581005586e-05, + "loss": 0.15, + "step": 12832 + }, + { + "epoch": 35.8463687150838, + "grad_norm": 0.7377099394798279, + "learning_rate": 2.8556052141527003e-05, + "loss": 0.1222, + "step": 12833 + }, + { + "epoch": 35.849162011173185, + "grad_norm": 0.5261802077293396, + "learning_rate": 2.855567970204842e-05, + "loss": 0.1326, + "step": 12834 + }, + { + "epoch": 35.85195530726257, + "grad_norm": 0.4714972972869873, + "learning_rate": 2.8555307262569836e-05, + "loss": 0.1246, + "step": 12835 + }, + { + "epoch": 35.85474860335196, + "grad_norm": 0.7476651668548584, + "learning_rate": 2.855493482309125e-05, + "loss": 0.1224, + "step": 12836 + }, + { + "epoch": 35.857541899441344, + "grad_norm": 0.5095197558403015, + "learning_rate": 2.8554562383612662e-05, + "loss": 0.1298, + "step": 12837 + }, + { + "epoch": 35.86033519553072, + "grad_norm": 0.7281497120857239, + "learning_rate": 2.855418994413408e-05, + "loss": 0.1114, + "step": 12838 + }, + { + "epoch": 35.86312849162011, + "grad_norm": 1.9302414655685425, + "learning_rate": 2.8553817504655495e-05, + "loss": 0.1184, + "step": 12839 + }, + { + "epoch": 35.865921787709496, + "grad_norm": 0.7788382768630981, + "learning_rate": 2.8553445065176908e-05, + "loss": 0.1311, + "step": 12840 + }, + { + "epoch": 35.86871508379888, + "grad_norm": 0.502640426158905, + "learning_rate": 2.8553072625698325e-05, + "loss": 0.1292, + "step": 12841 + }, + { + "epoch": 35.87150837988827, + "grad_norm": 0.7009037733078003, + "learning_rate": 2.855270018621974e-05, + "loss": 0.1, + "step": 12842 + }, + { + "epoch": 35.874301675977655, + "grad_norm": 0.5375313758850098, + "learning_rate": 2.8552327746741155e-05, + "loss": 0.1265, + "step": 12843 + }, + { + "epoch": 35.87709497206704, + "grad_norm": 0.9210308790206909, + "learning_rate": 2.855195530726257e-05, + "loss": 0.1128, + "step": 12844 + }, + { + "epoch": 35.87988826815643, + "grad_norm": 0.477958619594574, + "learning_rate": 2.8551582867783984e-05, + "loss": 0.1043, + "step": 12845 + }, + { + "epoch": 35.88268156424581, + "grad_norm": 0.6403813362121582, + "learning_rate": 2.85512104283054e-05, + "loss": 0.091, + "step": 12846 + }, + { + "epoch": 35.885474860335194, + "grad_norm": 0.7379566431045532, + "learning_rate": 2.8550837988826817e-05, + "loss": 0.1048, + "step": 12847 + }, + { + "epoch": 35.88826815642458, + "grad_norm": 0.7242845892906189, + "learning_rate": 2.8550465549348234e-05, + "loss": 0.1153, + "step": 12848 + }, + { + "epoch": 35.891061452513966, + "grad_norm": 0.6048550605773926, + "learning_rate": 2.8550093109869647e-05, + "loss": 0.093, + "step": 12849 + }, + { + "epoch": 35.89385474860335, + "grad_norm": 0.6468750238418579, + "learning_rate": 2.854972067039106e-05, + "loss": 0.098, + "step": 12850 + }, + { + "epoch": 35.89664804469274, + "grad_norm": 1.5503987073898315, + "learning_rate": 2.8549348230912477e-05, + "loss": 0.0998, + "step": 12851 + }, + { + "epoch": 35.899441340782126, + "grad_norm": 1.4842745065689087, + "learning_rate": 2.8548975791433893e-05, + "loss": 0.0988, + "step": 12852 + }, + { + "epoch": 35.90223463687151, + "grad_norm": 2.1299753189086914, + "learning_rate": 2.854860335195531e-05, + "loss": 0.0935, + "step": 12853 + }, + { + "epoch": 35.90502793296089, + "grad_norm": 0.7938839793205261, + "learning_rate": 2.8548230912476723e-05, + "loss": 0.0868, + "step": 12854 + }, + { + "epoch": 35.90782122905028, + "grad_norm": 1.2684820890426636, + "learning_rate": 2.854785847299814e-05, + "loss": 0.1238, + "step": 12855 + }, + { + "epoch": 35.910614525139664, + "grad_norm": 0.4250658452510834, + "learning_rate": 2.8547486033519556e-05, + "loss": 0.1297, + "step": 12856 + }, + { + "epoch": 35.91340782122905, + "grad_norm": 0.6026238203048706, + "learning_rate": 2.854711359404097e-05, + "loss": 0.1522, + "step": 12857 + }, + { + "epoch": 35.91620111731844, + "grad_norm": 0.7467273473739624, + "learning_rate": 2.8546741154562382e-05, + "loss": 0.138, + "step": 12858 + }, + { + "epoch": 35.91899441340782, + "grad_norm": 0.6978008151054382, + "learning_rate": 2.85463687150838e-05, + "loss": 0.1347, + "step": 12859 + }, + { + "epoch": 35.92178770949721, + "grad_norm": 0.6602522134780884, + "learning_rate": 2.8545996275605215e-05, + "loss": 0.1405, + "step": 12860 + }, + { + "epoch": 35.92458100558659, + "grad_norm": 0.40358373522758484, + "learning_rate": 2.8545623836126632e-05, + "loss": 0.1158, + "step": 12861 + }, + { + "epoch": 35.927374301675975, + "grad_norm": 0.47029733657836914, + "learning_rate": 2.8545251396648045e-05, + "loss": 0.1263, + "step": 12862 + }, + { + "epoch": 35.93016759776536, + "grad_norm": 0.7282479405403137, + "learning_rate": 2.8544878957169458e-05, + "loss": 0.1257, + "step": 12863 + }, + { + "epoch": 35.93296089385475, + "grad_norm": 0.4255654811859131, + "learning_rate": 2.8544506517690875e-05, + "loss": 0.1033, + "step": 12864 + }, + { + "epoch": 35.935754189944134, + "grad_norm": 0.5924680233001709, + "learning_rate": 2.854413407821229e-05, + "loss": 0.1242, + "step": 12865 + }, + { + "epoch": 35.93854748603352, + "grad_norm": 0.5686397552490234, + "learning_rate": 2.8543761638733708e-05, + "loss": 0.1305, + "step": 12866 + }, + { + "epoch": 35.94134078212291, + "grad_norm": 0.4644242227077484, + "learning_rate": 2.854338919925512e-05, + "loss": 0.0985, + "step": 12867 + }, + { + "epoch": 35.944134078212294, + "grad_norm": 0.7192767262458801, + "learning_rate": 2.8543016759776537e-05, + "loss": 0.0965, + "step": 12868 + }, + { + "epoch": 35.94692737430167, + "grad_norm": 0.8476279377937317, + "learning_rate": 2.8542644320297954e-05, + "loss": 0.1009, + "step": 12869 + }, + { + "epoch": 35.94972067039106, + "grad_norm": 0.7232741713523865, + "learning_rate": 2.8542271880819367e-05, + "loss": 0.1075, + "step": 12870 + }, + { + "epoch": 35.952513966480446, + "grad_norm": 0.4213355779647827, + "learning_rate": 2.854189944134078e-05, + "loss": 0.0973, + "step": 12871 + }, + { + "epoch": 35.95530726256983, + "grad_norm": 0.5588488578796387, + "learning_rate": 2.8541527001862197e-05, + "loss": 0.1027, + "step": 12872 + }, + { + "epoch": 35.95810055865922, + "grad_norm": 1.0680639743804932, + "learning_rate": 2.8541154562383613e-05, + "loss": 0.102, + "step": 12873 + }, + { + "epoch": 35.960893854748605, + "grad_norm": 0.6011932492256165, + "learning_rate": 2.854078212290503e-05, + "loss": 0.0924, + "step": 12874 + }, + { + "epoch": 35.96368715083799, + "grad_norm": 0.5648177862167358, + "learning_rate": 2.8540409683426446e-05, + "loss": 0.1097, + "step": 12875 + }, + { + "epoch": 35.96648044692738, + "grad_norm": 0.5351366400718689, + "learning_rate": 2.854003724394786e-05, + "loss": 0.077, + "step": 12876 + }, + { + "epoch": 35.96927374301676, + "grad_norm": 1.2414422035217285, + "learning_rate": 2.8539664804469273e-05, + "loss": 0.1098, + "step": 12877 + }, + { + "epoch": 35.97206703910614, + "grad_norm": 0.7010636925697327, + "learning_rate": 2.853929236499069e-05, + "loss": 0.0875, + "step": 12878 + }, + { + "epoch": 35.97486033519553, + "grad_norm": 0.877781093120575, + "learning_rate": 2.8538919925512106e-05, + "loss": 0.0852, + "step": 12879 + }, + { + "epoch": 35.977653631284916, + "grad_norm": 1.7339993715286255, + "learning_rate": 2.853854748603352e-05, + "loss": 0.12, + "step": 12880 + }, + { + "epoch": 35.9804469273743, + "grad_norm": 0.9633666276931763, + "learning_rate": 2.8538175046554935e-05, + "loss": 0.1415, + "step": 12881 + }, + { + "epoch": 35.98324022346369, + "grad_norm": 0.5213769674301147, + "learning_rate": 2.8537802607076352e-05, + "loss": 0.1367, + "step": 12882 + }, + { + "epoch": 35.986033519553075, + "grad_norm": 1.1626920700073242, + "learning_rate": 2.8537430167597765e-05, + "loss": 0.1061, + "step": 12883 + }, + { + "epoch": 35.98882681564246, + "grad_norm": 0.5910606384277344, + "learning_rate": 2.853705772811918e-05, + "loss": 0.1173, + "step": 12884 + }, + { + "epoch": 35.99162011173184, + "grad_norm": 0.8721029162406921, + "learning_rate": 2.8536685288640595e-05, + "loss": 0.0944, + "step": 12885 + }, + { + "epoch": 35.99441340782123, + "grad_norm": 0.5838152766227722, + "learning_rate": 2.853631284916201e-05, + "loss": 0.0921, + "step": 12886 + }, + { + "epoch": 35.997206703910614, + "grad_norm": 0.6435158848762512, + "learning_rate": 2.8535940409683428e-05, + "loss": 0.0985, + "step": 12887 + }, + { + "epoch": 36.0, + "grad_norm": 2.2878711223602295, + "learning_rate": 2.8535567970204844e-05, + "loss": 0.1276, + "step": 12888 + }, + { + "epoch": 36.002793296089386, + "grad_norm": 0.648641049861908, + "learning_rate": 2.8535195530726258e-05, + "loss": 0.1537, + "step": 12889 + }, + { + "epoch": 36.00558659217877, + "grad_norm": 0.8544210195541382, + "learning_rate": 2.853482309124767e-05, + "loss": 0.1192, + "step": 12890 + }, + { + "epoch": 36.00837988826816, + "grad_norm": 1.0910847187042236, + "learning_rate": 2.8534450651769087e-05, + "loss": 0.1507, + "step": 12891 + }, + { + "epoch": 36.01117318435754, + "grad_norm": 0.5895769000053406, + "learning_rate": 2.8534078212290504e-05, + "loss": 0.1154, + "step": 12892 + }, + { + "epoch": 36.013966480446925, + "grad_norm": 0.9862508773803711, + "learning_rate": 2.853370577281192e-05, + "loss": 0.114, + "step": 12893 + }, + { + "epoch": 36.01675977653631, + "grad_norm": 0.6444692611694336, + "learning_rate": 2.8533333333333333e-05, + "loss": 0.1065, + "step": 12894 + }, + { + "epoch": 36.0195530726257, + "grad_norm": 1.1897910833358765, + "learning_rate": 2.853296089385475e-05, + "loss": 0.0963, + "step": 12895 + }, + { + "epoch": 36.022346368715084, + "grad_norm": 1.2072213888168335, + "learning_rate": 2.8532588454376167e-05, + "loss": 0.1056, + "step": 12896 + }, + { + "epoch": 36.02513966480447, + "grad_norm": 0.370492160320282, + "learning_rate": 2.853221601489758e-05, + "loss": 0.1056, + "step": 12897 + }, + { + "epoch": 36.02793296089386, + "grad_norm": 0.4753597378730774, + "learning_rate": 2.8531843575418993e-05, + "loss": 0.0981, + "step": 12898 + }, + { + "epoch": 36.03072625698324, + "grad_norm": 0.8144912719726562, + "learning_rate": 2.853147113594041e-05, + "loss": 0.1023, + "step": 12899 + }, + { + "epoch": 36.03351955307262, + "grad_norm": 0.6275578141212463, + "learning_rate": 2.8531098696461826e-05, + "loss": 0.1094, + "step": 12900 + }, + { + "epoch": 36.03631284916201, + "grad_norm": 0.7182799577713013, + "learning_rate": 2.8530726256983242e-05, + "loss": 0.0955, + "step": 12901 + }, + { + "epoch": 36.039106145251395, + "grad_norm": 0.6165883541107178, + "learning_rate": 2.853035381750466e-05, + "loss": 0.1014, + "step": 12902 + }, + { + "epoch": 36.04189944134078, + "grad_norm": 0.5230837464332581, + "learning_rate": 2.852998137802607e-05, + "loss": 0.084, + "step": 12903 + }, + { + "epoch": 36.04469273743017, + "grad_norm": 0.8907346725463867, + "learning_rate": 2.8529608938547485e-05, + "loss": 0.086, + "step": 12904 + }, + { + "epoch": 36.047486033519554, + "grad_norm": 0.7332752346992493, + "learning_rate": 2.8529236499068902e-05, + "loss": 0.0941, + "step": 12905 + }, + { + "epoch": 36.05027932960894, + "grad_norm": 0.5982913374900818, + "learning_rate": 2.852886405959032e-05, + "loss": 0.0915, + "step": 12906 + }, + { + "epoch": 36.05307262569833, + "grad_norm": 0.5111281871795654, + "learning_rate": 2.852849162011173e-05, + "loss": 0.0752, + "step": 12907 + }, + { + "epoch": 36.055865921787706, + "grad_norm": 0.6880156397819519, + "learning_rate": 2.8528119180633148e-05, + "loss": 0.0837, + "step": 12908 + }, + { + "epoch": 36.05865921787709, + "grad_norm": 0.6235281825065613, + "learning_rate": 2.8527746741154565e-05, + "loss": 0.0833, + "step": 12909 + }, + { + "epoch": 36.06145251396648, + "grad_norm": 1.0251816511154175, + "learning_rate": 2.8527374301675978e-05, + "loss": 0.0877, + "step": 12910 + }, + { + "epoch": 36.064245810055866, + "grad_norm": 0.6626482605934143, + "learning_rate": 2.8527001862197394e-05, + "loss": 0.0739, + "step": 12911 + }, + { + "epoch": 36.06703910614525, + "grad_norm": 1.206688404083252, + "learning_rate": 2.8526629422718807e-05, + "loss": 0.0922, + "step": 12912 + }, + { + "epoch": 36.06983240223464, + "grad_norm": 0.8172720074653625, + "learning_rate": 2.8526256983240224e-05, + "loss": 0.0905, + "step": 12913 + }, + { + "epoch": 36.072625698324025, + "grad_norm": 0.8511542081832886, + "learning_rate": 2.852588454376164e-05, + "loss": 0.1296, + "step": 12914 + }, + { + "epoch": 36.07541899441341, + "grad_norm": 0.42571499943733215, + "learning_rate": 2.8525512104283057e-05, + "loss": 0.1351, + "step": 12915 + }, + { + "epoch": 36.07821229050279, + "grad_norm": 0.33394360542297363, + "learning_rate": 2.852513966480447e-05, + "loss": 0.1108, + "step": 12916 + }, + { + "epoch": 36.08100558659218, + "grad_norm": 0.35417550802230835, + "learning_rate": 2.8524767225325883e-05, + "loss": 0.1138, + "step": 12917 + }, + { + "epoch": 36.08379888268156, + "grad_norm": 0.6474469304084778, + "learning_rate": 2.85243947858473e-05, + "loss": 0.1093, + "step": 12918 + }, + { + "epoch": 36.08659217877095, + "grad_norm": 0.5933594703674316, + "learning_rate": 2.8524022346368716e-05, + "loss": 0.1263, + "step": 12919 + }, + { + "epoch": 36.089385474860336, + "grad_norm": 0.7326763272285461, + "learning_rate": 2.8523649906890133e-05, + "loss": 0.1049, + "step": 12920 + }, + { + "epoch": 36.09217877094972, + "grad_norm": 1.7910077571868896, + "learning_rate": 2.8523277467411546e-05, + "loss": 0.1245, + "step": 12921 + }, + { + "epoch": 36.09497206703911, + "grad_norm": 0.49482014775276184, + "learning_rate": 2.8522905027932963e-05, + "loss": 0.1024, + "step": 12922 + }, + { + "epoch": 36.097765363128495, + "grad_norm": 0.9543793797492981, + "learning_rate": 2.8522532588454376e-05, + "loss": 0.1005, + "step": 12923 + }, + { + "epoch": 36.100558659217874, + "grad_norm": 0.6018519997596741, + "learning_rate": 2.8522160148975792e-05, + "loss": 0.102, + "step": 12924 + }, + { + "epoch": 36.10335195530726, + "grad_norm": 0.4851997494697571, + "learning_rate": 2.8521787709497205e-05, + "loss": 0.1024, + "step": 12925 + }, + { + "epoch": 36.10614525139665, + "grad_norm": 0.456601619720459, + "learning_rate": 2.8521415270018622e-05, + "loss": 0.0895, + "step": 12926 + }, + { + "epoch": 36.108938547486034, + "grad_norm": 1.1100542545318604, + "learning_rate": 2.852104283054004e-05, + "loss": 0.09, + "step": 12927 + }, + { + "epoch": 36.11173184357542, + "grad_norm": 0.5304964184761047, + "learning_rate": 2.8520670391061455e-05, + "loss": 0.087, + "step": 12928 + }, + { + "epoch": 36.114525139664806, + "grad_norm": 0.5163066387176514, + "learning_rate": 2.852029795158287e-05, + "loss": 0.0924, + "step": 12929 + }, + { + "epoch": 36.11731843575419, + "grad_norm": 0.886943519115448, + "learning_rate": 2.851992551210428e-05, + "loss": 0.0931, + "step": 12930 + }, + { + "epoch": 36.12011173184357, + "grad_norm": 0.9616464376449585, + "learning_rate": 2.8519553072625698e-05, + "loss": 0.0798, + "step": 12931 + }, + { + "epoch": 36.12290502793296, + "grad_norm": 0.5473931431770325, + "learning_rate": 2.8519180633147114e-05, + "loss": 0.0989, + "step": 12932 + }, + { + "epoch": 36.125698324022345, + "grad_norm": 0.6164246201515198, + "learning_rate": 2.851880819366853e-05, + "loss": 0.0795, + "step": 12933 + }, + { + "epoch": 36.12849162011173, + "grad_norm": 0.5471236109733582, + "learning_rate": 2.8518435754189944e-05, + "loss": 0.0795, + "step": 12934 + }, + { + "epoch": 36.13128491620112, + "grad_norm": 0.9442342519760132, + "learning_rate": 2.851806331471136e-05, + "loss": 0.0996, + "step": 12935 + }, + { + "epoch": 36.134078212290504, + "grad_norm": 0.5754980444908142, + "learning_rate": 2.8517690875232777e-05, + "loss": 0.0824, + "step": 12936 + }, + { + "epoch": 36.13687150837989, + "grad_norm": 1.005746841430664, + "learning_rate": 2.851731843575419e-05, + "loss": 0.081, + "step": 12937 + }, + { + "epoch": 36.13966480446928, + "grad_norm": 1.3339556455612183, + "learning_rate": 2.8516945996275607e-05, + "loss": 0.1012, + "step": 12938 + }, + { + "epoch": 36.142458100558656, + "grad_norm": 0.8191988468170166, + "learning_rate": 2.851657355679702e-05, + "loss": 0.1449, + "step": 12939 + }, + { + "epoch": 36.14525139664804, + "grad_norm": 0.4437064230442047, + "learning_rate": 2.8516201117318436e-05, + "loss": 0.1351, + "step": 12940 + }, + { + "epoch": 36.14804469273743, + "grad_norm": 0.6032711863517761, + "learning_rate": 2.8515828677839853e-05, + "loss": 0.1128, + "step": 12941 + }, + { + "epoch": 36.150837988826815, + "grad_norm": 0.35685718059539795, + "learning_rate": 2.851545623836127e-05, + "loss": 0.1149, + "step": 12942 + }, + { + "epoch": 36.1536312849162, + "grad_norm": 1.5192835330963135, + "learning_rate": 2.851508379888268e-05, + "loss": 0.1166, + "step": 12943 + }, + { + "epoch": 36.15642458100559, + "grad_norm": 0.761411726474762, + "learning_rate": 2.8514711359404096e-05, + "loss": 0.0982, + "step": 12944 + }, + { + "epoch": 36.159217877094974, + "grad_norm": 0.5384179949760437, + "learning_rate": 2.8514338919925512e-05, + "loss": 0.1137, + "step": 12945 + }, + { + "epoch": 36.16201117318436, + "grad_norm": 0.5325331687927246, + "learning_rate": 2.851396648044693e-05, + "loss": 0.0953, + "step": 12946 + }, + { + "epoch": 36.16480446927374, + "grad_norm": 0.6533609628677368, + "learning_rate": 2.8513594040968345e-05, + "loss": 0.0855, + "step": 12947 + }, + { + "epoch": 36.167597765363126, + "grad_norm": 0.6269599795341492, + "learning_rate": 2.851322160148976e-05, + "loss": 0.1292, + "step": 12948 + }, + { + "epoch": 36.17039106145251, + "grad_norm": 0.6166209578514099, + "learning_rate": 2.8512849162011175e-05, + "loss": 0.0981, + "step": 12949 + }, + { + "epoch": 36.1731843575419, + "grad_norm": 1.446237325668335, + "learning_rate": 2.8512476722532588e-05, + "loss": 0.0945, + "step": 12950 + }, + { + "epoch": 36.175977653631286, + "grad_norm": 0.8645244836807251, + "learning_rate": 2.8512104283054005e-05, + "loss": 0.1015, + "step": 12951 + }, + { + "epoch": 36.17877094972067, + "grad_norm": 0.5276140570640564, + "learning_rate": 2.8511731843575418e-05, + "loss": 0.0883, + "step": 12952 + }, + { + "epoch": 36.18156424581006, + "grad_norm": 1.7483326196670532, + "learning_rate": 2.8511359404096835e-05, + "loss": 0.0882, + "step": 12953 + }, + { + "epoch": 36.184357541899445, + "grad_norm": 0.6467254161834717, + "learning_rate": 2.851098696461825e-05, + "loss": 0.0678, + "step": 12954 + }, + { + "epoch": 36.187150837988824, + "grad_norm": 0.5865139961242676, + "learning_rate": 2.8510614525139668e-05, + "loss": 0.0844, + "step": 12955 + }, + { + "epoch": 36.18994413407821, + "grad_norm": 0.4748874008655548, + "learning_rate": 2.851024208566108e-05, + "loss": 0.0694, + "step": 12956 + }, + { + "epoch": 36.1927374301676, + "grad_norm": 0.8655739426612854, + "learning_rate": 2.8509869646182494e-05, + "loss": 0.0855, + "step": 12957 + }, + { + "epoch": 36.19553072625698, + "grad_norm": 0.5284507274627686, + "learning_rate": 2.850949720670391e-05, + "loss": 0.0781, + "step": 12958 + }, + { + "epoch": 36.19832402234637, + "grad_norm": 0.7051674723625183, + "learning_rate": 2.8509124767225327e-05, + "loss": 0.0827, + "step": 12959 + }, + { + "epoch": 36.201117318435756, + "grad_norm": 1.7244144678115845, + "learning_rate": 2.8508752327746743e-05, + "loss": 0.0596, + "step": 12960 + }, + { + "epoch": 36.20391061452514, + "grad_norm": 0.9110144376754761, + "learning_rate": 2.8508379888268157e-05, + "loss": 0.126, + "step": 12961 + }, + { + "epoch": 36.20670391061452, + "grad_norm": 0.5209698677062988, + "learning_rate": 2.8508007448789573e-05, + "loss": 0.0718, + "step": 12962 + }, + { + "epoch": 36.20949720670391, + "grad_norm": 0.6912801861763, + "learning_rate": 2.8507635009310986e-05, + "loss": 0.0896, + "step": 12963 + }, + { + "epoch": 36.212290502793294, + "grad_norm": 0.6076659560203552, + "learning_rate": 2.8507262569832403e-05, + "loss": 0.1245, + "step": 12964 + }, + { + "epoch": 36.21508379888268, + "grad_norm": 0.5359377861022949, + "learning_rate": 2.8506890130353816e-05, + "loss": 0.1299, + "step": 12965 + }, + { + "epoch": 36.21787709497207, + "grad_norm": 0.5190545320510864, + "learning_rate": 2.8506517690875233e-05, + "loss": 0.1215, + "step": 12966 + }, + { + "epoch": 36.220670391061454, + "grad_norm": 0.4031839668750763, + "learning_rate": 2.850614525139665e-05, + "loss": 0.1068, + "step": 12967 + }, + { + "epoch": 36.22346368715084, + "grad_norm": 0.3847515285015106, + "learning_rate": 2.8505772811918066e-05, + "loss": 0.1102, + "step": 12968 + }, + { + "epoch": 36.226256983240226, + "grad_norm": 0.3372950255870819, + "learning_rate": 2.8505400372439482e-05, + "loss": 0.1108, + "step": 12969 + }, + { + "epoch": 36.229050279329606, + "grad_norm": 0.4942077398300171, + "learning_rate": 2.8505027932960892e-05, + "loss": 0.101, + "step": 12970 + }, + { + "epoch": 36.23184357541899, + "grad_norm": 0.7101955413818359, + "learning_rate": 2.850465549348231e-05, + "loss": 0.1033, + "step": 12971 + }, + { + "epoch": 36.23463687150838, + "grad_norm": 0.836714506149292, + "learning_rate": 2.8504283054003725e-05, + "loss": 0.1046, + "step": 12972 + }, + { + "epoch": 36.237430167597765, + "grad_norm": 0.6215742826461792, + "learning_rate": 2.850391061452514e-05, + "loss": 0.0962, + "step": 12973 + }, + { + "epoch": 36.24022346368715, + "grad_norm": 0.5017818212509155, + "learning_rate": 2.8503538175046555e-05, + "loss": 0.0958, + "step": 12974 + }, + { + "epoch": 36.24301675977654, + "grad_norm": 0.3836405873298645, + "learning_rate": 2.850316573556797e-05, + "loss": 0.09, + "step": 12975 + }, + { + "epoch": 36.245810055865924, + "grad_norm": 0.6300547122955322, + "learning_rate": 2.8502793296089388e-05, + "loss": 0.0855, + "step": 12976 + }, + { + "epoch": 36.24860335195531, + "grad_norm": 0.5787068009376526, + "learning_rate": 2.85024208566108e-05, + "loss": 0.1067, + "step": 12977 + }, + { + "epoch": 36.25139664804469, + "grad_norm": 0.7824311256408691, + "learning_rate": 2.8502048417132217e-05, + "loss": 0.0894, + "step": 12978 + }, + { + "epoch": 36.254189944134076, + "grad_norm": 0.43560871481895447, + "learning_rate": 2.850167597765363e-05, + "loss": 0.089, + "step": 12979 + }, + { + "epoch": 36.25698324022346, + "grad_norm": 0.6859767436981201, + "learning_rate": 2.8501303538175047e-05, + "loss": 0.0823, + "step": 12980 + }, + { + "epoch": 36.25977653631285, + "grad_norm": 0.9921254515647888, + "learning_rate": 2.8500931098696464e-05, + "loss": 0.0863, + "step": 12981 + }, + { + "epoch": 36.262569832402235, + "grad_norm": 0.6594396233558655, + "learning_rate": 2.850055865921788e-05, + "loss": 0.0953, + "step": 12982 + }, + { + "epoch": 36.26536312849162, + "grad_norm": 0.7425248026847839, + "learning_rate": 2.850018621973929e-05, + "loss": 0.0838, + "step": 12983 + }, + { + "epoch": 36.26815642458101, + "grad_norm": 0.81383216381073, + "learning_rate": 2.8499813780260706e-05, + "loss": 0.0677, + "step": 12984 + }, + { + "epoch": 36.270949720670394, + "grad_norm": 0.7734430432319641, + "learning_rate": 2.8499441340782123e-05, + "loss": 0.0844, + "step": 12985 + }, + { + "epoch": 36.273743016759774, + "grad_norm": 0.6407185196876526, + "learning_rate": 2.849906890130354e-05, + "loss": 0.0768, + "step": 12986 + }, + { + "epoch": 36.27653631284916, + "grad_norm": 1.0365118980407715, + "learning_rate": 2.8498696461824956e-05, + "loss": 0.0795, + "step": 12987 + }, + { + "epoch": 36.279329608938546, + "grad_norm": 1.0819549560546875, + "learning_rate": 2.849832402234637e-05, + "loss": 0.1175, + "step": 12988 + }, + { + "epoch": 36.28212290502793, + "grad_norm": 0.7784407734870911, + "learning_rate": 2.8497951582867786e-05, + "loss": 0.1483, + "step": 12989 + }, + { + "epoch": 36.28491620111732, + "grad_norm": 0.4731464087963104, + "learning_rate": 2.84975791433892e-05, + "loss": 0.1366, + "step": 12990 + }, + { + "epoch": 36.287709497206706, + "grad_norm": 0.4354664981365204, + "learning_rate": 2.8497206703910615e-05, + "loss": 0.124, + "step": 12991 + }, + { + "epoch": 36.29050279329609, + "grad_norm": 0.5147245526313782, + "learning_rate": 2.849683426443203e-05, + "loss": 0.1368, + "step": 12992 + }, + { + "epoch": 36.29329608938548, + "grad_norm": 0.47912850975990295, + "learning_rate": 2.8496461824953445e-05, + "loss": 0.1297, + "step": 12993 + }, + { + "epoch": 36.29608938547486, + "grad_norm": 0.6198381185531616, + "learning_rate": 2.849608938547486e-05, + "loss": 0.0994, + "step": 12994 + }, + { + "epoch": 36.298882681564244, + "grad_norm": 0.5819131135940552, + "learning_rate": 2.8495716945996278e-05, + "loss": 0.1218, + "step": 12995 + }, + { + "epoch": 36.30167597765363, + "grad_norm": 0.5696055889129639, + "learning_rate": 2.8495344506517695e-05, + "loss": 0.1289, + "step": 12996 + }, + { + "epoch": 36.30446927374302, + "grad_norm": 0.4840451180934906, + "learning_rate": 2.8494972067039104e-05, + "loss": 0.089, + "step": 12997 + }, + { + "epoch": 36.3072625698324, + "grad_norm": 0.5326617956161499, + "learning_rate": 2.849459962756052e-05, + "loss": 0.1083, + "step": 12998 + }, + { + "epoch": 36.31005586592179, + "grad_norm": 0.5152286291122437, + "learning_rate": 2.8494227188081938e-05, + "loss": 0.1092, + "step": 12999 + }, + { + "epoch": 36.312849162011176, + "grad_norm": 1.1104923486709595, + "learning_rate": 2.8493854748603354e-05, + "loss": 0.1045, + "step": 13000 + }, + { + "epoch": 36.312849162011176, + "eval_cer": 0.19295604071879807, + "eval_loss": 0.30565157532691956, + "eval_runtime": 23.8672, + "eval_samples_per_second": 190.136, + "eval_steps_per_second": 0.628, + "eval_wer": 0.34800703890952767, + "step": 13000 + }, + { + "epoch": 36.315642458100555, + "grad_norm": 0.5103163719177246, + "learning_rate": 2.8493482309124767e-05, + "loss": 0.1163, + "step": 13001 + }, + { + "epoch": 36.31843575418994, + "grad_norm": 0.45563486218452454, + "learning_rate": 2.8493109869646184e-05, + "loss": 0.0957, + "step": 13002 + }, + { + "epoch": 36.32122905027933, + "grad_norm": 0.6634702682495117, + "learning_rate": 2.8492737430167597e-05, + "loss": 0.086, + "step": 13003 + }, + { + "epoch": 36.324022346368714, + "grad_norm": 0.3813391327857971, + "learning_rate": 2.8492364990689013e-05, + "loss": 0.0841, + "step": 13004 + }, + { + "epoch": 36.3268156424581, + "grad_norm": 0.4647751748561859, + "learning_rate": 2.849199255121043e-05, + "loss": 0.0837, + "step": 13005 + }, + { + "epoch": 36.32960893854749, + "grad_norm": 0.5191593170166016, + "learning_rate": 2.8491620111731843e-05, + "loss": 0.0903, + "step": 13006 + }, + { + "epoch": 36.332402234636874, + "grad_norm": 2.6562979221343994, + "learning_rate": 2.849124767225326e-05, + "loss": 0.0852, + "step": 13007 + }, + { + "epoch": 36.33519553072626, + "grad_norm": 1.221608281135559, + "learning_rate": 2.8490875232774676e-05, + "loss": 0.102, + "step": 13008 + }, + { + "epoch": 36.33798882681564, + "grad_norm": 1.1749480962753296, + "learning_rate": 2.8490502793296093e-05, + "loss": 0.0789, + "step": 13009 + }, + { + "epoch": 36.340782122905026, + "grad_norm": 0.8152073621749878, + "learning_rate": 2.8490130353817502e-05, + "loss": 0.082, + "step": 13010 + }, + { + "epoch": 36.34357541899441, + "grad_norm": 0.6769988536834717, + "learning_rate": 2.848975791433892e-05, + "loss": 0.0732, + "step": 13011 + }, + { + "epoch": 36.3463687150838, + "grad_norm": 0.7155337929725647, + "learning_rate": 2.8489385474860336e-05, + "loss": 0.0835, + "step": 13012 + }, + { + "epoch": 36.349162011173185, + "grad_norm": 0.9940541386604309, + "learning_rate": 2.8489013035381752e-05, + "loss": 0.1017, + "step": 13013 + }, + { + "epoch": 36.35195530726257, + "grad_norm": 0.3431513011455536, + "learning_rate": 2.848864059590317e-05, + "loss": 0.1231, + "step": 13014 + }, + { + "epoch": 36.35474860335196, + "grad_norm": 0.4082512855529785, + "learning_rate": 2.8488268156424582e-05, + "loss": 0.1119, + "step": 13015 + }, + { + "epoch": 36.357541899441344, + "grad_norm": 0.6857700347900391, + "learning_rate": 2.8487895716946e-05, + "loss": 0.1125, + "step": 13016 + }, + { + "epoch": 36.36033519553072, + "grad_norm": 0.4404042661190033, + "learning_rate": 2.848752327746741e-05, + "loss": 0.1183, + "step": 13017 + }, + { + "epoch": 36.36312849162011, + "grad_norm": 0.4675848186016083, + "learning_rate": 2.8487150837988828e-05, + "loss": 0.122, + "step": 13018 + }, + { + "epoch": 36.365921787709496, + "grad_norm": 0.7972007393836975, + "learning_rate": 2.848677839851024e-05, + "loss": 0.1244, + "step": 13019 + }, + { + "epoch": 36.36871508379888, + "grad_norm": 0.8320142030715942, + "learning_rate": 2.8486405959031658e-05, + "loss": 0.1171, + "step": 13020 + }, + { + "epoch": 36.37150837988827, + "grad_norm": 0.9292674660682678, + "learning_rate": 2.8486033519553074e-05, + "loss": 0.1069, + "step": 13021 + }, + { + "epoch": 36.374301675977655, + "grad_norm": 0.4998623728752136, + "learning_rate": 2.848566108007449e-05, + "loss": 0.1004, + "step": 13022 + }, + { + "epoch": 36.37709497206704, + "grad_norm": 1.3461363315582275, + "learning_rate": 2.8485288640595904e-05, + "loss": 0.0973, + "step": 13023 + }, + { + "epoch": 36.37988826815643, + "grad_norm": 0.49873998761177063, + "learning_rate": 2.8484916201117317e-05, + "loss": 0.1177, + "step": 13024 + }, + { + "epoch": 36.38268156424581, + "grad_norm": 0.5666747689247131, + "learning_rate": 2.8484543761638734e-05, + "loss": 0.1083, + "step": 13025 + }, + { + "epoch": 36.385474860335194, + "grad_norm": 1.3581769466400146, + "learning_rate": 2.848417132216015e-05, + "loss": 0.0955, + "step": 13026 + }, + { + "epoch": 36.38826815642458, + "grad_norm": 0.6231361627578735, + "learning_rate": 2.8483798882681567e-05, + "loss": 0.0989, + "step": 13027 + }, + { + "epoch": 36.391061452513966, + "grad_norm": 0.5649336576461792, + "learning_rate": 2.848342644320298e-05, + "loss": 0.1005, + "step": 13028 + }, + { + "epoch": 36.39385474860335, + "grad_norm": 0.5340343117713928, + "learning_rate": 2.8483054003724396e-05, + "loss": 0.0914, + "step": 13029 + }, + { + "epoch": 36.39664804469274, + "grad_norm": 0.6382849216461182, + "learning_rate": 2.848268156424581e-05, + "loss": 0.0874, + "step": 13030 + }, + { + "epoch": 36.399441340782126, + "grad_norm": 0.7184431552886963, + "learning_rate": 2.8482309124767226e-05, + "loss": 0.0749, + "step": 13031 + }, + { + "epoch": 36.40223463687151, + "grad_norm": 0.7611850500106812, + "learning_rate": 2.8481936685288643e-05, + "loss": 0.064, + "step": 13032 + }, + { + "epoch": 36.40502793296089, + "grad_norm": 0.8529970645904541, + "learning_rate": 2.8481564245810056e-05, + "loss": 0.084, + "step": 13033 + }, + { + "epoch": 36.40782122905028, + "grad_norm": 0.6734018325805664, + "learning_rate": 2.8481191806331472e-05, + "loss": 0.0931, + "step": 13034 + }, + { + "epoch": 36.410614525139664, + "grad_norm": 0.7495433688163757, + "learning_rate": 2.848081936685289e-05, + "loss": 0.0937, + "step": 13035 + }, + { + "epoch": 36.41340782122905, + "grad_norm": 0.7531351447105408, + "learning_rate": 2.8480446927374305e-05, + "loss": 0.0887, + "step": 13036 + }, + { + "epoch": 36.41620111731844, + "grad_norm": 0.9327375888824463, + "learning_rate": 2.8480074487895715e-05, + "loss": 0.0894, + "step": 13037 + }, + { + "epoch": 36.41899441340782, + "grad_norm": 2.5498523712158203, + "learning_rate": 2.847970204841713e-05, + "loss": 0.1179, + "step": 13038 + }, + { + "epoch": 36.42178770949721, + "grad_norm": 0.732311487197876, + "learning_rate": 2.8479329608938548e-05, + "loss": 0.1288, + "step": 13039 + }, + { + "epoch": 36.42458100558659, + "grad_norm": 0.3687518239021301, + "learning_rate": 2.8478957169459965e-05, + "loss": 0.1117, + "step": 13040 + }, + { + "epoch": 36.427374301675975, + "grad_norm": 0.41522008180618286, + "learning_rate": 2.847858472998138e-05, + "loss": 0.0986, + "step": 13041 + }, + { + "epoch": 36.43016759776536, + "grad_norm": 1.1092604398727417, + "learning_rate": 2.8478212290502794e-05, + "loss": 0.1044, + "step": 13042 + }, + { + "epoch": 36.43296089385475, + "grad_norm": 0.5997103452682495, + "learning_rate": 2.8477839851024208e-05, + "loss": 0.1477, + "step": 13043 + }, + { + "epoch": 36.435754189944134, + "grad_norm": 1.1083749532699585, + "learning_rate": 2.8477467411545624e-05, + "loss": 0.1203, + "step": 13044 + }, + { + "epoch": 36.43854748603352, + "grad_norm": 0.6279087662696838, + "learning_rate": 2.847709497206704e-05, + "loss": 0.1205, + "step": 13045 + }, + { + "epoch": 36.44134078212291, + "grad_norm": 0.35380157828330994, + "learning_rate": 2.8476722532588454e-05, + "loss": 0.0955, + "step": 13046 + }, + { + "epoch": 36.444134078212294, + "grad_norm": 0.44671377539634705, + "learning_rate": 2.847635009310987e-05, + "loss": 0.1007, + "step": 13047 + }, + { + "epoch": 36.44692737430167, + "grad_norm": 0.5153653621673584, + "learning_rate": 2.8475977653631287e-05, + "loss": 0.0959, + "step": 13048 + }, + { + "epoch": 36.44972067039106, + "grad_norm": 0.7488757967948914, + "learning_rate": 2.8475605214152703e-05, + "loss": 0.0933, + "step": 13049 + }, + { + "epoch": 36.452513966480446, + "grad_norm": 1.1935946941375732, + "learning_rate": 2.8475232774674116e-05, + "loss": 0.1124, + "step": 13050 + }, + { + "epoch": 36.45530726256983, + "grad_norm": 0.9721536636352539, + "learning_rate": 2.847486033519553e-05, + "loss": 0.0827, + "step": 13051 + }, + { + "epoch": 36.45810055865922, + "grad_norm": 0.5015149712562561, + "learning_rate": 2.8474487895716946e-05, + "loss": 0.0967, + "step": 13052 + }, + { + "epoch": 36.460893854748605, + "grad_norm": 0.9974826574325562, + "learning_rate": 2.8474115456238363e-05, + "loss": 0.0924, + "step": 13053 + }, + { + "epoch": 36.46368715083799, + "grad_norm": 0.8104403614997864, + "learning_rate": 2.847374301675978e-05, + "loss": 0.0931, + "step": 13054 + }, + { + "epoch": 36.46648044692738, + "grad_norm": 0.632412314414978, + "learning_rate": 2.8473370577281192e-05, + "loss": 0.0825, + "step": 13055 + }, + { + "epoch": 36.46927374301676, + "grad_norm": 0.8117548227310181, + "learning_rate": 2.847299813780261e-05, + "loss": 0.084, + "step": 13056 + }, + { + "epoch": 36.47206703910614, + "grad_norm": 0.728563129901886, + "learning_rate": 2.8472625698324022e-05, + "loss": 0.091, + "step": 13057 + }, + { + "epoch": 36.47486033519553, + "grad_norm": 0.6499229669570923, + "learning_rate": 2.847225325884544e-05, + "loss": 0.1043, + "step": 13058 + }, + { + "epoch": 36.477653631284916, + "grad_norm": 0.5659500360488892, + "learning_rate": 2.8471880819366852e-05, + "loss": 0.0817, + "step": 13059 + }, + { + "epoch": 36.4804469273743, + "grad_norm": 1.1683993339538574, + "learning_rate": 2.8471508379888268e-05, + "loss": 0.0788, + "step": 13060 + }, + { + "epoch": 36.48324022346369, + "grad_norm": 1.722424030303955, + "learning_rate": 2.8471135940409685e-05, + "loss": 0.0754, + "step": 13061 + }, + { + "epoch": 36.486033519553075, + "grad_norm": 1.2399303913116455, + "learning_rate": 2.84707635009311e-05, + "loss": 0.117, + "step": 13062 + }, + { + "epoch": 36.48882681564246, + "grad_norm": 1.3291460275650024, + "learning_rate": 2.8470391061452514e-05, + "loss": 0.1093, + "step": 13063 + }, + { + "epoch": 36.49162011173184, + "grad_norm": 0.41691839694976807, + "learning_rate": 2.8470018621973928e-05, + "loss": 0.1385, + "step": 13064 + }, + { + "epoch": 36.49441340782123, + "grad_norm": 0.8629980087280273, + "learning_rate": 2.8469646182495344e-05, + "loss": 0.1391, + "step": 13065 + }, + { + "epoch": 36.497206703910614, + "grad_norm": 0.5989841818809509, + "learning_rate": 2.846927374301676e-05, + "loss": 0.1405, + "step": 13066 + }, + { + "epoch": 36.5, + "grad_norm": 0.4392365515232086, + "learning_rate": 2.8468901303538177e-05, + "loss": 0.1256, + "step": 13067 + }, + { + "epoch": 36.502793296089386, + "grad_norm": 0.41069284081459045, + "learning_rate": 2.846852886405959e-05, + "loss": 0.1267, + "step": 13068 + }, + { + "epoch": 36.50558659217877, + "grad_norm": 1.3699334859848022, + "learning_rate": 2.8468156424581007e-05, + "loss": 0.1088, + "step": 13069 + }, + { + "epoch": 36.50837988826816, + "grad_norm": 0.3773321807384491, + "learning_rate": 2.846778398510242e-05, + "loss": 0.1207, + "step": 13070 + }, + { + "epoch": 36.51117318435754, + "grad_norm": 0.3941558599472046, + "learning_rate": 2.8467411545623837e-05, + "loss": 0.0972, + "step": 13071 + }, + { + "epoch": 36.513966480446925, + "grad_norm": 0.5023883581161499, + "learning_rate": 2.8467039106145253e-05, + "loss": 0.0949, + "step": 13072 + }, + { + "epoch": 36.51675977653631, + "grad_norm": 0.6170898079872131, + "learning_rate": 2.8466666666666666e-05, + "loss": 0.1054, + "step": 13073 + }, + { + "epoch": 36.5195530726257, + "grad_norm": 0.4747655689716339, + "learning_rate": 2.8466294227188083e-05, + "loss": 0.1049, + "step": 13074 + }, + { + "epoch": 36.522346368715084, + "grad_norm": 0.6342079043388367, + "learning_rate": 2.84659217877095e-05, + "loss": 0.1078, + "step": 13075 + }, + { + "epoch": 36.52513966480447, + "grad_norm": 0.9232836961746216, + "learning_rate": 2.8465549348230916e-05, + "loss": 0.0895, + "step": 13076 + }, + { + "epoch": 36.52793296089386, + "grad_norm": 0.6280178427696228, + "learning_rate": 2.8465176908752326e-05, + "loss": 0.1007, + "step": 13077 + }, + { + "epoch": 36.53072625698324, + "grad_norm": 1.9618819952011108, + "learning_rate": 2.8464804469273742e-05, + "loss": 0.0875, + "step": 13078 + }, + { + "epoch": 36.53351955307262, + "grad_norm": 0.5735899209976196, + "learning_rate": 2.846443202979516e-05, + "loss": 0.0871, + "step": 13079 + }, + { + "epoch": 36.53631284916201, + "grad_norm": 0.7038599848747253, + "learning_rate": 2.8464059590316575e-05, + "loss": 0.0882, + "step": 13080 + }, + { + "epoch": 36.539106145251395, + "grad_norm": 0.5991548895835876, + "learning_rate": 2.8463687150837992e-05, + "loss": 0.0794, + "step": 13081 + }, + { + "epoch": 36.54189944134078, + "grad_norm": 0.7051090002059937, + "learning_rate": 2.8463314711359405e-05, + "loss": 0.0823, + "step": 13082 + }, + { + "epoch": 36.54469273743017, + "grad_norm": 0.5633767247200012, + "learning_rate": 2.8462942271880818e-05, + "loss": 0.0807, + "step": 13083 + }, + { + "epoch": 36.547486033519554, + "grad_norm": 0.7850005626678467, + "learning_rate": 2.8462569832402235e-05, + "loss": 0.0999, + "step": 13084 + }, + { + "epoch": 36.55027932960894, + "grad_norm": 0.6875097751617432, + "learning_rate": 2.846219739292365e-05, + "loss": 0.0797, + "step": 13085 + }, + { + "epoch": 36.55307262569833, + "grad_norm": 0.8078372478485107, + "learning_rate": 2.8461824953445064e-05, + "loss": 0.0774, + "step": 13086 + }, + { + "epoch": 36.555865921787706, + "grad_norm": 0.9205446243286133, + "learning_rate": 2.846145251396648e-05, + "loss": 0.0913, + "step": 13087 + }, + { + "epoch": 36.55865921787709, + "grad_norm": 1.0804953575134277, + "learning_rate": 2.8461080074487897e-05, + "loss": 0.1215, + "step": 13088 + }, + { + "epoch": 36.56145251396648, + "grad_norm": 0.5169636607170105, + "learning_rate": 2.8460707635009314e-05, + "loss": 0.12, + "step": 13089 + }, + { + "epoch": 36.564245810055866, + "grad_norm": 0.7846399545669556, + "learning_rate": 2.8460335195530727e-05, + "loss": 0.1233, + "step": 13090 + }, + { + "epoch": 36.56703910614525, + "grad_norm": 0.6060928702354431, + "learning_rate": 2.845996275605214e-05, + "loss": 0.1051, + "step": 13091 + }, + { + "epoch": 36.56983240223464, + "grad_norm": 0.9782649278640747, + "learning_rate": 2.8459590316573557e-05, + "loss": 0.1298, + "step": 13092 + }, + { + "epoch": 36.572625698324025, + "grad_norm": 0.5207920074462891, + "learning_rate": 2.8459217877094973e-05, + "loss": 0.1315, + "step": 13093 + }, + { + "epoch": 36.57541899441341, + "grad_norm": 0.5926817059516907, + "learning_rate": 2.845884543761639e-05, + "loss": 0.1254, + "step": 13094 + }, + { + "epoch": 36.57821229050279, + "grad_norm": 0.707974374294281, + "learning_rate": 2.8458472998137803e-05, + "loss": 0.1204, + "step": 13095 + }, + { + "epoch": 36.58100558659218, + "grad_norm": 0.5093036890029907, + "learning_rate": 2.845810055865922e-05, + "loss": 0.1044, + "step": 13096 + }, + { + "epoch": 36.58379888268156, + "grad_norm": 0.47131848335266113, + "learning_rate": 2.8457728119180633e-05, + "loss": 0.1029, + "step": 13097 + }, + { + "epoch": 36.58659217877095, + "grad_norm": 0.7507746815681458, + "learning_rate": 2.845735567970205e-05, + "loss": 0.1113, + "step": 13098 + }, + { + "epoch": 36.589385474860336, + "grad_norm": 0.445652037858963, + "learning_rate": 2.8456983240223466e-05, + "loss": 0.0843, + "step": 13099 + }, + { + "epoch": 36.59217877094972, + "grad_norm": 0.6210559606552124, + "learning_rate": 2.845661080074488e-05, + "loss": 0.0881, + "step": 13100 + }, + { + "epoch": 36.59497206703911, + "grad_norm": 0.5198919773101807, + "learning_rate": 2.8456238361266295e-05, + "loss": 0.099, + "step": 13101 + }, + { + "epoch": 36.59776536312849, + "grad_norm": 0.7758316397666931, + "learning_rate": 2.8455865921787712e-05, + "loss": 0.1087, + "step": 13102 + }, + { + "epoch": 36.600558659217874, + "grad_norm": 0.36870887875556946, + "learning_rate": 2.8455493482309125e-05, + "loss": 0.0796, + "step": 13103 + }, + { + "epoch": 36.60335195530726, + "grad_norm": 4.285307884216309, + "learning_rate": 2.8455121042830538e-05, + "loss": 0.0923, + "step": 13104 + }, + { + "epoch": 36.60614525139665, + "grad_norm": 1.537385106086731, + "learning_rate": 2.8454748603351955e-05, + "loss": 0.1028, + "step": 13105 + }, + { + "epoch": 36.608938547486034, + "grad_norm": 0.9062882661819458, + "learning_rate": 2.845437616387337e-05, + "loss": 0.0798, + "step": 13106 + }, + { + "epoch": 36.61173184357542, + "grad_norm": 1.2549960613250732, + "learning_rate": 2.8454003724394788e-05, + "loss": 0.0865, + "step": 13107 + }, + { + "epoch": 36.614525139664806, + "grad_norm": 0.4961925148963928, + "learning_rate": 2.8453631284916204e-05, + "loss": 0.0696, + "step": 13108 + }, + { + "epoch": 36.61731843575419, + "grad_norm": 1.2841949462890625, + "learning_rate": 2.8453258845437618e-05, + "loss": 0.0937, + "step": 13109 + }, + { + "epoch": 36.62011173184357, + "grad_norm": 0.7004602551460266, + "learning_rate": 2.845288640595903e-05, + "loss": 0.0846, + "step": 13110 + }, + { + "epoch": 36.62290502793296, + "grad_norm": 1.0388610363006592, + "learning_rate": 2.8452513966480447e-05, + "loss": 0.0798, + "step": 13111 + }, + { + "epoch": 36.625698324022345, + "grad_norm": 0.8924124240875244, + "learning_rate": 2.8452141527001864e-05, + "loss": 0.0911, + "step": 13112 + }, + { + "epoch": 36.62849162011173, + "grad_norm": 0.9572955369949341, + "learning_rate": 2.8451769087523277e-05, + "loss": 0.1039, + "step": 13113 + }, + { + "epoch": 36.63128491620112, + "grad_norm": 0.5310201644897461, + "learning_rate": 2.8451396648044693e-05, + "loss": 0.1167, + "step": 13114 + }, + { + "epoch": 36.634078212290504, + "grad_norm": 0.4729738235473633, + "learning_rate": 2.845102420856611e-05, + "loss": 0.1421, + "step": 13115 + }, + { + "epoch": 36.63687150837989, + "grad_norm": 0.3305386006832123, + "learning_rate": 2.8450651769087526e-05, + "loss": 0.1175, + "step": 13116 + }, + { + "epoch": 36.63966480446928, + "grad_norm": 0.49460217356681824, + "learning_rate": 2.845027932960894e-05, + "loss": 0.1306, + "step": 13117 + }, + { + "epoch": 36.642458100558656, + "grad_norm": 0.3999815881252289, + "learning_rate": 2.8449906890130353e-05, + "loss": 0.1112, + "step": 13118 + }, + { + "epoch": 36.64525139664804, + "grad_norm": 0.6515106558799744, + "learning_rate": 2.844953445065177e-05, + "loss": 0.1082, + "step": 13119 + }, + { + "epoch": 36.64804469273743, + "grad_norm": 1.4159810543060303, + "learning_rate": 2.8449162011173186e-05, + "loss": 0.1204, + "step": 13120 + }, + { + "epoch": 36.650837988826815, + "grad_norm": 0.7723098993301392, + "learning_rate": 2.8448789571694602e-05, + "loss": 0.1074, + "step": 13121 + }, + { + "epoch": 36.6536312849162, + "grad_norm": 0.9828838109970093, + "learning_rate": 2.8448417132216016e-05, + "loss": 0.1013, + "step": 13122 + }, + { + "epoch": 36.65642458100559, + "grad_norm": 0.5460968613624573, + "learning_rate": 2.844804469273743e-05, + "loss": 0.1136, + "step": 13123 + }, + { + "epoch": 36.659217877094974, + "grad_norm": 0.4706367552280426, + "learning_rate": 2.8447672253258845e-05, + "loss": 0.1149, + "step": 13124 + }, + { + "epoch": 36.66201117318436, + "grad_norm": 0.38748401403427124, + "learning_rate": 2.8447299813780262e-05, + "loss": 0.0912, + "step": 13125 + }, + { + "epoch": 36.66480446927374, + "grad_norm": 0.45856836438179016, + "learning_rate": 2.8446927374301678e-05, + "loss": 0.0865, + "step": 13126 + }, + { + "epoch": 36.667597765363126, + "grad_norm": 0.7570058703422546, + "learning_rate": 2.844655493482309e-05, + "loss": 0.1041, + "step": 13127 + }, + { + "epoch": 36.67039106145251, + "grad_norm": 0.821748673915863, + "learning_rate": 2.8446182495344508e-05, + "loss": 0.1071, + "step": 13128 + }, + { + "epoch": 36.6731843575419, + "grad_norm": 0.5184983611106873, + "learning_rate": 2.8445810055865925e-05, + "loss": 0.0797, + "step": 13129 + }, + { + "epoch": 36.675977653631286, + "grad_norm": 0.3808233439922333, + "learning_rate": 2.8445437616387338e-05, + "loss": 0.0871, + "step": 13130 + }, + { + "epoch": 36.67877094972067, + "grad_norm": 0.6422951221466064, + "learning_rate": 2.844506517690875e-05, + "loss": 0.0887, + "step": 13131 + }, + { + "epoch": 36.68156424581006, + "grad_norm": 0.684080958366394, + "learning_rate": 2.8444692737430167e-05, + "loss": 0.0795, + "step": 13132 + }, + { + "epoch": 36.684357541899445, + "grad_norm": 0.7174074053764343, + "learning_rate": 2.8444320297951584e-05, + "loss": 0.0931, + "step": 13133 + }, + { + "epoch": 36.687150837988824, + "grad_norm": 0.7364905476570129, + "learning_rate": 2.8443947858473e-05, + "loss": 0.0802, + "step": 13134 + }, + { + "epoch": 36.68994413407821, + "grad_norm": 0.7372745871543884, + "learning_rate": 2.8443575418994417e-05, + "loss": 0.0972, + "step": 13135 + }, + { + "epoch": 36.6927374301676, + "grad_norm": 0.7332994341850281, + "learning_rate": 2.844320297951583e-05, + "loss": 0.0594, + "step": 13136 + }, + { + "epoch": 36.69553072625698, + "grad_norm": 7.079324245452881, + "learning_rate": 2.8442830540037243e-05, + "loss": 0.0908, + "step": 13137 + }, + { + "epoch": 36.69832402234637, + "grad_norm": 1.1419892311096191, + "learning_rate": 2.844245810055866e-05, + "loss": 0.1062, + "step": 13138 + }, + { + "epoch": 36.701117318435756, + "grad_norm": 0.4915393590927124, + "learning_rate": 2.8442085661080076e-05, + "loss": 0.1254, + "step": 13139 + }, + { + "epoch": 36.70391061452514, + "grad_norm": 0.355114221572876, + "learning_rate": 2.844171322160149e-05, + "loss": 0.1235, + "step": 13140 + }, + { + "epoch": 36.70670391061452, + "grad_norm": 1.367867350578308, + "learning_rate": 2.8441340782122906e-05, + "loss": 0.1279, + "step": 13141 + }, + { + "epoch": 36.70949720670391, + "grad_norm": 0.5479897856712341, + "learning_rate": 2.8440968342644323e-05, + "loss": 0.1014, + "step": 13142 + }, + { + "epoch": 36.712290502793294, + "grad_norm": 0.44179767370224, + "learning_rate": 2.8440595903165736e-05, + "loss": 0.1249, + "step": 13143 + }, + { + "epoch": 36.71508379888268, + "grad_norm": 0.4458467960357666, + "learning_rate": 2.8440223463687152e-05, + "loss": 0.0995, + "step": 13144 + }, + { + "epoch": 36.71787709497207, + "grad_norm": 1.1193277835845947, + "learning_rate": 2.8439851024208565e-05, + "loss": 0.1329, + "step": 13145 + }, + { + "epoch": 36.720670391061454, + "grad_norm": 0.6375119686126709, + "learning_rate": 2.8439478584729982e-05, + "loss": 0.0915, + "step": 13146 + }, + { + "epoch": 36.72346368715084, + "grad_norm": 0.4403560161590576, + "learning_rate": 2.84391061452514e-05, + "loss": 0.1027, + "step": 13147 + }, + { + "epoch": 36.726256983240226, + "grad_norm": 0.6501517295837402, + "learning_rate": 2.8438733705772815e-05, + "loss": 0.1, + "step": 13148 + }, + { + "epoch": 36.729050279329606, + "grad_norm": 0.4135424792766571, + "learning_rate": 2.8438361266294228e-05, + "loss": 0.1057, + "step": 13149 + }, + { + "epoch": 36.73184357541899, + "grad_norm": 0.8687627911567688, + "learning_rate": 2.843798882681564e-05, + "loss": 0.1108, + "step": 13150 + }, + { + "epoch": 36.73463687150838, + "grad_norm": 0.4155297577381134, + "learning_rate": 2.8437616387337058e-05, + "loss": 0.0879, + "step": 13151 + }, + { + "epoch": 36.737430167597765, + "grad_norm": 0.5047573447227478, + "learning_rate": 2.8437243947858474e-05, + "loss": 0.101, + "step": 13152 + }, + { + "epoch": 36.74022346368715, + "grad_norm": 0.6942926645278931, + "learning_rate": 2.8436871508379887e-05, + "loss": 0.0892, + "step": 13153 + }, + { + "epoch": 36.74301675977654, + "grad_norm": 1.0379856824874878, + "learning_rate": 2.8436499068901304e-05, + "loss": 0.0922, + "step": 13154 + }, + { + "epoch": 36.745810055865924, + "grad_norm": 0.4566536247730255, + "learning_rate": 2.843612662942272e-05, + "loss": 0.0717, + "step": 13155 + }, + { + "epoch": 36.74860335195531, + "grad_norm": 0.9123262763023376, + "learning_rate": 2.8435754189944137e-05, + "loss": 0.0781, + "step": 13156 + }, + { + "epoch": 36.75139664804469, + "grad_norm": 0.4724482297897339, + "learning_rate": 2.843538175046555e-05, + "loss": 0.0946, + "step": 13157 + }, + { + "epoch": 36.754189944134076, + "grad_norm": 0.7829778790473938, + "learning_rate": 2.8435009310986963e-05, + "loss": 0.0967, + "step": 13158 + }, + { + "epoch": 36.75698324022346, + "grad_norm": 0.9873928427696228, + "learning_rate": 2.843463687150838e-05, + "loss": 0.0709, + "step": 13159 + }, + { + "epoch": 36.75977653631285, + "grad_norm": 0.6474883556365967, + "learning_rate": 2.8434264432029796e-05, + "loss": 0.0921, + "step": 13160 + }, + { + "epoch": 36.762569832402235, + "grad_norm": 0.6852738857269287, + "learning_rate": 2.8433891992551213e-05, + "loss": 0.0933, + "step": 13161 + }, + { + "epoch": 36.76536312849162, + "grad_norm": 0.5250356793403625, + "learning_rate": 2.8433519553072626e-05, + "loss": 0.0797, + "step": 13162 + }, + { + "epoch": 36.76815642458101, + "grad_norm": 1.3000695705413818, + "learning_rate": 2.843314711359404e-05, + "loss": 0.101, + "step": 13163 + }, + { + "epoch": 36.770949720670394, + "grad_norm": 0.4818575084209442, + "learning_rate": 2.8432774674115456e-05, + "loss": 0.132, + "step": 13164 + }, + { + "epoch": 36.773743016759774, + "grad_norm": 0.3751843273639679, + "learning_rate": 2.8432402234636872e-05, + "loss": 0.1106, + "step": 13165 + }, + { + "epoch": 36.77653631284916, + "grad_norm": 0.5293797254562378, + "learning_rate": 2.843202979515829e-05, + "loss": 0.1372, + "step": 13166 + }, + { + "epoch": 36.779329608938546, + "grad_norm": 0.47639548778533936, + "learning_rate": 2.8431657355679702e-05, + "loss": 0.1118, + "step": 13167 + }, + { + "epoch": 36.78212290502793, + "grad_norm": 0.4971713423728943, + "learning_rate": 2.843128491620112e-05, + "loss": 0.1132, + "step": 13168 + }, + { + "epoch": 36.78491620111732, + "grad_norm": 0.6892354488372803, + "learning_rate": 2.8430912476722535e-05, + "loss": 0.1185, + "step": 13169 + }, + { + "epoch": 36.787709497206706, + "grad_norm": 0.3397558629512787, + "learning_rate": 2.8430540037243948e-05, + "loss": 0.0944, + "step": 13170 + }, + { + "epoch": 36.79050279329609, + "grad_norm": 0.4257424473762512, + "learning_rate": 2.843016759776536e-05, + "loss": 0.1035, + "step": 13171 + }, + { + "epoch": 36.79329608938548, + "grad_norm": 0.412949800491333, + "learning_rate": 2.8429795158286778e-05, + "loss": 0.1024, + "step": 13172 + }, + { + "epoch": 36.79608938547486, + "grad_norm": 0.7454365491867065, + "learning_rate": 2.8429422718808194e-05, + "loss": 0.1367, + "step": 13173 + }, + { + "epoch": 36.798882681564244, + "grad_norm": 1.1098469495773315, + "learning_rate": 2.842905027932961e-05, + "loss": 0.1223, + "step": 13174 + }, + { + "epoch": 36.80167597765363, + "grad_norm": 0.5613962411880493, + "learning_rate": 2.8428677839851028e-05, + "loss": 0.1224, + "step": 13175 + }, + { + "epoch": 36.80446927374302, + "grad_norm": 0.46567198634147644, + "learning_rate": 2.842830540037244e-05, + "loss": 0.1028, + "step": 13176 + }, + { + "epoch": 36.8072625698324, + "grad_norm": 0.5464855432510376, + "learning_rate": 2.8427932960893854e-05, + "loss": 0.0952, + "step": 13177 + }, + { + "epoch": 36.81005586592179, + "grad_norm": 0.5916465520858765, + "learning_rate": 2.842756052141527e-05, + "loss": 0.0911, + "step": 13178 + }, + { + "epoch": 36.812849162011176, + "grad_norm": 0.6826441884040833, + "learning_rate": 2.8427188081936687e-05, + "loss": 0.1084, + "step": 13179 + }, + { + "epoch": 36.815642458100555, + "grad_norm": 0.7281197309494019, + "learning_rate": 2.84268156424581e-05, + "loss": 0.1074, + "step": 13180 + }, + { + "epoch": 36.81843575418994, + "grad_norm": 0.4663384258747101, + "learning_rate": 2.8426443202979517e-05, + "loss": 0.0977, + "step": 13181 + }, + { + "epoch": 36.82122905027933, + "grad_norm": 0.8397631645202637, + "learning_rate": 2.8426070763500933e-05, + "loss": 0.0837, + "step": 13182 + }, + { + "epoch": 36.824022346368714, + "grad_norm": 0.8874335885047913, + "learning_rate": 2.8425698324022346e-05, + "loss": 0.0988, + "step": 13183 + }, + { + "epoch": 36.8268156424581, + "grad_norm": 0.44335997104644775, + "learning_rate": 2.8425325884543763e-05, + "loss": 0.0917, + "step": 13184 + }, + { + "epoch": 36.82960893854749, + "grad_norm": 1.6380071640014648, + "learning_rate": 2.8424953445065176e-05, + "loss": 0.1009, + "step": 13185 + }, + { + "epoch": 36.832402234636874, + "grad_norm": 0.7539939284324646, + "learning_rate": 2.8424581005586593e-05, + "loss": 0.0974, + "step": 13186 + }, + { + "epoch": 36.83519553072626, + "grad_norm": 1.3203480243682861, + "learning_rate": 2.842420856610801e-05, + "loss": 0.109, + "step": 13187 + }, + { + "epoch": 36.83798882681564, + "grad_norm": 1.7456378936767578, + "learning_rate": 2.8423836126629426e-05, + "loss": 0.1004, + "step": 13188 + }, + { + "epoch": 36.840782122905026, + "grad_norm": 0.521722674369812, + "learning_rate": 2.842346368715084e-05, + "loss": 0.1203, + "step": 13189 + }, + { + "epoch": 36.84357541899441, + "grad_norm": 0.7139871120452881, + "learning_rate": 2.8423091247672252e-05, + "loss": 0.1281, + "step": 13190 + }, + { + "epoch": 36.8463687150838, + "grad_norm": 0.6199519038200378, + "learning_rate": 2.842271880819367e-05, + "loss": 0.1215, + "step": 13191 + }, + { + "epoch": 36.849162011173185, + "grad_norm": 0.37283897399902344, + "learning_rate": 2.8422346368715085e-05, + "loss": 0.1176, + "step": 13192 + }, + { + "epoch": 36.85195530726257, + "grad_norm": 0.3728958070278168, + "learning_rate": 2.84219739292365e-05, + "loss": 0.0929, + "step": 13193 + }, + { + "epoch": 36.85474860335196, + "grad_norm": 0.4494488537311554, + "learning_rate": 2.8421601489757915e-05, + "loss": 0.1108, + "step": 13194 + }, + { + "epoch": 36.857541899441344, + "grad_norm": 0.48986729979515076, + "learning_rate": 2.842122905027933e-05, + "loss": 0.1142, + "step": 13195 + }, + { + "epoch": 36.86033519553072, + "grad_norm": 0.43761172890663147, + "learning_rate": 2.8420856610800744e-05, + "loss": 0.1285, + "step": 13196 + }, + { + "epoch": 36.86312849162011, + "grad_norm": 0.674289345741272, + "learning_rate": 2.842048417132216e-05, + "loss": 0.1029, + "step": 13197 + }, + { + "epoch": 36.865921787709496, + "grad_norm": 0.5453643798828125, + "learning_rate": 2.8420111731843574e-05, + "loss": 0.1072, + "step": 13198 + }, + { + "epoch": 36.86871508379888, + "grad_norm": 0.4174542725086212, + "learning_rate": 2.841973929236499e-05, + "loss": 0.1048, + "step": 13199 + }, + { + "epoch": 36.87150837988827, + "grad_norm": 0.39227601885795593, + "learning_rate": 2.8419366852886407e-05, + "loss": 0.1043, + "step": 13200 + }, + { + "epoch": 36.874301675977655, + "grad_norm": 0.4821287989616394, + "learning_rate": 2.8418994413407824e-05, + "loss": 0.1172, + "step": 13201 + }, + { + "epoch": 36.87709497206704, + "grad_norm": 0.3642721474170685, + "learning_rate": 2.841862197392924e-05, + "loss": 0.0936, + "step": 13202 + }, + { + "epoch": 36.87988826815643, + "grad_norm": 0.5393512845039368, + "learning_rate": 2.841824953445065e-05, + "loss": 0.0887, + "step": 13203 + }, + { + "epoch": 36.88268156424581, + "grad_norm": 1.2354297637939453, + "learning_rate": 2.8417877094972066e-05, + "loss": 0.1027, + "step": 13204 + }, + { + "epoch": 36.885474860335194, + "grad_norm": 0.46873167157173157, + "learning_rate": 2.8417504655493483e-05, + "loss": 0.0888, + "step": 13205 + }, + { + "epoch": 36.88826815642458, + "grad_norm": 0.5780599117279053, + "learning_rate": 2.84171322160149e-05, + "loss": 0.1048, + "step": 13206 + }, + { + "epoch": 36.891061452513966, + "grad_norm": 0.7402818202972412, + "learning_rate": 2.8416759776536313e-05, + "loss": 0.1128, + "step": 13207 + }, + { + "epoch": 36.89385474860335, + "grad_norm": 0.5356085300445557, + "learning_rate": 2.841638733705773e-05, + "loss": 0.0932, + "step": 13208 + }, + { + "epoch": 36.89664804469274, + "grad_norm": 0.8292642831802368, + "learning_rate": 2.8416014897579146e-05, + "loss": 0.0829, + "step": 13209 + }, + { + "epoch": 36.899441340782126, + "grad_norm": 1.1126779317855835, + "learning_rate": 2.841564245810056e-05, + "loss": 0.097, + "step": 13210 + }, + { + "epoch": 36.90223463687151, + "grad_norm": 0.5330525636672974, + "learning_rate": 2.8415270018621975e-05, + "loss": 0.0742, + "step": 13211 + }, + { + "epoch": 36.90502793296089, + "grad_norm": 0.6607159376144409, + "learning_rate": 2.841489757914339e-05, + "loss": 0.0895, + "step": 13212 + }, + { + "epoch": 36.90782122905028, + "grad_norm": 1.2516320943832397, + "learning_rate": 2.8414525139664805e-05, + "loss": 0.1351, + "step": 13213 + }, + { + "epoch": 36.910614525139664, + "grad_norm": 0.37802550196647644, + "learning_rate": 2.841415270018622e-05, + "loss": 0.1214, + "step": 13214 + }, + { + "epoch": 36.91340782122905, + "grad_norm": 0.4019375443458557, + "learning_rate": 2.8413780260707638e-05, + "loss": 0.1308, + "step": 13215 + }, + { + "epoch": 36.91620111731844, + "grad_norm": 0.4334688186645508, + "learning_rate": 2.8413407821229048e-05, + "loss": 0.1254, + "step": 13216 + }, + { + "epoch": 36.91899441340782, + "grad_norm": 0.4634140133857727, + "learning_rate": 2.8413035381750464e-05, + "loss": 0.1257, + "step": 13217 + }, + { + "epoch": 36.92178770949721, + "grad_norm": 0.709736168384552, + "learning_rate": 2.841266294227188e-05, + "loss": 0.1236, + "step": 13218 + }, + { + "epoch": 36.92458100558659, + "grad_norm": 3.1148626804351807, + "learning_rate": 2.8412290502793298e-05, + "loss": 0.1265, + "step": 13219 + }, + { + "epoch": 36.927374301675975, + "grad_norm": 0.3892812430858612, + "learning_rate": 2.8411918063314714e-05, + "loss": 0.1122, + "step": 13220 + }, + { + "epoch": 36.93016759776536, + "grad_norm": 0.499705970287323, + "learning_rate": 2.8411545623836127e-05, + "loss": 0.0959, + "step": 13221 + }, + { + "epoch": 36.93296089385475, + "grad_norm": 0.6239756941795349, + "learning_rate": 2.8411173184357544e-05, + "loss": 0.1202, + "step": 13222 + }, + { + "epoch": 36.935754189944134, + "grad_norm": 0.6576144695281982, + "learning_rate": 2.8410800744878957e-05, + "loss": 0.0878, + "step": 13223 + }, + { + "epoch": 36.93854748603352, + "grad_norm": 0.7083747982978821, + "learning_rate": 2.8410428305400373e-05, + "loss": 0.0938, + "step": 13224 + }, + { + "epoch": 36.94134078212291, + "grad_norm": 1.140073299407959, + "learning_rate": 2.8410055865921787e-05, + "loss": 0.103, + "step": 13225 + }, + { + "epoch": 36.944134078212294, + "grad_norm": 0.5083725452423096, + "learning_rate": 2.8409683426443203e-05, + "loss": 0.0934, + "step": 13226 + }, + { + "epoch": 36.94692737430167, + "grad_norm": 0.7583737373352051, + "learning_rate": 2.840931098696462e-05, + "loss": 0.0941, + "step": 13227 + }, + { + "epoch": 36.94972067039106, + "grad_norm": 0.5554139018058777, + "learning_rate": 2.8408938547486036e-05, + "loss": 0.084, + "step": 13228 + }, + { + "epoch": 36.952513966480446, + "grad_norm": 0.7402909994125366, + "learning_rate": 2.8408566108007453e-05, + "loss": 0.0973, + "step": 13229 + }, + { + "epoch": 36.95530726256983, + "grad_norm": 0.7091984152793884, + "learning_rate": 2.8408193668528862e-05, + "loss": 0.0993, + "step": 13230 + }, + { + "epoch": 36.95810055865922, + "grad_norm": 0.7565969228744507, + "learning_rate": 2.840782122905028e-05, + "loss": 0.1019, + "step": 13231 + }, + { + "epoch": 36.960893854748605, + "grad_norm": 0.49960198998451233, + "learning_rate": 2.8407448789571696e-05, + "loss": 0.0865, + "step": 13232 + }, + { + "epoch": 36.96368715083799, + "grad_norm": 0.7727753520011902, + "learning_rate": 2.8407076350093112e-05, + "loss": 0.0817, + "step": 13233 + }, + { + "epoch": 36.96648044692738, + "grad_norm": 0.8217834234237671, + "learning_rate": 2.8406703910614525e-05, + "loss": 0.0905, + "step": 13234 + }, + { + "epoch": 36.96927374301676, + "grad_norm": 0.6883734464645386, + "learning_rate": 2.8406331471135942e-05, + "loss": 0.0864, + "step": 13235 + }, + { + "epoch": 36.97206703910614, + "grad_norm": 0.5650118589401245, + "learning_rate": 2.8405959031657355e-05, + "loss": 0.0866, + "step": 13236 + }, + { + "epoch": 36.97486033519553, + "grad_norm": 1.497057318687439, + "learning_rate": 2.840558659217877e-05, + "loss": 0.0874, + "step": 13237 + }, + { + "epoch": 36.977653631284916, + "grad_norm": 0.9132478833198547, + "learning_rate": 2.8405214152700188e-05, + "loss": 0.1133, + "step": 13238 + }, + { + "epoch": 36.9804469273743, + "grad_norm": 1.1537809371948242, + "learning_rate": 2.84048417132216e-05, + "loss": 0.133, + "step": 13239 + }, + { + "epoch": 36.98324022346369, + "grad_norm": 0.7350137829780579, + "learning_rate": 2.8404469273743018e-05, + "loss": 0.1315, + "step": 13240 + }, + { + "epoch": 36.986033519553075, + "grad_norm": 0.9472057223320007, + "learning_rate": 2.8404096834264434e-05, + "loss": 0.1178, + "step": 13241 + }, + { + "epoch": 36.98882681564246, + "grad_norm": 0.42533066868782043, + "learning_rate": 2.840372439478585e-05, + "loss": 0.0875, + "step": 13242 + }, + { + "epoch": 36.99162011173184, + "grad_norm": 0.6571608185768127, + "learning_rate": 2.840335195530726e-05, + "loss": 0.1159, + "step": 13243 + }, + { + "epoch": 36.99441340782123, + "grad_norm": 1.0682183504104614, + "learning_rate": 2.8402979515828677e-05, + "loss": 0.1039, + "step": 13244 + }, + { + "epoch": 36.997206703910614, + "grad_norm": 1.3828575611114502, + "learning_rate": 2.8402607076350094e-05, + "loss": 0.0998, + "step": 13245 + }, + { + "epoch": 37.0, + "grad_norm": 0.8961092233657837, + "learning_rate": 2.840223463687151e-05, + "loss": 0.0814, + "step": 13246 + }, + { + "epoch": 37.002793296089386, + "grad_norm": 0.74761563539505, + "learning_rate": 2.8401862197392923e-05, + "loss": 0.1385, + "step": 13247 + }, + { + "epoch": 37.00558659217877, + "grad_norm": 0.37932631373405457, + "learning_rate": 2.840148975791434e-05, + "loss": 0.1189, + "step": 13248 + }, + { + "epoch": 37.00837988826816, + "grad_norm": 0.43780866265296936, + "learning_rate": 2.8401117318435756e-05, + "loss": 0.1135, + "step": 13249 + }, + { + "epoch": 37.01117318435754, + "grad_norm": 0.3957724869251251, + "learning_rate": 2.840074487895717e-05, + "loss": 0.1077, + "step": 13250 + }, + { + "epoch": 37.013966480446925, + "grad_norm": 0.9262292981147766, + "learning_rate": 2.8400372439478586e-05, + "loss": 0.1378, + "step": 13251 + }, + { + "epoch": 37.01675977653631, + "grad_norm": 0.46259424090385437, + "learning_rate": 2.84e-05, + "loss": 0.1103, + "step": 13252 + }, + { + "epoch": 37.0195530726257, + "grad_norm": 0.5878164768218994, + "learning_rate": 2.8399627560521416e-05, + "loss": 0.1142, + "step": 13253 + }, + { + "epoch": 37.022346368715084, + "grad_norm": 0.6430202722549438, + "learning_rate": 2.8399255121042832e-05, + "loss": 0.1042, + "step": 13254 + }, + { + "epoch": 37.02513966480447, + "grad_norm": 0.4991042912006378, + "learning_rate": 2.839888268156425e-05, + "loss": 0.0864, + "step": 13255 + }, + { + "epoch": 37.02793296089386, + "grad_norm": 0.45447537302970886, + "learning_rate": 2.839851024208566e-05, + "loss": 0.123, + "step": 13256 + }, + { + "epoch": 37.03072625698324, + "grad_norm": 0.37202754616737366, + "learning_rate": 2.8398137802607075e-05, + "loss": 0.0912, + "step": 13257 + }, + { + "epoch": 37.03351955307262, + "grad_norm": 0.5839473605155945, + "learning_rate": 2.839776536312849e-05, + "loss": 0.0897, + "step": 13258 + }, + { + "epoch": 37.03631284916201, + "grad_norm": 0.3872879445552826, + "learning_rate": 2.8397392923649908e-05, + "loss": 0.0834, + "step": 13259 + }, + { + "epoch": 37.039106145251395, + "grad_norm": 0.910811185836792, + "learning_rate": 2.8397020484171325e-05, + "loss": 0.0976, + "step": 13260 + }, + { + "epoch": 37.04189944134078, + "grad_norm": 0.5208591222763062, + "learning_rate": 2.8396648044692738e-05, + "loss": 0.0856, + "step": 13261 + }, + { + "epoch": 37.04469273743017, + "grad_norm": 0.35912981629371643, + "learning_rate": 2.8396275605214154e-05, + "loss": 0.0799, + "step": 13262 + }, + { + "epoch": 37.047486033519554, + "grad_norm": 0.6543576717376709, + "learning_rate": 2.8395903165735567e-05, + "loss": 0.0988, + "step": 13263 + }, + { + "epoch": 37.05027932960894, + "grad_norm": 1.3767013549804688, + "learning_rate": 2.8395530726256984e-05, + "loss": 0.0887, + "step": 13264 + }, + { + "epoch": 37.05307262569833, + "grad_norm": 0.424280047416687, + "learning_rate": 2.8395158286778397e-05, + "loss": 0.0784, + "step": 13265 + }, + { + "epoch": 37.055865921787706, + "grad_norm": 0.49617505073547363, + "learning_rate": 2.8394785847299814e-05, + "loss": 0.0757, + "step": 13266 + }, + { + "epoch": 37.05865921787709, + "grad_norm": 0.46867772936820984, + "learning_rate": 2.839441340782123e-05, + "loss": 0.0662, + "step": 13267 + }, + { + "epoch": 37.06145251396648, + "grad_norm": 0.684806764125824, + "learning_rate": 2.8394040968342647e-05, + "loss": 0.0799, + "step": 13268 + }, + { + "epoch": 37.064245810055866, + "grad_norm": 0.6075620651245117, + "learning_rate": 2.8393668528864063e-05, + "loss": 0.0686, + "step": 13269 + }, + { + "epoch": 37.06703910614525, + "grad_norm": 1.639358401298523, + "learning_rate": 2.8393296089385473e-05, + "loss": 0.0766, + "step": 13270 + }, + { + "epoch": 37.06983240223464, + "grad_norm": 0.9062066674232483, + "learning_rate": 2.839292364990689e-05, + "loss": 0.0985, + "step": 13271 + }, + { + "epoch": 37.072625698324025, + "grad_norm": 0.5305954217910767, + "learning_rate": 2.8392551210428306e-05, + "loss": 0.143, + "step": 13272 + }, + { + "epoch": 37.07541899441341, + "grad_norm": 0.38107144832611084, + "learning_rate": 2.8392178770949723e-05, + "loss": 0.1104, + "step": 13273 + }, + { + "epoch": 37.07821229050279, + "grad_norm": 0.5925142168998718, + "learning_rate": 2.8391806331471136e-05, + "loss": 0.12, + "step": 13274 + }, + { + "epoch": 37.08100558659218, + "grad_norm": 0.45266959071159363, + "learning_rate": 2.8391433891992552e-05, + "loss": 0.1267, + "step": 13275 + }, + { + "epoch": 37.08379888268156, + "grad_norm": 0.48352015018463135, + "learning_rate": 2.8391061452513965e-05, + "loss": 0.0999, + "step": 13276 + }, + { + "epoch": 37.08659217877095, + "grad_norm": 0.388950377702713, + "learning_rate": 2.8390689013035382e-05, + "loss": 0.1228, + "step": 13277 + }, + { + "epoch": 37.089385474860336, + "grad_norm": 0.5266644358634949, + "learning_rate": 2.83903165735568e-05, + "loss": 0.1099, + "step": 13278 + }, + { + "epoch": 37.09217877094972, + "grad_norm": 0.5096754431724548, + "learning_rate": 2.8389944134078212e-05, + "loss": 0.1007, + "step": 13279 + }, + { + "epoch": 37.09497206703911, + "grad_norm": 0.6119211912155151, + "learning_rate": 2.8389571694599628e-05, + "loss": 0.1055, + "step": 13280 + }, + { + "epoch": 37.097765363128495, + "grad_norm": 0.44630804657936096, + "learning_rate": 2.8389199255121045e-05, + "loss": 0.1079, + "step": 13281 + }, + { + "epoch": 37.100558659217874, + "grad_norm": 0.4275634288787842, + "learning_rate": 2.838882681564246e-05, + "loss": 0.1094, + "step": 13282 + }, + { + "epoch": 37.10335195530726, + "grad_norm": 0.4561696946620941, + "learning_rate": 2.838845437616387e-05, + "loss": 0.0932, + "step": 13283 + }, + { + "epoch": 37.10614525139665, + "grad_norm": 0.5482856631278992, + "learning_rate": 2.8388081936685288e-05, + "loss": 0.0911, + "step": 13284 + }, + { + "epoch": 37.108938547486034, + "grad_norm": 0.9299963116645813, + "learning_rate": 2.8387709497206704e-05, + "loss": 0.0902, + "step": 13285 + }, + { + "epoch": 37.11173184357542, + "grad_norm": 0.4009720981121063, + "learning_rate": 2.838733705772812e-05, + "loss": 0.0824, + "step": 13286 + }, + { + "epoch": 37.114525139664806, + "grad_norm": 0.9100480079650879, + "learning_rate": 2.8386964618249537e-05, + "loss": 0.0713, + "step": 13287 + }, + { + "epoch": 37.11731843575419, + "grad_norm": 0.6844253540039062, + "learning_rate": 2.838659217877095e-05, + "loss": 0.0746, + "step": 13288 + }, + { + "epoch": 37.12011173184357, + "grad_norm": 0.5541043877601624, + "learning_rate": 2.8386219739292367e-05, + "loss": 0.075, + "step": 13289 + }, + { + "epoch": 37.12290502793296, + "grad_norm": 2.5420081615448, + "learning_rate": 2.838584729981378e-05, + "loss": 0.0753, + "step": 13290 + }, + { + "epoch": 37.125698324022345, + "grad_norm": 0.4890208840370178, + "learning_rate": 2.8385474860335197e-05, + "loss": 0.074, + "step": 13291 + }, + { + "epoch": 37.12849162011173, + "grad_norm": 0.865363359451294, + "learning_rate": 2.838510242085661e-05, + "loss": 0.082, + "step": 13292 + }, + { + "epoch": 37.13128491620112, + "grad_norm": 1.2129333019256592, + "learning_rate": 2.8384729981378026e-05, + "loss": 0.1015, + "step": 13293 + }, + { + "epoch": 37.134078212290504, + "grad_norm": 1.9394524097442627, + "learning_rate": 2.8384357541899443e-05, + "loss": 0.0688, + "step": 13294 + }, + { + "epoch": 37.13687150837989, + "grad_norm": 1.8177286386489868, + "learning_rate": 2.838398510242086e-05, + "loss": 0.0943, + "step": 13295 + }, + { + "epoch": 37.13966480446928, + "grad_norm": 0.7867361307144165, + "learning_rate": 2.8383612662942272e-05, + "loss": 0.1062, + "step": 13296 + }, + { + "epoch": 37.142458100558656, + "grad_norm": 0.5121132731437683, + "learning_rate": 2.8383240223463686e-05, + "loss": 0.1339, + "step": 13297 + }, + { + "epoch": 37.14525139664804, + "grad_norm": 0.46477118134498596, + "learning_rate": 2.8382867783985102e-05, + "loss": 0.1315, + "step": 13298 + }, + { + "epoch": 37.14804469273743, + "grad_norm": 0.4613207280635834, + "learning_rate": 2.838249534450652e-05, + "loss": 0.1268, + "step": 13299 + }, + { + "epoch": 37.150837988826815, + "grad_norm": 1.3940175771713257, + "learning_rate": 2.8382122905027935e-05, + "loss": 0.1207, + "step": 13300 + }, + { + "epoch": 37.1536312849162, + "grad_norm": 0.4325193762779236, + "learning_rate": 2.838175046554935e-05, + "loss": 0.0965, + "step": 13301 + }, + { + "epoch": 37.15642458100559, + "grad_norm": 0.415004163980484, + "learning_rate": 2.8381378026070765e-05, + "loss": 0.1054, + "step": 13302 + }, + { + "epoch": 37.159217877094974, + "grad_norm": 0.3888610005378723, + "learning_rate": 2.8381005586592178e-05, + "loss": 0.1094, + "step": 13303 + }, + { + "epoch": 37.16201117318436, + "grad_norm": 0.5196507573127747, + "learning_rate": 2.8380633147113595e-05, + "loss": 0.112, + "step": 13304 + }, + { + "epoch": 37.16480446927374, + "grad_norm": 0.49852314591407776, + "learning_rate": 2.838026070763501e-05, + "loss": 0.0795, + "step": 13305 + }, + { + "epoch": 37.167597765363126, + "grad_norm": 0.7088412046432495, + "learning_rate": 2.8379888268156424e-05, + "loss": 0.0973, + "step": 13306 + }, + { + "epoch": 37.17039106145251, + "grad_norm": 0.3519684076309204, + "learning_rate": 2.837951582867784e-05, + "loss": 0.0835, + "step": 13307 + }, + { + "epoch": 37.1731843575419, + "grad_norm": 0.6783422827720642, + "learning_rate": 2.8379143389199257e-05, + "loss": 0.0806, + "step": 13308 + }, + { + "epoch": 37.175977653631286, + "grad_norm": 0.7427796125411987, + "learning_rate": 2.8378770949720674e-05, + "loss": 0.0869, + "step": 13309 + }, + { + "epoch": 37.17877094972067, + "grad_norm": 0.6161222457885742, + "learning_rate": 2.8378398510242084e-05, + "loss": 0.0739, + "step": 13310 + }, + { + "epoch": 37.18156424581006, + "grad_norm": 0.5002378225326538, + "learning_rate": 2.83780260707635e-05, + "loss": 0.0853, + "step": 13311 + }, + { + "epoch": 37.184357541899445, + "grad_norm": 0.4288215637207031, + "learning_rate": 2.8377653631284917e-05, + "loss": 0.0827, + "step": 13312 + }, + { + "epoch": 37.187150837988824, + "grad_norm": 0.5347750782966614, + "learning_rate": 2.8377281191806333e-05, + "loss": 0.0902, + "step": 13313 + }, + { + "epoch": 37.18994413407821, + "grad_norm": 1.5278944969177246, + "learning_rate": 2.837690875232775e-05, + "loss": 0.0627, + "step": 13314 + }, + { + "epoch": 37.1927374301676, + "grad_norm": 0.489658385515213, + "learning_rate": 2.8376536312849163e-05, + "loss": 0.0789, + "step": 13315 + }, + { + "epoch": 37.19553072625698, + "grad_norm": 0.5824589133262634, + "learning_rate": 2.8376163873370576e-05, + "loss": 0.0842, + "step": 13316 + }, + { + "epoch": 37.19832402234637, + "grad_norm": 0.7069962620735168, + "learning_rate": 2.8375791433891993e-05, + "loss": 0.0673, + "step": 13317 + }, + { + "epoch": 37.201117318435756, + "grad_norm": 0.6756222248077393, + "learning_rate": 2.837541899441341e-05, + "loss": 0.0781, + "step": 13318 + }, + { + "epoch": 37.20391061452514, + "grad_norm": 0.5991140007972717, + "learning_rate": 2.8375046554934822e-05, + "loss": 0.0648, + "step": 13319 + }, + { + "epoch": 37.20670391061452, + "grad_norm": 1.3049763441085815, + "learning_rate": 2.837467411545624e-05, + "loss": 0.0709, + "step": 13320 + }, + { + "epoch": 37.20949720670391, + "grad_norm": 1.0473449230194092, + "learning_rate": 2.8374301675977655e-05, + "loss": 0.0869, + "step": 13321 + }, + { + "epoch": 37.212290502793294, + "grad_norm": 0.6897624135017395, + "learning_rate": 2.8373929236499072e-05, + "loss": 0.1259, + "step": 13322 + }, + { + "epoch": 37.21508379888268, + "grad_norm": 0.5800973176956177, + "learning_rate": 2.8373556797020485e-05, + "loss": 0.1242, + "step": 13323 + }, + { + "epoch": 37.21787709497207, + "grad_norm": 0.3458525538444519, + "learning_rate": 2.8373184357541898e-05, + "loss": 0.1159, + "step": 13324 + }, + { + "epoch": 37.220670391061454, + "grad_norm": 0.36441630125045776, + "learning_rate": 2.8372811918063315e-05, + "loss": 0.1088, + "step": 13325 + }, + { + "epoch": 37.22346368715084, + "grad_norm": 0.48520901799201965, + "learning_rate": 2.837243947858473e-05, + "loss": 0.111, + "step": 13326 + }, + { + "epoch": 37.226256983240226, + "grad_norm": 0.5218999981880188, + "learning_rate": 2.8372067039106148e-05, + "loss": 0.0923, + "step": 13327 + }, + { + "epoch": 37.229050279329606, + "grad_norm": 0.5178213715553284, + "learning_rate": 2.837169459962756e-05, + "loss": 0.1068, + "step": 13328 + }, + { + "epoch": 37.23184357541899, + "grad_norm": 0.4273589849472046, + "learning_rate": 2.8371322160148978e-05, + "loss": 0.086, + "step": 13329 + }, + { + "epoch": 37.23463687150838, + "grad_norm": 0.32568174600601196, + "learning_rate": 2.837094972067039e-05, + "loss": 0.0961, + "step": 13330 + }, + { + "epoch": 37.237430167597765, + "grad_norm": 0.43901607394218445, + "learning_rate": 2.8370577281191807e-05, + "loss": 0.0957, + "step": 13331 + }, + { + "epoch": 37.24022346368715, + "grad_norm": 0.4905693233013153, + "learning_rate": 2.8370204841713224e-05, + "loss": 0.1086, + "step": 13332 + }, + { + "epoch": 37.24301675977654, + "grad_norm": 0.6996617317199707, + "learning_rate": 2.8369832402234637e-05, + "loss": 0.0883, + "step": 13333 + }, + { + "epoch": 37.245810055865924, + "grad_norm": 0.424746572971344, + "learning_rate": 2.8369459962756053e-05, + "loss": 0.1, + "step": 13334 + }, + { + "epoch": 37.24860335195531, + "grad_norm": 0.6246015429496765, + "learning_rate": 2.836908752327747e-05, + "loss": 0.0867, + "step": 13335 + }, + { + "epoch": 37.25139664804469, + "grad_norm": 0.7625470161437988, + "learning_rate": 2.8368715083798883e-05, + "loss": 0.091, + "step": 13336 + }, + { + "epoch": 37.254189944134076, + "grad_norm": 0.7978317141532898, + "learning_rate": 2.8368342644320296e-05, + "loss": 0.0841, + "step": 13337 + }, + { + "epoch": 37.25698324022346, + "grad_norm": 0.6554155945777893, + "learning_rate": 2.8367970204841713e-05, + "loss": 0.092, + "step": 13338 + }, + { + "epoch": 37.25977653631285, + "grad_norm": 0.9054871797561646, + "learning_rate": 2.836759776536313e-05, + "loss": 0.1006, + "step": 13339 + }, + { + "epoch": 37.262569832402235, + "grad_norm": 0.7611244916915894, + "learning_rate": 2.8367225325884546e-05, + "loss": 0.0789, + "step": 13340 + }, + { + "epoch": 37.26536312849162, + "grad_norm": 0.748970091342926, + "learning_rate": 2.836685288640596e-05, + "loss": 0.0634, + "step": 13341 + }, + { + "epoch": 37.26815642458101, + "grad_norm": 0.8926631808280945, + "learning_rate": 2.8366480446927376e-05, + "loss": 0.0833, + "step": 13342 + }, + { + "epoch": 37.270949720670394, + "grad_norm": 0.6337838768959045, + "learning_rate": 2.836610800744879e-05, + "loss": 0.0817, + "step": 13343 + }, + { + "epoch": 37.273743016759774, + "grad_norm": 0.4261005222797394, + "learning_rate": 2.8365735567970205e-05, + "loss": 0.0568, + "step": 13344 + }, + { + "epoch": 37.27653631284916, + "grad_norm": 0.7921914458274841, + "learning_rate": 2.8365363128491622e-05, + "loss": 0.0658, + "step": 13345 + }, + { + "epoch": 37.279329608938546, + "grad_norm": 1.3355543613433838, + "learning_rate": 2.8364990689013035e-05, + "loss": 0.1009, + "step": 13346 + }, + { + "epoch": 37.28212290502793, + "grad_norm": 0.6163201332092285, + "learning_rate": 2.836461824953445e-05, + "loss": 0.1288, + "step": 13347 + }, + { + "epoch": 37.28491620111732, + "grad_norm": 0.60809725522995, + "learning_rate": 2.8364245810055868e-05, + "loss": 0.1338, + "step": 13348 + }, + { + "epoch": 37.287709497206706, + "grad_norm": 0.4188610315322876, + "learning_rate": 2.8363873370577284e-05, + "loss": 0.1319, + "step": 13349 + }, + { + "epoch": 37.29050279329609, + "grad_norm": 0.6678484678268433, + "learning_rate": 2.8363500931098694e-05, + "loss": 0.1344, + "step": 13350 + }, + { + "epoch": 37.29329608938548, + "grad_norm": 0.3436428904533386, + "learning_rate": 2.836312849162011e-05, + "loss": 0.1105, + "step": 13351 + }, + { + "epoch": 37.29608938547486, + "grad_norm": 0.4123501479625702, + "learning_rate": 2.8362756052141527e-05, + "loss": 0.0983, + "step": 13352 + }, + { + "epoch": 37.298882681564244, + "grad_norm": 0.6006595492362976, + "learning_rate": 2.8362383612662944e-05, + "loss": 0.1112, + "step": 13353 + }, + { + "epoch": 37.30167597765363, + "grad_norm": 0.46118563413619995, + "learning_rate": 2.836201117318436e-05, + "loss": 0.0929, + "step": 13354 + }, + { + "epoch": 37.30446927374302, + "grad_norm": 0.3846932053565979, + "learning_rate": 2.8361638733705774e-05, + "loss": 0.1011, + "step": 13355 + }, + { + "epoch": 37.3072625698324, + "grad_norm": 0.4042539596557617, + "learning_rate": 2.8361266294227187e-05, + "loss": 0.0863, + "step": 13356 + }, + { + "epoch": 37.31005586592179, + "grad_norm": 0.6703089475631714, + "learning_rate": 2.8360893854748603e-05, + "loss": 0.089, + "step": 13357 + }, + { + "epoch": 37.312849162011176, + "grad_norm": 0.8210147619247437, + "learning_rate": 2.836052141527002e-05, + "loss": 0.0928, + "step": 13358 + }, + { + "epoch": 37.315642458100555, + "grad_norm": 0.48430782556533813, + "learning_rate": 2.8360148975791433e-05, + "loss": 0.089, + "step": 13359 + }, + { + "epoch": 37.31843575418994, + "grad_norm": 0.5108228921890259, + "learning_rate": 2.835977653631285e-05, + "loss": 0.1011, + "step": 13360 + }, + { + "epoch": 37.32122905027933, + "grad_norm": 0.4103139340877533, + "learning_rate": 2.8359404096834266e-05, + "loss": 0.0889, + "step": 13361 + }, + { + "epoch": 37.324022346368714, + "grad_norm": 0.5112071633338928, + "learning_rate": 2.8359031657355683e-05, + "loss": 0.0867, + "step": 13362 + }, + { + "epoch": 37.3268156424581, + "grad_norm": 0.6700937747955322, + "learning_rate": 2.8358659217877096e-05, + "loss": 0.081, + "step": 13363 + }, + { + "epoch": 37.32960893854749, + "grad_norm": 1.1024699211120605, + "learning_rate": 2.835828677839851e-05, + "loss": 0.093, + "step": 13364 + }, + { + "epoch": 37.332402234636874, + "grad_norm": 0.5729321837425232, + "learning_rate": 2.8357914338919925e-05, + "loss": 0.072, + "step": 13365 + }, + { + "epoch": 37.33519553072626, + "grad_norm": 0.5340880155563354, + "learning_rate": 2.8357541899441342e-05, + "loss": 0.0751, + "step": 13366 + }, + { + "epoch": 37.33798882681564, + "grad_norm": 0.6916103363037109, + "learning_rate": 2.835716945996276e-05, + "loss": 0.0762, + "step": 13367 + }, + { + "epoch": 37.340782122905026, + "grad_norm": 0.693001925945282, + "learning_rate": 2.835679702048417e-05, + "loss": 0.0707, + "step": 13368 + }, + { + "epoch": 37.34357541899441, + "grad_norm": 0.8475944995880127, + "learning_rate": 2.8356424581005588e-05, + "loss": 0.0819, + "step": 13369 + }, + { + "epoch": 37.3463687150838, + "grad_norm": 1.7849116325378418, + "learning_rate": 2.8356052141527e-05, + "loss": 0.0717, + "step": 13370 + }, + { + "epoch": 37.349162011173185, + "grad_norm": 1.116263747215271, + "learning_rate": 2.8355679702048418e-05, + "loss": 0.1079, + "step": 13371 + }, + { + "epoch": 37.35195530726257, + "grad_norm": 0.4863429069519043, + "learning_rate": 2.8355307262569834e-05, + "loss": 0.1154, + "step": 13372 + }, + { + "epoch": 37.35474860335196, + "grad_norm": 0.4517359435558319, + "learning_rate": 2.8354934823091247e-05, + "loss": 0.0943, + "step": 13373 + }, + { + "epoch": 37.357541899441344, + "grad_norm": 0.36375024914741516, + "learning_rate": 2.8354562383612664e-05, + "loss": 0.1108, + "step": 13374 + }, + { + "epoch": 37.36033519553072, + "grad_norm": 0.685482919216156, + "learning_rate": 2.835418994413408e-05, + "loss": 0.0988, + "step": 13375 + }, + { + "epoch": 37.36312849162011, + "grad_norm": 0.4224201738834381, + "learning_rate": 2.8353817504655494e-05, + "loss": 0.1146, + "step": 13376 + }, + { + "epoch": 37.365921787709496, + "grad_norm": 0.5666427612304688, + "learning_rate": 2.8353445065176907e-05, + "loss": 0.1056, + "step": 13377 + }, + { + "epoch": 37.36871508379888, + "grad_norm": 0.4976658225059509, + "learning_rate": 2.8353072625698323e-05, + "loss": 0.1166, + "step": 13378 + }, + { + "epoch": 37.37150837988827, + "grad_norm": 1.321623682975769, + "learning_rate": 2.835270018621974e-05, + "loss": 0.0955, + "step": 13379 + }, + { + "epoch": 37.374301675977655, + "grad_norm": 0.47414878010749817, + "learning_rate": 2.8352327746741156e-05, + "loss": 0.0955, + "step": 13380 + }, + { + "epoch": 37.37709497206704, + "grad_norm": 0.6215054392814636, + "learning_rate": 2.8351955307262573e-05, + "loss": 0.11, + "step": 13381 + }, + { + "epoch": 37.37988826815643, + "grad_norm": 0.4070114493370056, + "learning_rate": 2.8351582867783986e-05, + "loss": 0.1018, + "step": 13382 + }, + { + "epoch": 37.38268156424581, + "grad_norm": 0.38315749168395996, + "learning_rate": 2.83512104283054e-05, + "loss": 0.0971, + "step": 13383 + }, + { + "epoch": 37.385474860335194, + "grad_norm": 0.522167444229126, + "learning_rate": 2.8350837988826816e-05, + "loss": 0.0904, + "step": 13384 + }, + { + "epoch": 37.38826815642458, + "grad_norm": 0.3863961696624756, + "learning_rate": 2.8350465549348232e-05, + "loss": 0.0726, + "step": 13385 + }, + { + "epoch": 37.391061452513966, + "grad_norm": 1.890020489692688, + "learning_rate": 2.8350093109869645e-05, + "loss": 0.0855, + "step": 13386 + }, + { + "epoch": 37.39385474860335, + "grad_norm": 0.49436914920806885, + "learning_rate": 2.8349720670391062e-05, + "loss": 0.0852, + "step": 13387 + }, + { + "epoch": 37.39664804469274, + "grad_norm": 0.6537649035453796, + "learning_rate": 2.834934823091248e-05, + "loss": 0.068, + "step": 13388 + }, + { + "epoch": 37.399441340782126, + "grad_norm": 0.5225786566734314, + "learning_rate": 2.8348975791433895e-05, + "loss": 0.0778, + "step": 13389 + }, + { + "epoch": 37.40223463687151, + "grad_norm": 0.6574114561080933, + "learning_rate": 2.8348603351955308e-05, + "loss": 0.068, + "step": 13390 + }, + { + "epoch": 37.40502793296089, + "grad_norm": 0.6743513941764832, + "learning_rate": 2.834823091247672e-05, + "loss": 0.0831, + "step": 13391 + }, + { + "epoch": 37.40782122905028, + "grad_norm": 0.46237507462501526, + "learning_rate": 2.8347858472998138e-05, + "loss": 0.0859, + "step": 13392 + }, + { + "epoch": 37.410614525139664, + "grad_norm": 0.529401957988739, + "learning_rate": 2.8347486033519554e-05, + "loss": 0.0712, + "step": 13393 + }, + { + "epoch": 37.41340782122905, + "grad_norm": 0.7206148505210876, + "learning_rate": 2.834711359404097e-05, + "loss": 0.088, + "step": 13394 + }, + { + "epoch": 37.41620111731844, + "grad_norm": 0.673270583152771, + "learning_rate": 2.8346741154562384e-05, + "loss": 0.0656, + "step": 13395 + }, + { + "epoch": 37.41899441340782, + "grad_norm": 3.8371360301971436, + "learning_rate": 2.8346368715083797e-05, + "loss": 0.1028, + "step": 13396 + }, + { + "epoch": 37.42178770949721, + "grad_norm": 1.7300827503204346, + "learning_rate": 2.8345996275605214e-05, + "loss": 0.1199, + "step": 13397 + }, + { + "epoch": 37.42458100558659, + "grad_norm": 0.36664679646492004, + "learning_rate": 2.834562383612663e-05, + "loss": 0.1263, + "step": 13398 + }, + { + "epoch": 37.427374301675975, + "grad_norm": 0.3068758547306061, + "learning_rate": 2.8345251396648047e-05, + "loss": 0.0956, + "step": 13399 + }, + { + "epoch": 37.43016759776536, + "grad_norm": 0.503147304058075, + "learning_rate": 2.834487895716946e-05, + "loss": 0.1094, + "step": 13400 + }, + { + "epoch": 37.43296089385475, + "grad_norm": 0.7151728868484497, + "learning_rate": 2.8344506517690877e-05, + "loss": 0.1028, + "step": 13401 + }, + { + "epoch": 37.435754189944134, + "grad_norm": 0.7435734868049622, + "learning_rate": 2.8344134078212293e-05, + "loss": 0.1255, + "step": 13402 + }, + { + "epoch": 37.43854748603352, + "grad_norm": 0.472522109746933, + "learning_rate": 2.8343761638733706e-05, + "loss": 0.1107, + "step": 13403 + }, + { + "epoch": 37.44134078212291, + "grad_norm": 0.37457749247550964, + "learning_rate": 2.834338919925512e-05, + "loss": 0.0905, + "step": 13404 + }, + { + "epoch": 37.444134078212294, + "grad_norm": 0.4437308609485626, + "learning_rate": 2.8343016759776536e-05, + "loss": 0.0931, + "step": 13405 + }, + { + "epoch": 37.44692737430167, + "grad_norm": 0.4241126775741577, + "learning_rate": 2.8342644320297952e-05, + "loss": 0.0853, + "step": 13406 + }, + { + "epoch": 37.44972067039106, + "grad_norm": 0.7503301501274109, + "learning_rate": 2.834227188081937e-05, + "loss": 0.1058, + "step": 13407 + }, + { + "epoch": 37.452513966480446, + "grad_norm": 0.43189507722854614, + "learning_rate": 2.8341899441340786e-05, + "loss": 0.0872, + "step": 13408 + }, + { + "epoch": 37.45530726256983, + "grad_norm": 0.5886490941047668, + "learning_rate": 2.83415270018622e-05, + "loss": 0.0919, + "step": 13409 + }, + { + "epoch": 37.45810055865922, + "grad_norm": 0.6694846153259277, + "learning_rate": 2.8341154562383612e-05, + "loss": 0.1075, + "step": 13410 + }, + { + "epoch": 37.460893854748605, + "grad_norm": 0.5880787968635559, + "learning_rate": 2.834078212290503e-05, + "loss": 0.0863, + "step": 13411 + }, + { + "epoch": 37.46368715083799, + "grad_norm": 0.6287997364997864, + "learning_rate": 2.8340409683426445e-05, + "loss": 0.0732, + "step": 13412 + }, + { + "epoch": 37.46648044692738, + "grad_norm": 0.43379098176956177, + "learning_rate": 2.8340037243947858e-05, + "loss": 0.0837, + "step": 13413 + }, + { + "epoch": 37.46927374301676, + "grad_norm": 0.432059645652771, + "learning_rate": 2.8339664804469275e-05, + "loss": 0.0781, + "step": 13414 + }, + { + "epoch": 37.47206703910614, + "grad_norm": 0.5733720660209656, + "learning_rate": 2.833929236499069e-05, + "loss": 0.0967, + "step": 13415 + }, + { + "epoch": 37.47486033519553, + "grad_norm": 0.5086501240730286, + "learning_rate": 2.8338919925512104e-05, + "loss": 0.079, + "step": 13416 + }, + { + "epoch": 37.477653631284916, + "grad_norm": 1.2448936700820923, + "learning_rate": 2.833854748603352e-05, + "loss": 0.0589, + "step": 13417 + }, + { + "epoch": 37.4804469273743, + "grad_norm": 0.6683579683303833, + "learning_rate": 2.8338175046554934e-05, + "loss": 0.0611, + "step": 13418 + }, + { + "epoch": 37.48324022346369, + "grad_norm": 0.5866195559501648, + "learning_rate": 2.833780260707635e-05, + "loss": 0.0747, + "step": 13419 + }, + { + "epoch": 37.486033519553075, + "grad_norm": 1.6199088096618652, + "learning_rate": 2.8337430167597767e-05, + "loss": 0.0792, + "step": 13420 + }, + { + "epoch": 37.48882681564246, + "grad_norm": 1.4266573190689087, + "learning_rate": 2.8337057728119184e-05, + "loss": 0.0934, + "step": 13421 + }, + { + "epoch": 37.49162011173184, + "grad_norm": 1.9309886693954468, + "learning_rate": 2.8336685288640597e-05, + "loss": 0.1278, + "step": 13422 + }, + { + "epoch": 37.49441340782123, + "grad_norm": 0.704800009727478, + "learning_rate": 2.833631284916201e-05, + "loss": 0.1105, + "step": 13423 + }, + { + "epoch": 37.497206703910614, + "grad_norm": 0.5257163047790527, + "learning_rate": 2.8335940409683426e-05, + "loss": 0.122, + "step": 13424 + }, + { + "epoch": 37.5, + "grad_norm": 0.5658948421478271, + "learning_rate": 2.8335567970204843e-05, + "loss": 0.102, + "step": 13425 + }, + { + "epoch": 37.502793296089386, + "grad_norm": 1.2854669094085693, + "learning_rate": 2.833519553072626e-05, + "loss": 0.1175, + "step": 13426 + }, + { + "epoch": 37.50558659217877, + "grad_norm": 0.4650685787200928, + "learning_rate": 2.8334823091247673e-05, + "loss": 0.1125, + "step": 13427 + }, + { + "epoch": 37.50837988826816, + "grad_norm": 0.41859397292137146, + "learning_rate": 2.833445065176909e-05, + "loss": 0.1228, + "step": 13428 + }, + { + "epoch": 37.51117318435754, + "grad_norm": 0.8538311719894409, + "learning_rate": 2.8334078212290506e-05, + "loss": 0.0994, + "step": 13429 + }, + { + "epoch": 37.513966480446925, + "grad_norm": 0.7266781330108643, + "learning_rate": 2.833370577281192e-05, + "loss": 0.0919, + "step": 13430 + }, + { + "epoch": 37.51675977653631, + "grad_norm": 0.46352332830429077, + "learning_rate": 2.8333333333333332e-05, + "loss": 0.0963, + "step": 13431 + }, + { + "epoch": 37.5195530726257, + "grad_norm": 0.6529415249824524, + "learning_rate": 2.833296089385475e-05, + "loss": 0.0961, + "step": 13432 + }, + { + "epoch": 37.522346368715084, + "grad_norm": 0.618046224117279, + "learning_rate": 2.8332588454376165e-05, + "loss": 0.0979, + "step": 13433 + }, + { + "epoch": 37.52513966480447, + "grad_norm": 0.5989570021629333, + "learning_rate": 2.833221601489758e-05, + "loss": 0.096, + "step": 13434 + }, + { + "epoch": 37.52793296089386, + "grad_norm": 0.5621271729469299, + "learning_rate": 2.8331843575418995e-05, + "loss": 0.086, + "step": 13435 + }, + { + "epoch": 37.53072625698324, + "grad_norm": 0.4390662908554077, + "learning_rate": 2.8331471135940408e-05, + "loss": 0.0627, + "step": 13436 + }, + { + "epoch": 37.53351955307262, + "grad_norm": 0.6179614663124084, + "learning_rate": 2.8331098696461824e-05, + "loss": 0.0813, + "step": 13437 + }, + { + "epoch": 37.53631284916201, + "grad_norm": 0.4908084571361542, + "learning_rate": 2.833072625698324e-05, + "loss": 0.0697, + "step": 13438 + }, + { + "epoch": 37.539106145251395, + "grad_norm": 0.4476589262485504, + "learning_rate": 2.8330353817504657e-05, + "loss": 0.0832, + "step": 13439 + }, + { + "epoch": 37.54189944134078, + "grad_norm": 0.47580498456954956, + "learning_rate": 2.832998137802607e-05, + "loss": 0.0648, + "step": 13440 + }, + { + "epoch": 37.54469273743017, + "grad_norm": 0.5279530882835388, + "learning_rate": 2.8329608938547487e-05, + "loss": 0.081, + "step": 13441 + }, + { + "epoch": 37.547486033519554, + "grad_norm": 0.5034967064857483, + "learning_rate": 2.8329236499068904e-05, + "loss": 0.0704, + "step": 13442 + }, + { + "epoch": 37.55027932960894, + "grad_norm": 0.7681204080581665, + "learning_rate": 2.8328864059590317e-05, + "loss": 0.0901, + "step": 13443 + }, + { + "epoch": 37.55307262569833, + "grad_norm": 1.0770174264907837, + "learning_rate": 2.832849162011173e-05, + "loss": 0.0817, + "step": 13444 + }, + { + "epoch": 37.555865921787706, + "grad_norm": 0.6479112505912781, + "learning_rate": 2.8328119180633147e-05, + "loss": 0.0739, + "step": 13445 + }, + { + "epoch": 37.55865921787709, + "grad_norm": 2.0961380004882812, + "learning_rate": 2.8327746741154563e-05, + "loss": 0.1462, + "step": 13446 + }, + { + "epoch": 37.56145251396648, + "grad_norm": 0.4542543888092041, + "learning_rate": 2.832737430167598e-05, + "loss": 0.1317, + "step": 13447 + }, + { + "epoch": 37.564245810055866, + "grad_norm": 0.42304617166519165, + "learning_rate": 2.8327001862197396e-05, + "loss": 0.1034, + "step": 13448 + }, + { + "epoch": 37.56703910614525, + "grad_norm": 2.2587661743164062, + "learning_rate": 2.832662942271881e-05, + "loss": 0.1152, + "step": 13449 + }, + { + "epoch": 37.56983240223464, + "grad_norm": 0.6085516214370728, + "learning_rate": 2.8326256983240222e-05, + "loss": 0.1089, + "step": 13450 + }, + { + "epoch": 37.572625698324025, + "grad_norm": 0.7562211751937866, + "learning_rate": 2.832588454376164e-05, + "loss": 0.139, + "step": 13451 + }, + { + "epoch": 37.57541899441341, + "grad_norm": 0.7050812244415283, + "learning_rate": 2.8325512104283056e-05, + "loss": 0.1198, + "step": 13452 + }, + { + "epoch": 37.57821229050279, + "grad_norm": 2.613920211791992, + "learning_rate": 2.832513966480447e-05, + "loss": 0.0951, + "step": 13453 + }, + { + "epoch": 37.58100558659218, + "grad_norm": 0.5703511238098145, + "learning_rate": 2.8324767225325885e-05, + "loss": 0.1243, + "step": 13454 + }, + { + "epoch": 37.58379888268156, + "grad_norm": 0.6454934477806091, + "learning_rate": 2.8324394785847302e-05, + "loss": 0.1023, + "step": 13455 + }, + { + "epoch": 37.58659217877095, + "grad_norm": 0.364760160446167, + "learning_rate": 2.8324022346368715e-05, + "loss": 0.1019, + "step": 13456 + }, + { + "epoch": 37.589385474860336, + "grad_norm": 0.5709313154220581, + "learning_rate": 2.832364990689013e-05, + "loss": 0.098, + "step": 13457 + }, + { + "epoch": 37.59217877094972, + "grad_norm": 0.5588530898094177, + "learning_rate": 2.8323277467411545e-05, + "loss": 0.0932, + "step": 13458 + }, + { + "epoch": 37.59497206703911, + "grad_norm": 0.7134435176849365, + "learning_rate": 2.832290502793296e-05, + "loss": 0.0909, + "step": 13459 + }, + { + "epoch": 37.59776536312849, + "grad_norm": 0.5087964534759521, + "learning_rate": 2.8322532588454378e-05, + "loss": 0.0866, + "step": 13460 + }, + { + "epoch": 37.600558659217874, + "grad_norm": 2.167469024658203, + "learning_rate": 2.8322160148975794e-05, + "loss": 0.0873, + "step": 13461 + }, + { + "epoch": 37.60335195530726, + "grad_norm": 0.5431972146034241, + "learning_rate": 2.8321787709497207e-05, + "loss": 0.0714, + "step": 13462 + }, + { + "epoch": 37.60614525139665, + "grad_norm": 0.6218196153640747, + "learning_rate": 2.832141527001862e-05, + "loss": 0.0896, + "step": 13463 + }, + { + "epoch": 37.608938547486034, + "grad_norm": 0.6366216540336609, + "learning_rate": 2.8321042830540037e-05, + "loss": 0.0856, + "step": 13464 + }, + { + "epoch": 37.61173184357542, + "grad_norm": 0.6926112771034241, + "learning_rate": 2.8320670391061454e-05, + "loss": 0.083, + "step": 13465 + }, + { + "epoch": 37.614525139664806, + "grad_norm": 1.5821130275726318, + "learning_rate": 2.832029795158287e-05, + "loss": 0.0878, + "step": 13466 + }, + { + "epoch": 37.61731843575419, + "grad_norm": 0.8636652827262878, + "learning_rate": 2.8319925512104283e-05, + "loss": 0.0942, + "step": 13467 + }, + { + "epoch": 37.62011173184357, + "grad_norm": 0.61641526222229, + "learning_rate": 2.83195530726257e-05, + "loss": 0.0624, + "step": 13468 + }, + { + "epoch": 37.62290502793296, + "grad_norm": 0.6826140880584717, + "learning_rate": 2.8319180633147116e-05, + "loss": 0.0932, + "step": 13469 + }, + { + "epoch": 37.625698324022345, + "grad_norm": 1.7921136617660522, + "learning_rate": 2.831880819366853e-05, + "loss": 0.0746, + "step": 13470 + }, + { + "epoch": 37.62849162011173, + "grad_norm": 1.4691777229309082, + "learning_rate": 2.8318435754189943e-05, + "loss": 0.0733, + "step": 13471 + }, + { + "epoch": 37.63128491620112, + "grad_norm": 2.1894166469573975, + "learning_rate": 2.831806331471136e-05, + "loss": 0.1993, + "step": 13472 + }, + { + "epoch": 37.634078212290504, + "grad_norm": 0.7527551054954529, + "learning_rate": 2.8317690875232776e-05, + "loss": 0.1275, + "step": 13473 + }, + { + "epoch": 37.63687150837989, + "grad_norm": 0.49783098697662354, + "learning_rate": 2.8317318435754192e-05, + "loss": 0.1285, + "step": 13474 + }, + { + "epoch": 37.63966480446928, + "grad_norm": 0.5186768174171448, + "learning_rate": 2.831694599627561e-05, + "loss": 0.1161, + "step": 13475 + }, + { + "epoch": 37.642458100558656, + "grad_norm": 0.4457460641860962, + "learning_rate": 2.831657355679702e-05, + "loss": 0.1206, + "step": 13476 + }, + { + "epoch": 37.64525139664804, + "grad_norm": 0.4691607356071472, + "learning_rate": 2.8316201117318435e-05, + "loss": 0.1025, + "step": 13477 + }, + { + "epoch": 37.64804469273743, + "grad_norm": 0.4640813171863556, + "learning_rate": 2.831582867783985e-05, + "loss": 0.0891, + "step": 13478 + }, + { + "epoch": 37.650837988826815, + "grad_norm": 0.5140835642814636, + "learning_rate": 2.8315456238361268e-05, + "loss": 0.0927, + "step": 13479 + }, + { + "epoch": 37.6536312849162, + "grad_norm": 0.53755122423172, + "learning_rate": 2.831508379888268e-05, + "loss": 0.1013, + "step": 13480 + }, + { + "epoch": 37.65642458100559, + "grad_norm": 0.5389484167098999, + "learning_rate": 2.8314711359404098e-05, + "loss": 0.0797, + "step": 13481 + }, + { + "epoch": 37.659217877094974, + "grad_norm": 0.45718449354171753, + "learning_rate": 2.8314338919925514e-05, + "loss": 0.0992, + "step": 13482 + }, + { + "epoch": 37.66201117318436, + "grad_norm": 0.4195377230644226, + "learning_rate": 2.8313966480446927e-05, + "loss": 0.0825, + "step": 13483 + }, + { + "epoch": 37.66480446927374, + "grad_norm": 0.6377631425857544, + "learning_rate": 2.8313594040968344e-05, + "loss": 0.088, + "step": 13484 + }, + { + "epoch": 37.667597765363126, + "grad_norm": 0.6691291332244873, + "learning_rate": 2.8313221601489757e-05, + "loss": 0.121, + "step": 13485 + }, + { + "epoch": 37.67039106145251, + "grad_norm": 0.5260879993438721, + "learning_rate": 2.8312849162011174e-05, + "loss": 0.0961, + "step": 13486 + }, + { + "epoch": 37.6731843575419, + "grad_norm": 0.5460242033004761, + "learning_rate": 2.831247672253259e-05, + "loss": 0.0994, + "step": 13487 + }, + { + "epoch": 37.675977653631286, + "grad_norm": 2.0270771980285645, + "learning_rate": 2.8312104283054007e-05, + "loss": 0.0847, + "step": 13488 + }, + { + "epoch": 37.67877094972067, + "grad_norm": 0.4517759084701538, + "learning_rate": 2.831173184357542e-05, + "loss": 0.0899, + "step": 13489 + }, + { + "epoch": 37.68156424581006, + "grad_norm": 0.40351763367652893, + "learning_rate": 2.8311359404096833e-05, + "loss": 0.0745, + "step": 13490 + }, + { + "epoch": 37.684357541899445, + "grad_norm": 0.5966194272041321, + "learning_rate": 2.831098696461825e-05, + "loss": 0.0867, + "step": 13491 + }, + { + "epoch": 37.687150837988824, + "grad_norm": 0.5589469075202942, + "learning_rate": 2.8310614525139666e-05, + "loss": 0.0737, + "step": 13492 + }, + { + "epoch": 37.68994413407821, + "grad_norm": 1.3126776218414307, + "learning_rate": 2.8310242085661083e-05, + "loss": 0.0851, + "step": 13493 + }, + { + "epoch": 37.6927374301676, + "grad_norm": 1.3814119100570679, + "learning_rate": 2.8309869646182496e-05, + "loss": 0.0877, + "step": 13494 + }, + { + "epoch": 37.69553072625698, + "grad_norm": 1.1588068008422852, + "learning_rate": 2.8309497206703912e-05, + "loss": 0.0776, + "step": 13495 + }, + { + "epoch": 37.69832402234637, + "grad_norm": 0.6683679223060608, + "learning_rate": 2.8309124767225325e-05, + "loss": 0.0942, + "step": 13496 + }, + { + "epoch": 37.701117318435756, + "grad_norm": 0.5763810873031616, + "learning_rate": 2.8308752327746742e-05, + "loss": 0.1245, + "step": 13497 + }, + { + "epoch": 37.70391061452514, + "grad_norm": 0.3733951449394226, + "learning_rate": 2.8308379888268155e-05, + "loss": 0.1116, + "step": 13498 + }, + { + "epoch": 37.70670391061452, + "grad_norm": 0.5441080331802368, + "learning_rate": 2.830800744878957e-05, + "loss": 0.1028, + "step": 13499 + }, + { + "epoch": 37.70949720670391, + "grad_norm": 0.721950888633728, + "learning_rate": 2.8307635009310988e-05, + "loss": 0.1203, + "step": 13500 + }, + { + "epoch": 37.712290502793294, + "grad_norm": 1.3386579751968384, + "learning_rate": 2.8307262569832405e-05, + "loss": 0.1179, + "step": 13501 + }, + { + "epoch": 37.71508379888268, + "grad_norm": 0.4122866988182068, + "learning_rate": 2.830689013035382e-05, + "loss": 0.0993, + "step": 13502 + }, + { + "epoch": 37.71787709497207, + "grad_norm": 0.8496657609939575, + "learning_rate": 2.830651769087523e-05, + "loss": 0.1042, + "step": 13503 + }, + { + "epoch": 37.720670391061454, + "grad_norm": 0.42184126377105713, + "learning_rate": 2.8306145251396648e-05, + "loss": 0.1155, + "step": 13504 + }, + { + "epoch": 37.72346368715084, + "grad_norm": 0.4997349977493286, + "learning_rate": 2.8305772811918064e-05, + "loss": 0.0942, + "step": 13505 + }, + { + "epoch": 37.726256983240226, + "grad_norm": 0.8263510465621948, + "learning_rate": 2.830540037243948e-05, + "loss": 0.0947, + "step": 13506 + }, + { + "epoch": 37.729050279329606, + "grad_norm": 0.4532104730606079, + "learning_rate": 2.8305027932960894e-05, + "loss": 0.093, + "step": 13507 + }, + { + "epoch": 37.73184357541899, + "grad_norm": 0.6759839653968811, + "learning_rate": 2.830465549348231e-05, + "loss": 0.101, + "step": 13508 + }, + { + "epoch": 37.73463687150838, + "grad_norm": 0.47681787610054016, + "learning_rate": 2.8304283054003727e-05, + "loss": 0.0886, + "step": 13509 + }, + { + "epoch": 37.737430167597765, + "grad_norm": 0.7991171479225159, + "learning_rate": 2.830391061452514e-05, + "loss": 0.0772, + "step": 13510 + }, + { + "epoch": 37.74022346368715, + "grad_norm": 0.5240347981452942, + "learning_rate": 2.8303538175046557e-05, + "loss": 0.0871, + "step": 13511 + }, + { + "epoch": 37.74301675977654, + "grad_norm": 1.0193067789077759, + "learning_rate": 2.830316573556797e-05, + "loss": 0.0934, + "step": 13512 + }, + { + "epoch": 37.745810055865924, + "grad_norm": 0.5378578305244446, + "learning_rate": 2.8302793296089386e-05, + "loss": 0.074, + "step": 13513 + }, + { + "epoch": 37.74860335195531, + "grad_norm": 0.5372198820114136, + "learning_rate": 2.8302420856610803e-05, + "loss": 0.0699, + "step": 13514 + }, + { + "epoch": 37.75139664804469, + "grad_norm": 1.436480164527893, + "learning_rate": 2.830204841713222e-05, + "loss": 0.084, + "step": 13515 + }, + { + "epoch": 37.754189944134076, + "grad_norm": 0.9659476280212402, + "learning_rate": 2.830167597765363e-05, + "loss": 0.0992, + "step": 13516 + }, + { + "epoch": 37.75698324022346, + "grad_norm": 0.6202499270439148, + "learning_rate": 2.8301303538175046e-05, + "loss": 0.089, + "step": 13517 + }, + { + "epoch": 37.75977653631285, + "grad_norm": 0.5778876543045044, + "learning_rate": 2.8300931098696462e-05, + "loss": 0.0689, + "step": 13518 + }, + { + "epoch": 37.762569832402235, + "grad_norm": 0.6211645007133484, + "learning_rate": 2.830055865921788e-05, + "loss": 0.0678, + "step": 13519 + }, + { + "epoch": 37.76536312849162, + "grad_norm": 1.2204548120498657, + "learning_rate": 2.8300186219739295e-05, + "loss": 0.0869, + "step": 13520 + }, + { + "epoch": 37.76815642458101, + "grad_norm": 1.3301259279251099, + "learning_rate": 2.829981378026071e-05, + "loss": 0.1025, + "step": 13521 + }, + { + "epoch": 37.770949720670394, + "grad_norm": 0.6454889178276062, + "learning_rate": 2.8299441340782125e-05, + "loss": 0.1185, + "step": 13522 + }, + { + "epoch": 37.773743016759774, + "grad_norm": 0.678033709526062, + "learning_rate": 2.8299068901303538e-05, + "loss": 0.1458, + "step": 13523 + }, + { + "epoch": 37.77653631284916, + "grad_norm": 1.3174045085906982, + "learning_rate": 2.8298696461824955e-05, + "loss": 0.1204, + "step": 13524 + }, + { + "epoch": 37.779329608938546, + "grad_norm": 0.5161803960800171, + "learning_rate": 2.8298324022346368e-05, + "loss": 0.0941, + "step": 13525 + }, + { + "epoch": 37.78212290502793, + "grad_norm": 3.0800626277923584, + "learning_rate": 2.8297951582867784e-05, + "loss": 0.1096, + "step": 13526 + }, + { + "epoch": 37.78491620111732, + "grad_norm": 0.6972345113754272, + "learning_rate": 2.82975791433892e-05, + "loss": 0.1123, + "step": 13527 + }, + { + "epoch": 37.787709497206706, + "grad_norm": 1.100494384765625, + "learning_rate": 2.8297206703910617e-05, + "loss": 0.1191, + "step": 13528 + }, + { + "epoch": 37.79050279329609, + "grad_norm": 0.4266567826271057, + "learning_rate": 2.829683426443203e-05, + "loss": 0.1074, + "step": 13529 + }, + { + "epoch": 37.79329608938548, + "grad_norm": 1.1610013246536255, + "learning_rate": 2.8296461824953444e-05, + "loss": 0.1241, + "step": 13530 + }, + { + "epoch": 37.79608938547486, + "grad_norm": 0.9091783165931702, + "learning_rate": 2.829608938547486e-05, + "loss": 0.097, + "step": 13531 + }, + { + "epoch": 37.798882681564244, + "grad_norm": 0.601504385471344, + "learning_rate": 2.8295716945996277e-05, + "loss": 0.1166, + "step": 13532 + }, + { + "epoch": 37.80167597765363, + "grad_norm": 0.9661040306091309, + "learning_rate": 2.8295344506517693e-05, + "loss": 0.098, + "step": 13533 + }, + { + "epoch": 37.80446927374302, + "grad_norm": 0.47214701771736145, + "learning_rate": 2.8294972067039106e-05, + "loss": 0.0988, + "step": 13534 + }, + { + "epoch": 37.8072625698324, + "grad_norm": 0.49803289771080017, + "learning_rate": 2.8294599627560523e-05, + "loss": 0.0838, + "step": 13535 + }, + { + "epoch": 37.81005586592179, + "grad_norm": 0.5302631258964539, + "learning_rate": 2.8294227188081936e-05, + "loss": 0.0729, + "step": 13536 + }, + { + "epoch": 37.812849162011176, + "grad_norm": 0.8785151839256287, + "learning_rate": 2.8293854748603353e-05, + "loss": 0.0772, + "step": 13537 + }, + { + "epoch": 37.815642458100555, + "grad_norm": 0.7076191306114197, + "learning_rate": 2.8293482309124766e-05, + "loss": 0.0923, + "step": 13538 + }, + { + "epoch": 37.81843575418994, + "grad_norm": 0.60430508852005, + "learning_rate": 2.8293109869646182e-05, + "loss": 0.0891, + "step": 13539 + }, + { + "epoch": 37.82122905027933, + "grad_norm": 1.816399335861206, + "learning_rate": 2.82927374301676e-05, + "loss": 0.0803, + "step": 13540 + }, + { + "epoch": 37.824022346368714, + "grad_norm": 0.838579535484314, + "learning_rate": 2.8292364990689015e-05, + "loss": 0.1021, + "step": 13541 + }, + { + "epoch": 37.8268156424581, + "grad_norm": 0.48321452736854553, + "learning_rate": 2.8291992551210432e-05, + "loss": 0.0738, + "step": 13542 + }, + { + "epoch": 37.82960893854749, + "grad_norm": 2.8692708015441895, + "learning_rate": 2.829162011173184e-05, + "loss": 0.0862, + "step": 13543 + }, + { + "epoch": 37.832402234636874, + "grad_norm": 0.587476909160614, + "learning_rate": 2.8291247672253258e-05, + "loss": 0.0766, + "step": 13544 + }, + { + "epoch": 37.83519553072626, + "grad_norm": 1.1105986833572388, + "learning_rate": 2.8290875232774675e-05, + "loss": 0.072, + "step": 13545 + }, + { + "epoch": 37.83798882681564, + "grad_norm": 2.053555488586426, + "learning_rate": 2.829050279329609e-05, + "loss": 0.115, + "step": 13546 + }, + { + "epoch": 37.840782122905026, + "grad_norm": 0.5315178632736206, + "learning_rate": 2.8290130353817504e-05, + "loss": 0.1307, + "step": 13547 + }, + { + "epoch": 37.84357541899441, + "grad_norm": 0.47529634833335876, + "learning_rate": 2.828975791433892e-05, + "loss": 0.1158, + "step": 13548 + }, + { + "epoch": 37.8463687150838, + "grad_norm": 0.3512713313102722, + "learning_rate": 2.8289385474860337e-05, + "loss": 0.1268, + "step": 13549 + }, + { + "epoch": 37.849162011173185, + "grad_norm": 0.44737356901168823, + "learning_rate": 2.828901303538175e-05, + "loss": 0.1351, + "step": 13550 + }, + { + "epoch": 37.85195530726257, + "grad_norm": 0.6788381338119507, + "learning_rate": 2.8288640595903167e-05, + "loss": 0.1074, + "step": 13551 + }, + { + "epoch": 37.85474860335196, + "grad_norm": 0.9020258188247681, + "learning_rate": 2.828826815642458e-05, + "loss": 0.1038, + "step": 13552 + }, + { + "epoch": 37.857541899441344, + "grad_norm": 0.4978260397911072, + "learning_rate": 2.8287895716945997e-05, + "loss": 0.1096, + "step": 13553 + }, + { + "epoch": 37.86033519553072, + "grad_norm": 0.9137853980064392, + "learning_rate": 2.8287523277467413e-05, + "loss": 0.0982, + "step": 13554 + }, + { + "epoch": 37.86312849162011, + "grad_norm": 0.5086469650268555, + "learning_rate": 2.828715083798883e-05, + "loss": 0.1034, + "step": 13555 + }, + { + "epoch": 37.865921787709496, + "grad_norm": 0.7379629015922546, + "learning_rate": 2.828677839851024e-05, + "loss": 0.1085, + "step": 13556 + }, + { + "epoch": 37.86871508379888, + "grad_norm": 1.3744231462478638, + "learning_rate": 2.8286405959031656e-05, + "loss": 0.0843, + "step": 13557 + }, + { + "epoch": 37.87150837988827, + "grad_norm": 0.49118366837501526, + "learning_rate": 2.8286033519553073e-05, + "loss": 0.1005, + "step": 13558 + }, + { + "epoch": 37.874301675977655, + "grad_norm": 0.5433399081230164, + "learning_rate": 2.828566108007449e-05, + "loss": 0.0903, + "step": 13559 + }, + { + "epoch": 37.87709497206704, + "grad_norm": 0.4455433487892151, + "learning_rate": 2.8285288640595906e-05, + "loss": 0.0804, + "step": 13560 + }, + { + "epoch": 37.87988826815643, + "grad_norm": 0.52919602394104, + "learning_rate": 2.828491620111732e-05, + "loss": 0.087, + "step": 13561 + }, + { + "epoch": 37.88268156424581, + "grad_norm": 0.5937462449073792, + "learning_rate": 2.8284543761638735e-05, + "loss": 0.0739, + "step": 13562 + }, + { + "epoch": 37.885474860335194, + "grad_norm": 0.5406915545463562, + "learning_rate": 2.828417132216015e-05, + "loss": 0.0722, + "step": 13563 + }, + { + "epoch": 37.88826815642458, + "grad_norm": 0.6284484267234802, + "learning_rate": 2.8283798882681565e-05, + "loss": 0.0586, + "step": 13564 + }, + { + "epoch": 37.891061452513966, + "grad_norm": 1.0538344383239746, + "learning_rate": 2.828342644320298e-05, + "loss": 0.1021, + "step": 13565 + }, + { + "epoch": 37.89385474860335, + "grad_norm": 0.4975854754447937, + "learning_rate": 2.8283054003724395e-05, + "loss": 0.08, + "step": 13566 + }, + { + "epoch": 37.89664804469274, + "grad_norm": 0.5422156453132629, + "learning_rate": 2.828268156424581e-05, + "loss": 0.0784, + "step": 13567 + }, + { + "epoch": 37.899441340782126, + "grad_norm": 2.4349567890167236, + "learning_rate": 2.8282309124767228e-05, + "loss": 0.0901, + "step": 13568 + }, + { + "epoch": 37.90223463687151, + "grad_norm": 6.845480918884277, + "learning_rate": 2.8281936685288644e-05, + "loss": 0.0814, + "step": 13569 + }, + { + "epoch": 37.90502793296089, + "grad_norm": 1.056230902671814, + "learning_rate": 2.8281564245810054e-05, + "loss": 0.1114, + "step": 13570 + }, + { + "epoch": 37.90782122905028, + "grad_norm": 1.1407275199890137, + "learning_rate": 2.828119180633147e-05, + "loss": 0.1104, + "step": 13571 + }, + { + "epoch": 37.910614525139664, + "grad_norm": 1.0816618204116821, + "learning_rate": 2.8280819366852887e-05, + "loss": 0.1578, + "step": 13572 + }, + { + "epoch": 37.91340782122905, + "grad_norm": 0.6514893770217896, + "learning_rate": 2.8280446927374304e-05, + "loss": 0.1263, + "step": 13573 + }, + { + "epoch": 37.91620111731844, + "grad_norm": 0.3552793562412262, + "learning_rate": 2.8280074487895717e-05, + "loss": 0.1194, + "step": 13574 + }, + { + "epoch": 37.91899441340782, + "grad_norm": 0.46572446823120117, + "learning_rate": 2.8279702048417134e-05, + "loss": 0.1245, + "step": 13575 + }, + { + "epoch": 37.92178770949721, + "grad_norm": 0.40307238698005676, + "learning_rate": 2.8279329608938547e-05, + "loss": 0.1126, + "step": 13576 + }, + { + "epoch": 37.92458100558659, + "grad_norm": 0.931151270866394, + "learning_rate": 2.8278957169459963e-05, + "loss": 0.1095, + "step": 13577 + }, + { + "epoch": 37.927374301675975, + "grad_norm": 0.5486778616905212, + "learning_rate": 2.827858472998138e-05, + "loss": 0.1035, + "step": 13578 + }, + { + "epoch": 37.93016759776536, + "grad_norm": 0.2694908380508423, + "learning_rate": 2.8278212290502793e-05, + "loss": 0.0905, + "step": 13579 + }, + { + "epoch": 37.93296089385475, + "grad_norm": 0.449146568775177, + "learning_rate": 2.827783985102421e-05, + "loss": 0.1, + "step": 13580 + }, + { + "epoch": 37.935754189944134, + "grad_norm": 0.754401445388794, + "learning_rate": 2.8277467411545626e-05, + "loss": 0.1009, + "step": 13581 + }, + { + "epoch": 37.93854748603352, + "grad_norm": 0.639654815196991, + "learning_rate": 2.8277094972067042e-05, + "loss": 0.0942, + "step": 13582 + }, + { + "epoch": 37.94134078212291, + "grad_norm": 0.9236147403717041, + "learning_rate": 2.8276722532588452e-05, + "loss": 0.0889, + "step": 13583 + }, + { + "epoch": 37.944134078212294, + "grad_norm": 1.026513695716858, + "learning_rate": 2.827635009310987e-05, + "loss": 0.0894, + "step": 13584 + }, + { + "epoch": 37.94692737430167, + "grad_norm": 0.43466776609420776, + "learning_rate": 2.8275977653631285e-05, + "loss": 0.0806, + "step": 13585 + }, + { + "epoch": 37.94972067039106, + "grad_norm": 0.4979175925254822, + "learning_rate": 2.8275605214152702e-05, + "loss": 0.0805, + "step": 13586 + }, + { + "epoch": 37.952513966480446, + "grad_norm": 0.5099448561668396, + "learning_rate": 2.827523277467412e-05, + "loss": 0.0756, + "step": 13587 + }, + { + "epoch": 37.95530726256983, + "grad_norm": 0.818507969379425, + "learning_rate": 2.827486033519553e-05, + "loss": 0.0806, + "step": 13588 + }, + { + "epoch": 37.95810055865922, + "grad_norm": 0.5272806882858276, + "learning_rate": 2.8274487895716948e-05, + "loss": 0.0785, + "step": 13589 + }, + { + "epoch": 37.960893854748605, + "grad_norm": 0.7429022192955017, + "learning_rate": 2.827411545623836e-05, + "loss": 0.0851, + "step": 13590 + }, + { + "epoch": 37.96368715083799, + "grad_norm": 1.2446386814117432, + "learning_rate": 2.8273743016759778e-05, + "loss": 0.0571, + "step": 13591 + }, + { + "epoch": 37.96648044692738, + "grad_norm": 0.5699859857559204, + "learning_rate": 2.827337057728119e-05, + "loss": 0.0644, + "step": 13592 + }, + { + "epoch": 37.96927374301676, + "grad_norm": 0.9957475662231445, + "learning_rate": 2.8272998137802607e-05, + "loss": 0.0865, + "step": 13593 + }, + { + "epoch": 37.97206703910614, + "grad_norm": 0.5944711565971375, + "learning_rate": 2.8272625698324024e-05, + "loss": 0.0766, + "step": 13594 + }, + { + "epoch": 37.97486033519553, + "grad_norm": 0.7629641890525818, + "learning_rate": 2.827225325884544e-05, + "loss": 0.0716, + "step": 13595 + }, + { + "epoch": 37.977653631284916, + "grad_norm": 1.5692315101623535, + "learning_rate": 2.8271880819366854e-05, + "loss": 0.0926, + "step": 13596 + }, + { + "epoch": 37.9804469273743, + "grad_norm": 0.5939427018165588, + "learning_rate": 2.8271508379888267e-05, + "loss": 0.1423, + "step": 13597 + }, + { + "epoch": 37.98324022346369, + "grad_norm": 0.5194096565246582, + "learning_rate": 2.8271135940409683e-05, + "loss": 0.1203, + "step": 13598 + }, + { + "epoch": 37.986033519553075, + "grad_norm": 0.6489980220794678, + "learning_rate": 2.82707635009311e-05, + "loss": 0.099, + "step": 13599 + }, + { + "epoch": 37.98882681564246, + "grad_norm": 1.8672524690628052, + "learning_rate": 2.8270391061452516e-05, + "loss": 0.1208, + "step": 13600 + }, + { + "epoch": 37.99162011173184, + "grad_norm": 0.8116322755813599, + "learning_rate": 2.827001862197393e-05, + "loss": 0.0923, + "step": 13601 + }, + { + "epoch": 37.99441340782123, + "grad_norm": 0.7780147194862366, + "learning_rate": 2.8269646182495346e-05, + "loss": 0.0892, + "step": 13602 + }, + { + "epoch": 37.997206703910614, + "grad_norm": 1.1590795516967773, + "learning_rate": 2.826927374301676e-05, + "loss": 0.0803, + "step": 13603 + }, + { + "epoch": 38.0, + "grad_norm": 0.8958903551101685, + "learning_rate": 2.8268901303538176e-05, + "loss": 0.0854, + "step": 13604 + }, + { + "epoch": 38.002793296089386, + "grad_norm": 0.7198182344436646, + "learning_rate": 2.8268528864059592e-05, + "loss": 0.1597, + "step": 13605 + }, + { + "epoch": 38.00558659217877, + "grad_norm": 0.893574059009552, + "learning_rate": 2.8268156424581005e-05, + "loss": 0.1118, + "step": 13606 + }, + { + "epoch": 38.00837988826816, + "grad_norm": 0.37087881565093994, + "learning_rate": 2.8267783985102422e-05, + "loss": 0.1334, + "step": 13607 + }, + { + "epoch": 38.01117318435754, + "grad_norm": 0.3942200839519501, + "learning_rate": 2.826741154562384e-05, + "loss": 0.1025, + "step": 13608 + }, + { + "epoch": 38.013966480446925, + "grad_norm": 0.6616092920303345, + "learning_rate": 2.8267039106145255e-05, + "loss": 0.1067, + "step": 13609 + }, + { + "epoch": 38.01675977653631, + "grad_norm": 1.5576893091201782, + "learning_rate": 2.8266666666666665e-05, + "loss": 0.1003, + "step": 13610 + }, + { + "epoch": 38.0195530726257, + "grad_norm": 0.5423663258552551, + "learning_rate": 2.826629422718808e-05, + "loss": 0.0944, + "step": 13611 + }, + { + "epoch": 38.022346368715084, + "grad_norm": 0.7358835339546204, + "learning_rate": 2.8265921787709498e-05, + "loss": 0.1091, + "step": 13612 + }, + { + "epoch": 38.02513966480447, + "grad_norm": 1.841209053993225, + "learning_rate": 2.8265549348230914e-05, + "loss": 0.1033, + "step": 13613 + }, + { + "epoch": 38.02793296089386, + "grad_norm": 0.5449988842010498, + "learning_rate": 2.826517690875233e-05, + "loss": 0.1002, + "step": 13614 + }, + { + "epoch": 38.03072625698324, + "grad_norm": 1.0527795553207397, + "learning_rate": 2.8264804469273744e-05, + "loss": 0.0782, + "step": 13615 + }, + { + "epoch": 38.03351955307262, + "grad_norm": 0.8708153367042542, + "learning_rate": 2.8264432029795157e-05, + "loss": 0.0923, + "step": 13616 + }, + { + "epoch": 38.03631284916201, + "grad_norm": 0.5480820536613464, + "learning_rate": 2.8264059590316574e-05, + "loss": 0.0742, + "step": 13617 + }, + { + "epoch": 38.039106145251395, + "grad_norm": 0.9960919618606567, + "learning_rate": 2.826368715083799e-05, + "loss": 0.0832, + "step": 13618 + }, + { + "epoch": 38.04189944134078, + "grad_norm": 0.5023010969161987, + "learning_rate": 2.8263314711359403e-05, + "loss": 0.0854, + "step": 13619 + }, + { + "epoch": 38.04469273743017, + "grad_norm": 0.4977894127368927, + "learning_rate": 2.826294227188082e-05, + "loss": 0.08, + "step": 13620 + }, + { + "epoch": 38.047486033519554, + "grad_norm": 0.5609042048454285, + "learning_rate": 2.8262569832402237e-05, + "loss": 0.063, + "step": 13621 + }, + { + "epoch": 38.05027932960894, + "grad_norm": 1.672320008277893, + "learning_rate": 2.8262197392923653e-05, + "loss": 0.0659, + "step": 13622 + }, + { + "epoch": 38.05307262569833, + "grad_norm": 0.5528787970542908, + "learning_rate": 2.8261824953445066e-05, + "loss": 0.0887, + "step": 13623 + }, + { + "epoch": 38.055865921787706, + "grad_norm": 1.5771676301956177, + "learning_rate": 2.826145251396648e-05, + "loss": 0.0774, + "step": 13624 + }, + { + "epoch": 38.05865921787709, + "grad_norm": 0.7770093083381653, + "learning_rate": 2.8261080074487896e-05, + "loss": 0.0618, + "step": 13625 + }, + { + "epoch": 38.06145251396648, + "grad_norm": 0.7451122403144836, + "learning_rate": 2.8260707635009312e-05, + "loss": 0.0577, + "step": 13626 + }, + { + "epoch": 38.064245810055866, + "grad_norm": 0.6932182312011719, + "learning_rate": 2.826033519553073e-05, + "loss": 0.0737, + "step": 13627 + }, + { + "epoch": 38.06703910614525, + "grad_norm": 1.5139777660369873, + "learning_rate": 2.8259962756052142e-05, + "loss": 0.0694, + "step": 13628 + }, + { + "epoch": 38.06983240223464, + "grad_norm": 1.4735448360443115, + "learning_rate": 2.8259590316573555e-05, + "loss": 0.0989, + "step": 13629 + }, + { + "epoch": 38.072625698324025, + "grad_norm": 1.1085277795791626, + "learning_rate": 2.8259217877094972e-05, + "loss": 0.1209, + "step": 13630 + }, + { + "epoch": 38.07541899441341, + "grad_norm": 0.3986848294734955, + "learning_rate": 2.825884543761639e-05, + "loss": 0.1104, + "step": 13631 + }, + { + "epoch": 38.07821229050279, + "grad_norm": 0.44649389386177063, + "learning_rate": 2.82584729981378e-05, + "loss": 0.1067, + "step": 13632 + }, + { + "epoch": 38.08100558659218, + "grad_norm": 0.689566433429718, + "learning_rate": 2.8258100558659218e-05, + "loss": 0.1, + "step": 13633 + }, + { + "epoch": 38.08379888268156, + "grad_norm": 1.3156629800796509, + "learning_rate": 2.8257728119180635e-05, + "loss": 0.117, + "step": 13634 + }, + { + "epoch": 38.08659217877095, + "grad_norm": 0.5990111827850342, + "learning_rate": 2.825735567970205e-05, + "loss": 0.0831, + "step": 13635 + }, + { + "epoch": 38.089385474860336, + "grad_norm": 0.5798774361610413, + "learning_rate": 2.8256983240223464e-05, + "loss": 0.0903, + "step": 13636 + }, + { + "epoch": 38.09217877094972, + "grad_norm": 0.5133525729179382, + "learning_rate": 2.8256610800744877e-05, + "loss": 0.0883, + "step": 13637 + }, + { + "epoch": 38.09497206703911, + "grad_norm": 0.4218067526817322, + "learning_rate": 2.8256238361266294e-05, + "loss": 0.0781, + "step": 13638 + }, + { + "epoch": 38.097765363128495, + "grad_norm": 0.4693346917629242, + "learning_rate": 2.825586592178771e-05, + "loss": 0.0987, + "step": 13639 + }, + { + "epoch": 38.100558659217874, + "grad_norm": 0.5241656303405762, + "learning_rate": 2.8255493482309127e-05, + "loss": 0.0797, + "step": 13640 + }, + { + "epoch": 38.10335195530726, + "grad_norm": 1.5795722007751465, + "learning_rate": 2.825512104283054e-05, + "loss": 0.0955, + "step": 13641 + }, + { + "epoch": 38.10614525139665, + "grad_norm": 1.8565741777420044, + "learning_rate": 2.8254748603351957e-05, + "loss": 0.0908, + "step": 13642 + }, + { + "epoch": 38.108938547486034, + "grad_norm": 0.6266420483589172, + "learning_rate": 2.825437616387337e-05, + "loss": 0.0917, + "step": 13643 + }, + { + "epoch": 38.11173184357542, + "grad_norm": 0.49125924706459045, + "learning_rate": 2.8254003724394786e-05, + "loss": 0.0922, + "step": 13644 + }, + { + "epoch": 38.114525139664806, + "grad_norm": 1.1133333444595337, + "learning_rate": 2.8253631284916203e-05, + "loss": 0.0682, + "step": 13645 + }, + { + "epoch": 38.11731843575419, + "grad_norm": 0.5136995911598206, + "learning_rate": 2.8253258845437616e-05, + "loss": 0.0777, + "step": 13646 + }, + { + "epoch": 38.12011173184357, + "grad_norm": 0.6037970185279846, + "learning_rate": 2.8252886405959033e-05, + "loss": 0.0809, + "step": 13647 + }, + { + "epoch": 38.12290502793296, + "grad_norm": 0.4830288290977478, + "learning_rate": 2.825251396648045e-05, + "loss": 0.0784, + "step": 13648 + }, + { + "epoch": 38.125698324022345, + "grad_norm": 2.425668478012085, + "learning_rate": 2.8252141527001862e-05, + "loss": 0.0558, + "step": 13649 + }, + { + "epoch": 38.12849162011173, + "grad_norm": 0.6688857674598694, + "learning_rate": 2.8251769087523275e-05, + "loss": 0.078, + "step": 13650 + }, + { + "epoch": 38.13128491620112, + "grad_norm": 1.0375055074691772, + "learning_rate": 2.8251396648044692e-05, + "loss": 0.0747, + "step": 13651 + }, + { + "epoch": 38.134078212290504, + "grad_norm": 0.7346288561820984, + "learning_rate": 2.825102420856611e-05, + "loss": 0.0718, + "step": 13652 + }, + { + "epoch": 38.13687150837989, + "grad_norm": 0.8996039628982544, + "learning_rate": 2.8250651769087525e-05, + "loss": 0.0792, + "step": 13653 + }, + { + "epoch": 38.13966480446928, + "grad_norm": 0.8932623863220215, + "learning_rate": 2.825027932960894e-05, + "loss": 0.0857, + "step": 13654 + }, + { + "epoch": 38.142458100558656, + "grad_norm": 0.378711462020874, + "learning_rate": 2.8249906890130355e-05, + "loss": 0.1138, + "step": 13655 + }, + { + "epoch": 38.14525139664804, + "grad_norm": 0.49095889925956726, + "learning_rate": 2.8249534450651768e-05, + "loss": 0.1177, + "step": 13656 + }, + { + "epoch": 38.14804469273743, + "grad_norm": 0.3813093304634094, + "learning_rate": 2.8249162011173184e-05, + "loss": 0.1115, + "step": 13657 + }, + { + "epoch": 38.150837988826815, + "grad_norm": 1.2772905826568604, + "learning_rate": 2.82487895716946e-05, + "loss": 0.1208, + "step": 13658 + }, + { + "epoch": 38.1536312849162, + "grad_norm": 0.4416792392730713, + "learning_rate": 2.8248417132216014e-05, + "loss": 0.0938, + "step": 13659 + }, + { + "epoch": 38.15642458100559, + "grad_norm": 0.6753383278846741, + "learning_rate": 2.824804469273743e-05, + "loss": 0.105, + "step": 13660 + }, + { + "epoch": 38.159217877094974, + "grad_norm": 0.6458865404129028, + "learning_rate": 2.8247672253258847e-05, + "loss": 0.1056, + "step": 13661 + }, + { + "epoch": 38.16201117318436, + "grad_norm": 0.559935986995697, + "learning_rate": 2.8247299813780264e-05, + "loss": 0.101, + "step": 13662 + }, + { + "epoch": 38.16480446927374, + "grad_norm": 0.5054547786712646, + "learning_rate": 2.8246927374301677e-05, + "loss": 0.0761, + "step": 13663 + }, + { + "epoch": 38.167597765363126, + "grad_norm": 0.45961886644363403, + "learning_rate": 2.824655493482309e-05, + "loss": 0.0965, + "step": 13664 + }, + { + "epoch": 38.17039106145251, + "grad_norm": 0.4697854518890381, + "learning_rate": 2.8246182495344507e-05, + "loss": 0.0898, + "step": 13665 + }, + { + "epoch": 38.1731843575419, + "grad_norm": 0.577907145023346, + "learning_rate": 2.8245810055865923e-05, + "loss": 0.0934, + "step": 13666 + }, + { + "epoch": 38.175977653631286, + "grad_norm": 0.4840467870235443, + "learning_rate": 2.824543761638734e-05, + "loss": 0.0776, + "step": 13667 + }, + { + "epoch": 38.17877094972067, + "grad_norm": 0.8989623785018921, + "learning_rate": 2.8245065176908753e-05, + "loss": 0.0881, + "step": 13668 + }, + { + "epoch": 38.18156424581006, + "grad_norm": 0.5477580428123474, + "learning_rate": 2.8244692737430166e-05, + "loss": 0.0836, + "step": 13669 + }, + { + "epoch": 38.184357541899445, + "grad_norm": 0.7397028803825378, + "learning_rate": 2.8244320297951582e-05, + "loss": 0.068, + "step": 13670 + }, + { + "epoch": 38.187150837988824, + "grad_norm": 0.5579710602760315, + "learning_rate": 2.8243947858473e-05, + "loss": 0.0811, + "step": 13671 + }, + { + "epoch": 38.18994413407821, + "grad_norm": 0.40642082691192627, + "learning_rate": 2.8243575418994415e-05, + "loss": 0.0666, + "step": 13672 + }, + { + "epoch": 38.1927374301676, + "grad_norm": 0.541372537612915, + "learning_rate": 2.824320297951583e-05, + "loss": 0.062, + "step": 13673 + }, + { + "epoch": 38.19553072625698, + "grad_norm": 0.4557761251926422, + "learning_rate": 2.8242830540037245e-05, + "loss": 0.0749, + "step": 13674 + }, + { + "epoch": 38.19832402234637, + "grad_norm": 2.782489538192749, + "learning_rate": 2.8242458100558662e-05, + "loss": 0.0729, + "step": 13675 + }, + { + "epoch": 38.201117318435756, + "grad_norm": 0.8448912501335144, + "learning_rate": 2.8242085661080075e-05, + "loss": 0.075, + "step": 13676 + }, + { + "epoch": 38.20391061452514, + "grad_norm": 0.615928053855896, + "learning_rate": 2.8241713221601488e-05, + "loss": 0.057, + "step": 13677 + }, + { + "epoch": 38.20670391061452, + "grad_norm": 0.7912046313285828, + "learning_rate": 2.8241340782122905e-05, + "loss": 0.0759, + "step": 13678 + }, + { + "epoch": 38.20949720670391, + "grad_norm": 1.0558451414108276, + "learning_rate": 2.824096834264432e-05, + "loss": 0.0853, + "step": 13679 + }, + { + "epoch": 38.212290502793294, + "grad_norm": 0.37920260429382324, + "learning_rate": 2.8240595903165738e-05, + "loss": 0.1139, + "step": 13680 + }, + { + "epoch": 38.21508379888268, + "grad_norm": 0.694470226764679, + "learning_rate": 2.8240223463687154e-05, + "loss": 0.1109, + "step": 13681 + }, + { + "epoch": 38.21787709497207, + "grad_norm": 0.4372653663158417, + "learning_rate": 2.8239851024208567e-05, + "loss": 0.1277, + "step": 13682 + }, + { + "epoch": 38.220670391061454, + "grad_norm": 0.5218306183815002, + "learning_rate": 2.823947858472998e-05, + "loss": 0.1122, + "step": 13683 + }, + { + "epoch": 38.22346368715084, + "grad_norm": 0.39648476243019104, + "learning_rate": 2.8239106145251397e-05, + "loss": 0.1024, + "step": 13684 + }, + { + "epoch": 38.226256983240226, + "grad_norm": 0.49528464674949646, + "learning_rate": 2.8238733705772814e-05, + "loss": 0.107, + "step": 13685 + }, + { + "epoch": 38.229050279329606, + "grad_norm": 0.4020237326622009, + "learning_rate": 2.8238361266294227e-05, + "loss": 0.09, + "step": 13686 + }, + { + "epoch": 38.23184357541899, + "grad_norm": 0.9500442743301392, + "learning_rate": 2.8237988826815643e-05, + "loss": 0.0947, + "step": 13687 + }, + { + "epoch": 38.23463687150838, + "grad_norm": 1.2254416942596436, + "learning_rate": 2.823761638733706e-05, + "loss": 0.1017, + "step": 13688 + }, + { + "epoch": 38.237430167597765, + "grad_norm": 0.573433518409729, + "learning_rate": 2.8237243947858473e-05, + "loss": 0.0968, + "step": 13689 + }, + { + "epoch": 38.24022346368715, + "grad_norm": 0.4361385405063629, + "learning_rate": 2.823687150837989e-05, + "loss": 0.0667, + "step": 13690 + }, + { + "epoch": 38.24301675977654, + "grad_norm": 0.9446371793746948, + "learning_rate": 2.8236499068901303e-05, + "loss": 0.0832, + "step": 13691 + }, + { + "epoch": 38.245810055865924, + "grad_norm": 0.9464927911758423, + "learning_rate": 2.823612662942272e-05, + "loss": 0.0936, + "step": 13692 + }, + { + "epoch": 38.24860335195531, + "grad_norm": 0.7366499900817871, + "learning_rate": 2.8235754189944136e-05, + "loss": 0.0939, + "step": 13693 + }, + { + "epoch": 38.25139664804469, + "grad_norm": 0.6689229011535645, + "learning_rate": 2.8235381750465552e-05, + "loss": 0.0875, + "step": 13694 + }, + { + "epoch": 38.254189944134076, + "grad_norm": 0.6350197792053223, + "learning_rate": 2.8235009310986965e-05, + "loss": 0.0712, + "step": 13695 + }, + { + "epoch": 38.25698324022346, + "grad_norm": 0.38538724184036255, + "learning_rate": 2.823463687150838e-05, + "loss": 0.0752, + "step": 13696 + }, + { + "epoch": 38.25977653631285, + "grad_norm": 1.7887818813323975, + "learning_rate": 2.8234264432029795e-05, + "loss": 0.0693, + "step": 13697 + }, + { + "epoch": 38.262569832402235, + "grad_norm": 1.0628087520599365, + "learning_rate": 2.823389199255121e-05, + "loss": 0.0703, + "step": 13698 + }, + { + "epoch": 38.26536312849162, + "grad_norm": 0.6115090847015381, + "learning_rate": 2.8233519553072628e-05, + "loss": 0.0882, + "step": 13699 + }, + { + "epoch": 38.26815642458101, + "grad_norm": 0.5248382091522217, + "learning_rate": 2.823314711359404e-05, + "loss": 0.0687, + "step": 13700 + }, + { + "epoch": 38.270949720670394, + "grad_norm": 0.67684406042099, + "learning_rate": 2.8232774674115458e-05, + "loss": 0.0637, + "step": 13701 + }, + { + "epoch": 38.273743016759774, + "grad_norm": 1.2188458442687988, + "learning_rate": 2.8232402234636874e-05, + "loss": 0.066, + "step": 13702 + }, + { + "epoch": 38.27653631284916, + "grad_norm": 0.8558314442634583, + "learning_rate": 2.8232029795158287e-05, + "loss": 0.0658, + "step": 13703 + }, + { + "epoch": 38.279329608938546, + "grad_norm": 0.9805386066436768, + "learning_rate": 2.82316573556797e-05, + "loss": 0.0978, + "step": 13704 + }, + { + "epoch": 38.28212290502793, + "grad_norm": 0.578081488609314, + "learning_rate": 2.8231284916201117e-05, + "loss": 0.1263, + "step": 13705 + }, + { + "epoch": 38.28491620111732, + "grad_norm": 0.49524828791618347, + "learning_rate": 2.8230912476722534e-05, + "loss": 0.0998, + "step": 13706 + }, + { + "epoch": 38.287709497206706, + "grad_norm": 0.8349334001541138, + "learning_rate": 2.823054003724395e-05, + "loss": 0.1322, + "step": 13707 + }, + { + "epoch": 38.29050279329609, + "grad_norm": 0.9149228930473328, + "learning_rate": 2.8230167597765367e-05, + "loss": 0.1065, + "step": 13708 + }, + { + "epoch": 38.29329608938548, + "grad_norm": 1.0362051725387573, + "learning_rate": 2.8229795158286776e-05, + "loss": 0.1195, + "step": 13709 + }, + { + "epoch": 38.29608938547486, + "grad_norm": 0.8216106295585632, + "learning_rate": 2.8229422718808193e-05, + "loss": 0.1023, + "step": 13710 + }, + { + "epoch": 38.298882681564244, + "grad_norm": 0.5101133584976196, + "learning_rate": 2.822905027932961e-05, + "loss": 0.1119, + "step": 13711 + }, + { + "epoch": 38.30167597765363, + "grad_norm": 1.2717500925064087, + "learning_rate": 2.8228677839851026e-05, + "loss": 0.1024, + "step": 13712 + }, + { + "epoch": 38.30446927374302, + "grad_norm": 0.4609566032886505, + "learning_rate": 2.822830540037244e-05, + "loss": 0.0885, + "step": 13713 + }, + { + "epoch": 38.3072625698324, + "grad_norm": 0.8391607999801636, + "learning_rate": 2.8227932960893856e-05, + "loss": 0.0964, + "step": 13714 + }, + { + "epoch": 38.31005586592179, + "grad_norm": 1.4443435668945312, + "learning_rate": 2.8227560521415272e-05, + "loss": 0.081, + "step": 13715 + }, + { + "epoch": 38.312849162011176, + "grad_norm": 0.8448764085769653, + "learning_rate": 2.8227188081936685e-05, + "loss": 0.1022, + "step": 13716 + }, + { + "epoch": 38.315642458100555, + "grad_norm": 0.4799998104572296, + "learning_rate": 2.8226815642458102e-05, + "loss": 0.0798, + "step": 13717 + }, + { + "epoch": 38.31843575418994, + "grad_norm": 0.45563554763793945, + "learning_rate": 2.8226443202979515e-05, + "loss": 0.0934, + "step": 13718 + }, + { + "epoch": 38.32122905027933, + "grad_norm": 0.5778693556785583, + "learning_rate": 2.822607076350093e-05, + "loss": 0.0816, + "step": 13719 + }, + { + "epoch": 38.324022346368714, + "grad_norm": 0.4454435110092163, + "learning_rate": 2.8225698324022348e-05, + "loss": 0.0824, + "step": 13720 + }, + { + "epoch": 38.3268156424581, + "grad_norm": 0.5092951655387878, + "learning_rate": 2.8225325884543765e-05, + "loss": 0.0774, + "step": 13721 + }, + { + "epoch": 38.32960893854749, + "grad_norm": 2.418661117553711, + "learning_rate": 2.8224953445065178e-05, + "loss": 0.0668, + "step": 13722 + }, + { + "epoch": 38.332402234636874, + "grad_norm": 0.5884087681770325, + "learning_rate": 2.822458100558659e-05, + "loss": 0.0876, + "step": 13723 + }, + { + "epoch": 38.33519553072626, + "grad_norm": 0.46387264132499695, + "learning_rate": 2.8224208566108008e-05, + "loss": 0.0781, + "step": 13724 + }, + { + "epoch": 38.33798882681564, + "grad_norm": 0.5995131134986877, + "learning_rate": 2.8223836126629424e-05, + "loss": 0.0644, + "step": 13725 + }, + { + "epoch": 38.340782122905026, + "grad_norm": 0.6904237866401672, + "learning_rate": 2.8223463687150837e-05, + "loss": 0.0881, + "step": 13726 + }, + { + "epoch": 38.34357541899441, + "grad_norm": 1.140262484550476, + "learning_rate": 2.8223091247672254e-05, + "loss": 0.0847, + "step": 13727 + }, + { + "epoch": 38.3463687150838, + "grad_norm": 0.8256446123123169, + "learning_rate": 2.822271880819367e-05, + "loss": 0.0815, + "step": 13728 + }, + { + "epoch": 38.349162011173185, + "grad_norm": 1.4070671796798706, + "learning_rate": 2.8222346368715083e-05, + "loss": 0.092, + "step": 13729 + }, + { + "epoch": 38.35195530726257, + "grad_norm": 0.3839598298072815, + "learning_rate": 2.82219739292365e-05, + "loss": 0.1188, + "step": 13730 + }, + { + "epoch": 38.35474860335196, + "grad_norm": 0.4576537609100342, + "learning_rate": 2.8221601489757913e-05, + "loss": 0.1068, + "step": 13731 + }, + { + "epoch": 38.357541899441344, + "grad_norm": 0.454610675573349, + "learning_rate": 2.822122905027933e-05, + "loss": 0.117, + "step": 13732 + }, + { + "epoch": 38.36033519553072, + "grad_norm": 0.3502713143825531, + "learning_rate": 2.8220856610800746e-05, + "loss": 0.1086, + "step": 13733 + }, + { + "epoch": 38.36312849162011, + "grad_norm": 0.4660640060901642, + "learning_rate": 2.8220484171322163e-05, + "loss": 0.101, + "step": 13734 + }, + { + "epoch": 38.365921787709496, + "grad_norm": 0.5614672303199768, + "learning_rate": 2.8220111731843576e-05, + "loss": 0.1131, + "step": 13735 + }, + { + "epoch": 38.36871508379888, + "grad_norm": 0.5471348166465759, + "learning_rate": 2.821973929236499e-05, + "loss": 0.1126, + "step": 13736 + }, + { + "epoch": 38.37150837988827, + "grad_norm": 0.4048161208629608, + "learning_rate": 2.8219366852886406e-05, + "loss": 0.1018, + "step": 13737 + }, + { + "epoch": 38.374301675977655, + "grad_norm": 0.4353809952735901, + "learning_rate": 2.8218994413407822e-05, + "loss": 0.0969, + "step": 13738 + }, + { + "epoch": 38.37709497206704, + "grad_norm": 0.6568073034286499, + "learning_rate": 2.821862197392924e-05, + "loss": 0.0982, + "step": 13739 + }, + { + "epoch": 38.37988826815643, + "grad_norm": 0.6703651547431946, + "learning_rate": 2.8218249534450652e-05, + "loss": 0.0996, + "step": 13740 + }, + { + "epoch": 38.38268156424581, + "grad_norm": 0.5202404260635376, + "learning_rate": 2.821787709497207e-05, + "loss": 0.077, + "step": 13741 + }, + { + "epoch": 38.385474860335194, + "grad_norm": 1.2995223999023438, + "learning_rate": 2.8217504655493485e-05, + "loss": 0.0766, + "step": 13742 + }, + { + "epoch": 38.38826815642458, + "grad_norm": 0.3537974953651428, + "learning_rate": 2.8217132216014898e-05, + "loss": 0.0695, + "step": 13743 + }, + { + "epoch": 38.391061452513966, + "grad_norm": 0.6345534324645996, + "learning_rate": 2.821675977653631e-05, + "loss": 0.0658, + "step": 13744 + }, + { + "epoch": 38.39385474860335, + "grad_norm": 0.5241303443908691, + "learning_rate": 2.8216387337057728e-05, + "loss": 0.0929, + "step": 13745 + }, + { + "epoch": 38.39664804469274, + "grad_norm": 0.687155544757843, + "learning_rate": 2.8216014897579144e-05, + "loss": 0.0966, + "step": 13746 + }, + { + "epoch": 38.399441340782126, + "grad_norm": 0.9251924753189087, + "learning_rate": 2.821564245810056e-05, + "loss": 0.0783, + "step": 13747 + }, + { + "epoch": 38.40223463687151, + "grad_norm": 0.5184628963470459, + "learning_rate": 2.8215270018621977e-05, + "loss": 0.0647, + "step": 13748 + }, + { + "epoch": 38.40502793296089, + "grad_norm": 0.8536320328712463, + "learning_rate": 2.8214897579143387e-05, + "loss": 0.0717, + "step": 13749 + }, + { + "epoch": 38.40782122905028, + "grad_norm": 0.48444679379463196, + "learning_rate": 2.8214525139664804e-05, + "loss": 0.0751, + "step": 13750 + }, + { + "epoch": 38.410614525139664, + "grad_norm": 0.9786247611045837, + "learning_rate": 2.821415270018622e-05, + "loss": 0.0753, + "step": 13751 + }, + { + "epoch": 38.41340782122905, + "grad_norm": 0.5685248374938965, + "learning_rate": 2.8213780260707637e-05, + "loss": 0.0634, + "step": 13752 + }, + { + "epoch": 38.41620111731844, + "grad_norm": 0.9955393671989441, + "learning_rate": 2.821340782122905e-05, + "loss": 0.0654, + "step": 13753 + }, + { + "epoch": 38.41899441340782, + "grad_norm": 0.8188115954399109, + "learning_rate": 2.8213035381750466e-05, + "loss": 0.0809, + "step": 13754 + }, + { + "epoch": 38.42178770949721, + "grad_norm": 0.38793066143989563, + "learning_rate": 2.8212662942271883e-05, + "loss": 0.1298, + "step": 13755 + }, + { + "epoch": 38.42458100558659, + "grad_norm": 0.7562322020530701, + "learning_rate": 2.8212290502793296e-05, + "loss": 0.1211, + "step": 13756 + }, + { + "epoch": 38.427374301675975, + "grad_norm": 0.9052409529685974, + "learning_rate": 2.8211918063314713e-05, + "loss": 0.0966, + "step": 13757 + }, + { + "epoch": 38.43016759776536, + "grad_norm": 0.6667492389678955, + "learning_rate": 2.8211545623836126e-05, + "loss": 0.1246, + "step": 13758 + }, + { + "epoch": 38.43296089385475, + "grad_norm": 0.618563711643219, + "learning_rate": 2.8211173184357542e-05, + "loss": 0.1239, + "step": 13759 + }, + { + "epoch": 38.435754189944134, + "grad_norm": 0.48272815346717834, + "learning_rate": 2.821080074487896e-05, + "loss": 0.0999, + "step": 13760 + }, + { + "epoch": 38.43854748603352, + "grad_norm": 0.4821487367153168, + "learning_rate": 2.8210428305400375e-05, + "loss": 0.1096, + "step": 13761 + }, + { + "epoch": 38.44134078212291, + "grad_norm": 0.41666385531425476, + "learning_rate": 2.821005586592179e-05, + "loss": 0.0921, + "step": 13762 + }, + { + "epoch": 38.444134078212294, + "grad_norm": 0.5778366327285767, + "learning_rate": 2.82096834264432e-05, + "loss": 0.1046, + "step": 13763 + }, + { + "epoch": 38.44692737430167, + "grad_norm": 0.4795728027820587, + "learning_rate": 2.8209310986964618e-05, + "loss": 0.0929, + "step": 13764 + }, + { + "epoch": 38.44972067039106, + "grad_norm": 0.627427875995636, + "learning_rate": 2.8208938547486035e-05, + "loss": 0.0938, + "step": 13765 + }, + { + "epoch": 38.452513966480446, + "grad_norm": 1.396582841873169, + "learning_rate": 2.820856610800745e-05, + "loss": 0.0865, + "step": 13766 + }, + { + "epoch": 38.45530726256983, + "grad_norm": 0.6990799307823181, + "learning_rate": 2.8208193668528864e-05, + "loss": 0.0727, + "step": 13767 + }, + { + "epoch": 38.45810055865922, + "grad_norm": 0.41566789150238037, + "learning_rate": 2.820782122905028e-05, + "loss": 0.0746, + "step": 13768 + }, + { + "epoch": 38.460893854748605, + "grad_norm": 1.025338888168335, + "learning_rate": 2.8207448789571694e-05, + "loss": 0.0679, + "step": 13769 + }, + { + "epoch": 38.46368715083799, + "grad_norm": 0.9662831425666809, + "learning_rate": 2.820707635009311e-05, + "loss": 0.084, + "step": 13770 + }, + { + "epoch": 38.46648044692738, + "grad_norm": 0.534064769744873, + "learning_rate": 2.8206703910614524e-05, + "loss": 0.067, + "step": 13771 + }, + { + "epoch": 38.46927374301676, + "grad_norm": 0.6542330384254456, + "learning_rate": 2.820633147113594e-05, + "loss": 0.0653, + "step": 13772 + }, + { + "epoch": 38.47206703910614, + "grad_norm": 0.7994427680969238, + "learning_rate": 2.8205959031657357e-05, + "loss": 0.0705, + "step": 13773 + }, + { + "epoch": 38.47486033519553, + "grad_norm": 0.8987044095993042, + "learning_rate": 2.8205586592178773e-05, + "loss": 0.0683, + "step": 13774 + }, + { + "epoch": 38.477653631284916, + "grad_norm": 1.035936951637268, + "learning_rate": 2.820521415270019e-05, + "loss": 0.0574, + "step": 13775 + }, + { + "epoch": 38.4804469273743, + "grad_norm": 0.48982343077659607, + "learning_rate": 2.82048417132216e-05, + "loss": 0.0704, + "step": 13776 + }, + { + "epoch": 38.48324022346369, + "grad_norm": 1.3583160638809204, + "learning_rate": 2.8204469273743016e-05, + "loss": 0.0734, + "step": 13777 + }, + { + "epoch": 38.486033519553075, + "grad_norm": 2.6069440841674805, + "learning_rate": 2.8204096834264433e-05, + "loss": 0.0759, + "step": 13778 + }, + { + "epoch": 38.48882681564246, + "grad_norm": 0.9855948686599731, + "learning_rate": 2.820372439478585e-05, + "loss": 0.0954, + "step": 13779 + }, + { + "epoch": 38.49162011173184, + "grad_norm": 0.4216324985027313, + "learning_rate": 2.8203351955307262e-05, + "loss": 0.1209, + "step": 13780 + }, + { + "epoch": 38.49441340782123, + "grad_norm": 0.5342094898223877, + "learning_rate": 2.820297951582868e-05, + "loss": 0.1108, + "step": 13781 + }, + { + "epoch": 38.497206703910614, + "grad_norm": 0.45024871826171875, + "learning_rate": 2.8202607076350095e-05, + "loss": 0.109, + "step": 13782 + }, + { + "epoch": 38.5, + "grad_norm": 0.807569682598114, + "learning_rate": 2.820223463687151e-05, + "loss": 0.1157, + "step": 13783 + }, + { + "epoch": 38.502793296089386, + "grad_norm": 0.4081205129623413, + "learning_rate": 2.8201862197392925e-05, + "loss": 0.1132, + "step": 13784 + }, + { + "epoch": 38.50558659217877, + "grad_norm": 0.5807958245277405, + "learning_rate": 2.8201489757914338e-05, + "loss": 0.0973, + "step": 13785 + }, + { + "epoch": 38.50837988826816, + "grad_norm": 0.4793674647808075, + "learning_rate": 2.8201117318435755e-05, + "loss": 0.0997, + "step": 13786 + }, + { + "epoch": 38.51117318435754, + "grad_norm": 0.5050674676895142, + "learning_rate": 2.820074487895717e-05, + "loss": 0.0974, + "step": 13787 + }, + { + "epoch": 38.513966480446925, + "grad_norm": 0.39423036575317383, + "learning_rate": 2.8200372439478588e-05, + "loss": 0.0904, + "step": 13788 + }, + { + "epoch": 38.51675977653631, + "grad_norm": 1.1291871070861816, + "learning_rate": 2.8199999999999998e-05, + "loss": 0.104, + "step": 13789 + }, + { + "epoch": 38.5195530726257, + "grad_norm": 0.47780680656433105, + "learning_rate": 2.8199627560521414e-05, + "loss": 0.0994, + "step": 13790 + }, + { + "epoch": 38.522346368715084, + "grad_norm": 1.103088140487671, + "learning_rate": 2.819925512104283e-05, + "loss": 0.0869, + "step": 13791 + }, + { + "epoch": 38.52513966480447, + "grad_norm": 0.7918819189071655, + "learning_rate": 2.8198882681564247e-05, + "loss": 0.0948, + "step": 13792 + }, + { + "epoch": 38.52793296089386, + "grad_norm": 1.0708072185516357, + "learning_rate": 2.8198510242085664e-05, + "loss": 0.0765, + "step": 13793 + }, + { + "epoch": 38.53072625698324, + "grad_norm": 0.7510428428649902, + "learning_rate": 2.8198137802607077e-05, + "loss": 0.0889, + "step": 13794 + }, + { + "epoch": 38.53351955307262, + "grad_norm": 0.503798246383667, + "learning_rate": 2.8197765363128493e-05, + "loss": 0.0804, + "step": 13795 + }, + { + "epoch": 38.53631284916201, + "grad_norm": 0.4742206633090973, + "learning_rate": 2.8197392923649907e-05, + "loss": 0.081, + "step": 13796 + }, + { + "epoch": 38.539106145251395, + "grad_norm": 1.0509443283081055, + "learning_rate": 2.8197020484171323e-05, + "loss": 0.076, + "step": 13797 + }, + { + "epoch": 38.54189944134078, + "grad_norm": 1.1594491004943848, + "learning_rate": 2.8196648044692736e-05, + "loss": 0.0838, + "step": 13798 + }, + { + "epoch": 38.54469273743017, + "grad_norm": 0.6252346634864807, + "learning_rate": 2.8196275605214153e-05, + "loss": 0.0907, + "step": 13799 + }, + { + "epoch": 38.547486033519554, + "grad_norm": 8.080602645874023, + "learning_rate": 2.819590316573557e-05, + "loss": 0.0713, + "step": 13800 + }, + { + "epoch": 38.55027932960894, + "grad_norm": 0.5671552419662476, + "learning_rate": 2.8195530726256986e-05, + "loss": 0.0737, + "step": 13801 + }, + { + "epoch": 38.55307262569833, + "grad_norm": 0.8001420497894287, + "learning_rate": 2.8195158286778402e-05, + "loss": 0.0651, + "step": 13802 + }, + { + "epoch": 38.555865921787706, + "grad_norm": 0.5933485627174377, + "learning_rate": 2.8194785847299812e-05, + "loss": 0.0728, + "step": 13803 + }, + { + "epoch": 38.55865921787709, + "grad_norm": 1.5367540121078491, + "learning_rate": 2.819441340782123e-05, + "loss": 0.0816, + "step": 13804 + }, + { + "epoch": 38.56145251396648, + "grad_norm": 0.5064272880554199, + "learning_rate": 2.8194040968342645e-05, + "loss": 0.1266, + "step": 13805 + }, + { + "epoch": 38.564245810055866, + "grad_norm": 0.4853374660015106, + "learning_rate": 2.8193668528864062e-05, + "loss": 0.134, + "step": 13806 + }, + { + "epoch": 38.56703910614525, + "grad_norm": 0.6735977530479431, + "learning_rate": 2.8193296089385475e-05, + "loss": 0.1075, + "step": 13807 + }, + { + "epoch": 38.56983240223464, + "grad_norm": 0.6326329708099365, + "learning_rate": 2.819292364990689e-05, + "loss": 0.1032, + "step": 13808 + }, + { + "epoch": 38.572625698324025, + "grad_norm": 0.4763261675834656, + "learning_rate": 2.8192551210428305e-05, + "loss": 0.1095, + "step": 13809 + }, + { + "epoch": 38.57541899441341, + "grad_norm": 0.5325655341148376, + "learning_rate": 2.819217877094972e-05, + "loss": 0.0995, + "step": 13810 + }, + { + "epoch": 38.57821229050279, + "grad_norm": 0.8144444823265076, + "learning_rate": 2.8191806331471138e-05, + "loss": 0.1076, + "step": 13811 + }, + { + "epoch": 38.58100558659218, + "grad_norm": 1.1911715269088745, + "learning_rate": 2.819143389199255e-05, + "loss": 0.1014, + "step": 13812 + }, + { + "epoch": 38.58379888268156, + "grad_norm": 0.39264824986457825, + "learning_rate": 2.8191061452513967e-05, + "loss": 0.0828, + "step": 13813 + }, + { + "epoch": 38.58659217877095, + "grad_norm": 0.7393624782562256, + "learning_rate": 2.8190689013035384e-05, + "loss": 0.1007, + "step": 13814 + }, + { + "epoch": 38.589385474860336, + "grad_norm": 0.6280959248542786, + "learning_rate": 2.81903165735568e-05, + "loss": 0.0924, + "step": 13815 + }, + { + "epoch": 38.59217877094972, + "grad_norm": 0.6185868978500366, + "learning_rate": 2.818994413407821e-05, + "loss": 0.0936, + "step": 13816 + }, + { + "epoch": 38.59497206703911, + "grad_norm": 0.678554356098175, + "learning_rate": 2.8189571694599627e-05, + "loss": 0.0812, + "step": 13817 + }, + { + "epoch": 38.59776536312849, + "grad_norm": 0.6086649894714355, + "learning_rate": 2.8189199255121043e-05, + "loss": 0.0808, + "step": 13818 + }, + { + "epoch": 38.600558659217874, + "grad_norm": 0.5977902412414551, + "learning_rate": 2.818882681564246e-05, + "loss": 0.0703, + "step": 13819 + }, + { + "epoch": 38.60335195530726, + "grad_norm": 0.729066014289856, + "learning_rate": 2.8188454376163873e-05, + "loss": 0.0817, + "step": 13820 + }, + { + "epoch": 38.60614525139665, + "grad_norm": 0.4781682789325714, + "learning_rate": 2.818808193668529e-05, + "loss": 0.0712, + "step": 13821 + }, + { + "epoch": 38.608938547486034, + "grad_norm": 0.64860999584198, + "learning_rate": 2.8187709497206706e-05, + "loss": 0.0797, + "step": 13822 + }, + { + "epoch": 38.61173184357542, + "grad_norm": 1.1487290859222412, + "learning_rate": 2.818733705772812e-05, + "loss": 0.0681, + "step": 13823 + }, + { + "epoch": 38.614525139664806, + "grad_norm": 0.7879559993743896, + "learning_rate": 2.8186964618249536e-05, + "loss": 0.0826, + "step": 13824 + }, + { + "epoch": 38.61731843575419, + "grad_norm": 0.5300081372261047, + "learning_rate": 2.818659217877095e-05, + "loss": 0.0761, + "step": 13825 + }, + { + "epoch": 38.62011173184357, + "grad_norm": 0.8959120512008667, + "learning_rate": 2.8186219739292365e-05, + "loss": 0.0746, + "step": 13826 + }, + { + "epoch": 38.62290502793296, + "grad_norm": 1.5305004119873047, + "learning_rate": 2.8185847299813782e-05, + "loss": 0.0786, + "step": 13827 + }, + { + "epoch": 38.625698324022345, + "grad_norm": 0.9396858215332031, + "learning_rate": 2.81854748603352e-05, + "loss": 0.0861, + "step": 13828 + }, + { + "epoch": 38.62849162011173, + "grad_norm": 6.650175094604492, + "learning_rate": 2.8185102420856608e-05, + "loss": 0.0934, + "step": 13829 + }, + { + "epoch": 38.63128491620112, + "grad_norm": 0.46983224153518677, + "learning_rate": 2.8184729981378025e-05, + "loss": 0.1453, + "step": 13830 + }, + { + "epoch": 38.634078212290504, + "grad_norm": 0.42784979939460754, + "learning_rate": 2.818435754189944e-05, + "loss": 0.1195, + "step": 13831 + }, + { + "epoch": 38.63687150837989, + "grad_norm": 0.42154332995414734, + "learning_rate": 2.8183985102420858e-05, + "loss": 0.1004, + "step": 13832 + }, + { + "epoch": 38.63966480446928, + "grad_norm": 0.3716309070587158, + "learning_rate": 2.8183612662942274e-05, + "loss": 0.1004, + "step": 13833 + }, + { + "epoch": 38.642458100558656, + "grad_norm": 0.4281313419342041, + "learning_rate": 2.8183240223463688e-05, + "loss": 0.1035, + "step": 13834 + }, + { + "epoch": 38.64525139664804, + "grad_norm": 0.7026533484458923, + "learning_rate": 2.8182867783985104e-05, + "loss": 0.1159, + "step": 13835 + }, + { + "epoch": 38.64804469273743, + "grad_norm": 0.47632020711898804, + "learning_rate": 2.8182495344506517e-05, + "loss": 0.1036, + "step": 13836 + }, + { + "epoch": 38.650837988826815, + "grad_norm": 0.3790722191333771, + "learning_rate": 2.8182122905027934e-05, + "loss": 0.0978, + "step": 13837 + }, + { + "epoch": 38.6536312849162, + "grad_norm": 1.067626714706421, + "learning_rate": 2.8181750465549347e-05, + "loss": 0.0944, + "step": 13838 + }, + { + "epoch": 38.65642458100559, + "grad_norm": 0.43713054060935974, + "learning_rate": 2.8181378026070763e-05, + "loss": 0.0967, + "step": 13839 + }, + { + "epoch": 38.659217877094974, + "grad_norm": 0.4869667887687683, + "learning_rate": 2.818100558659218e-05, + "loss": 0.1297, + "step": 13840 + }, + { + "epoch": 38.66201117318436, + "grad_norm": 0.6247203946113586, + "learning_rate": 2.8180633147113597e-05, + "loss": 0.083, + "step": 13841 + }, + { + "epoch": 38.66480446927374, + "grad_norm": 0.5846632122993469, + "learning_rate": 2.8180260707635013e-05, + "loss": 0.0756, + "step": 13842 + }, + { + "epoch": 38.667597765363126, + "grad_norm": 0.569110631942749, + "learning_rate": 2.8179888268156423e-05, + "loss": 0.0884, + "step": 13843 + }, + { + "epoch": 38.67039106145251, + "grad_norm": 0.5299766063690186, + "learning_rate": 2.817951582867784e-05, + "loss": 0.079, + "step": 13844 + }, + { + "epoch": 38.6731843575419, + "grad_norm": 0.4069921672344208, + "learning_rate": 2.8179143389199256e-05, + "loss": 0.0792, + "step": 13845 + }, + { + "epoch": 38.675977653631286, + "grad_norm": 0.5567750334739685, + "learning_rate": 2.8178770949720672e-05, + "loss": 0.0842, + "step": 13846 + }, + { + "epoch": 38.67877094972067, + "grad_norm": 1.068578839302063, + "learning_rate": 2.8178398510242086e-05, + "loss": 0.0707, + "step": 13847 + }, + { + "epoch": 38.68156424581006, + "grad_norm": 0.6553998589515686, + "learning_rate": 2.8178026070763502e-05, + "loss": 0.0668, + "step": 13848 + }, + { + "epoch": 38.684357541899445, + "grad_norm": 1.1579440832138062, + "learning_rate": 2.8177653631284915e-05, + "loss": 0.0767, + "step": 13849 + }, + { + "epoch": 38.687150837988824, + "grad_norm": 0.6024009585380554, + "learning_rate": 2.8177281191806332e-05, + "loss": 0.0665, + "step": 13850 + }, + { + "epoch": 38.68994413407821, + "grad_norm": 0.5937240719795227, + "learning_rate": 2.817690875232775e-05, + "loss": 0.0685, + "step": 13851 + }, + { + "epoch": 38.6927374301676, + "grad_norm": 0.9592479467391968, + "learning_rate": 2.817653631284916e-05, + "loss": 0.0723, + "step": 13852 + }, + { + "epoch": 38.69553072625698, + "grad_norm": 0.8652563095092773, + "learning_rate": 2.8176163873370578e-05, + "loss": 0.0864, + "step": 13853 + }, + { + "epoch": 38.69832402234637, + "grad_norm": 1.0788688659667969, + "learning_rate": 2.8175791433891995e-05, + "loss": 0.0815, + "step": 13854 + }, + { + "epoch": 38.701117318435756, + "grad_norm": 0.6012893319129944, + "learning_rate": 2.817541899441341e-05, + "loss": 0.125, + "step": 13855 + }, + { + "epoch": 38.70391061452514, + "grad_norm": 0.5100256204605103, + "learning_rate": 2.817504655493482e-05, + "loss": 0.1371, + "step": 13856 + }, + { + "epoch": 38.70670391061452, + "grad_norm": 0.3431745171546936, + "learning_rate": 2.8174674115456237e-05, + "loss": 0.1087, + "step": 13857 + }, + { + "epoch": 38.70949720670391, + "grad_norm": 0.6963807940483093, + "learning_rate": 2.8174301675977654e-05, + "loss": 0.1034, + "step": 13858 + }, + { + "epoch": 38.712290502793294, + "grad_norm": 0.7737045884132385, + "learning_rate": 2.817392923649907e-05, + "loss": 0.1246, + "step": 13859 + }, + { + "epoch": 38.71508379888268, + "grad_norm": 0.7375789284706116, + "learning_rate": 2.8173556797020487e-05, + "loss": 0.1079, + "step": 13860 + }, + { + "epoch": 38.71787709497207, + "grad_norm": 0.6823984980583191, + "learning_rate": 2.81731843575419e-05, + "loss": 0.1127, + "step": 13861 + }, + { + "epoch": 38.720670391061454, + "grad_norm": 0.37735646963119507, + "learning_rate": 2.8172811918063317e-05, + "loss": 0.0995, + "step": 13862 + }, + { + "epoch": 38.72346368715084, + "grad_norm": 0.5369113683700562, + "learning_rate": 2.817243947858473e-05, + "loss": 0.0871, + "step": 13863 + }, + { + "epoch": 38.726256983240226, + "grad_norm": 0.7061443328857422, + "learning_rate": 2.8172067039106146e-05, + "loss": 0.0957, + "step": 13864 + }, + { + "epoch": 38.729050279329606, + "grad_norm": 0.6290752291679382, + "learning_rate": 2.817169459962756e-05, + "loss": 0.1146, + "step": 13865 + }, + { + "epoch": 38.73184357541899, + "grad_norm": 0.6431031823158264, + "learning_rate": 2.8171322160148976e-05, + "loss": 0.1022, + "step": 13866 + }, + { + "epoch": 38.73463687150838, + "grad_norm": 0.5127524733543396, + "learning_rate": 2.8170949720670393e-05, + "loss": 0.0906, + "step": 13867 + }, + { + "epoch": 38.737430167597765, + "grad_norm": 0.6714181303977966, + "learning_rate": 2.817057728119181e-05, + "loss": 0.0906, + "step": 13868 + }, + { + "epoch": 38.74022346368715, + "grad_norm": 0.6430686116218567, + "learning_rate": 2.8170204841713222e-05, + "loss": 0.0862, + "step": 13869 + }, + { + "epoch": 38.74301675977654, + "grad_norm": 0.5403818488121033, + "learning_rate": 2.8169832402234635e-05, + "loss": 0.0742, + "step": 13870 + }, + { + "epoch": 38.745810055865924, + "grad_norm": 0.678941547870636, + "learning_rate": 2.8169459962756052e-05, + "loss": 0.0797, + "step": 13871 + }, + { + "epoch": 38.74860335195531, + "grad_norm": 0.6076549291610718, + "learning_rate": 2.816908752327747e-05, + "loss": 0.0872, + "step": 13872 + }, + { + "epoch": 38.75139664804469, + "grad_norm": 1.3735816478729248, + "learning_rate": 2.8168715083798885e-05, + "loss": 0.1057, + "step": 13873 + }, + { + "epoch": 38.754189944134076, + "grad_norm": 1.2291306257247925, + "learning_rate": 2.8168342644320298e-05, + "loss": 0.0818, + "step": 13874 + }, + { + "epoch": 38.75698324022346, + "grad_norm": 0.5107596516609192, + "learning_rate": 2.8167970204841715e-05, + "loss": 0.064, + "step": 13875 + }, + { + "epoch": 38.75977653631285, + "grad_norm": 0.7181755900382996, + "learning_rate": 2.8167597765363128e-05, + "loss": 0.0756, + "step": 13876 + }, + { + "epoch": 38.762569832402235, + "grad_norm": 0.790803074836731, + "learning_rate": 2.8167225325884544e-05, + "loss": 0.092, + "step": 13877 + }, + { + "epoch": 38.76536312849162, + "grad_norm": 0.9664952158927917, + "learning_rate": 2.816685288640596e-05, + "loss": 0.0977, + "step": 13878 + }, + { + "epoch": 38.76815642458101, + "grad_norm": 2.558032989501953, + "learning_rate": 2.8166480446927374e-05, + "loss": 0.0898, + "step": 13879 + }, + { + "epoch": 38.770949720670394, + "grad_norm": 0.33781322836875916, + "learning_rate": 2.816610800744879e-05, + "loss": 0.1236, + "step": 13880 + }, + { + "epoch": 38.773743016759774, + "grad_norm": 0.3987351357936859, + "learning_rate": 2.8165735567970207e-05, + "loss": 0.1282, + "step": 13881 + }, + { + "epoch": 38.77653631284916, + "grad_norm": 0.7068341374397278, + "learning_rate": 2.8165363128491624e-05, + "loss": 0.1016, + "step": 13882 + }, + { + "epoch": 38.779329608938546, + "grad_norm": 0.35489025712013245, + "learning_rate": 2.8164990689013033e-05, + "loss": 0.1154, + "step": 13883 + }, + { + "epoch": 38.78212290502793, + "grad_norm": 0.7438069581985474, + "learning_rate": 2.816461824953445e-05, + "loss": 0.1131, + "step": 13884 + }, + { + "epoch": 38.78491620111732, + "grad_norm": 0.9405038952827454, + "learning_rate": 2.8164245810055866e-05, + "loss": 0.117, + "step": 13885 + }, + { + "epoch": 38.787709497206706, + "grad_norm": 0.49241313338279724, + "learning_rate": 2.8163873370577283e-05, + "loss": 0.1054, + "step": 13886 + }, + { + "epoch": 38.79050279329609, + "grad_norm": 0.4505404531955719, + "learning_rate": 2.81635009310987e-05, + "loss": 0.0912, + "step": 13887 + }, + { + "epoch": 38.79329608938548, + "grad_norm": 0.539735734462738, + "learning_rate": 2.8163128491620113e-05, + "loss": 0.0984, + "step": 13888 + }, + { + "epoch": 38.79608938547486, + "grad_norm": 0.45133382081985474, + "learning_rate": 2.8162756052141526e-05, + "loss": 0.0825, + "step": 13889 + }, + { + "epoch": 38.798882681564244, + "grad_norm": 1.4024449586868286, + "learning_rate": 2.8162383612662942e-05, + "loss": 0.097, + "step": 13890 + }, + { + "epoch": 38.80167597765363, + "grad_norm": 0.7537744641304016, + "learning_rate": 2.816201117318436e-05, + "loss": 0.0805, + "step": 13891 + }, + { + "epoch": 38.80446927374302, + "grad_norm": 0.401835173368454, + "learning_rate": 2.8161638733705772e-05, + "loss": 0.0813, + "step": 13892 + }, + { + "epoch": 38.8072625698324, + "grad_norm": 0.42641526460647583, + "learning_rate": 2.816126629422719e-05, + "loss": 0.0802, + "step": 13893 + }, + { + "epoch": 38.81005586592179, + "grad_norm": 0.6978179216384888, + "learning_rate": 2.8160893854748605e-05, + "loss": 0.0812, + "step": 13894 + }, + { + "epoch": 38.812849162011176, + "grad_norm": 0.34789130091667175, + "learning_rate": 2.816052141527002e-05, + "loss": 0.0638, + "step": 13895 + }, + { + "epoch": 38.815642458100555, + "grad_norm": 0.6026569604873657, + "learning_rate": 2.8160148975791435e-05, + "loss": 0.064, + "step": 13896 + }, + { + "epoch": 38.81843575418994, + "grad_norm": 0.44745948910713196, + "learning_rate": 2.8159776536312848e-05, + "loss": 0.0839, + "step": 13897 + }, + { + "epoch": 38.82122905027933, + "grad_norm": 0.6541919112205505, + "learning_rate": 2.8159404096834265e-05, + "loss": 0.0789, + "step": 13898 + }, + { + "epoch": 38.824022346368714, + "grad_norm": 0.5804952383041382, + "learning_rate": 2.815903165735568e-05, + "loss": 0.0668, + "step": 13899 + }, + { + "epoch": 38.8268156424581, + "grad_norm": 0.7965481877326965, + "learning_rate": 2.8158659217877098e-05, + "loss": 0.0809, + "step": 13900 + }, + { + "epoch": 38.82960893854749, + "grad_norm": 0.732953667640686, + "learning_rate": 2.815828677839851e-05, + "loss": 0.0766, + "step": 13901 + }, + { + "epoch": 38.832402234636874, + "grad_norm": 0.8889634013175964, + "learning_rate": 2.8157914338919927e-05, + "loss": 0.0866, + "step": 13902 + }, + { + "epoch": 38.83519553072626, + "grad_norm": 0.752293050289154, + "learning_rate": 2.815754189944134e-05, + "loss": 0.074, + "step": 13903 + }, + { + "epoch": 38.83798882681564, + "grad_norm": 1.1037317514419556, + "learning_rate": 2.8157169459962757e-05, + "loss": 0.088, + "step": 13904 + }, + { + "epoch": 38.840782122905026, + "grad_norm": 0.6492994427680969, + "learning_rate": 2.8156797020484173e-05, + "loss": 0.1333, + "step": 13905 + }, + { + "epoch": 38.84357541899441, + "grad_norm": 0.7365188002586365, + "learning_rate": 2.8156424581005587e-05, + "loss": 0.1376, + "step": 13906 + }, + { + "epoch": 38.8463687150838, + "grad_norm": NaN, + "learning_rate": 2.8156424581005587e-05, + "loss": 0.1063, + "step": 13907 + }, + { + "epoch": 38.849162011173185, + "grad_norm": 0.850227415561676, + "learning_rate": 2.8156052141527003e-05, + "loss": 0.1292, + "step": 13908 + }, + { + "epoch": 38.85195530726257, + "grad_norm": 0.6937223672866821, + "learning_rate": 2.815567970204842e-05, + "loss": 0.1142, + "step": 13909 + }, + { + "epoch": 38.85474860335196, + "grad_norm": 0.5119539499282837, + "learning_rate": 2.8155307262569833e-05, + "loss": 0.1076, + "step": 13910 + }, + { + "epoch": 38.857541899441344, + "grad_norm": 0.6546431183815002, + "learning_rate": 2.8154934823091246e-05, + "loss": 0.1022, + "step": 13911 + }, + { + "epoch": 38.86033519553072, + "grad_norm": 1.2861254215240479, + "learning_rate": 2.8154562383612663e-05, + "loss": 0.0916, + "step": 13912 + }, + { + "epoch": 38.86312849162011, + "grad_norm": 0.3550238609313965, + "learning_rate": 2.815418994413408e-05, + "loss": 0.093, + "step": 13913 + }, + { + "epoch": 38.865921787709496, + "grad_norm": 0.595964252948761, + "learning_rate": 2.8153817504655496e-05, + "loss": 0.0908, + "step": 13914 + }, + { + "epoch": 38.86871508379888, + "grad_norm": 0.5790852308273315, + "learning_rate": 2.815344506517691e-05, + "loss": 0.1004, + "step": 13915 + }, + { + "epoch": 38.87150837988827, + "grad_norm": 0.5495722889900208, + "learning_rate": 2.8153072625698325e-05, + "loss": 0.0834, + "step": 13916 + }, + { + "epoch": 38.874301675977655, + "grad_norm": 0.9984388947486877, + "learning_rate": 2.815270018621974e-05, + "loss": 0.0855, + "step": 13917 + }, + { + "epoch": 38.87709497206704, + "grad_norm": 0.5843257308006287, + "learning_rate": 2.8152327746741155e-05, + "loss": 0.0896, + "step": 13918 + }, + { + "epoch": 38.87988826815643, + "grad_norm": 0.6011370420455933, + "learning_rate": 2.815195530726257e-05, + "loss": 0.0821, + "step": 13919 + }, + { + "epoch": 38.88268156424581, + "grad_norm": 0.4790153205394745, + "learning_rate": 2.8151582867783985e-05, + "loss": 0.0833, + "step": 13920 + }, + { + "epoch": 38.885474860335194, + "grad_norm": 0.6771031022071838, + "learning_rate": 2.81512104283054e-05, + "loss": 0.0998, + "step": 13921 + }, + { + "epoch": 38.88826815642458, + "grad_norm": 0.9446739554405212, + "learning_rate": 2.8150837988826818e-05, + "loss": 0.0745, + "step": 13922 + }, + { + "epoch": 38.891061452513966, + "grad_norm": 0.49410757422447205, + "learning_rate": 2.8150465549348234e-05, + "loss": 0.0633, + "step": 13923 + }, + { + "epoch": 38.89385474860335, + "grad_norm": 0.48964133858680725, + "learning_rate": 2.8150093109869644e-05, + "loss": 0.0612, + "step": 13924 + }, + { + "epoch": 38.89664804469274, + "grad_norm": 0.581567645072937, + "learning_rate": 2.814972067039106e-05, + "loss": 0.0676, + "step": 13925 + }, + { + "epoch": 38.899441340782126, + "grad_norm": 1.5354546308517456, + "learning_rate": 2.8149348230912477e-05, + "loss": 0.0603, + "step": 13926 + }, + { + "epoch": 38.90223463687151, + "grad_norm": 1.0759246349334717, + "learning_rate": 2.8148975791433894e-05, + "loss": 0.0595, + "step": 13927 + }, + { + "epoch": 38.90502793296089, + "grad_norm": 0.7140375971794128, + "learning_rate": 2.814860335195531e-05, + "loss": 0.0801, + "step": 13928 + }, + { + "epoch": 38.90782122905028, + "grad_norm": 0.9975626468658447, + "learning_rate": 2.8148230912476723e-05, + "loss": 0.0932, + "step": 13929 + }, + { + "epoch": 38.910614525139664, + "grad_norm": 0.6339572668075562, + "learning_rate": 2.8147858472998136e-05, + "loss": 0.1189, + "step": 13930 + }, + { + "epoch": 38.91340782122905, + "grad_norm": 0.43222346901893616, + "learning_rate": 2.8147486033519553e-05, + "loss": 0.115, + "step": 13931 + }, + { + "epoch": 38.91620111731844, + "grad_norm": 0.7280413508415222, + "learning_rate": 2.814711359404097e-05, + "loss": 0.0914, + "step": 13932 + }, + { + "epoch": 38.91899441340782, + "grad_norm": 0.5688527822494507, + "learning_rate": 2.8146741154562383e-05, + "loss": 0.1085, + "step": 13933 + }, + { + "epoch": 38.92178770949721, + "grad_norm": 0.7641273140907288, + "learning_rate": 2.81463687150838e-05, + "loss": 0.1243, + "step": 13934 + }, + { + "epoch": 38.92458100558659, + "grad_norm": 0.4794342815876007, + "learning_rate": 2.8145996275605216e-05, + "loss": 0.1112, + "step": 13935 + }, + { + "epoch": 38.927374301675975, + "grad_norm": 1.7034844160079956, + "learning_rate": 2.8145623836126632e-05, + "loss": 0.1209, + "step": 13936 + }, + { + "epoch": 38.93016759776536, + "grad_norm": 0.4324362576007843, + "learning_rate": 2.8145251396648045e-05, + "loss": 0.1074, + "step": 13937 + }, + { + "epoch": 38.93296089385475, + "grad_norm": 1.1344329118728638, + "learning_rate": 2.814487895716946e-05, + "loss": 0.1, + "step": 13938 + }, + { + "epoch": 38.935754189944134, + "grad_norm": 0.7757636308670044, + "learning_rate": 2.8144506517690875e-05, + "loss": 0.1044, + "step": 13939 + }, + { + "epoch": 38.93854748603352, + "grad_norm": 0.5023086071014404, + "learning_rate": 2.814413407821229e-05, + "loss": 0.1031, + "step": 13940 + }, + { + "epoch": 38.94134078212291, + "grad_norm": 0.4075654149055481, + "learning_rate": 2.8143761638733708e-05, + "loss": 0.1082, + "step": 13941 + }, + { + "epoch": 38.944134078212294, + "grad_norm": 0.48283904790878296, + "learning_rate": 2.814338919925512e-05, + "loss": 0.0905, + "step": 13942 + }, + { + "epoch": 38.94692737430167, + "grad_norm": 0.8485958576202393, + "learning_rate": 2.8143016759776538e-05, + "loss": 0.1032, + "step": 13943 + }, + { + "epoch": 38.94972067039106, + "grad_norm": 0.6102892756462097, + "learning_rate": 2.814264432029795e-05, + "loss": 0.0903, + "step": 13944 + }, + { + "epoch": 38.952513966480446, + "grad_norm": 0.5736841559410095, + "learning_rate": 2.8142271880819368e-05, + "loss": 0.0699, + "step": 13945 + }, + { + "epoch": 38.95530726256983, + "grad_norm": 0.7148399353027344, + "learning_rate": 2.8141899441340784e-05, + "loss": 0.0906, + "step": 13946 + }, + { + "epoch": 38.95810055865922, + "grad_norm": 1.0956953763961792, + "learning_rate": 2.8141527001862197e-05, + "loss": 0.0947, + "step": 13947 + }, + { + "epoch": 38.960893854748605, + "grad_norm": 0.6016318202018738, + "learning_rate": 2.8141154562383614e-05, + "loss": 0.0669, + "step": 13948 + }, + { + "epoch": 38.96368715083799, + "grad_norm": 0.7711247801780701, + "learning_rate": 2.814078212290503e-05, + "loss": 0.0866, + "step": 13949 + }, + { + "epoch": 38.96648044692738, + "grad_norm": 0.5966675281524658, + "learning_rate": 2.8140409683426443e-05, + "loss": 0.0763, + "step": 13950 + }, + { + "epoch": 38.96927374301676, + "grad_norm": 0.6873340606689453, + "learning_rate": 2.8140037243947857e-05, + "loss": 0.0739, + "step": 13951 + }, + { + "epoch": 38.97206703910614, + "grad_norm": 1.223775029182434, + "learning_rate": 2.8139664804469273e-05, + "loss": 0.0636, + "step": 13952 + }, + { + "epoch": 38.97486033519553, + "grad_norm": 0.8691296577453613, + "learning_rate": 2.813929236499069e-05, + "loss": 0.0678, + "step": 13953 + }, + { + "epoch": 38.977653631284916, + "grad_norm": 1.0352729558944702, + "learning_rate": 2.8138919925512106e-05, + "loss": 0.0936, + "step": 13954 + }, + { + "epoch": 38.9804469273743, + "grad_norm": 0.8331159353256226, + "learning_rate": 2.8138547486033523e-05, + "loss": 0.1105, + "step": 13955 + }, + { + "epoch": 38.98324022346369, + "grad_norm": 0.5664482712745667, + "learning_rate": 2.8138175046554936e-05, + "loss": 0.1048, + "step": 13956 + }, + { + "epoch": 38.986033519553075, + "grad_norm": 0.41837936639785767, + "learning_rate": 2.813780260707635e-05, + "loss": 0.0925, + "step": 13957 + }, + { + "epoch": 38.98882681564246, + "grad_norm": 3.002607583999634, + "learning_rate": 2.8137430167597766e-05, + "loss": 0.0986, + "step": 13958 + }, + { + "epoch": 38.99162011173184, + "grad_norm": 0.7762014865875244, + "learning_rate": 2.8137057728119182e-05, + "loss": 0.0654, + "step": 13959 + }, + { + "epoch": 38.99441340782123, + "grad_norm": 1.0032436847686768, + "learning_rate": 2.8136685288640595e-05, + "loss": 0.0858, + "step": 13960 + }, + { + "epoch": 38.997206703910614, + "grad_norm": 1.0900036096572876, + "learning_rate": 2.8136312849162012e-05, + "loss": 0.0764, + "step": 13961 + }, + { + "epoch": 39.0, + "grad_norm": 0.7696616649627686, + "learning_rate": 2.813594040968343e-05, + "loss": 0.089, + "step": 13962 + }, + { + "epoch": 39.002793296089386, + "grad_norm": 0.6514900922775269, + "learning_rate": 2.8135567970204845e-05, + "loss": 0.1198, + "step": 13963 + }, + { + "epoch": 39.00558659217877, + "grad_norm": 0.35125869512557983, + "learning_rate": 2.8135195530726258e-05, + "loss": 0.1265, + "step": 13964 + }, + { + "epoch": 39.00837988826816, + "grad_norm": 0.33939993381500244, + "learning_rate": 2.813482309124767e-05, + "loss": 0.1047, + "step": 13965 + }, + { + "epoch": 39.01117318435754, + "grad_norm": 0.4291524291038513, + "learning_rate": 2.8134450651769088e-05, + "loss": 0.1084, + "step": 13966 + }, + { + "epoch": 39.013966480446925, + "grad_norm": 1.307497501373291, + "learning_rate": 2.8134078212290504e-05, + "loss": 0.1234, + "step": 13967 + }, + { + "epoch": 39.01675977653631, + "grad_norm": 0.4758478105068207, + "learning_rate": 2.813370577281192e-05, + "loss": 0.0846, + "step": 13968 + }, + { + "epoch": 39.0195530726257, + "grad_norm": 0.7991107702255249, + "learning_rate": 2.8133333333333334e-05, + "loss": 0.0974, + "step": 13969 + }, + { + "epoch": 39.022346368715084, + "grad_norm": 0.39669328927993774, + "learning_rate": 2.8132960893854747e-05, + "loss": 0.0913, + "step": 13970 + }, + { + "epoch": 39.02513966480447, + "grad_norm": 0.5983617305755615, + "learning_rate": 2.8132588454376164e-05, + "loss": 0.0935, + "step": 13971 + }, + { + "epoch": 39.02793296089386, + "grad_norm": 1.2277920246124268, + "learning_rate": 2.813221601489758e-05, + "loss": 0.0805, + "step": 13972 + }, + { + "epoch": 39.03072625698324, + "grad_norm": 0.3706749975681305, + "learning_rate": 2.8131843575418997e-05, + "loss": 0.0657, + "step": 13973 + }, + { + "epoch": 39.03351955307262, + "grad_norm": 1.1726256608963013, + "learning_rate": 2.813147113594041e-05, + "loss": 0.0962, + "step": 13974 + }, + { + "epoch": 39.03631284916201, + "grad_norm": 0.4656217396259308, + "learning_rate": 2.8131098696461826e-05, + "loss": 0.0754, + "step": 13975 + }, + { + "epoch": 39.039106145251395, + "grad_norm": 0.6613001823425293, + "learning_rate": 2.8130726256983243e-05, + "loss": 0.0779, + "step": 13976 + }, + { + "epoch": 39.04189944134078, + "grad_norm": 0.6982211470603943, + "learning_rate": 2.8130353817504656e-05, + "loss": 0.0752, + "step": 13977 + }, + { + "epoch": 39.04469273743017, + "grad_norm": 0.5721521377563477, + "learning_rate": 2.812998137802607e-05, + "loss": 0.0859, + "step": 13978 + }, + { + "epoch": 39.047486033519554, + "grad_norm": 0.9946912527084351, + "learning_rate": 2.8129608938547486e-05, + "loss": 0.0697, + "step": 13979 + }, + { + "epoch": 39.05027932960894, + "grad_norm": 0.7086393237113953, + "learning_rate": 2.8129236499068902e-05, + "loss": 0.0738, + "step": 13980 + }, + { + "epoch": 39.05307262569833, + "grad_norm": 0.6088959574699402, + "learning_rate": 2.812886405959032e-05, + "loss": 0.0728, + "step": 13981 + }, + { + "epoch": 39.055865921787706, + "grad_norm": 0.47745397686958313, + "learning_rate": 2.8128491620111735e-05, + "loss": 0.0543, + "step": 13982 + }, + { + "epoch": 39.05865921787709, + "grad_norm": 0.475727915763855, + "learning_rate": 2.812811918063315e-05, + "loss": 0.0632, + "step": 13983 + }, + { + "epoch": 39.06145251396648, + "grad_norm": 0.6240501403808594, + "learning_rate": 2.812774674115456e-05, + "loss": 0.0712, + "step": 13984 + }, + { + "epoch": 39.064245810055866, + "grad_norm": 0.5536040663719177, + "learning_rate": 2.8127374301675978e-05, + "loss": 0.0542, + "step": 13985 + }, + { + "epoch": 39.06703910614525, + "grad_norm": 0.9809906482696533, + "learning_rate": 2.8127001862197395e-05, + "loss": 0.0697, + "step": 13986 + }, + { + "epoch": 39.06983240223464, + "grad_norm": 5.530771732330322, + "learning_rate": 2.8126629422718808e-05, + "loss": 0.0735, + "step": 13987 + }, + { + "epoch": 39.072625698324025, + "grad_norm": 0.5011076927185059, + "learning_rate": 2.8126256983240224e-05, + "loss": 0.1168, + "step": 13988 + }, + { + "epoch": 39.07541899441341, + "grad_norm": 0.763650119304657, + "learning_rate": 2.812588454376164e-05, + "loss": 0.1218, + "step": 13989 + }, + { + "epoch": 39.07821229050279, + "grad_norm": 0.36942729353904724, + "learning_rate": 2.8125512104283054e-05, + "loss": 0.1008, + "step": 13990 + }, + { + "epoch": 39.08100558659218, + "grad_norm": 0.9547336101531982, + "learning_rate": 2.812513966480447e-05, + "loss": 0.1045, + "step": 13991 + }, + { + "epoch": 39.08379888268156, + "grad_norm": 1.6202845573425293, + "learning_rate": 2.8124767225325884e-05, + "loss": 0.0881, + "step": 13992 + }, + { + "epoch": 39.08659217877095, + "grad_norm": 0.40806204080581665, + "learning_rate": 2.81243947858473e-05, + "loss": 0.1054, + "step": 13993 + }, + { + "epoch": 39.089385474860336, + "grad_norm": 0.73966383934021, + "learning_rate": 2.8124022346368717e-05, + "loss": 0.094, + "step": 13994 + }, + { + "epoch": 39.09217877094972, + "grad_norm": 0.43949905037879944, + "learning_rate": 2.8123649906890133e-05, + "loss": 0.1135, + "step": 13995 + }, + { + "epoch": 39.09497206703911, + "grad_norm": 0.4388486444950104, + "learning_rate": 2.8123277467411546e-05, + "loss": 0.0768, + "step": 13996 + }, + { + "epoch": 39.097765363128495, + "grad_norm": 1.4556699991226196, + "learning_rate": 2.812290502793296e-05, + "loss": 0.1005, + "step": 13997 + }, + { + "epoch": 39.100558659217874, + "grad_norm": 0.4349512755870819, + "learning_rate": 2.8122532588454376e-05, + "loss": 0.0869, + "step": 13998 + }, + { + "epoch": 39.10335195530726, + "grad_norm": 0.5730288624763489, + "learning_rate": 2.8122160148975793e-05, + "loss": 0.1073, + "step": 13999 + }, + { + "epoch": 39.10614525139665, + "grad_norm": 0.6371434926986694, + "learning_rate": 2.812178770949721e-05, + "loss": 0.066, + "step": 14000 + }, + { + "epoch": 39.10614525139665, + "eval_cer": 0.19417258573096352, + "eval_loss": 0.33140841126441956, + "eval_runtime": 24.1764, + "eval_samples_per_second": 187.704, + "eval_steps_per_second": 0.62, + "eval_wer": 0.3525879165386442, + "step": 14000 + }, + { + "epoch": 39.108938547486034, + "grad_norm": 0.6971572637557983, + "learning_rate": 2.8121415270018622e-05, + "loss": 0.0865, + "step": 14001 + }, + { + "epoch": 39.11173184357542, + "grad_norm": 0.6072068214416504, + "learning_rate": 2.812104283054004e-05, + "loss": 0.0748, + "step": 14002 + }, + { + "epoch": 39.114525139664806, + "grad_norm": 0.7059956789016724, + "learning_rate": 2.8120670391061455e-05, + "loss": 0.0632, + "step": 14003 + }, + { + "epoch": 39.11731843575419, + "grad_norm": 0.5159445405006409, + "learning_rate": 2.812029795158287e-05, + "loss": 0.0765, + "step": 14004 + }, + { + "epoch": 39.12011173184357, + "grad_norm": 0.5925543308258057, + "learning_rate": 2.8119925512104282e-05, + "loss": 0.0844, + "step": 14005 + }, + { + "epoch": 39.12290502793296, + "grad_norm": 0.5741701722145081, + "learning_rate": 2.8119553072625698e-05, + "loss": 0.0686, + "step": 14006 + }, + { + "epoch": 39.125698324022345, + "grad_norm": 0.7267081141471863, + "learning_rate": 2.8119180633147115e-05, + "loss": 0.0819, + "step": 14007 + }, + { + "epoch": 39.12849162011173, + "grad_norm": 0.576292872428894, + "learning_rate": 2.811880819366853e-05, + "loss": 0.0647, + "step": 14008 + }, + { + "epoch": 39.13128491620112, + "grad_norm": 0.8595332503318787, + "learning_rate": 2.8118435754189944e-05, + "loss": 0.0606, + "step": 14009 + }, + { + "epoch": 39.134078212290504, + "grad_norm": 0.546978771686554, + "learning_rate": 2.8118063314711358e-05, + "loss": 0.0788, + "step": 14010 + }, + { + "epoch": 39.13687150837989, + "grad_norm": 3.2389917373657227, + "learning_rate": 2.8117690875232774e-05, + "loss": 0.0581, + "step": 14011 + }, + { + "epoch": 39.13966480446928, + "grad_norm": 1.1062215566635132, + "learning_rate": 2.811731843575419e-05, + "loss": 0.091, + "step": 14012 + }, + { + "epoch": 39.142458100558656, + "grad_norm": 0.538170337677002, + "learning_rate": 2.8116945996275607e-05, + "loss": 0.1187, + "step": 14013 + }, + { + "epoch": 39.14525139664804, + "grad_norm": 0.5342308878898621, + "learning_rate": 2.811657355679702e-05, + "loss": 0.1048, + "step": 14014 + }, + { + "epoch": 39.14804469273743, + "grad_norm": 0.5564582347869873, + "learning_rate": 2.8116201117318437e-05, + "loss": 0.0921, + "step": 14015 + }, + { + "epoch": 39.150837988826815, + "grad_norm": 0.6443662047386169, + "learning_rate": 2.8115828677839853e-05, + "loss": 0.1035, + "step": 14016 + }, + { + "epoch": 39.1536312849162, + "grad_norm": 1.4874422550201416, + "learning_rate": 2.8115456238361267e-05, + "loss": 0.0925, + "step": 14017 + }, + { + "epoch": 39.15642458100559, + "grad_norm": 0.6134142279624939, + "learning_rate": 2.811508379888268e-05, + "loss": 0.0962, + "step": 14018 + }, + { + "epoch": 39.159217877094974, + "grad_norm": 0.35554879903793335, + "learning_rate": 2.8114711359404096e-05, + "loss": 0.0855, + "step": 14019 + }, + { + "epoch": 39.16201117318436, + "grad_norm": 0.34897568821907043, + "learning_rate": 2.8114338919925513e-05, + "loss": 0.0965, + "step": 14020 + }, + { + "epoch": 39.16480446927374, + "grad_norm": 0.5209454894065857, + "learning_rate": 2.811396648044693e-05, + "loss": 0.0918, + "step": 14021 + }, + { + "epoch": 39.167597765363126, + "grad_norm": 0.7637613415718079, + "learning_rate": 2.8113594040968346e-05, + "loss": 0.0924, + "step": 14022 + }, + { + "epoch": 39.17039106145251, + "grad_norm": 0.786975085735321, + "learning_rate": 2.811322160148976e-05, + "loss": 0.091, + "step": 14023 + }, + { + "epoch": 39.1731843575419, + "grad_norm": 1.267651081085205, + "learning_rate": 2.8112849162011172e-05, + "loss": 0.0903, + "step": 14024 + }, + { + "epoch": 39.175977653631286, + "grad_norm": 0.5136451125144958, + "learning_rate": 2.811247672253259e-05, + "loss": 0.0776, + "step": 14025 + }, + { + "epoch": 39.17877094972067, + "grad_norm": 0.6962524652481079, + "learning_rate": 2.8112104283054005e-05, + "loss": 0.0755, + "step": 14026 + }, + { + "epoch": 39.18156424581006, + "grad_norm": 1.0299972295761108, + "learning_rate": 2.811173184357542e-05, + "loss": 0.0694, + "step": 14027 + }, + { + "epoch": 39.184357541899445, + "grad_norm": 0.6340653896331787, + "learning_rate": 2.8111359404096835e-05, + "loss": 0.0644, + "step": 14028 + }, + { + "epoch": 39.187150837988824, + "grad_norm": 1.1818031072616577, + "learning_rate": 2.811098696461825e-05, + "loss": 0.0688, + "step": 14029 + }, + { + "epoch": 39.18994413407821, + "grad_norm": 0.9673193693161011, + "learning_rate": 2.8110614525139665e-05, + "loss": 0.0652, + "step": 14030 + }, + { + "epoch": 39.1927374301676, + "grad_norm": 0.4876945912837982, + "learning_rate": 2.811024208566108e-05, + "loss": 0.076, + "step": 14031 + }, + { + "epoch": 39.19553072625698, + "grad_norm": 0.9110379219055176, + "learning_rate": 2.8109869646182494e-05, + "loss": 0.0776, + "step": 14032 + }, + { + "epoch": 39.19832402234637, + "grad_norm": 0.9087245464324951, + "learning_rate": 2.810949720670391e-05, + "loss": 0.0692, + "step": 14033 + }, + { + "epoch": 39.201117318435756, + "grad_norm": 0.6021818518638611, + "learning_rate": 2.8109124767225327e-05, + "loss": 0.0639, + "step": 14034 + }, + { + "epoch": 39.20391061452514, + "grad_norm": 0.6411771774291992, + "learning_rate": 2.8108752327746744e-05, + "loss": 0.0611, + "step": 14035 + }, + { + "epoch": 39.20670391061452, + "grad_norm": 0.8078656792640686, + "learning_rate": 2.8108379888268157e-05, + "loss": 0.0685, + "step": 14036 + }, + { + "epoch": 39.20949720670391, + "grad_norm": 1.8257516622543335, + "learning_rate": 2.810800744878957e-05, + "loss": 0.0939, + "step": 14037 + }, + { + "epoch": 39.212290502793294, + "grad_norm": 0.898176908493042, + "learning_rate": 2.8107635009310987e-05, + "loss": 0.135, + "step": 14038 + }, + { + "epoch": 39.21508379888268, + "grad_norm": 0.5809824466705322, + "learning_rate": 2.8107262569832403e-05, + "loss": 0.1338, + "step": 14039 + }, + { + "epoch": 39.21787709497207, + "grad_norm": 0.473695307970047, + "learning_rate": 2.810689013035382e-05, + "loss": 0.1053, + "step": 14040 + }, + { + "epoch": 39.220670391061454, + "grad_norm": 0.4969199299812317, + "learning_rate": 2.8106517690875233e-05, + "loss": 0.0975, + "step": 14041 + }, + { + "epoch": 39.22346368715084, + "grad_norm": 0.4871613681316376, + "learning_rate": 2.810614525139665e-05, + "loss": 0.1139, + "step": 14042 + }, + { + "epoch": 39.226256983240226, + "grad_norm": 0.4796241819858551, + "learning_rate": 2.8105772811918066e-05, + "loss": 0.1128, + "step": 14043 + }, + { + "epoch": 39.229050279329606, + "grad_norm": 0.48692968487739563, + "learning_rate": 2.810540037243948e-05, + "loss": 0.1031, + "step": 14044 + }, + { + "epoch": 39.23184357541899, + "grad_norm": 0.5723931789398193, + "learning_rate": 2.8105027932960892e-05, + "loss": 0.0873, + "step": 14045 + }, + { + "epoch": 39.23463687150838, + "grad_norm": 0.4131653904914856, + "learning_rate": 2.810465549348231e-05, + "loss": 0.0882, + "step": 14046 + }, + { + "epoch": 39.237430167597765, + "grad_norm": 0.42187732458114624, + "learning_rate": 2.8104283054003725e-05, + "loss": 0.0901, + "step": 14047 + }, + { + "epoch": 39.24022346368715, + "grad_norm": 0.41817548871040344, + "learning_rate": 2.8103910614525142e-05, + "loss": 0.0881, + "step": 14048 + }, + { + "epoch": 39.24301675977654, + "grad_norm": 0.9153322577476501, + "learning_rate": 2.810353817504656e-05, + "loss": 0.0737, + "step": 14049 + }, + { + "epoch": 39.245810055865924, + "grad_norm": 0.4398233890533447, + "learning_rate": 2.8103165735567968e-05, + "loss": 0.0854, + "step": 14050 + }, + { + "epoch": 39.24860335195531, + "grad_norm": 1.3926239013671875, + "learning_rate": 2.8102793296089385e-05, + "loss": 0.0922, + "step": 14051 + }, + { + "epoch": 39.25139664804469, + "grad_norm": 0.44430455565452576, + "learning_rate": 2.81024208566108e-05, + "loss": 0.0797, + "step": 14052 + }, + { + "epoch": 39.254189944134076, + "grad_norm": 0.7253776788711548, + "learning_rate": 2.8102048417132218e-05, + "loss": 0.0634, + "step": 14053 + }, + { + "epoch": 39.25698324022346, + "grad_norm": 0.5427342057228088, + "learning_rate": 2.810167597765363e-05, + "loss": 0.0688, + "step": 14054 + }, + { + "epoch": 39.25977653631285, + "grad_norm": 0.4682829678058624, + "learning_rate": 2.8101303538175048e-05, + "loss": 0.0767, + "step": 14055 + }, + { + "epoch": 39.262569832402235, + "grad_norm": 0.6315123438835144, + "learning_rate": 2.8100931098696464e-05, + "loss": 0.0767, + "step": 14056 + }, + { + "epoch": 39.26536312849162, + "grad_norm": 0.6270127892494202, + "learning_rate": 2.8100558659217877e-05, + "loss": 0.0678, + "step": 14057 + }, + { + "epoch": 39.26815642458101, + "grad_norm": 0.7695331573486328, + "learning_rate": 2.8100186219739294e-05, + "loss": 0.0581, + "step": 14058 + }, + { + "epoch": 39.270949720670394, + "grad_norm": 0.5316011309623718, + "learning_rate": 2.8099813780260707e-05, + "loss": 0.0441, + "step": 14059 + }, + { + "epoch": 39.273743016759774, + "grad_norm": 1.906714677810669, + "learning_rate": 2.8099441340782123e-05, + "loss": 0.0508, + "step": 14060 + }, + { + "epoch": 39.27653631284916, + "grad_norm": 1.0087958574295044, + "learning_rate": 2.809906890130354e-05, + "loss": 0.0654, + "step": 14061 + }, + { + "epoch": 39.279329608938546, + "grad_norm": 0.7182826995849609, + "learning_rate": 2.8098696461824956e-05, + "loss": 0.0691, + "step": 14062 + }, + { + "epoch": 39.28212290502793, + "grad_norm": 0.5427562594413757, + "learning_rate": 2.8098324022346366e-05, + "loss": 0.1368, + "step": 14063 + }, + { + "epoch": 39.28491620111732, + "grad_norm": 0.4124605357646942, + "learning_rate": 2.8097951582867783e-05, + "loss": 0.1261, + "step": 14064 + }, + { + "epoch": 39.287709497206706, + "grad_norm": 0.4254513382911682, + "learning_rate": 2.80975791433892e-05, + "loss": 0.1088, + "step": 14065 + }, + { + "epoch": 39.29050279329609, + "grad_norm": 0.5419891476631165, + "learning_rate": 2.8097206703910616e-05, + "loss": 0.0907, + "step": 14066 + }, + { + "epoch": 39.29329608938548, + "grad_norm": 0.6975832581520081, + "learning_rate": 2.8096834264432032e-05, + "loss": 0.1305, + "step": 14067 + }, + { + "epoch": 39.29608938547486, + "grad_norm": 0.8800750970840454, + "learning_rate": 2.8096461824953446e-05, + "loss": 0.0896, + "step": 14068 + }, + { + "epoch": 39.298882681564244, + "grad_norm": 0.8170784115791321, + "learning_rate": 2.8096089385474862e-05, + "loss": 0.0979, + "step": 14069 + }, + { + "epoch": 39.30167597765363, + "grad_norm": 0.482838898897171, + "learning_rate": 2.8095716945996275e-05, + "loss": 0.0924, + "step": 14070 + }, + { + "epoch": 39.30446927374302, + "grad_norm": 1.0014489889144897, + "learning_rate": 2.8095344506517692e-05, + "loss": 0.1035, + "step": 14071 + }, + { + "epoch": 39.3072625698324, + "grad_norm": 0.9670178890228271, + "learning_rate": 2.8094972067039105e-05, + "loss": 0.0795, + "step": 14072 + }, + { + "epoch": 39.31005586592179, + "grad_norm": 0.5662729740142822, + "learning_rate": 2.809459962756052e-05, + "loss": 0.0938, + "step": 14073 + }, + { + "epoch": 39.312849162011176, + "grad_norm": 0.4208625853061676, + "learning_rate": 2.8094227188081938e-05, + "loss": 0.0889, + "step": 14074 + }, + { + "epoch": 39.315642458100555, + "grad_norm": 0.5477436184883118, + "learning_rate": 2.8093854748603355e-05, + "loss": 0.0723, + "step": 14075 + }, + { + "epoch": 39.31843575418994, + "grad_norm": 0.7096989154815674, + "learning_rate": 2.809348230912477e-05, + "loss": 0.087, + "step": 14076 + }, + { + "epoch": 39.32122905027933, + "grad_norm": 0.7472924590110779, + "learning_rate": 2.809310986964618e-05, + "loss": 0.084, + "step": 14077 + }, + { + "epoch": 39.324022346368714, + "grad_norm": 0.5588383674621582, + "learning_rate": 2.8092737430167597e-05, + "loss": 0.0663, + "step": 14078 + }, + { + "epoch": 39.3268156424581, + "grad_norm": 0.5221477150917053, + "learning_rate": 2.8092364990689014e-05, + "loss": 0.0759, + "step": 14079 + }, + { + "epoch": 39.32960893854749, + "grad_norm": 0.3937850296497345, + "learning_rate": 2.809199255121043e-05, + "loss": 0.0716, + "step": 14080 + }, + { + "epoch": 39.332402234636874, + "grad_norm": 0.7107058167457581, + "learning_rate": 2.8091620111731844e-05, + "loss": 0.0602, + "step": 14081 + }, + { + "epoch": 39.33519553072626, + "grad_norm": 0.4628162086009979, + "learning_rate": 2.809124767225326e-05, + "loss": 0.0652, + "step": 14082 + }, + { + "epoch": 39.33798882681564, + "grad_norm": 1.3661383390426636, + "learning_rate": 2.8090875232774673e-05, + "loss": 0.0666, + "step": 14083 + }, + { + "epoch": 39.340782122905026, + "grad_norm": 0.8093706965446472, + "learning_rate": 2.809050279329609e-05, + "loss": 0.0653, + "step": 14084 + }, + { + "epoch": 39.34357541899441, + "grad_norm": 0.6508257389068604, + "learning_rate": 2.8090130353817506e-05, + "loss": 0.0558, + "step": 14085 + }, + { + "epoch": 39.3463687150838, + "grad_norm": 0.9563724994659424, + "learning_rate": 2.808975791433892e-05, + "loss": 0.085, + "step": 14086 + }, + { + "epoch": 39.349162011173185, + "grad_norm": 2.588078498840332, + "learning_rate": 2.8089385474860336e-05, + "loss": 0.0651, + "step": 14087 + }, + { + "epoch": 39.35195530726257, + "grad_norm": 0.45743268728256226, + "learning_rate": 2.8089013035381753e-05, + "loss": 0.1059, + "step": 14088 + }, + { + "epoch": 39.35474860335196, + "grad_norm": 0.3402569890022278, + "learning_rate": 2.808864059590317e-05, + "loss": 0.1082, + "step": 14089 + }, + { + "epoch": 39.357541899441344, + "grad_norm": 0.4471811354160309, + "learning_rate": 2.808826815642458e-05, + "loss": 0.1244, + "step": 14090 + }, + { + "epoch": 39.36033519553072, + "grad_norm": 0.5271093249320984, + "learning_rate": 2.8087895716945995e-05, + "loss": 0.1043, + "step": 14091 + }, + { + "epoch": 39.36312849162011, + "grad_norm": 0.7452728152275085, + "learning_rate": 2.8087523277467412e-05, + "loss": 0.099, + "step": 14092 + }, + { + "epoch": 39.365921787709496, + "grad_norm": 0.36608049273490906, + "learning_rate": 2.808715083798883e-05, + "loss": 0.0975, + "step": 14093 + }, + { + "epoch": 39.36871508379888, + "grad_norm": 0.6358633041381836, + "learning_rate": 2.8086778398510245e-05, + "loss": 0.0935, + "step": 14094 + }, + { + "epoch": 39.37150837988827, + "grad_norm": 2.0319418907165527, + "learning_rate": 2.8086405959031658e-05, + "loss": 0.0833, + "step": 14095 + }, + { + "epoch": 39.374301675977655, + "grad_norm": 0.5779388546943665, + "learning_rate": 2.8086033519553075e-05, + "loss": 0.0712, + "step": 14096 + }, + { + "epoch": 39.37709497206704, + "grad_norm": 0.5047096610069275, + "learning_rate": 2.8085661080074488e-05, + "loss": 0.0938, + "step": 14097 + }, + { + "epoch": 39.37988826815643, + "grad_norm": 0.8044900894165039, + "learning_rate": 2.8085288640595904e-05, + "loss": 0.0866, + "step": 14098 + }, + { + "epoch": 39.38268156424581, + "grad_norm": 0.4790472388267517, + "learning_rate": 2.8084916201117317e-05, + "loss": 0.0746, + "step": 14099 + }, + { + "epoch": 39.385474860335194, + "grad_norm": 0.41471508145332336, + "learning_rate": 2.8084543761638734e-05, + "loss": 0.0782, + "step": 14100 + }, + { + "epoch": 39.38826815642458, + "grad_norm": 0.705610454082489, + "learning_rate": 2.808417132216015e-05, + "loss": 0.0863, + "step": 14101 + }, + { + "epoch": 39.391061452513966, + "grad_norm": 0.4514126181602478, + "learning_rate": 2.8083798882681567e-05, + "loss": 0.0611, + "step": 14102 + }, + { + "epoch": 39.39385474860335, + "grad_norm": 0.45585376024246216, + "learning_rate": 2.808342644320298e-05, + "loss": 0.0714, + "step": 14103 + }, + { + "epoch": 39.39664804469274, + "grad_norm": 0.53668212890625, + "learning_rate": 2.8083054003724393e-05, + "loss": 0.0644, + "step": 14104 + }, + { + "epoch": 39.399441340782126, + "grad_norm": 0.8949863910675049, + "learning_rate": 2.808268156424581e-05, + "loss": 0.0743, + "step": 14105 + }, + { + "epoch": 39.40223463687151, + "grad_norm": 0.6802737712860107, + "learning_rate": 2.8082309124767226e-05, + "loss": 0.0834, + "step": 14106 + }, + { + "epoch": 39.40502793296089, + "grad_norm": 0.6090199947357178, + "learning_rate": 2.8081936685288643e-05, + "loss": 0.0682, + "step": 14107 + }, + { + "epoch": 39.40782122905028, + "grad_norm": 1.3451402187347412, + "learning_rate": 2.8081564245810056e-05, + "loss": 0.0809, + "step": 14108 + }, + { + "epoch": 39.410614525139664, + "grad_norm": 0.7989690899848938, + "learning_rate": 2.8081191806331473e-05, + "loss": 0.0728, + "step": 14109 + }, + { + "epoch": 39.41340782122905, + "grad_norm": 0.620924174785614, + "learning_rate": 2.8080819366852886e-05, + "loss": 0.0685, + "step": 14110 + }, + { + "epoch": 39.41620111731844, + "grad_norm": 0.8403414487838745, + "learning_rate": 2.8080446927374302e-05, + "loss": 0.0782, + "step": 14111 + }, + { + "epoch": 39.41899441340782, + "grad_norm": 0.5945524573326111, + "learning_rate": 2.8080074487895716e-05, + "loss": 0.0979, + "step": 14112 + }, + { + "epoch": 39.42178770949721, + "grad_norm": 0.3319501578807831, + "learning_rate": 2.8079702048417132e-05, + "loss": 0.1112, + "step": 14113 + }, + { + "epoch": 39.42458100558659, + "grad_norm": 0.835342288017273, + "learning_rate": 2.807932960893855e-05, + "loss": 0.1252, + "step": 14114 + }, + { + "epoch": 39.427374301675975, + "grad_norm": 0.39194944500923157, + "learning_rate": 2.8078957169459965e-05, + "loss": 0.1014, + "step": 14115 + }, + { + "epoch": 39.43016759776536, + "grad_norm": 0.3819669485092163, + "learning_rate": 2.807858472998138e-05, + "loss": 0.1122, + "step": 14116 + }, + { + "epoch": 39.43296089385475, + "grad_norm": 0.4715390205383301, + "learning_rate": 2.807821229050279e-05, + "loss": 0.1079, + "step": 14117 + }, + { + "epoch": 39.435754189944134, + "grad_norm": 0.45563575625419617, + "learning_rate": 2.8077839851024208e-05, + "loss": 0.0958, + "step": 14118 + }, + { + "epoch": 39.43854748603352, + "grad_norm": 1.964948296546936, + "learning_rate": 2.8077467411545624e-05, + "loss": 0.104, + "step": 14119 + }, + { + "epoch": 39.44134078212291, + "grad_norm": 0.6076468229293823, + "learning_rate": 2.807709497206704e-05, + "loss": 0.1077, + "step": 14120 + }, + { + "epoch": 39.444134078212294, + "grad_norm": 0.6097455620765686, + "learning_rate": 2.8076722532588454e-05, + "loss": 0.0816, + "step": 14121 + }, + { + "epoch": 39.44692737430167, + "grad_norm": 0.39880454540252686, + "learning_rate": 2.807635009310987e-05, + "loss": 0.0907, + "step": 14122 + }, + { + "epoch": 39.44972067039106, + "grad_norm": 0.6842474341392517, + "learning_rate": 2.8075977653631284e-05, + "loss": 0.0836, + "step": 14123 + }, + { + "epoch": 39.452513966480446, + "grad_norm": 0.6509041786193848, + "learning_rate": 2.80756052141527e-05, + "loss": 0.0875, + "step": 14124 + }, + { + "epoch": 39.45530726256983, + "grad_norm": 0.4081242084503174, + "learning_rate": 2.8075232774674117e-05, + "loss": 0.0872, + "step": 14125 + }, + { + "epoch": 39.45810055865922, + "grad_norm": 0.6554751396179199, + "learning_rate": 2.807486033519553e-05, + "loss": 0.0683, + "step": 14126 + }, + { + "epoch": 39.460893854748605, + "grad_norm": 0.6362093687057495, + "learning_rate": 2.8074487895716947e-05, + "loss": 0.0859, + "step": 14127 + }, + { + "epoch": 39.46368715083799, + "grad_norm": 0.5444076061248779, + "learning_rate": 2.8074115456238363e-05, + "loss": 0.0696, + "step": 14128 + }, + { + "epoch": 39.46648044692738, + "grad_norm": 0.5742375254631042, + "learning_rate": 2.807374301675978e-05, + "loss": 0.0611, + "step": 14129 + }, + { + "epoch": 39.46927374301676, + "grad_norm": 0.7663331031799316, + "learning_rate": 2.807337057728119e-05, + "loss": 0.0657, + "step": 14130 + }, + { + "epoch": 39.47206703910614, + "grad_norm": 0.7069665193557739, + "learning_rate": 2.8072998137802606e-05, + "loss": 0.0605, + "step": 14131 + }, + { + "epoch": 39.47486033519553, + "grad_norm": 0.5457783341407776, + "learning_rate": 2.8072625698324022e-05, + "loss": 0.064, + "step": 14132 + }, + { + "epoch": 39.477653631284916, + "grad_norm": 0.7250013947486877, + "learning_rate": 2.807225325884544e-05, + "loss": 0.0716, + "step": 14133 + }, + { + "epoch": 39.4804469273743, + "grad_norm": 1.0458954572677612, + "learning_rate": 2.8071880819366856e-05, + "loss": 0.0792, + "step": 14134 + }, + { + "epoch": 39.48324022346369, + "grad_norm": 6.449680328369141, + "learning_rate": 2.807150837988827e-05, + "loss": 0.0729, + "step": 14135 + }, + { + "epoch": 39.486033519553075, + "grad_norm": 0.9771922826766968, + "learning_rate": 2.8071135940409685e-05, + "loss": 0.054, + "step": 14136 + }, + { + "epoch": 39.48882681564246, + "grad_norm": 2.305572032928467, + "learning_rate": 2.80707635009311e-05, + "loss": 0.0798, + "step": 14137 + }, + { + "epoch": 39.49162011173184, + "grad_norm": 0.4115498661994934, + "learning_rate": 2.8070391061452515e-05, + "loss": 0.1246, + "step": 14138 + }, + { + "epoch": 39.49441340782123, + "grad_norm": 0.3628709018230438, + "learning_rate": 2.8070018621973928e-05, + "loss": 0.1026, + "step": 14139 + }, + { + "epoch": 39.497206703910614, + "grad_norm": 0.4918837249279022, + "learning_rate": 2.8069646182495345e-05, + "loss": 0.1112, + "step": 14140 + }, + { + "epoch": 39.5, + "grad_norm": 0.41584405303001404, + "learning_rate": 2.806927374301676e-05, + "loss": 0.1129, + "step": 14141 + }, + { + "epoch": 39.502793296089386, + "grad_norm": 0.3921940326690674, + "learning_rate": 2.8068901303538178e-05, + "loss": 0.1063, + "step": 14142 + }, + { + "epoch": 39.50558659217877, + "grad_norm": 0.6264927983283997, + "learning_rate": 2.806852886405959e-05, + "loss": 0.0859, + "step": 14143 + }, + { + "epoch": 39.50837988826816, + "grad_norm": 0.6804631948471069, + "learning_rate": 2.8068156424581004e-05, + "loss": 0.1155, + "step": 14144 + }, + { + "epoch": 39.51117318435754, + "grad_norm": 0.742387056350708, + "learning_rate": 2.806778398510242e-05, + "loss": 0.1074, + "step": 14145 + }, + { + "epoch": 39.513966480446925, + "grad_norm": 0.7848478555679321, + "learning_rate": 2.8067411545623837e-05, + "loss": 0.092, + "step": 14146 + }, + { + "epoch": 39.51675977653631, + "grad_norm": 0.5721147060394287, + "learning_rate": 2.8067039106145254e-05, + "loss": 0.0832, + "step": 14147 + }, + { + "epoch": 39.5195530726257, + "grad_norm": 0.41486915946006775, + "learning_rate": 2.8066666666666667e-05, + "loss": 0.0876, + "step": 14148 + }, + { + "epoch": 39.522346368715084, + "grad_norm": 0.6503865718841553, + "learning_rate": 2.8066294227188083e-05, + "loss": 0.0898, + "step": 14149 + }, + { + "epoch": 39.52513966480447, + "grad_norm": 0.5623265504837036, + "learning_rate": 2.8065921787709496e-05, + "loss": 0.0692, + "step": 14150 + }, + { + "epoch": 39.52793296089386, + "grad_norm": 0.47943446040153503, + "learning_rate": 2.8065549348230913e-05, + "loss": 0.0587, + "step": 14151 + }, + { + "epoch": 39.53072625698324, + "grad_norm": 0.5485885739326477, + "learning_rate": 2.806517690875233e-05, + "loss": 0.0678, + "step": 14152 + }, + { + "epoch": 39.53351955307262, + "grad_norm": 1.0156785249710083, + "learning_rate": 2.8064804469273743e-05, + "loss": 0.0783, + "step": 14153 + }, + { + "epoch": 39.53631284916201, + "grad_norm": 0.6572557091712952, + "learning_rate": 2.806443202979516e-05, + "loss": 0.0673, + "step": 14154 + }, + { + "epoch": 39.539106145251395, + "grad_norm": 0.7374659776687622, + "learning_rate": 2.8064059590316576e-05, + "loss": 0.0684, + "step": 14155 + }, + { + "epoch": 39.54189944134078, + "grad_norm": 2.1345183849334717, + "learning_rate": 2.8063687150837992e-05, + "loss": 0.0613, + "step": 14156 + }, + { + "epoch": 39.54469273743017, + "grad_norm": 0.5760238170623779, + "learning_rate": 2.8063314711359402e-05, + "loss": 0.0638, + "step": 14157 + }, + { + "epoch": 39.547486033519554, + "grad_norm": 0.727780282497406, + "learning_rate": 2.806294227188082e-05, + "loss": 0.0833, + "step": 14158 + }, + { + "epoch": 39.55027932960894, + "grad_norm": 1.3523720502853394, + "learning_rate": 2.8062569832402235e-05, + "loss": 0.0769, + "step": 14159 + }, + { + "epoch": 39.55307262569833, + "grad_norm": 3.398127317428589, + "learning_rate": 2.806219739292365e-05, + "loss": 0.0701, + "step": 14160 + }, + { + "epoch": 39.555865921787706, + "grad_norm": 1.3665070533752441, + "learning_rate": 2.8061824953445068e-05, + "loss": 0.0677, + "step": 14161 + }, + { + "epoch": 39.55865921787709, + "grad_norm": 1.4449412822723389, + "learning_rate": 2.806145251396648e-05, + "loss": 0.0688, + "step": 14162 + }, + { + "epoch": 39.56145251396648, + "grad_norm": 0.4667634069919586, + "learning_rate": 2.8061080074487894e-05, + "loss": 0.1122, + "step": 14163 + }, + { + "epoch": 39.564245810055866, + "grad_norm": 1.8011445999145508, + "learning_rate": 2.806070763500931e-05, + "loss": 0.1067, + "step": 14164 + }, + { + "epoch": 39.56703910614525, + "grad_norm": 0.8155667781829834, + "learning_rate": 2.8060335195530728e-05, + "loss": 0.107, + "step": 14165 + }, + { + "epoch": 39.56983240223464, + "grad_norm": 0.4566032588481903, + "learning_rate": 2.805996275605214e-05, + "loss": 0.1196, + "step": 14166 + }, + { + "epoch": 39.572625698324025, + "grad_norm": 0.5658359527587891, + "learning_rate": 2.8059590316573557e-05, + "loss": 0.1058, + "step": 14167 + }, + { + "epoch": 39.57541899441341, + "grad_norm": 0.5597356557846069, + "learning_rate": 2.8059217877094974e-05, + "loss": 0.0945, + "step": 14168 + }, + { + "epoch": 39.57821229050279, + "grad_norm": 0.4629966616630554, + "learning_rate": 2.805884543761639e-05, + "loss": 0.0886, + "step": 14169 + }, + { + "epoch": 39.58100558659218, + "grad_norm": 0.5495848655700684, + "learning_rate": 2.8058472998137803e-05, + "loss": 0.0824, + "step": 14170 + }, + { + "epoch": 39.58379888268156, + "grad_norm": 0.7070400714874268, + "learning_rate": 2.8058100558659217e-05, + "loss": 0.0944, + "step": 14171 + }, + { + "epoch": 39.58659217877095, + "grad_norm": 0.47436705231666565, + "learning_rate": 2.8057728119180633e-05, + "loss": 0.1053, + "step": 14172 + }, + { + "epoch": 39.589385474860336, + "grad_norm": 0.6575849056243896, + "learning_rate": 2.805735567970205e-05, + "loss": 0.0967, + "step": 14173 + }, + { + "epoch": 39.59217877094972, + "grad_norm": 0.3452988862991333, + "learning_rate": 2.8056983240223466e-05, + "loss": 0.0702, + "step": 14174 + }, + { + "epoch": 39.59497206703911, + "grad_norm": 0.48231810331344604, + "learning_rate": 2.805661080074488e-05, + "loss": 0.0891, + "step": 14175 + }, + { + "epoch": 39.59776536312849, + "grad_norm": 0.4622657299041748, + "learning_rate": 2.8056238361266296e-05, + "loss": 0.0781, + "step": 14176 + }, + { + "epoch": 39.600558659217874, + "grad_norm": 0.5737977027893066, + "learning_rate": 2.805586592178771e-05, + "loss": 0.0725, + "step": 14177 + }, + { + "epoch": 39.60335195530726, + "grad_norm": 0.5904179215431213, + "learning_rate": 2.8055493482309126e-05, + "loss": 0.0682, + "step": 14178 + }, + { + "epoch": 39.60614525139665, + "grad_norm": 0.598842978477478, + "learning_rate": 2.8055121042830542e-05, + "loss": 0.0713, + "step": 14179 + }, + { + "epoch": 39.608938547486034, + "grad_norm": 1.2096911668777466, + "learning_rate": 2.8054748603351955e-05, + "loss": 0.0689, + "step": 14180 + }, + { + "epoch": 39.61173184357542, + "grad_norm": 0.6495121717453003, + "learning_rate": 2.8054376163873372e-05, + "loss": 0.0592, + "step": 14181 + }, + { + "epoch": 39.614525139664806, + "grad_norm": 0.7364746928215027, + "learning_rate": 2.8054003724394788e-05, + "loss": 0.0676, + "step": 14182 + }, + { + "epoch": 39.61731843575419, + "grad_norm": 0.8187002539634705, + "learning_rate": 2.80536312849162e-05, + "loss": 0.0713, + "step": 14183 + }, + { + "epoch": 39.62011173184357, + "grad_norm": 1.7137912511825562, + "learning_rate": 2.8053258845437615e-05, + "loss": 0.0681, + "step": 14184 + }, + { + "epoch": 39.62290502793296, + "grad_norm": 0.659242570400238, + "learning_rate": 2.805288640595903e-05, + "loss": 0.0643, + "step": 14185 + }, + { + "epoch": 39.625698324022345, + "grad_norm": 0.7001863121986389, + "learning_rate": 2.8052513966480448e-05, + "loss": 0.0719, + "step": 14186 + }, + { + "epoch": 39.62849162011173, + "grad_norm": 0.9085218906402588, + "learning_rate": 2.8052141527001864e-05, + "loss": 0.069, + "step": 14187 + }, + { + "epoch": 39.63128491620112, + "grad_norm": 0.4690159261226654, + "learning_rate": 2.805176908752328e-05, + "loss": 0.1228, + "step": 14188 + }, + { + "epoch": 39.634078212290504, + "grad_norm": 1.9306825399398804, + "learning_rate": 2.8051396648044694e-05, + "loss": 0.1132, + "step": 14189 + }, + { + "epoch": 39.63687150837989, + "grad_norm": 0.4652332663536072, + "learning_rate": 2.8051024208566107e-05, + "loss": 0.1105, + "step": 14190 + }, + { + "epoch": 39.63966480446928, + "grad_norm": 0.9801783561706543, + "learning_rate": 2.8050651769087524e-05, + "loss": 0.1193, + "step": 14191 + }, + { + "epoch": 39.642458100558656, + "grad_norm": 0.5541779398918152, + "learning_rate": 2.805027932960894e-05, + "loss": 0.0895, + "step": 14192 + }, + { + "epoch": 39.64525139664804, + "grad_norm": 0.556161105632782, + "learning_rate": 2.8049906890130353e-05, + "loss": 0.1158, + "step": 14193 + }, + { + "epoch": 39.64804469273743, + "grad_norm": 0.5025841593742371, + "learning_rate": 2.804953445065177e-05, + "loss": 0.0957, + "step": 14194 + }, + { + "epoch": 39.650837988826815, + "grad_norm": 0.44033291935920715, + "learning_rate": 2.8049162011173186e-05, + "loss": 0.1012, + "step": 14195 + }, + { + "epoch": 39.6536312849162, + "grad_norm": 0.7681562304496765, + "learning_rate": 2.8048789571694603e-05, + "loss": 0.0897, + "step": 14196 + }, + { + "epoch": 39.65642458100559, + "grad_norm": 0.5846547484397888, + "learning_rate": 2.8048417132216016e-05, + "loss": 0.096, + "step": 14197 + }, + { + "epoch": 39.659217877094974, + "grad_norm": 0.6225091218948364, + "learning_rate": 2.804804469273743e-05, + "loss": 0.0858, + "step": 14198 + }, + { + "epoch": 39.66201117318436, + "grad_norm": 1.0970350503921509, + "learning_rate": 2.8047672253258846e-05, + "loss": 0.0784, + "step": 14199 + }, + { + "epoch": 39.66480446927374, + "grad_norm": 0.49973392486572266, + "learning_rate": 2.8047299813780262e-05, + "loss": 0.0879, + "step": 14200 + }, + { + "epoch": 39.667597765363126, + "grad_norm": 0.6278055906295776, + "learning_rate": 2.804692737430168e-05, + "loss": 0.0761, + "step": 14201 + }, + { + "epoch": 39.67039106145251, + "grad_norm": 1.7215722799301147, + "learning_rate": 2.8046554934823092e-05, + "loss": 0.0636, + "step": 14202 + }, + { + "epoch": 39.6731843575419, + "grad_norm": 0.9092133045196533, + "learning_rate": 2.8046182495344505e-05, + "loss": 0.0693, + "step": 14203 + }, + { + "epoch": 39.675977653631286, + "grad_norm": 1.6165531873703003, + "learning_rate": 2.804581005586592e-05, + "loss": 0.0615, + "step": 14204 + }, + { + "epoch": 39.67877094972067, + "grad_norm": 2.356397867202759, + "learning_rate": 2.8045437616387338e-05, + "loss": 0.0687, + "step": 14205 + }, + { + "epoch": 39.68156424581006, + "grad_norm": 0.9183029532432556, + "learning_rate": 2.804506517690875e-05, + "loss": 0.0867, + "step": 14206 + }, + { + "epoch": 39.684357541899445, + "grad_norm": 0.8334953784942627, + "learning_rate": 2.8044692737430168e-05, + "loss": 0.0972, + "step": 14207 + }, + { + "epoch": 39.687150837988824, + "grad_norm": 0.6918975114822388, + "learning_rate": 2.8044320297951584e-05, + "loss": 0.0632, + "step": 14208 + }, + { + "epoch": 39.68994413407821, + "grad_norm": 0.6519749760627747, + "learning_rate": 2.8043947858473e-05, + "loss": 0.0667, + "step": 14209 + }, + { + "epoch": 39.6927374301676, + "grad_norm": 0.7133312225341797, + "learning_rate": 2.8043575418994414e-05, + "loss": 0.0558, + "step": 14210 + }, + { + "epoch": 39.69553072625698, + "grad_norm": 0.9435344934463501, + "learning_rate": 2.8043202979515827e-05, + "loss": 0.0692, + "step": 14211 + }, + { + "epoch": 39.69832402234637, + "grad_norm": 1.1047242879867554, + "learning_rate": 2.8042830540037244e-05, + "loss": 0.1021, + "step": 14212 + }, + { + "epoch": 39.701117318435756, + "grad_norm": 0.41885241866111755, + "learning_rate": 2.804245810055866e-05, + "loss": 0.1193, + "step": 14213 + }, + { + "epoch": 39.70391061452514, + "grad_norm": 0.674311101436615, + "learning_rate": 2.8042085661080077e-05, + "loss": 0.1036, + "step": 14214 + }, + { + "epoch": 39.70670391061452, + "grad_norm": 0.7689853310585022, + "learning_rate": 2.804171322160149e-05, + "loss": 0.1153, + "step": 14215 + }, + { + "epoch": 39.70949720670391, + "grad_norm": 0.4734978675842285, + "learning_rate": 2.8041340782122906e-05, + "loss": 0.104, + "step": 14216 + }, + { + "epoch": 39.712290502793294, + "grad_norm": 0.37448063492774963, + "learning_rate": 2.804096834264432e-05, + "loss": 0.0837, + "step": 14217 + }, + { + "epoch": 39.71508379888268, + "grad_norm": 0.4446708559989929, + "learning_rate": 2.8040595903165736e-05, + "loss": 0.1221, + "step": 14218 + }, + { + "epoch": 39.71787709497207, + "grad_norm": 0.38524436950683594, + "learning_rate": 2.8040223463687153e-05, + "loss": 0.0892, + "step": 14219 + }, + { + "epoch": 39.720670391061454, + "grad_norm": 0.4627881348133087, + "learning_rate": 2.8039851024208566e-05, + "loss": 0.078, + "step": 14220 + }, + { + "epoch": 39.72346368715084, + "grad_norm": 0.5576992630958557, + "learning_rate": 2.8039478584729982e-05, + "loss": 0.1032, + "step": 14221 + }, + { + "epoch": 39.726256983240226, + "grad_norm": 0.6829530000686646, + "learning_rate": 2.80391061452514e-05, + "loss": 0.0995, + "step": 14222 + }, + { + "epoch": 39.729050279329606, + "grad_norm": 0.6657201051712036, + "learning_rate": 2.8038733705772812e-05, + "loss": 0.0738, + "step": 14223 + }, + { + "epoch": 39.73184357541899, + "grad_norm": 0.5690224766731262, + "learning_rate": 2.8038361266294225e-05, + "loss": 0.0789, + "step": 14224 + }, + { + "epoch": 39.73463687150838, + "grad_norm": 0.39684414863586426, + "learning_rate": 2.8037988826815642e-05, + "loss": 0.0735, + "step": 14225 + }, + { + "epoch": 39.737430167597765, + "grad_norm": 0.6344386339187622, + "learning_rate": 2.8037616387337058e-05, + "loss": 0.09, + "step": 14226 + }, + { + "epoch": 39.74022346368715, + "grad_norm": 1.0169410705566406, + "learning_rate": 2.8037243947858475e-05, + "loss": 0.0632, + "step": 14227 + }, + { + "epoch": 39.74301675977654, + "grad_norm": 1.0585399866104126, + "learning_rate": 2.803687150837989e-05, + "loss": 0.071, + "step": 14228 + }, + { + "epoch": 39.745810055865924, + "grad_norm": 0.5339102149009705, + "learning_rate": 2.8036499068901304e-05, + "loss": 0.0761, + "step": 14229 + }, + { + "epoch": 39.74860335195531, + "grad_norm": 0.6914986968040466, + "learning_rate": 2.8036126629422718e-05, + "loss": 0.0841, + "step": 14230 + }, + { + "epoch": 39.75139664804469, + "grad_norm": 0.8863212466239929, + "learning_rate": 2.8035754189944134e-05, + "loss": 0.0625, + "step": 14231 + }, + { + "epoch": 39.754189944134076, + "grad_norm": 0.5183951258659363, + "learning_rate": 2.803538175046555e-05, + "loss": 0.0507, + "step": 14232 + }, + { + "epoch": 39.75698324022346, + "grad_norm": 1.7011725902557373, + "learning_rate": 2.8035009310986964e-05, + "loss": 0.0828, + "step": 14233 + }, + { + "epoch": 39.75977653631285, + "grad_norm": 0.800396203994751, + "learning_rate": 2.803463687150838e-05, + "loss": 0.0728, + "step": 14234 + }, + { + "epoch": 39.762569832402235, + "grad_norm": 0.538485050201416, + "learning_rate": 2.8034264432029797e-05, + "loss": 0.0608, + "step": 14235 + }, + { + "epoch": 39.76536312849162, + "grad_norm": 2.7083613872528076, + "learning_rate": 2.8033891992551213e-05, + "loss": 0.0735, + "step": 14236 + }, + { + "epoch": 39.76815642458101, + "grad_norm": 0.8871488571166992, + "learning_rate": 2.8033519553072627e-05, + "loss": 0.09, + "step": 14237 + }, + { + "epoch": 39.770949720670394, + "grad_norm": 0.46429383754730225, + "learning_rate": 2.803314711359404e-05, + "loss": 0.1277, + "step": 14238 + }, + { + "epoch": 39.773743016759774, + "grad_norm": 4.347440242767334, + "learning_rate": 2.8032774674115456e-05, + "loss": 0.112, + "step": 14239 + }, + { + "epoch": 39.77653631284916, + "grad_norm": 0.5897279977798462, + "learning_rate": 2.8032402234636873e-05, + "loss": 0.1218, + "step": 14240 + }, + { + "epoch": 39.779329608938546, + "grad_norm": 0.32862409949302673, + "learning_rate": 2.803202979515829e-05, + "loss": 0.0982, + "step": 14241 + }, + { + "epoch": 39.78212290502793, + "grad_norm": 0.4493812322616577, + "learning_rate": 2.8031657355679702e-05, + "loss": 0.0899, + "step": 14242 + }, + { + "epoch": 39.78491620111732, + "grad_norm": 1.6584150791168213, + "learning_rate": 2.8031284916201116e-05, + "loss": 0.0986, + "step": 14243 + }, + { + "epoch": 39.787709497206706, + "grad_norm": 0.4811403453350067, + "learning_rate": 2.8030912476722532e-05, + "loss": 0.1088, + "step": 14244 + }, + { + "epoch": 39.79050279329609, + "grad_norm": 0.49481475353240967, + "learning_rate": 2.803054003724395e-05, + "loss": 0.0902, + "step": 14245 + }, + { + "epoch": 39.79329608938548, + "grad_norm": 0.5636563897132874, + "learning_rate": 2.8030167597765365e-05, + "loss": 0.0763, + "step": 14246 + }, + { + "epoch": 39.79608938547486, + "grad_norm": 1.6484318971633911, + "learning_rate": 2.802979515828678e-05, + "loss": 0.1107, + "step": 14247 + }, + { + "epoch": 39.798882681564244, + "grad_norm": 0.5521343946456909, + "learning_rate": 2.8029422718808195e-05, + "loss": 0.1059, + "step": 14248 + }, + { + "epoch": 39.80167597765363, + "grad_norm": 0.49966639280319214, + "learning_rate": 2.802905027932961e-05, + "loss": 0.0761, + "step": 14249 + }, + { + "epoch": 39.80446927374302, + "grad_norm": 0.507246196269989, + "learning_rate": 2.8028677839851025e-05, + "loss": 0.0742, + "step": 14250 + }, + { + "epoch": 39.8072625698324, + "grad_norm": 0.4819926917552948, + "learning_rate": 2.8028305400372438e-05, + "loss": 0.0768, + "step": 14251 + }, + { + "epoch": 39.81005586592179, + "grad_norm": 0.47581082582473755, + "learning_rate": 2.8027932960893854e-05, + "loss": 0.0818, + "step": 14252 + }, + { + "epoch": 39.812849162011176, + "grad_norm": 0.5876019597053528, + "learning_rate": 2.802756052141527e-05, + "loss": 0.0665, + "step": 14253 + }, + { + "epoch": 39.815642458100555, + "grad_norm": 0.6665530204772949, + "learning_rate": 2.8027188081936687e-05, + "loss": 0.0861, + "step": 14254 + }, + { + "epoch": 39.81843575418994, + "grad_norm": 0.5753546953201294, + "learning_rate": 2.8026815642458104e-05, + "loss": 0.071, + "step": 14255 + }, + { + "epoch": 39.82122905027933, + "grad_norm": 0.5827332139015198, + "learning_rate": 2.8026443202979517e-05, + "loss": 0.0823, + "step": 14256 + }, + { + "epoch": 39.824022346368714, + "grad_norm": 0.9817180633544922, + "learning_rate": 2.802607076350093e-05, + "loss": 0.0733, + "step": 14257 + }, + { + "epoch": 39.8268156424581, + "grad_norm": 0.6878452301025391, + "learning_rate": 2.8025698324022347e-05, + "loss": 0.0683, + "step": 14258 + }, + { + "epoch": 39.82960893854749, + "grad_norm": 0.5927018523216248, + "learning_rate": 2.8025325884543763e-05, + "loss": 0.056, + "step": 14259 + }, + { + "epoch": 39.832402234636874, + "grad_norm": 0.995243489742279, + "learning_rate": 2.8024953445065176e-05, + "loss": 0.0651, + "step": 14260 + }, + { + "epoch": 39.83519553072626, + "grad_norm": 0.7964522242546082, + "learning_rate": 2.8024581005586593e-05, + "loss": 0.0708, + "step": 14261 + }, + { + "epoch": 39.83798882681564, + "grad_norm": 0.777323305606842, + "learning_rate": 2.802420856610801e-05, + "loss": 0.0842, + "step": 14262 + }, + { + "epoch": 39.840782122905026, + "grad_norm": 0.4395795464515686, + "learning_rate": 2.8023836126629423e-05, + "loss": 0.1242, + "step": 14263 + }, + { + "epoch": 39.84357541899441, + "grad_norm": 0.4282305836677551, + "learning_rate": 2.802346368715084e-05, + "loss": 0.11, + "step": 14264 + }, + { + "epoch": 39.8463687150838, + "grad_norm": 1.6273667812347412, + "learning_rate": 2.8023091247672252e-05, + "loss": 0.1049, + "step": 14265 + }, + { + "epoch": 39.849162011173185, + "grad_norm": 0.3908649981021881, + "learning_rate": 2.802271880819367e-05, + "loss": 0.1083, + "step": 14266 + }, + { + "epoch": 39.85195530726257, + "grad_norm": 1.3554387092590332, + "learning_rate": 2.8022346368715085e-05, + "loss": 0.0912, + "step": 14267 + }, + { + "epoch": 39.85474860335196, + "grad_norm": 0.5367942452430725, + "learning_rate": 2.8021973929236502e-05, + "loss": 0.0993, + "step": 14268 + }, + { + "epoch": 39.857541899441344, + "grad_norm": 0.5035430788993835, + "learning_rate": 2.8021601489757915e-05, + "loss": 0.1065, + "step": 14269 + }, + { + "epoch": 39.86033519553072, + "grad_norm": 0.45717984437942505, + "learning_rate": 2.8021229050279328e-05, + "loss": 0.0988, + "step": 14270 + }, + { + "epoch": 39.86312849162011, + "grad_norm": 0.5185608863830566, + "learning_rate": 2.8020856610800745e-05, + "loss": 0.0888, + "step": 14271 + }, + { + "epoch": 39.865921787709496, + "grad_norm": 0.5018408894538879, + "learning_rate": 2.802048417132216e-05, + "loss": 0.0837, + "step": 14272 + }, + { + "epoch": 39.86871508379888, + "grad_norm": 0.744564950466156, + "learning_rate": 2.8020111731843578e-05, + "loss": 0.1047, + "step": 14273 + }, + { + "epoch": 39.87150837988827, + "grad_norm": 2.836118698120117, + "learning_rate": 2.801973929236499e-05, + "loss": 0.093, + "step": 14274 + }, + { + "epoch": 39.874301675977655, + "grad_norm": 0.4738086462020874, + "learning_rate": 2.8019366852886407e-05, + "loss": 0.0875, + "step": 14275 + }, + { + "epoch": 39.87709497206704, + "grad_norm": 0.7673394083976746, + "learning_rate": 2.8018994413407824e-05, + "loss": 0.0956, + "step": 14276 + }, + { + "epoch": 39.87988826815643, + "grad_norm": 0.8053054213523865, + "learning_rate": 2.8018621973929237e-05, + "loss": 0.1007, + "step": 14277 + }, + { + "epoch": 39.88268156424581, + "grad_norm": 0.4581826627254486, + "learning_rate": 2.801824953445065e-05, + "loss": 0.0634, + "step": 14278 + }, + { + "epoch": 39.885474860335194, + "grad_norm": 1.1644536256790161, + "learning_rate": 2.8017877094972067e-05, + "loss": 0.0881, + "step": 14279 + }, + { + "epoch": 39.88826815642458, + "grad_norm": 0.5722439885139465, + "learning_rate": 2.8017504655493483e-05, + "loss": 0.0643, + "step": 14280 + }, + { + "epoch": 39.891061452513966, + "grad_norm": 0.8460761308670044, + "learning_rate": 2.80171322160149e-05, + "loss": 0.0649, + "step": 14281 + }, + { + "epoch": 39.89385474860335, + "grad_norm": 0.6839624643325806, + "learning_rate": 2.8016759776536316e-05, + "loss": 0.0627, + "step": 14282 + }, + { + "epoch": 39.89664804469274, + "grad_norm": 0.5937744379043579, + "learning_rate": 2.8016387337057726e-05, + "loss": 0.0873, + "step": 14283 + }, + { + "epoch": 39.899441340782126, + "grad_norm": 1.0784815549850464, + "learning_rate": 2.8016014897579143e-05, + "loss": 0.0702, + "step": 14284 + }, + { + "epoch": 39.90223463687151, + "grad_norm": 1.2640727758407593, + "learning_rate": 2.801564245810056e-05, + "loss": 0.0674, + "step": 14285 + }, + { + "epoch": 39.90502793296089, + "grad_norm": 1.0117549896240234, + "learning_rate": 2.8015270018621976e-05, + "loss": 0.0865, + "step": 14286 + }, + { + "epoch": 39.90782122905028, + "grad_norm": 1.3263020515441895, + "learning_rate": 2.801489757914339e-05, + "loss": 0.1109, + "step": 14287 + }, + { + "epoch": 39.910614525139664, + "grad_norm": 0.49168261885643005, + "learning_rate": 2.8014525139664806e-05, + "loss": 0.1222, + "step": 14288 + }, + { + "epoch": 39.91340782122905, + "grad_norm": 1.0296025276184082, + "learning_rate": 2.8014152700186222e-05, + "loss": 0.0977, + "step": 14289 + }, + { + "epoch": 39.91620111731844, + "grad_norm": 0.38111355900764465, + "learning_rate": 2.8013780260707635e-05, + "loss": 0.1103, + "step": 14290 + }, + { + "epoch": 39.91899441340782, + "grad_norm": 0.5509767532348633, + "learning_rate": 2.8013407821229052e-05, + "loss": 0.1052, + "step": 14291 + }, + { + "epoch": 39.92178770949721, + "grad_norm": 0.6078382134437561, + "learning_rate": 2.8013035381750465e-05, + "loss": 0.1179, + "step": 14292 + }, + { + "epoch": 39.92458100558659, + "grad_norm": 0.6375477313995361, + "learning_rate": 2.801266294227188e-05, + "loss": 0.0974, + "step": 14293 + }, + { + "epoch": 39.927374301675975, + "grad_norm": 0.4404497444629669, + "learning_rate": 2.8012290502793298e-05, + "loss": 0.1004, + "step": 14294 + }, + { + "epoch": 39.93016759776536, + "grad_norm": 0.431220680475235, + "learning_rate": 2.8011918063314714e-05, + "loss": 0.086, + "step": 14295 + }, + { + "epoch": 39.93296089385475, + "grad_norm": 0.4747678339481354, + "learning_rate": 2.8011545623836128e-05, + "loss": 0.0872, + "step": 14296 + }, + { + "epoch": 39.935754189944134, + "grad_norm": 0.45655983686447144, + "learning_rate": 2.801117318435754e-05, + "loss": 0.0823, + "step": 14297 + }, + { + "epoch": 39.93854748603352, + "grad_norm": 0.5093713998794556, + "learning_rate": 2.8010800744878957e-05, + "loss": 0.1011, + "step": 14298 + }, + { + "epoch": 39.94134078212291, + "grad_norm": 1.0064916610717773, + "learning_rate": 2.8010428305400374e-05, + "loss": 0.071, + "step": 14299 + }, + { + "epoch": 39.944134078212294, + "grad_norm": 0.6916133761405945, + "learning_rate": 2.8010055865921787e-05, + "loss": 0.0954, + "step": 14300 + }, + { + "epoch": 39.94692737430167, + "grad_norm": 0.5857235789299011, + "learning_rate": 2.8009683426443204e-05, + "loss": 0.0667, + "step": 14301 + }, + { + "epoch": 39.94972067039106, + "grad_norm": 0.8640013337135315, + "learning_rate": 2.800931098696462e-05, + "loss": 0.0913, + "step": 14302 + }, + { + "epoch": 39.952513966480446, + "grad_norm": 0.6610291004180908, + "learning_rate": 2.8008938547486033e-05, + "loss": 0.0762, + "step": 14303 + }, + { + "epoch": 39.95530726256983, + "grad_norm": 0.5654322504997253, + "learning_rate": 2.800856610800745e-05, + "loss": 0.0852, + "step": 14304 + }, + { + "epoch": 39.95810055865922, + "grad_norm": 0.6809983849525452, + "learning_rate": 2.8008193668528863e-05, + "loss": 0.07, + "step": 14305 + }, + { + "epoch": 39.960893854748605, + "grad_norm": 1.1601518392562866, + "learning_rate": 2.800782122905028e-05, + "loss": 0.0683, + "step": 14306 + }, + { + "epoch": 39.96368715083799, + "grad_norm": 0.5202375054359436, + "learning_rate": 2.8007448789571696e-05, + "loss": 0.0589, + "step": 14307 + }, + { + "epoch": 39.96648044692738, + "grad_norm": 0.7650030255317688, + "learning_rate": 2.8007076350093113e-05, + "loss": 0.0526, + "step": 14308 + }, + { + "epoch": 39.96927374301676, + "grad_norm": 0.6923040747642517, + "learning_rate": 2.8006703910614526e-05, + "loss": 0.0627, + "step": 14309 + }, + { + "epoch": 39.97206703910614, + "grad_norm": 1.1617201566696167, + "learning_rate": 2.800633147113594e-05, + "loss": 0.0793, + "step": 14310 + }, + { + "epoch": 39.97486033519553, + "grad_norm": 0.6925303936004639, + "learning_rate": 2.8005959031657355e-05, + "loss": 0.0679, + "step": 14311 + }, + { + "epoch": 39.977653631284916, + "grad_norm": 0.9245287775993347, + "learning_rate": 2.8005586592178772e-05, + "loss": 0.0968, + "step": 14312 + }, + { + "epoch": 39.9804469273743, + "grad_norm": 0.6803935766220093, + "learning_rate": 2.800521415270019e-05, + "loss": 0.1276, + "step": 14313 + }, + { + "epoch": 39.98324022346369, + "grad_norm": 0.7780025601387024, + "learning_rate": 2.80048417132216e-05, + "loss": 0.0874, + "step": 14314 + }, + { + "epoch": 39.986033519553075, + "grad_norm": 0.5871025919914246, + "learning_rate": 2.8004469273743018e-05, + "loss": 0.0881, + "step": 14315 + }, + { + "epoch": 39.98882681564246, + "grad_norm": 1.0268954038619995, + "learning_rate": 2.8004096834264435e-05, + "loss": 0.0937, + "step": 14316 + }, + { + "epoch": 39.99162011173184, + "grad_norm": 0.5539591908454895, + "learning_rate": 2.8003724394785848e-05, + "loss": 0.0747, + "step": 14317 + }, + { + "epoch": 39.99441340782123, + "grad_norm": 0.6209248304367065, + "learning_rate": 2.800335195530726e-05, + "loss": 0.0772, + "step": 14318 + }, + { + "epoch": 39.997206703910614, + "grad_norm": 1.1217749118804932, + "learning_rate": 2.8002979515828677e-05, + "loss": 0.0669, + "step": 14319 + }, + { + "epoch": 40.0, + "grad_norm": 2.176927328109741, + "learning_rate": 2.8002607076350094e-05, + "loss": 0.0753, + "step": 14320 + }, + { + "epoch": 40.002793296089386, + "grad_norm": 0.5678653120994568, + "learning_rate": 2.800223463687151e-05, + "loss": 0.1228, + "step": 14321 + }, + { + "epoch": 40.00558659217877, + "grad_norm": 0.5473582148551941, + "learning_rate": 2.8001862197392927e-05, + "loss": 0.1258, + "step": 14322 + }, + { + "epoch": 40.00837988826816, + "grad_norm": 0.5535367727279663, + "learning_rate": 2.8001489757914337e-05, + "loss": 0.1184, + "step": 14323 + }, + { + "epoch": 40.01117318435754, + "grad_norm": 0.45651867985725403, + "learning_rate": 2.8001117318435753e-05, + "loss": 0.1059, + "step": 14324 + }, + { + "epoch": 40.013966480446925, + "grad_norm": 0.5261795520782471, + "learning_rate": 2.800074487895717e-05, + "loss": 0.099, + "step": 14325 + }, + { + "epoch": 40.01675977653631, + "grad_norm": 0.5381410717964172, + "learning_rate": 2.8000372439478586e-05, + "loss": 0.0939, + "step": 14326 + }, + { + "epoch": 40.0195530726257, + "grad_norm": 0.6642642617225647, + "learning_rate": 2.8e-05, + "loss": 0.1063, + "step": 14327 + }, + { + "epoch": 40.022346368715084, + "grad_norm": 0.670744776725769, + "learning_rate": 2.7999627560521416e-05, + "loss": 0.0833, + "step": 14328 + }, + { + "epoch": 40.02513966480447, + "grad_norm": 0.46698614954948425, + "learning_rate": 2.7999255121042833e-05, + "loss": 0.0869, + "step": 14329 + }, + { + "epoch": 40.02793296089386, + "grad_norm": 0.7870785593986511, + "learning_rate": 2.7998882681564246e-05, + "loss": 0.0841, + "step": 14330 + }, + { + "epoch": 40.03072625698324, + "grad_norm": 0.9782555103302002, + "learning_rate": 2.7998510242085662e-05, + "loss": 0.0904, + "step": 14331 + }, + { + "epoch": 40.03351955307262, + "grad_norm": 0.4644381105899811, + "learning_rate": 2.7998137802607075e-05, + "loss": 0.0845, + "step": 14332 + }, + { + "epoch": 40.03631284916201, + "grad_norm": 0.4506870210170746, + "learning_rate": 2.7997765363128492e-05, + "loss": 0.0743, + "step": 14333 + }, + { + "epoch": 40.039106145251395, + "grad_norm": 0.858180046081543, + "learning_rate": 2.799739292364991e-05, + "loss": 0.0618, + "step": 14334 + }, + { + "epoch": 40.04189944134078, + "grad_norm": 0.6270216107368469, + "learning_rate": 2.7997020484171325e-05, + "loss": 0.0827, + "step": 14335 + }, + { + "epoch": 40.04469273743017, + "grad_norm": 0.5544313788414001, + "learning_rate": 2.7996648044692738e-05, + "loss": 0.0632, + "step": 14336 + }, + { + "epoch": 40.047486033519554, + "grad_norm": 0.4814247488975525, + "learning_rate": 2.799627560521415e-05, + "loss": 0.0757, + "step": 14337 + }, + { + "epoch": 40.05027932960894, + "grad_norm": 0.802367627620697, + "learning_rate": 2.7995903165735568e-05, + "loss": 0.0631, + "step": 14338 + }, + { + "epoch": 40.05307262569833, + "grad_norm": 0.6439089179039001, + "learning_rate": 2.7995530726256984e-05, + "loss": 0.0662, + "step": 14339 + }, + { + "epoch": 40.055865921787706, + "grad_norm": 0.9100235104560852, + "learning_rate": 2.79951582867784e-05, + "loss": 0.0829, + "step": 14340 + }, + { + "epoch": 40.05865921787709, + "grad_norm": 0.6533300876617432, + "learning_rate": 2.7994785847299814e-05, + "loss": 0.058, + "step": 14341 + }, + { + "epoch": 40.06145251396648, + "grad_norm": 1.2383813858032227, + "learning_rate": 2.799441340782123e-05, + "loss": 0.0726, + "step": 14342 + }, + { + "epoch": 40.064245810055866, + "grad_norm": 0.5637646913528442, + "learning_rate": 2.7994040968342644e-05, + "loss": 0.0492, + "step": 14343 + }, + { + "epoch": 40.06703910614525, + "grad_norm": 0.9562973976135254, + "learning_rate": 2.799366852886406e-05, + "loss": 0.0568, + "step": 14344 + }, + { + "epoch": 40.06983240223464, + "grad_norm": 0.7770678400993347, + "learning_rate": 2.7993296089385474e-05, + "loss": 0.0772, + "step": 14345 + }, + { + "epoch": 40.072625698324025, + "grad_norm": 0.5435793399810791, + "learning_rate": 2.799292364990689e-05, + "loss": 0.1256, + "step": 14346 + }, + { + "epoch": 40.07541899441341, + "grad_norm": 1.7920631170272827, + "learning_rate": 2.7992551210428307e-05, + "loss": 0.098, + "step": 14347 + }, + { + "epoch": 40.07821229050279, + "grad_norm": 0.39850175380706787, + "learning_rate": 2.7992178770949723e-05, + "loss": 0.0998, + "step": 14348 + }, + { + "epoch": 40.08100558659218, + "grad_norm": 0.6519312262535095, + "learning_rate": 2.799180633147114e-05, + "loss": 0.0952, + "step": 14349 + }, + { + "epoch": 40.08379888268156, + "grad_norm": 0.7915263175964355, + "learning_rate": 2.799143389199255e-05, + "loss": 0.1013, + "step": 14350 + }, + { + "epoch": 40.08659217877095, + "grad_norm": 0.4706772565841675, + "learning_rate": 2.7991061452513966e-05, + "loss": 0.0888, + "step": 14351 + }, + { + "epoch": 40.089385474860336, + "grad_norm": 0.6955475807189941, + "learning_rate": 2.7990689013035382e-05, + "loss": 0.1043, + "step": 14352 + }, + { + "epoch": 40.09217877094972, + "grad_norm": 0.7222599983215332, + "learning_rate": 2.79903165735568e-05, + "loss": 0.0761, + "step": 14353 + }, + { + "epoch": 40.09497206703911, + "grad_norm": 0.6320319175720215, + "learning_rate": 2.7989944134078212e-05, + "loss": 0.0912, + "step": 14354 + }, + { + "epoch": 40.097765363128495, + "grad_norm": 1.4133728742599487, + "learning_rate": 2.798957169459963e-05, + "loss": 0.0945, + "step": 14355 + }, + { + "epoch": 40.100558659217874, + "grad_norm": 1.0008569955825806, + "learning_rate": 2.7989199255121045e-05, + "loss": 0.0726, + "step": 14356 + }, + { + "epoch": 40.10335195530726, + "grad_norm": 0.4394350051879883, + "learning_rate": 2.798882681564246e-05, + "loss": 0.084, + "step": 14357 + }, + { + "epoch": 40.10614525139665, + "grad_norm": 0.3773570656776428, + "learning_rate": 2.7988454376163875e-05, + "loss": 0.0732, + "step": 14358 + }, + { + "epoch": 40.108938547486034, + "grad_norm": 0.4384825825691223, + "learning_rate": 2.7988081936685288e-05, + "loss": 0.0789, + "step": 14359 + }, + { + "epoch": 40.11173184357542, + "grad_norm": 0.7717735171318054, + "learning_rate": 2.7987709497206705e-05, + "loss": 0.0739, + "step": 14360 + }, + { + "epoch": 40.114525139664806, + "grad_norm": 0.5386891961097717, + "learning_rate": 2.798733705772812e-05, + "loss": 0.0608, + "step": 14361 + }, + { + "epoch": 40.11731843575419, + "grad_norm": 2.1493372917175293, + "learning_rate": 2.7986964618249538e-05, + "loss": 0.0664, + "step": 14362 + }, + { + "epoch": 40.12011173184357, + "grad_norm": 0.4856381118297577, + "learning_rate": 2.7986592178770947e-05, + "loss": 0.0518, + "step": 14363 + }, + { + "epoch": 40.12290502793296, + "grad_norm": 0.7893396019935608, + "learning_rate": 2.7986219739292364e-05, + "loss": 0.0706, + "step": 14364 + }, + { + "epoch": 40.125698324022345, + "grad_norm": 0.5926256775856018, + "learning_rate": 2.798584729981378e-05, + "loss": 0.0508, + "step": 14365 + }, + { + "epoch": 40.12849162011173, + "grad_norm": 1.3735241889953613, + "learning_rate": 2.7985474860335197e-05, + "loss": 0.047, + "step": 14366 + }, + { + "epoch": 40.13128491620112, + "grad_norm": 0.5357168912887573, + "learning_rate": 2.7985102420856614e-05, + "loss": 0.0542, + "step": 14367 + }, + { + "epoch": 40.134078212290504, + "grad_norm": 1.2910338640213013, + "learning_rate": 2.7984729981378027e-05, + "loss": 0.0524, + "step": 14368 + }, + { + "epoch": 40.13687150837989, + "grad_norm": 1.250328779220581, + "learning_rate": 2.7984357541899443e-05, + "loss": 0.0719, + "step": 14369 + }, + { + "epoch": 40.13966480446928, + "grad_norm": 0.7681555151939392, + "learning_rate": 2.7983985102420856e-05, + "loss": 0.0585, + "step": 14370 + }, + { + "epoch": 40.142458100558656, + "grad_norm": 0.6215242743492126, + "learning_rate": 2.7983612662942273e-05, + "loss": 0.1108, + "step": 14371 + }, + { + "epoch": 40.14525139664804, + "grad_norm": 0.5097048282623291, + "learning_rate": 2.7983240223463686e-05, + "loss": 0.1084, + "step": 14372 + }, + { + "epoch": 40.14804469273743, + "grad_norm": 0.5649687051773071, + "learning_rate": 2.7982867783985103e-05, + "loss": 0.0893, + "step": 14373 + }, + { + "epoch": 40.150837988826815, + "grad_norm": 0.3667728304862976, + "learning_rate": 2.798249534450652e-05, + "loss": 0.0917, + "step": 14374 + }, + { + "epoch": 40.1536312849162, + "grad_norm": 0.6583181023597717, + "learning_rate": 2.7982122905027936e-05, + "loss": 0.116, + "step": 14375 + }, + { + "epoch": 40.15642458100559, + "grad_norm": 0.32655197381973267, + "learning_rate": 2.7981750465549352e-05, + "loss": 0.0803, + "step": 14376 + }, + { + "epoch": 40.159217877094974, + "grad_norm": 0.3498465120792389, + "learning_rate": 2.7981378026070762e-05, + "loss": 0.0804, + "step": 14377 + }, + { + "epoch": 40.16201117318436, + "grad_norm": 0.5142444968223572, + "learning_rate": 2.798100558659218e-05, + "loss": 0.0842, + "step": 14378 + }, + { + "epoch": 40.16480446927374, + "grad_norm": 0.3818032443523407, + "learning_rate": 2.7980633147113595e-05, + "loss": 0.0755, + "step": 14379 + }, + { + "epoch": 40.167597765363126, + "grad_norm": 0.33751991391181946, + "learning_rate": 2.798026070763501e-05, + "loss": 0.0748, + "step": 14380 + }, + { + "epoch": 40.17039106145251, + "grad_norm": 0.5693769454956055, + "learning_rate": 2.7979888268156425e-05, + "loss": 0.09, + "step": 14381 + }, + { + "epoch": 40.1731843575419, + "grad_norm": 1.0851978063583374, + "learning_rate": 2.797951582867784e-05, + "loss": 0.0818, + "step": 14382 + }, + { + "epoch": 40.175977653631286, + "grad_norm": 0.44379207491874695, + "learning_rate": 2.7979143389199254e-05, + "loss": 0.0567, + "step": 14383 + }, + { + "epoch": 40.17877094972067, + "grad_norm": 0.4657677114009857, + "learning_rate": 2.797877094972067e-05, + "loss": 0.0682, + "step": 14384 + }, + { + "epoch": 40.18156424581006, + "grad_norm": 0.752565860748291, + "learning_rate": 2.7978398510242087e-05, + "loss": 0.0659, + "step": 14385 + }, + { + "epoch": 40.184357541899445, + "grad_norm": 0.3436073362827301, + "learning_rate": 2.79780260707635e-05, + "loss": 0.0481, + "step": 14386 + }, + { + "epoch": 40.187150837988824, + "grad_norm": 0.47406312823295593, + "learning_rate": 2.7977653631284917e-05, + "loss": 0.0684, + "step": 14387 + }, + { + "epoch": 40.18994413407821, + "grad_norm": 0.3645632863044739, + "learning_rate": 2.7977281191806334e-05, + "loss": 0.056, + "step": 14388 + }, + { + "epoch": 40.1927374301676, + "grad_norm": 0.5328596830368042, + "learning_rate": 2.797690875232775e-05, + "loss": 0.0826, + "step": 14389 + }, + { + "epoch": 40.19553072625698, + "grad_norm": 0.599498450756073, + "learning_rate": 2.797653631284916e-05, + "loss": 0.0528, + "step": 14390 + }, + { + "epoch": 40.19832402234637, + "grad_norm": 0.48396867513656616, + "learning_rate": 2.7976163873370577e-05, + "loss": 0.0501, + "step": 14391 + }, + { + "epoch": 40.201117318435756, + "grad_norm": 0.5428760647773743, + "learning_rate": 2.7975791433891993e-05, + "loss": 0.0593, + "step": 14392 + }, + { + "epoch": 40.20391061452514, + "grad_norm": 0.500650942325592, + "learning_rate": 2.797541899441341e-05, + "loss": 0.0567, + "step": 14393 + }, + { + "epoch": 40.20670391061452, + "grad_norm": 0.5861997008323669, + "learning_rate": 2.7975046554934823e-05, + "loss": 0.0499, + "step": 14394 + }, + { + "epoch": 40.20949720670391, + "grad_norm": 0.7839285731315613, + "learning_rate": 2.797467411545624e-05, + "loss": 0.0783, + "step": 14395 + }, + { + "epoch": 40.212290502793294, + "grad_norm": 0.4657648801803589, + "learning_rate": 2.7974301675977656e-05, + "loss": 0.112, + "step": 14396 + }, + { + "epoch": 40.21508379888268, + "grad_norm": 0.5099153518676758, + "learning_rate": 2.797392923649907e-05, + "loss": 0.1139, + "step": 14397 + }, + { + "epoch": 40.21787709497207, + "grad_norm": 0.42119988799095154, + "learning_rate": 2.7973556797020486e-05, + "loss": 0.1025, + "step": 14398 + }, + { + "epoch": 40.220670391061454, + "grad_norm": 0.42016634345054626, + "learning_rate": 2.79731843575419e-05, + "loss": 0.1136, + "step": 14399 + }, + { + "epoch": 40.22346368715084, + "grad_norm": 0.5976691246032715, + "learning_rate": 2.7972811918063315e-05, + "loss": 0.0839, + "step": 14400 + }, + { + "epoch": 40.226256983240226, + "grad_norm": 0.5471959114074707, + "learning_rate": 2.7972439478584732e-05, + "loss": 0.0925, + "step": 14401 + }, + { + "epoch": 40.229050279329606, + "grad_norm": 0.42388787865638733, + "learning_rate": 2.7972067039106148e-05, + "loss": 0.0783, + "step": 14402 + }, + { + "epoch": 40.23184357541899, + "grad_norm": 1.051491379737854, + "learning_rate": 2.7971694599627558e-05, + "loss": 0.0788, + "step": 14403 + }, + { + "epoch": 40.23463687150838, + "grad_norm": 0.45184585452079773, + "learning_rate": 2.7971322160148975e-05, + "loss": 0.0794, + "step": 14404 + }, + { + "epoch": 40.237430167597765, + "grad_norm": 0.42618072032928467, + "learning_rate": 2.797094972067039e-05, + "loss": 0.0745, + "step": 14405 + }, + { + "epoch": 40.24022346368715, + "grad_norm": 0.64252769947052, + "learning_rate": 2.7970577281191808e-05, + "loss": 0.089, + "step": 14406 + }, + { + "epoch": 40.24301675977654, + "grad_norm": 0.3849390745162964, + "learning_rate": 2.7970204841713224e-05, + "loss": 0.0719, + "step": 14407 + }, + { + "epoch": 40.245810055865924, + "grad_norm": 0.6026856303215027, + "learning_rate": 2.7969832402234637e-05, + "loss": 0.085, + "step": 14408 + }, + { + "epoch": 40.24860335195531, + "grad_norm": 0.4292961657047272, + "learning_rate": 2.7969459962756054e-05, + "loss": 0.0612, + "step": 14409 + }, + { + "epoch": 40.25139664804469, + "grad_norm": 0.5009925961494446, + "learning_rate": 2.7969087523277467e-05, + "loss": 0.0689, + "step": 14410 + }, + { + "epoch": 40.254189944134076, + "grad_norm": 0.642277717590332, + "learning_rate": 2.7968715083798884e-05, + "loss": 0.061, + "step": 14411 + }, + { + "epoch": 40.25698324022346, + "grad_norm": 0.5778630971908569, + "learning_rate": 2.7968342644320297e-05, + "loss": 0.0685, + "step": 14412 + }, + { + "epoch": 40.25977653631285, + "grad_norm": 0.4615555703639984, + "learning_rate": 2.7967970204841713e-05, + "loss": 0.0555, + "step": 14413 + }, + { + "epoch": 40.262569832402235, + "grad_norm": 0.7726988196372986, + "learning_rate": 2.796759776536313e-05, + "loss": 0.0588, + "step": 14414 + }, + { + "epoch": 40.26536312849162, + "grad_norm": 3.481968402862549, + "learning_rate": 2.7967225325884546e-05, + "loss": 0.0512, + "step": 14415 + }, + { + "epoch": 40.26815642458101, + "grad_norm": 0.6325606107711792, + "learning_rate": 2.7966852886405963e-05, + "loss": 0.0612, + "step": 14416 + }, + { + "epoch": 40.270949720670394, + "grad_norm": 0.653575599193573, + "learning_rate": 2.7966480446927373e-05, + "loss": 0.0583, + "step": 14417 + }, + { + "epoch": 40.273743016759774, + "grad_norm": 0.5182580947875977, + "learning_rate": 2.796610800744879e-05, + "loss": 0.0494, + "step": 14418 + }, + { + "epoch": 40.27653631284916, + "grad_norm": 0.47365811467170715, + "learning_rate": 2.7965735567970206e-05, + "loss": 0.0543, + "step": 14419 + }, + { + "epoch": 40.279329608938546, + "grad_norm": 0.5956838130950928, + "learning_rate": 2.7965363128491622e-05, + "loss": 0.0553, + "step": 14420 + }, + { + "epoch": 40.28212290502793, + "grad_norm": 0.28832823038101196, + "learning_rate": 2.7964990689013035e-05, + "loss": 0.1122, + "step": 14421 + }, + { + "epoch": 40.28491620111732, + "grad_norm": 0.4647994041442871, + "learning_rate": 2.7964618249534452e-05, + "loss": 0.0936, + "step": 14422 + }, + { + "epoch": 40.287709497206706, + "grad_norm": 0.4172978699207306, + "learning_rate": 2.7964245810055865e-05, + "loss": 0.1262, + "step": 14423 + }, + { + "epoch": 40.29050279329609, + "grad_norm": 0.3757588565349579, + "learning_rate": 2.796387337057728e-05, + "loss": 0.0863, + "step": 14424 + }, + { + "epoch": 40.29329608938548, + "grad_norm": 0.37633031606674194, + "learning_rate": 2.7963500931098698e-05, + "loss": 0.1049, + "step": 14425 + }, + { + "epoch": 40.29608938547486, + "grad_norm": 0.6368955373764038, + "learning_rate": 2.796312849162011e-05, + "loss": 0.0953, + "step": 14426 + }, + { + "epoch": 40.298882681564244, + "grad_norm": 0.4368513226509094, + "learning_rate": 2.7962756052141528e-05, + "loss": 0.0892, + "step": 14427 + }, + { + "epoch": 40.30167597765363, + "grad_norm": 0.32566121220588684, + "learning_rate": 2.7962383612662944e-05, + "loss": 0.0631, + "step": 14428 + }, + { + "epoch": 40.30446927374302, + "grad_norm": 0.5550962686538696, + "learning_rate": 2.796201117318436e-05, + "loss": 0.0729, + "step": 14429 + }, + { + "epoch": 40.3072625698324, + "grad_norm": 0.5800783634185791, + "learning_rate": 2.796163873370577e-05, + "loss": 0.0814, + "step": 14430 + }, + { + "epoch": 40.31005586592179, + "grad_norm": 0.8175288438796997, + "learning_rate": 2.7961266294227187e-05, + "loss": 0.0848, + "step": 14431 + }, + { + "epoch": 40.312849162011176, + "grad_norm": 0.8370891213417053, + "learning_rate": 2.7960893854748604e-05, + "loss": 0.0752, + "step": 14432 + }, + { + "epoch": 40.315642458100555, + "grad_norm": 1.0616742372512817, + "learning_rate": 2.796052141527002e-05, + "loss": 0.0812, + "step": 14433 + }, + { + "epoch": 40.31843575418994, + "grad_norm": 0.5681815147399902, + "learning_rate": 2.7960148975791437e-05, + "loss": 0.0702, + "step": 14434 + }, + { + "epoch": 40.32122905027933, + "grad_norm": 0.5669416189193726, + "learning_rate": 2.795977653631285e-05, + "loss": 0.0609, + "step": 14435 + }, + { + "epoch": 40.324022346368714, + "grad_norm": 0.4177617132663727, + "learning_rate": 2.7959404096834266e-05, + "loss": 0.0593, + "step": 14436 + }, + { + "epoch": 40.3268156424581, + "grad_norm": 0.5505834221839905, + "learning_rate": 2.795903165735568e-05, + "loss": 0.0596, + "step": 14437 + }, + { + "epoch": 40.32960893854749, + "grad_norm": 4.969182014465332, + "learning_rate": 2.7958659217877096e-05, + "loss": 0.0519, + "step": 14438 + }, + { + "epoch": 40.332402234636874, + "grad_norm": 0.5033639073371887, + "learning_rate": 2.795828677839851e-05, + "loss": 0.0539, + "step": 14439 + }, + { + "epoch": 40.33519553072626, + "grad_norm": 0.5225487351417542, + "learning_rate": 2.7957914338919926e-05, + "loss": 0.063, + "step": 14440 + }, + { + "epoch": 40.33798882681564, + "grad_norm": 0.4726100564002991, + "learning_rate": 2.7957541899441342e-05, + "loss": 0.0489, + "step": 14441 + }, + { + "epoch": 40.340782122905026, + "grad_norm": 0.5585299134254456, + "learning_rate": 2.795716945996276e-05, + "loss": 0.063, + "step": 14442 + }, + { + "epoch": 40.34357541899441, + "grad_norm": 0.5827820301055908, + "learning_rate": 2.7956797020484172e-05, + "loss": 0.0558, + "step": 14443 + }, + { + "epoch": 40.3463687150838, + "grad_norm": 0.8305608630180359, + "learning_rate": 2.7956424581005585e-05, + "loss": 0.0703, + "step": 14444 + }, + { + "epoch": 40.349162011173185, + "grad_norm": 0.937005341053009, + "learning_rate": 2.7956052141527e-05, + "loss": 0.0648, + "step": 14445 + }, + { + "epoch": 40.35195530726257, + "grad_norm": 0.4663710594177246, + "learning_rate": 2.7955679702048418e-05, + "loss": 0.1125, + "step": 14446 + }, + { + "epoch": 40.35474860335196, + "grad_norm": 0.48535189032554626, + "learning_rate": 2.7955307262569835e-05, + "loss": 0.1212, + "step": 14447 + }, + { + "epoch": 40.357541899441344, + "grad_norm": 0.41850247979164124, + "learning_rate": 2.7954934823091248e-05, + "loss": 0.0953, + "step": 14448 + }, + { + "epoch": 40.36033519553072, + "grad_norm": 0.4255189597606659, + "learning_rate": 2.7954562383612664e-05, + "loss": 0.0872, + "step": 14449 + }, + { + "epoch": 40.36312849162011, + "grad_norm": 0.4751981198787689, + "learning_rate": 2.7954189944134078e-05, + "loss": 0.0822, + "step": 14450 + }, + { + "epoch": 40.365921787709496, + "grad_norm": 0.4691271185874939, + "learning_rate": 2.7953817504655494e-05, + "loss": 0.0849, + "step": 14451 + }, + { + "epoch": 40.36871508379888, + "grad_norm": 0.3597653806209564, + "learning_rate": 2.795344506517691e-05, + "loss": 0.07, + "step": 14452 + }, + { + "epoch": 40.37150837988827, + "grad_norm": 0.6811707615852356, + "learning_rate": 2.7953072625698324e-05, + "loss": 0.0875, + "step": 14453 + }, + { + "epoch": 40.374301675977655, + "grad_norm": 2.1243433952331543, + "learning_rate": 2.795270018621974e-05, + "loss": 0.0804, + "step": 14454 + }, + { + "epoch": 40.37709497206704, + "grad_norm": 0.46102091670036316, + "learning_rate": 2.7952327746741157e-05, + "loss": 0.0682, + "step": 14455 + }, + { + "epoch": 40.37988826815643, + "grad_norm": 0.4449564218521118, + "learning_rate": 2.7951955307262573e-05, + "loss": 0.0889, + "step": 14456 + }, + { + "epoch": 40.38268156424581, + "grad_norm": 0.4207201898097992, + "learning_rate": 2.7951582867783983e-05, + "loss": 0.0683, + "step": 14457 + }, + { + "epoch": 40.385474860335194, + "grad_norm": 0.5252768993377686, + "learning_rate": 2.79512104283054e-05, + "loss": 0.0772, + "step": 14458 + }, + { + "epoch": 40.38826815642458, + "grad_norm": 0.42456692457199097, + "learning_rate": 2.7950837988826816e-05, + "loss": 0.0872, + "step": 14459 + }, + { + "epoch": 40.391061452513966, + "grad_norm": 0.4476633369922638, + "learning_rate": 2.7950465549348233e-05, + "loss": 0.0624, + "step": 14460 + }, + { + "epoch": 40.39385474860335, + "grad_norm": 0.43775421380996704, + "learning_rate": 2.795009310986965e-05, + "loss": 0.0569, + "step": 14461 + }, + { + "epoch": 40.39664804469274, + "grad_norm": 0.6744902729988098, + "learning_rate": 2.7949720670391062e-05, + "loss": 0.0745, + "step": 14462 + }, + { + "epoch": 40.399441340782126, + "grad_norm": 0.5445318222045898, + "learning_rate": 2.7949348230912476e-05, + "loss": 0.0621, + "step": 14463 + }, + { + "epoch": 40.40223463687151, + "grad_norm": 0.5824902057647705, + "learning_rate": 2.7948975791433892e-05, + "loss": 0.0565, + "step": 14464 + }, + { + "epoch": 40.40502793296089, + "grad_norm": 0.493397057056427, + "learning_rate": 2.794860335195531e-05, + "loss": 0.0728, + "step": 14465 + }, + { + "epoch": 40.40782122905028, + "grad_norm": 0.4022001028060913, + "learning_rate": 2.7948230912476722e-05, + "loss": 0.0395, + "step": 14466 + }, + { + "epoch": 40.410614525139664, + "grad_norm": 1.2229799032211304, + "learning_rate": 2.794785847299814e-05, + "loss": 0.0436, + "step": 14467 + }, + { + "epoch": 40.41340782122905, + "grad_norm": 0.49946653842926025, + "learning_rate": 2.7947486033519555e-05, + "loss": 0.0421, + "step": 14468 + }, + { + "epoch": 40.41620111731844, + "grad_norm": 3.8544399738311768, + "learning_rate": 2.794711359404097e-05, + "loss": 0.071, + "step": 14469 + }, + { + "epoch": 40.41899441340782, + "grad_norm": 1.3584914207458496, + "learning_rate": 2.7946741154562385e-05, + "loss": 0.0672, + "step": 14470 + }, + { + "epoch": 40.42178770949721, + "grad_norm": 0.46967747807502747, + "learning_rate": 2.7946368715083798e-05, + "loss": 0.1106, + "step": 14471 + }, + { + "epoch": 40.42458100558659, + "grad_norm": 0.5531519651412964, + "learning_rate": 2.7945996275605214e-05, + "loss": 0.1038, + "step": 14472 + }, + { + "epoch": 40.427374301675975, + "grad_norm": 0.39342841506004333, + "learning_rate": 2.794562383612663e-05, + "loss": 0.0982, + "step": 14473 + }, + { + "epoch": 40.43016759776536, + "grad_norm": 0.44272753596305847, + "learning_rate": 2.7945251396648047e-05, + "loss": 0.1056, + "step": 14474 + }, + { + "epoch": 40.43296089385475, + "grad_norm": 0.4127972424030304, + "learning_rate": 2.794487895716946e-05, + "loss": 0.0891, + "step": 14475 + }, + { + "epoch": 40.435754189944134, + "grad_norm": 0.5239284634590149, + "learning_rate": 2.7944506517690874e-05, + "loss": 0.0843, + "step": 14476 + }, + { + "epoch": 40.43854748603352, + "grad_norm": 0.38211050629615784, + "learning_rate": 2.794413407821229e-05, + "loss": 0.093, + "step": 14477 + }, + { + "epoch": 40.44134078212291, + "grad_norm": 5.0251665115356445, + "learning_rate": 2.7943761638733707e-05, + "loss": 0.0865, + "step": 14478 + }, + { + "epoch": 40.444134078212294, + "grad_norm": 0.45448026061058044, + "learning_rate": 2.7943389199255123e-05, + "loss": 0.0901, + "step": 14479 + }, + { + "epoch": 40.44692737430167, + "grad_norm": 0.37671929597854614, + "learning_rate": 2.7943016759776536e-05, + "loss": 0.0817, + "step": 14480 + }, + { + "epoch": 40.44972067039106, + "grad_norm": 0.4009794294834137, + "learning_rate": 2.7942644320297953e-05, + "loss": 0.0694, + "step": 14481 + }, + { + "epoch": 40.452513966480446, + "grad_norm": 0.9435110092163086, + "learning_rate": 2.794227188081937e-05, + "loss": 0.0907, + "step": 14482 + }, + { + "epoch": 40.45530726256983, + "grad_norm": 0.9389351606369019, + "learning_rate": 2.7941899441340783e-05, + "loss": 0.0748, + "step": 14483 + }, + { + "epoch": 40.45810055865922, + "grad_norm": 0.6485225558280945, + "learning_rate": 2.7941527001862196e-05, + "loss": 0.0669, + "step": 14484 + }, + { + "epoch": 40.460893854748605, + "grad_norm": 0.582301914691925, + "learning_rate": 2.7941154562383612e-05, + "loss": 0.0605, + "step": 14485 + }, + { + "epoch": 40.46368715083799, + "grad_norm": 0.5514672994613647, + "learning_rate": 2.794078212290503e-05, + "loss": 0.0804, + "step": 14486 + }, + { + "epoch": 40.46648044692738, + "grad_norm": 0.5276500582695007, + "learning_rate": 2.7940409683426445e-05, + "loss": 0.0702, + "step": 14487 + }, + { + "epoch": 40.46927374301676, + "grad_norm": 0.8701039552688599, + "learning_rate": 2.794003724394786e-05, + "loss": 0.0541, + "step": 14488 + }, + { + "epoch": 40.47206703910614, + "grad_norm": 0.5142228007316589, + "learning_rate": 2.7939664804469275e-05, + "loss": 0.0597, + "step": 14489 + }, + { + "epoch": 40.47486033519553, + "grad_norm": 0.6884022951126099, + "learning_rate": 2.7939292364990688e-05, + "loss": 0.078, + "step": 14490 + }, + { + "epoch": 40.477653631284916, + "grad_norm": 0.5889853239059448, + "learning_rate": 2.7938919925512105e-05, + "loss": 0.0625, + "step": 14491 + }, + { + "epoch": 40.4804469273743, + "grad_norm": 0.6778228282928467, + "learning_rate": 2.793854748603352e-05, + "loss": 0.047, + "step": 14492 + }, + { + "epoch": 40.48324022346369, + "grad_norm": 0.7762649655342102, + "learning_rate": 2.7938175046554934e-05, + "loss": 0.0599, + "step": 14493 + }, + { + "epoch": 40.486033519553075, + "grad_norm": 0.826934814453125, + "learning_rate": 2.793780260707635e-05, + "loss": 0.0671, + "step": 14494 + }, + { + "epoch": 40.48882681564246, + "grad_norm": 1.0442043542861938, + "learning_rate": 2.7937430167597767e-05, + "loss": 0.0821, + "step": 14495 + }, + { + "epoch": 40.49162011173184, + "grad_norm": 0.43911314010620117, + "learning_rate": 2.793705772811918e-05, + "loss": 0.1057, + "step": 14496 + }, + { + "epoch": 40.49441340782123, + "grad_norm": 0.35671278834342957, + "learning_rate": 2.7936685288640594e-05, + "loss": 0.1034, + "step": 14497 + }, + { + "epoch": 40.497206703910614, + "grad_norm": 0.4185168147087097, + "learning_rate": 2.793631284916201e-05, + "loss": 0.0982, + "step": 14498 + }, + { + "epoch": 40.5, + "grad_norm": 0.6661487817764282, + "learning_rate": 2.7935940409683427e-05, + "loss": 0.1043, + "step": 14499 + }, + { + "epoch": 40.502793296089386, + "grad_norm": 0.47618383169174194, + "learning_rate": 2.7935567970204843e-05, + "loss": 0.0795, + "step": 14500 + }, + { + "epoch": 40.50558659217877, + "grad_norm": 0.6199933290481567, + "learning_rate": 2.793519553072626e-05, + "loss": 0.0956, + "step": 14501 + }, + { + "epoch": 40.50837988826816, + "grad_norm": 0.5050598978996277, + "learning_rate": 2.7934823091247673e-05, + "loss": 0.0903, + "step": 14502 + }, + { + "epoch": 40.51117318435754, + "grad_norm": 0.44571369886398315, + "learning_rate": 2.7934450651769086e-05, + "loss": 0.0989, + "step": 14503 + }, + { + "epoch": 40.513966480446925, + "grad_norm": 0.6096507906913757, + "learning_rate": 2.7934078212290503e-05, + "loss": 0.0992, + "step": 14504 + }, + { + "epoch": 40.51675977653631, + "grad_norm": 0.4174920320510864, + "learning_rate": 2.793370577281192e-05, + "loss": 0.0811, + "step": 14505 + }, + { + "epoch": 40.5195530726257, + "grad_norm": 0.45605355501174927, + "learning_rate": 2.7933333333333332e-05, + "loss": 0.0709, + "step": 14506 + }, + { + "epoch": 40.522346368715084, + "grad_norm": 0.6448119282722473, + "learning_rate": 2.793296089385475e-05, + "loss": 0.0744, + "step": 14507 + }, + { + "epoch": 40.52513966480447, + "grad_norm": 0.42011839151382446, + "learning_rate": 2.7932588454376165e-05, + "loss": 0.0625, + "step": 14508 + }, + { + "epoch": 40.52793296089386, + "grad_norm": 0.35459408164024353, + "learning_rate": 2.7932216014897582e-05, + "loss": 0.0648, + "step": 14509 + }, + { + "epoch": 40.53072625698324, + "grad_norm": 0.6403807997703552, + "learning_rate": 2.7931843575418995e-05, + "loss": 0.0633, + "step": 14510 + }, + { + "epoch": 40.53351955307262, + "grad_norm": 0.5211617350578308, + "learning_rate": 2.793147113594041e-05, + "loss": 0.0634, + "step": 14511 + }, + { + "epoch": 40.53631284916201, + "grad_norm": 0.6971527934074402, + "learning_rate": 2.7931098696461825e-05, + "loss": 0.0705, + "step": 14512 + }, + { + "epoch": 40.539106145251395, + "grad_norm": 0.4454783797264099, + "learning_rate": 2.793072625698324e-05, + "loss": 0.0612, + "step": 14513 + }, + { + "epoch": 40.54189944134078, + "grad_norm": 0.6056893467903137, + "learning_rate": 2.7930353817504658e-05, + "loss": 0.0581, + "step": 14514 + }, + { + "epoch": 40.54469273743017, + "grad_norm": 0.7049122452735901, + "learning_rate": 2.792998137802607e-05, + "loss": 0.0668, + "step": 14515 + }, + { + "epoch": 40.547486033519554, + "grad_norm": 0.5253912210464478, + "learning_rate": 2.7929608938547484e-05, + "loss": 0.0573, + "step": 14516 + }, + { + "epoch": 40.55027932960894, + "grad_norm": 0.7046807408332825, + "learning_rate": 2.79292364990689e-05, + "loss": 0.0731, + "step": 14517 + }, + { + "epoch": 40.55307262569833, + "grad_norm": 0.6289392113685608, + "learning_rate": 2.7928864059590317e-05, + "loss": 0.0561, + "step": 14518 + }, + { + "epoch": 40.555865921787706, + "grad_norm": 0.7007555961608887, + "learning_rate": 2.7928491620111734e-05, + "loss": 0.0856, + "step": 14519 + }, + { + "epoch": 40.55865921787709, + "grad_norm": 0.6680285930633545, + "learning_rate": 2.7928119180633147e-05, + "loss": 0.0735, + "step": 14520 + }, + { + "epoch": 40.56145251396648, + "grad_norm": 0.4536433815956116, + "learning_rate": 2.7927746741154564e-05, + "loss": 0.1123, + "step": 14521 + }, + { + "epoch": 40.564245810055866, + "grad_norm": 0.38425010442733765, + "learning_rate": 2.792737430167598e-05, + "loss": 0.0893, + "step": 14522 + }, + { + "epoch": 40.56703910614525, + "grad_norm": 0.4154813885688782, + "learning_rate": 2.7927001862197393e-05, + "loss": 0.0943, + "step": 14523 + }, + { + "epoch": 40.56983240223464, + "grad_norm": 0.37830376625061035, + "learning_rate": 2.7926629422718806e-05, + "loss": 0.0907, + "step": 14524 + }, + { + "epoch": 40.572625698324025, + "grad_norm": 0.47841203212738037, + "learning_rate": 2.7926256983240223e-05, + "loss": 0.0934, + "step": 14525 + }, + { + "epoch": 40.57541899441341, + "grad_norm": 0.46583595871925354, + "learning_rate": 2.792588454376164e-05, + "loss": 0.091, + "step": 14526 + }, + { + "epoch": 40.57821229050279, + "grad_norm": 0.43896111845970154, + "learning_rate": 2.7925512104283056e-05, + "loss": 0.09, + "step": 14527 + }, + { + "epoch": 40.58100558659218, + "grad_norm": 0.44908177852630615, + "learning_rate": 2.7925139664804472e-05, + "loss": 0.0777, + "step": 14528 + }, + { + "epoch": 40.58379888268156, + "grad_norm": 0.3676010072231293, + "learning_rate": 2.7924767225325886e-05, + "loss": 0.0721, + "step": 14529 + }, + { + "epoch": 40.58659217877095, + "grad_norm": 0.46356382966041565, + "learning_rate": 2.79243947858473e-05, + "loss": 0.0784, + "step": 14530 + }, + { + "epoch": 40.589385474860336, + "grad_norm": 0.5510007739067078, + "learning_rate": 2.7924022346368715e-05, + "loss": 0.0676, + "step": 14531 + }, + { + "epoch": 40.59217877094972, + "grad_norm": 0.44155868887901306, + "learning_rate": 2.7923649906890132e-05, + "loss": 0.0682, + "step": 14532 + }, + { + "epoch": 40.59497206703911, + "grad_norm": 0.5362675189971924, + "learning_rate": 2.7923277467411545e-05, + "loss": 0.0648, + "step": 14533 + }, + { + "epoch": 40.59776536312849, + "grad_norm": 0.419712096452713, + "learning_rate": 2.792290502793296e-05, + "loss": 0.0596, + "step": 14534 + }, + { + "epoch": 40.600558659217874, + "grad_norm": 0.486835241317749, + "learning_rate": 2.7922532588454378e-05, + "loss": 0.0729, + "step": 14535 + }, + { + "epoch": 40.60335195530726, + "grad_norm": 0.7811875939369202, + "learning_rate": 2.792216014897579e-05, + "loss": 0.0536, + "step": 14536 + }, + { + "epoch": 40.60614525139665, + "grad_norm": 0.4854688346385956, + "learning_rate": 2.7921787709497208e-05, + "loss": 0.0623, + "step": 14537 + }, + { + "epoch": 40.608938547486034, + "grad_norm": 0.5020999312400818, + "learning_rate": 2.792141527001862e-05, + "loss": 0.0613, + "step": 14538 + }, + { + "epoch": 40.61173184357542, + "grad_norm": 0.57589191198349, + "learning_rate": 2.7921042830540037e-05, + "loss": 0.0588, + "step": 14539 + }, + { + "epoch": 40.614525139664806, + "grad_norm": 4.15870475769043, + "learning_rate": 2.7920670391061454e-05, + "loss": 0.0704, + "step": 14540 + }, + { + "epoch": 40.61731843575419, + "grad_norm": 0.9325683116912842, + "learning_rate": 2.792029795158287e-05, + "loss": 0.0503, + "step": 14541 + }, + { + "epoch": 40.62011173184357, + "grad_norm": 0.6491597890853882, + "learning_rate": 2.7919925512104284e-05, + "loss": 0.0649, + "step": 14542 + }, + { + "epoch": 40.62290502793296, + "grad_norm": 0.5965604186058044, + "learning_rate": 2.7919553072625697e-05, + "loss": 0.0567, + "step": 14543 + }, + { + "epoch": 40.625698324022345, + "grad_norm": 1.9754986763000488, + "learning_rate": 2.7919180633147113e-05, + "loss": 0.0649, + "step": 14544 + }, + { + "epoch": 40.62849162011173, + "grad_norm": 0.8857296705245972, + "learning_rate": 2.791880819366853e-05, + "loss": 0.0754, + "step": 14545 + }, + { + "epoch": 40.63128491620112, + "grad_norm": 0.3801395297050476, + "learning_rate": 2.7918435754189946e-05, + "loss": 0.1142, + "step": 14546 + }, + { + "epoch": 40.634078212290504, + "grad_norm": 0.5717987418174744, + "learning_rate": 2.791806331471136e-05, + "loss": 0.1133, + "step": 14547 + }, + { + "epoch": 40.63687150837989, + "grad_norm": 0.39118239283561707, + "learning_rate": 2.7917690875232776e-05, + "loss": 0.1203, + "step": 14548 + }, + { + "epoch": 40.63966480446928, + "grad_norm": 0.5031203031539917, + "learning_rate": 2.7917318435754193e-05, + "loss": 0.0949, + "step": 14549 + }, + { + "epoch": 40.642458100558656, + "grad_norm": 0.8269062042236328, + "learning_rate": 2.7916945996275606e-05, + "loss": 0.0994, + "step": 14550 + }, + { + "epoch": 40.64525139664804, + "grad_norm": 0.5697177052497864, + "learning_rate": 2.791657355679702e-05, + "loss": 0.0816, + "step": 14551 + }, + { + "epoch": 40.64804469273743, + "grad_norm": 0.6117486953735352, + "learning_rate": 2.7916201117318435e-05, + "loss": 0.0902, + "step": 14552 + }, + { + "epoch": 40.650837988826815, + "grad_norm": 0.6950986385345459, + "learning_rate": 2.7915828677839852e-05, + "loss": 0.0915, + "step": 14553 + }, + { + "epoch": 40.6536312849162, + "grad_norm": 0.3879424035549164, + "learning_rate": 2.791545623836127e-05, + "loss": 0.0828, + "step": 14554 + }, + { + "epoch": 40.65642458100559, + "grad_norm": 0.7181951403617859, + "learning_rate": 2.7915083798882685e-05, + "loss": 0.0943, + "step": 14555 + }, + { + "epoch": 40.659217877094974, + "grad_norm": 0.7685832977294922, + "learning_rate": 2.7914711359404095e-05, + "loss": 0.0964, + "step": 14556 + }, + { + "epoch": 40.66201117318436, + "grad_norm": 0.4591868817806244, + "learning_rate": 2.791433891992551e-05, + "loss": 0.0708, + "step": 14557 + }, + { + "epoch": 40.66480446927374, + "grad_norm": 0.4427039325237274, + "learning_rate": 2.7913966480446928e-05, + "loss": 0.0619, + "step": 14558 + }, + { + "epoch": 40.667597765363126, + "grad_norm": 0.6571880578994751, + "learning_rate": 2.7913594040968344e-05, + "loss": 0.0806, + "step": 14559 + }, + { + "epoch": 40.67039106145251, + "grad_norm": 0.5900469422340393, + "learning_rate": 2.7913221601489758e-05, + "loss": 0.0786, + "step": 14560 + }, + { + "epoch": 40.6731843575419, + "grad_norm": 1.4742902517318726, + "learning_rate": 2.7912849162011174e-05, + "loss": 0.0886, + "step": 14561 + }, + { + "epoch": 40.675977653631286, + "grad_norm": 0.6321272850036621, + "learning_rate": 2.791247672253259e-05, + "loss": 0.0676, + "step": 14562 + }, + { + "epoch": 40.67877094972067, + "grad_norm": 0.5920112133026123, + "learning_rate": 2.7912104283054004e-05, + "loss": 0.0605, + "step": 14563 + }, + { + "epoch": 40.68156424581006, + "grad_norm": 0.5011624693870544, + "learning_rate": 2.791173184357542e-05, + "loss": 0.0636, + "step": 14564 + }, + { + "epoch": 40.684357541899445, + "grad_norm": 0.6942429542541504, + "learning_rate": 2.7911359404096833e-05, + "loss": 0.0695, + "step": 14565 + }, + { + "epoch": 40.687150837988824, + "grad_norm": 1.290750503540039, + "learning_rate": 2.791098696461825e-05, + "loss": 0.0561, + "step": 14566 + }, + { + "epoch": 40.68994413407821, + "grad_norm": 0.6236220002174377, + "learning_rate": 2.7910614525139667e-05, + "loss": 0.0674, + "step": 14567 + }, + { + "epoch": 40.6927374301676, + "grad_norm": 1.2290533781051636, + "learning_rate": 2.7910242085661083e-05, + "loss": 0.0828, + "step": 14568 + }, + { + "epoch": 40.69553072625698, + "grad_norm": 0.9133752584457397, + "learning_rate": 2.7909869646182496e-05, + "loss": 0.0604, + "step": 14569 + }, + { + "epoch": 40.69832402234637, + "grad_norm": 0.9128588438034058, + "learning_rate": 2.790949720670391e-05, + "loss": 0.0736, + "step": 14570 + }, + { + "epoch": 40.701117318435756, + "grad_norm": 0.35948440432548523, + "learning_rate": 2.7909124767225326e-05, + "loss": 0.1269, + "step": 14571 + }, + { + "epoch": 40.70391061452514, + "grad_norm": 0.4382905662059784, + "learning_rate": 2.7908752327746742e-05, + "loss": 0.1037, + "step": 14572 + }, + { + "epoch": 40.70670391061452, + "grad_norm": 0.43199819326400757, + "learning_rate": 2.790837988826816e-05, + "loss": 0.0929, + "step": 14573 + }, + { + "epoch": 40.70949720670391, + "grad_norm": 0.5664367079734802, + "learning_rate": 2.7908007448789572e-05, + "loss": 0.0909, + "step": 14574 + }, + { + "epoch": 40.712290502793294, + "grad_norm": 0.4947103261947632, + "learning_rate": 2.790763500931099e-05, + "loss": 0.1143, + "step": 14575 + }, + { + "epoch": 40.71508379888268, + "grad_norm": 0.41548848152160645, + "learning_rate": 2.7907262569832402e-05, + "loss": 0.1068, + "step": 14576 + }, + { + "epoch": 40.71787709497207, + "grad_norm": 0.5549097061157227, + "learning_rate": 2.790689013035382e-05, + "loss": 0.0869, + "step": 14577 + }, + { + "epoch": 40.720670391061454, + "grad_norm": 0.49775880575180054, + "learning_rate": 2.790651769087523e-05, + "loss": 0.0873, + "step": 14578 + }, + { + "epoch": 40.72346368715084, + "grad_norm": 0.5756362080574036, + "learning_rate": 2.7906145251396648e-05, + "loss": 0.0794, + "step": 14579 + }, + { + "epoch": 40.726256983240226, + "grad_norm": 1.2377746105194092, + "learning_rate": 2.7905772811918065e-05, + "loss": 0.093, + "step": 14580 + }, + { + "epoch": 40.729050279329606, + "grad_norm": 0.51951664686203, + "learning_rate": 2.790540037243948e-05, + "loss": 0.085, + "step": 14581 + }, + { + "epoch": 40.73184357541899, + "grad_norm": 0.6841007471084595, + "learning_rate": 2.7905027932960894e-05, + "loss": 0.0772, + "step": 14582 + }, + { + "epoch": 40.73463687150838, + "grad_norm": 0.7512128949165344, + "learning_rate": 2.7904655493482307e-05, + "loss": 0.0704, + "step": 14583 + }, + { + "epoch": 40.737430167597765, + "grad_norm": 0.5525809526443481, + "learning_rate": 2.7904283054003724e-05, + "loss": 0.0709, + "step": 14584 + }, + { + "epoch": 40.74022346368715, + "grad_norm": 0.7639769315719604, + "learning_rate": 2.790391061452514e-05, + "loss": 0.0719, + "step": 14585 + }, + { + "epoch": 40.74301675977654, + "grad_norm": 1.0180376768112183, + "learning_rate": 2.7903538175046557e-05, + "loss": 0.0557, + "step": 14586 + }, + { + "epoch": 40.745810055865924, + "grad_norm": 0.40605321526527405, + "learning_rate": 2.790316573556797e-05, + "loss": 0.0549, + "step": 14587 + }, + { + "epoch": 40.74860335195531, + "grad_norm": 0.6973661184310913, + "learning_rate": 2.7902793296089387e-05, + "loss": 0.0639, + "step": 14588 + }, + { + "epoch": 40.75139664804469, + "grad_norm": 0.7087168097496033, + "learning_rate": 2.7902420856610803e-05, + "loss": 0.0482, + "step": 14589 + }, + { + "epoch": 40.754189944134076, + "grad_norm": 0.531726598739624, + "learning_rate": 2.7902048417132216e-05, + "loss": 0.0805, + "step": 14590 + }, + { + "epoch": 40.75698324022346, + "grad_norm": 0.5857186317443848, + "learning_rate": 2.790167597765363e-05, + "loss": 0.0562, + "step": 14591 + }, + { + "epoch": 40.75977653631285, + "grad_norm": 0.6027331948280334, + "learning_rate": 2.7901303538175046e-05, + "loss": 0.0566, + "step": 14592 + }, + { + "epoch": 40.762569832402235, + "grad_norm": 0.8034844398498535, + "learning_rate": 2.7900931098696463e-05, + "loss": 0.0761, + "step": 14593 + }, + { + "epoch": 40.76536312849162, + "grad_norm": 0.9171918034553528, + "learning_rate": 2.790055865921788e-05, + "loss": 0.0566, + "step": 14594 + }, + { + "epoch": 40.76815642458101, + "grad_norm": 1.3798469305038452, + "learning_rate": 2.7900186219739296e-05, + "loss": 0.0858, + "step": 14595 + }, + { + "epoch": 40.770949720670394, + "grad_norm": 0.36340096592903137, + "learning_rate": 2.7899813780260705e-05, + "loss": 0.1105, + "step": 14596 + }, + { + "epoch": 40.773743016759774, + "grad_norm": 0.4096601903438568, + "learning_rate": 2.7899441340782122e-05, + "loss": 0.111, + "step": 14597 + }, + { + "epoch": 40.77653631284916, + "grad_norm": 0.364399790763855, + "learning_rate": 2.789906890130354e-05, + "loss": 0.0952, + "step": 14598 + }, + { + "epoch": 40.779329608938546, + "grad_norm": 0.4748511016368866, + "learning_rate": 2.7898696461824955e-05, + "loss": 0.0942, + "step": 14599 + }, + { + "epoch": 40.78212290502793, + "grad_norm": 0.5478412508964539, + "learning_rate": 2.7898324022346368e-05, + "loss": 0.0948, + "step": 14600 + }, + { + "epoch": 40.78491620111732, + "grad_norm": 1.1447967290878296, + "learning_rate": 2.7897951582867785e-05, + "loss": 0.0876, + "step": 14601 + }, + { + "epoch": 40.787709497206706, + "grad_norm": 0.3725445568561554, + "learning_rate": 2.78975791433892e-05, + "loss": 0.0803, + "step": 14602 + }, + { + "epoch": 40.79050279329609, + "grad_norm": 0.6547613739967346, + "learning_rate": 2.7897206703910614e-05, + "loss": 0.0856, + "step": 14603 + }, + { + "epoch": 40.79329608938548, + "grad_norm": 0.5171564221382141, + "learning_rate": 2.789683426443203e-05, + "loss": 0.0728, + "step": 14604 + }, + { + "epoch": 40.79608938547486, + "grad_norm": 0.4760761857032776, + "learning_rate": 2.7896461824953444e-05, + "loss": 0.0834, + "step": 14605 + }, + { + "epoch": 40.798882681564244, + "grad_norm": 0.3625076115131378, + "learning_rate": 2.789608938547486e-05, + "loss": 0.0785, + "step": 14606 + }, + { + "epoch": 40.80167597765363, + "grad_norm": 0.4390624165534973, + "learning_rate": 2.7895716945996277e-05, + "loss": 0.0606, + "step": 14607 + }, + { + "epoch": 40.80446927374302, + "grad_norm": 0.5994507670402527, + "learning_rate": 2.7895344506517694e-05, + "loss": 0.0739, + "step": 14608 + }, + { + "epoch": 40.8072625698324, + "grad_norm": 1.1417317390441895, + "learning_rate": 2.7894972067039107e-05, + "loss": 0.0718, + "step": 14609 + }, + { + "epoch": 40.81005586592179, + "grad_norm": 0.49603936076164246, + "learning_rate": 2.789459962756052e-05, + "loss": 0.0643, + "step": 14610 + }, + { + "epoch": 40.812849162011176, + "grad_norm": 0.4498858153820038, + "learning_rate": 2.7894227188081937e-05, + "loss": 0.0641, + "step": 14611 + }, + { + "epoch": 40.815642458100555, + "grad_norm": 0.6053182482719421, + "learning_rate": 2.7893854748603353e-05, + "loss": 0.083, + "step": 14612 + }, + { + "epoch": 40.81843575418994, + "grad_norm": 0.47660768032073975, + "learning_rate": 2.789348230912477e-05, + "loss": 0.0651, + "step": 14613 + }, + { + "epoch": 40.82122905027933, + "grad_norm": 0.9234516024589539, + "learning_rate": 2.7893109869646183e-05, + "loss": 0.0572, + "step": 14614 + }, + { + "epoch": 40.824022346368714, + "grad_norm": 0.4281456768512726, + "learning_rate": 2.78927374301676e-05, + "loss": 0.0598, + "step": 14615 + }, + { + "epoch": 40.8268156424581, + "grad_norm": 0.7843309044837952, + "learning_rate": 2.7892364990689012e-05, + "loss": 0.0701, + "step": 14616 + }, + { + "epoch": 40.82960893854749, + "grad_norm": 0.6239994764328003, + "learning_rate": 2.789199255121043e-05, + "loss": 0.0551, + "step": 14617 + }, + { + "epoch": 40.832402234636874, + "grad_norm": 1.3414064645767212, + "learning_rate": 2.7891620111731842e-05, + "loss": 0.0644, + "step": 14618 + }, + { + "epoch": 40.83519553072626, + "grad_norm": 0.9046507477760315, + "learning_rate": 2.789124767225326e-05, + "loss": 0.0603, + "step": 14619 + }, + { + "epoch": 40.83798882681564, + "grad_norm": 2.760960817337036, + "learning_rate": 2.7890875232774675e-05, + "loss": 0.0804, + "step": 14620 + }, + { + "epoch": 40.840782122905026, + "grad_norm": 0.5460548400878906, + "learning_rate": 2.7890502793296092e-05, + "loss": 0.1343, + "step": 14621 + }, + { + "epoch": 40.84357541899441, + "grad_norm": 3.355820417404175, + "learning_rate": 2.7890130353817508e-05, + "loss": 0.1169, + "step": 14622 + }, + { + "epoch": 40.8463687150838, + "grad_norm": 0.6353903412818909, + "learning_rate": 2.7889757914338918e-05, + "loss": 0.1038, + "step": 14623 + }, + { + "epoch": 40.849162011173185, + "grad_norm": 0.6150625348091125, + "learning_rate": 2.7889385474860335e-05, + "loss": 0.1228, + "step": 14624 + }, + { + "epoch": 40.85195530726257, + "grad_norm": 1.1271939277648926, + "learning_rate": 2.788901303538175e-05, + "loss": 0.1036, + "step": 14625 + }, + { + "epoch": 40.85474860335196, + "grad_norm": 0.4333701431751251, + "learning_rate": 2.7888640595903168e-05, + "loss": 0.0952, + "step": 14626 + }, + { + "epoch": 40.857541899441344, + "grad_norm": 0.8575129508972168, + "learning_rate": 2.788826815642458e-05, + "loss": 0.1089, + "step": 14627 + }, + { + "epoch": 40.86033519553072, + "grad_norm": 0.6449485421180725, + "learning_rate": 2.7887895716945997e-05, + "loss": 0.1026, + "step": 14628 + }, + { + "epoch": 40.86312849162011, + "grad_norm": 0.6939660310745239, + "learning_rate": 2.7887523277467414e-05, + "loss": 0.0795, + "step": 14629 + }, + { + "epoch": 40.865921787709496, + "grad_norm": 0.5396578311920166, + "learning_rate": 2.7887150837988827e-05, + "loss": 0.0819, + "step": 14630 + }, + { + "epoch": 40.86871508379888, + "grad_norm": 0.839833676815033, + "learning_rate": 2.7886778398510244e-05, + "loss": 0.0825, + "step": 14631 + }, + { + "epoch": 40.87150837988827, + "grad_norm": 0.7828419804573059, + "learning_rate": 2.7886405959031657e-05, + "loss": 0.0778, + "step": 14632 + }, + { + "epoch": 40.874301675977655, + "grad_norm": 0.8061816096305847, + "learning_rate": 2.7886033519553073e-05, + "loss": 0.0721, + "step": 14633 + }, + { + "epoch": 40.87709497206704, + "grad_norm": 0.6840987205505371, + "learning_rate": 2.788566108007449e-05, + "loss": 0.0719, + "step": 14634 + }, + { + "epoch": 40.87988826815643, + "grad_norm": 0.45203396677970886, + "learning_rate": 2.7885288640595906e-05, + "loss": 0.0654, + "step": 14635 + }, + { + "epoch": 40.88268156424581, + "grad_norm": 0.5765841007232666, + "learning_rate": 2.7884916201117316e-05, + "loss": 0.0524, + "step": 14636 + }, + { + "epoch": 40.885474860335194, + "grad_norm": 0.43038612604141235, + "learning_rate": 2.7884543761638733e-05, + "loss": 0.06, + "step": 14637 + }, + { + "epoch": 40.88826815642458, + "grad_norm": 0.8130972981452942, + "learning_rate": 2.788417132216015e-05, + "loss": 0.0752, + "step": 14638 + }, + { + "epoch": 40.891061452513966, + "grad_norm": 0.7436705827713013, + "learning_rate": 2.7883798882681566e-05, + "loss": 0.0655, + "step": 14639 + }, + { + "epoch": 40.89385474860335, + "grad_norm": 0.5607637763023376, + "learning_rate": 2.7883426443202982e-05, + "loss": 0.0632, + "step": 14640 + }, + { + "epoch": 40.89664804469274, + "grad_norm": 0.5693575143814087, + "learning_rate": 2.7883054003724395e-05, + "loss": 0.0668, + "step": 14641 + }, + { + "epoch": 40.899441340782126, + "grad_norm": 0.7615234851837158, + "learning_rate": 2.7882681564245812e-05, + "loss": 0.0543, + "step": 14642 + }, + { + "epoch": 40.90223463687151, + "grad_norm": 0.7442018985748291, + "learning_rate": 2.7882309124767225e-05, + "loss": 0.0603, + "step": 14643 + }, + { + "epoch": 40.90502793296089, + "grad_norm": 0.8504787683486938, + "learning_rate": 2.788193668528864e-05, + "loss": 0.0611, + "step": 14644 + }, + { + "epoch": 40.90782122905028, + "grad_norm": 1.3431705236434937, + "learning_rate": 2.7881564245810055e-05, + "loss": 0.0889, + "step": 14645 + }, + { + "epoch": 40.910614525139664, + "grad_norm": 0.4658811092376709, + "learning_rate": 2.788119180633147e-05, + "loss": 0.12, + "step": 14646 + }, + { + "epoch": 40.91340782122905, + "grad_norm": 1.0010968446731567, + "learning_rate": 2.7880819366852888e-05, + "loss": 0.1074, + "step": 14647 + }, + { + "epoch": 40.91620111731844, + "grad_norm": 0.4536726176738739, + "learning_rate": 2.7880446927374304e-05, + "loss": 0.0979, + "step": 14648 + }, + { + "epoch": 40.91899441340782, + "grad_norm": 0.39666202664375305, + "learning_rate": 2.788007448789572e-05, + "loss": 0.0961, + "step": 14649 + }, + { + "epoch": 40.92178770949721, + "grad_norm": 0.5305911898612976, + "learning_rate": 2.787970204841713e-05, + "loss": 0.0957, + "step": 14650 + }, + { + "epoch": 40.92458100558659, + "grad_norm": 0.5663063526153564, + "learning_rate": 2.7879329608938547e-05, + "loss": 0.1055, + "step": 14651 + }, + { + "epoch": 40.927374301675975, + "grad_norm": 0.46840327978134155, + "learning_rate": 2.7878957169459964e-05, + "loss": 0.0929, + "step": 14652 + }, + { + "epoch": 40.93016759776536, + "grad_norm": 0.4518483281135559, + "learning_rate": 2.787858472998138e-05, + "loss": 0.0952, + "step": 14653 + }, + { + "epoch": 40.93296089385475, + "grad_norm": 0.5788100957870483, + "learning_rate": 2.7878212290502793e-05, + "loss": 0.092, + "step": 14654 + }, + { + "epoch": 40.935754189944134, + "grad_norm": 0.4993632733821869, + "learning_rate": 2.787783985102421e-05, + "loss": 0.0937, + "step": 14655 + }, + { + "epoch": 40.93854748603352, + "grad_norm": 0.47988271713256836, + "learning_rate": 2.7877467411545623e-05, + "loss": 0.0756, + "step": 14656 + }, + { + "epoch": 40.94134078212291, + "grad_norm": 0.5549359321594238, + "learning_rate": 2.787709497206704e-05, + "loss": 0.0886, + "step": 14657 + }, + { + "epoch": 40.944134078212294, + "grad_norm": 0.8118703961372375, + "learning_rate": 2.7876722532588456e-05, + "loss": 0.077, + "step": 14658 + }, + { + "epoch": 40.94692737430167, + "grad_norm": 0.5645464658737183, + "learning_rate": 2.787635009310987e-05, + "loss": 0.0829, + "step": 14659 + }, + { + "epoch": 40.94972067039106, + "grad_norm": 1.7529387474060059, + "learning_rate": 2.7875977653631286e-05, + "loss": 0.0828, + "step": 14660 + }, + { + "epoch": 40.952513966480446, + "grad_norm": 0.5295884013175964, + "learning_rate": 2.7875605214152702e-05, + "loss": 0.0687, + "step": 14661 + }, + { + "epoch": 40.95530726256983, + "grad_norm": 0.6344390511512756, + "learning_rate": 2.787523277467412e-05, + "loss": 0.074, + "step": 14662 + }, + { + "epoch": 40.95810055865922, + "grad_norm": 0.5011096596717834, + "learning_rate": 2.787486033519553e-05, + "loss": 0.0577, + "step": 14663 + }, + { + "epoch": 40.960893854748605, + "grad_norm": 0.43451863527297974, + "learning_rate": 2.7874487895716945e-05, + "loss": 0.0582, + "step": 14664 + }, + { + "epoch": 40.96368715083799, + "grad_norm": 0.40815645456314087, + "learning_rate": 2.787411545623836e-05, + "loss": 0.0537, + "step": 14665 + }, + { + "epoch": 40.96648044692738, + "grad_norm": 0.5409014225006104, + "learning_rate": 2.7873743016759778e-05, + "loss": 0.0622, + "step": 14666 + }, + { + "epoch": 40.96927374301676, + "grad_norm": 1.0386098623275757, + "learning_rate": 2.7873370577281195e-05, + "loss": 0.0785, + "step": 14667 + }, + { + "epoch": 40.97206703910614, + "grad_norm": 1.1590218544006348, + "learning_rate": 2.7872998137802608e-05, + "loss": 0.0564, + "step": 14668 + }, + { + "epoch": 40.97486033519553, + "grad_norm": 1.0375707149505615, + "learning_rate": 2.7872625698324024e-05, + "loss": 0.0561, + "step": 14669 + }, + { + "epoch": 40.977653631284916, + "grad_norm": 1.1281620264053345, + "learning_rate": 2.7872253258845438e-05, + "loss": 0.1009, + "step": 14670 + }, + { + "epoch": 40.9804469273743, + "grad_norm": 0.5173236131668091, + "learning_rate": 2.7871880819366854e-05, + "loss": 0.0997, + "step": 14671 + }, + { + "epoch": 40.98324022346369, + "grad_norm": 0.48880478739738464, + "learning_rate": 2.7871508379888267e-05, + "loss": 0.0861, + "step": 14672 + }, + { + "epoch": 40.986033519553075, + "grad_norm": 0.5137539505958557, + "learning_rate": 2.7871135940409684e-05, + "loss": 0.0761, + "step": 14673 + }, + { + "epoch": 40.98882681564246, + "grad_norm": 0.7893472909927368, + "learning_rate": 2.78707635009311e-05, + "loss": 0.0769, + "step": 14674 + }, + { + "epoch": 40.99162011173184, + "grad_norm": 0.6086484789848328, + "learning_rate": 2.7870391061452517e-05, + "loss": 0.0614, + "step": 14675 + }, + { + "epoch": 40.99441340782123, + "grad_norm": 0.5154939293861389, + "learning_rate": 2.787001862197393e-05, + "loss": 0.076, + "step": 14676 + }, + { + "epoch": 40.997206703910614, + "grad_norm": 0.41545847058296204, + "learning_rate": 2.7869646182495343e-05, + "loss": 0.0524, + "step": 14677 + }, + { + "epoch": 41.0, + "grad_norm": 0.6128211617469788, + "learning_rate": 2.786927374301676e-05, + "loss": 0.0578, + "step": 14678 + }, + { + "epoch": 41.002793296089386, + "grad_norm": 0.4900100529193878, + "learning_rate": 2.7868901303538176e-05, + "loss": 0.1245, + "step": 14679 + }, + { + "epoch": 41.00558659217877, + "grad_norm": 0.3723195791244507, + "learning_rate": 2.7868528864059593e-05, + "loss": 0.0912, + "step": 14680 + }, + { + "epoch": 41.00837988826816, + "grad_norm": 0.43763288855552673, + "learning_rate": 2.7868156424581006e-05, + "loss": 0.0919, + "step": 14681 + }, + { + "epoch": 41.01117318435754, + "grad_norm": 2.799816131591797, + "learning_rate": 2.7867783985102422e-05, + "loss": 0.1061, + "step": 14682 + }, + { + "epoch": 41.013966480446925, + "grad_norm": 0.6542422771453857, + "learning_rate": 2.7867411545623836e-05, + "loss": 0.115, + "step": 14683 + }, + { + "epoch": 41.01675977653631, + "grad_norm": 0.40656599402427673, + "learning_rate": 2.7867039106145252e-05, + "loss": 0.09, + "step": 14684 + }, + { + "epoch": 41.0195530726257, + "grad_norm": 0.37568140029907227, + "learning_rate": 2.7866666666666665e-05, + "loss": 0.0707, + "step": 14685 + }, + { + "epoch": 41.022346368715084, + "grad_norm": 0.7936986088752747, + "learning_rate": 2.7866294227188082e-05, + "loss": 0.0882, + "step": 14686 + }, + { + "epoch": 41.02513966480447, + "grad_norm": 0.4303237795829773, + "learning_rate": 2.78659217877095e-05, + "loss": 0.0781, + "step": 14687 + }, + { + "epoch": 41.02793296089386, + "grad_norm": 0.5066516995429993, + "learning_rate": 2.7865549348230915e-05, + "loss": 0.0884, + "step": 14688 + }, + { + "epoch": 41.03072625698324, + "grad_norm": 0.6076153516769409, + "learning_rate": 2.786517690875233e-05, + "loss": 0.0697, + "step": 14689 + }, + { + "epoch": 41.03351955307262, + "grad_norm": 0.6061561107635498, + "learning_rate": 2.786480446927374e-05, + "loss": 0.0674, + "step": 14690 + }, + { + "epoch": 41.03631284916201, + "grad_norm": 0.6678930521011353, + "learning_rate": 2.7864432029795158e-05, + "loss": 0.066, + "step": 14691 + }, + { + "epoch": 41.039106145251395, + "grad_norm": 0.3917797803878784, + "learning_rate": 2.7864059590316574e-05, + "loss": 0.0788, + "step": 14692 + }, + { + "epoch": 41.04189944134078, + "grad_norm": 0.4100882411003113, + "learning_rate": 2.786368715083799e-05, + "loss": 0.0695, + "step": 14693 + }, + { + "epoch": 41.04469273743017, + "grad_norm": 0.5277134776115417, + "learning_rate": 2.7863314711359404e-05, + "loss": 0.0641, + "step": 14694 + }, + { + "epoch": 41.047486033519554, + "grad_norm": 0.44914713501930237, + "learning_rate": 2.786294227188082e-05, + "loss": 0.0512, + "step": 14695 + }, + { + "epoch": 41.05027932960894, + "grad_norm": 0.5408116579055786, + "learning_rate": 2.7862569832402234e-05, + "loss": 0.0438, + "step": 14696 + }, + { + "epoch": 41.05307262569833, + "grad_norm": 0.5342534780502319, + "learning_rate": 2.786219739292365e-05, + "loss": 0.0502, + "step": 14697 + }, + { + "epoch": 41.055865921787706, + "grad_norm": 0.49865981936454773, + "learning_rate": 2.7861824953445067e-05, + "loss": 0.0616, + "step": 14698 + }, + { + "epoch": 41.05865921787709, + "grad_norm": 0.6132954359054565, + "learning_rate": 2.786145251396648e-05, + "loss": 0.0613, + "step": 14699 + }, + { + "epoch": 41.06145251396648, + "grad_norm": 0.4548164904117584, + "learning_rate": 2.7861080074487896e-05, + "loss": 0.04, + "step": 14700 + }, + { + "epoch": 41.064245810055866, + "grad_norm": 0.4129049479961395, + "learning_rate": 2.7860707635009313e-05, + "loss": 0.0376, + "step": 14701 + }, + { + "epoch": 41.06703910614525, + "grad_norm": 0.7624076008796692, + "learning_rate": 2.786033519553073e-05, + "loss": 0.0524, + "step": 14702 + }, + { + "epoch": 41.06983240223464, + "grad_norm": 0.9274603128433228, + "learning_rate": 2.785996275605214e-05, + "loss": 0.0833, + "step": 14703 + }, + { + "epoch": 41.072625698324025, + "grad_norm": 0.4091751277446747, + "learning_rate": 2.7859590316573556e-05, + "loss": 0.101, + "step": 14704 + }, + { + "epoch": 41.07541899441341, + "grad_norm": 0.45989635586738586, + "learning_rate": 2.7859217877094972e-05, + "loss": 0.0884, + "step": 14705 + }, + { + "epoch": 41.07821229050279, + "grad_norm": 0.39060357213020325, + "learning_rate": 2.785884543761639e-05, + "loss": 0.0931, + "step": 14706 + }, + { + "epoch": 41.08100558659218, + "grad_norm": 0.36533117294311523, + "learning_rate": 2.7858472998137805e-05, + "loss": 0.0791, + "step": 14707 + }, + { + "epoch": 41.08379888268156, + "grad_norm": 0.38319122791290283, + "learning_rate": 2.785810055865922e-05, + "loss": 0.0818, + "step": 14708 + }, + { + "epoch": 41.08659217877095, + "grad_norm": 0.3377833664417267, + "learning_rate": 2.7857728119180635e-05, + "loss": 0.0692, + "step": 14709 + }, + { + "epoch": 41.089385474860336, + "grad_norm": 0.5383819341659546, + "learning_rate": 2.7857355679702048e-05, + "loss": 0.085, + "step": 14710 + }, + { + "epoch": 41.09217877094972, + "grad_norm": 0.4759708344936371, + "learning_rate": 2.7856983240223465e-05, + "loss": 0.0733, + "step": 14711 + }, + { + "epoch": 41.09497206703911, + "grad_norm": 0.8149548172950745, + "learning_rate": 2.7856610800744878e-05, + "loss": 0.0753, + "step": 14712 + }, + { + "epoch": 41.097765363128495, + "grad_norm": 0.41559454798698425, + "learning_rate": 2.7856238361266294e-05, + "loss": 0.0757, + "step": 14713 + }, + { + "epoch": 41.100558659217874, + "grad_norm": 1.1481742858886719, + "learning_rate": 2.785586592178771e-05, + "loss": 0.0625, + "step": 14714 + }, + { + "epoch": 41.10335195530726, + "grad_norm": 0.4176693856716156, + "learning_rate": 2.7855493482309127e-05, + "loss": 0.0639, + "step": 14715 + }, + { + "epoch": 41.10614525139665, + "grad_norm": 0.2994093894958496, + "learning_rate": 2.785512104283054e-05, + "loss": 0.0531, + "step": 14716 + }, + { + "epoch": 41.108938547486034, + "grad_norm": 0.5096707344055176, + "learning_rate": 2.7854748603351954e-05, + "loss": 0.0613, + "step": 14717 + }, + { + "epoch": 41.11173184357542, + "grad_norm": 0.5281067490577698, + "learning_rate": 2.785437616387337e-05, + "loss": 0.0507, + "step": 14718 + }, + { + "epoch": 41.114525139664806, + "grad_norm": 0.666096031665802, + "learning_rate": 2.7854003724394787e-05, + "loss": 0.0549, + "step": 14719 + }, + { + "epoch": 41.11731843575419, + "grad_norm": 0.5702342391014099, + "learning_rate": 2.7853631284916203e-05, + "loss": 0.0451, + "step": 14720 + }, + { + "epoch": 41.12011173184357, + "grad_norm": 0.7825790643692017, + "learning_rate": 2.7853258845437616e-05, + "loss": 0.0543, + "step": 14721 + }, + { + "epoch": 41.12290502793296, + "grad_norm": 0.6974553465843201, + "learning_rate": 2.7852886405959033e-05, + "loss": 0.0576, + "step": 14722 + }, + { + "epoch": 41.125698324022345, + "grad_norm": 0.5903391242027283, + "learning_rate": 2.7852513966480446e-05, + "loss": 0.0527, + "step": 14723 + }, + { + "epoch": 41.12849162011173, + "grad_norm": 0.5248437523841858, + "learning_rate": 2.7852141527001863e-05, + "loss": 0.0542, + "step": 14724 + }, + { + "epoch": 41.13128491620112, + "grad_norm": 0.8205862641334534, + "learning_rate": 2.785176908752328e-05, + "loss": 0.0544, + "step": 14725 + }, + { + "epoch": 41.134078212290504, + "grad_norm": 0.7282425761222839, + "learning_rate": 2.7851396648044692e-05, + "loss": 0.0467, + "step": 14726 + }, + { + "epoch": 41.13687150837989, + "grad_norm": 1.1748319864273071, + "learning_rate": 2.785102420856611e-05, + "loss": 0.0498, + "step": 14727 + }, + { + "epoch": 41.13966480446928, + "grad_norm": 0.9652423858642578, + "learning_rate": 2.7850651769087525e-05, + "loss": 0.0688, + "step": 14728 + }, + { + "epoch": 41.142458100558656, + "grad_norm": 0.426304429769516, + "learning_rate": 2.7850279329608942e-05, + "loss": 0.113, + "step": 14729 + }, + { + "epoch": 41.14525139664804, + "grad_norm": 0.38239341974258423, + "learning_rate": 2.7849906890130352e-05, + "loss": 0.1137, + "step": 14730 + }, + { + "epoch": 41.14804469273743, + "grad_norm": 0.48191624879837036, + "learning_rate": 2.7849534450651768e-05, + "loss": 0.099, + "step": 14731 + }, + { + "epoch": 41.150837988826815, + "grad_norm": 0.6918531656265259, + "learning_rate": 2.7849162011173185e-05, + "loss": 0.0973, + "step": 14732 + }, + { + "epoch": 41.1536312849162, + "grad_norm": 0.4797912836074829, + "learning_rate": 2.78487895716946e-05, + "loss": 0.1098, + "step": 14733 + }, + { + "epoch": 41.15642458100559, + "grad_norm": 0.5439945459365845, + "learning_rate": 2.7848417132216018e-05, + "loss": 0.081, + "step": 14734 + }, + { + "epoch": 41.159217877094974, + "grad_norm": 0.4403104782104492, + "learning_rate": 2.784804469273743e-05, + "loss": 0.0913, + "step": 14735 + }, + { + "epoch": 41.16201117318436, + "grad_norm": 0.3833204507827759, + "learning_rate": 2.7847672253258844e-05, + "loss": 0.0715, + "step": 14736 + }, + { + "epoch": 41.16480446927374, + "grad_norm": 0.6090783476829529, + "learning_rate": 2.784729981378026e-05, + "loss": 0.0723, + "step": 14737 + }, + { + "epoch": 41.167597765363126, + "grad_norm": 0.49020543694496155, + "learning_rate": 2.7846927374301677e-05, + "loss": 0.0797, + "step": 14738 + }, + { + "epoch": 41.17039106145251, + "grad_norm": 0.4901595413684845, + "learning_rate": 2.784655493482309e-05, + "loss": 0.0651, + "step": 14739 + }, + { + "epoch": 41.1731843575419, + "grad_norm": 0.4309602975845337, + "learning_rate": 2.7846182495344507e-05, + "loss": 0.0581, + "step": 14740 + }, + { + "epoch": 41.175977653631286, + "grad_norm": 0.41323333978652954, + "learning_rate": 2.7845810055865923e-05, + "loss": 0.0583, + "step": 14741 + }, + { + "epoch": 41.17877094972067, + "grad_norm": 0.7987499237060547, + "learning_rate": 2.784543761638734e-05, + "loss": 0.0655, + "step": 14742 + }, + { + "epoch": 41.18156424581006, + "grad_norm": 0.662294864654541, + "learning_rate": 2.7845065176908753e-05, + "loss": 0.0675, + "step": 14743 + }, + { + "epoch": 41.184357541899445, + "grad_norm": 0.5393799543380737, + "learning_rate": 2.7844692737430166e-05, + "loss": 0.0559, + "step": 14744 + }, + { + "epoch": 41.187150837988824, + "grad_norm": 0.46662837266921997, + "learning_rate": 2.7844320297951583e-05, + "loss": 0.0606, + "step": 14745 + }, + { + "epoch": 41.18994413407821, + "grad_norm": 0.4699210822582245, + "learning_rate": 2.7843947858473e-05, + "loss": 0.0503, + "step": 14746 + }, + { + "epoch": 41.1927374301676, + "grad_norm": 0.8150215148925781, + "learning_rate": 2.7843575418994416e-05, + "loss": 0.0577, + "step": 14747 + }, + { + "epoch": 41.19553072625698, + "grad_norm": 0.6451020836830139, + "learning_rate": 2.784320297951583e-05, + "loss": 0.0516, + "step": 14748 + }, + { + "epoch": 41.19832402234637, + "grad_norm": 0.5426979660987854, + "learning_rate": 2.7842830540037246e-05, + "loss": 0.0713, + "step": 14749 + }, + { + "epoch": 41.201117318435756, + "grad_norm": 0.65524822473526, + "learning_rate": 2.784245810055866e-05, + "loss": 0.0563, + "step": 14750 + }, + { + "epoch": 41.20391061452514, + "grad_norm": 1.6307145357131958, + "learning_rate": 2.7842085661080075e-05, + "loss": 0.0608, + "step": 14751 + }, + { + "epoch": 41.20670391061452, + "grad_norm": 1.6736966371536255, + "learning_rate": 2.7841713221601492e-05, + "loss": 0.0576, + "step": 14752 + }, + { + "epoch": 41.20949720670391, + "grad_norm": 0.8141931295394897, + "learning_rate": 2.7841340782122905e-05, + "loss": 0.0623, + "step": 14753 + }, + { + "epoch": 41.212290502793294, + "grad_norm": 0.43874529004096985, + "learning_rate": 2.784096834264432e-05, + "loss": 0.1115, + "step": 14754 + }, + { + "epoch": 41.21508379888268, + "grad_norm": 0.6305559873580933, + "learning_rate": 2.7840595903165738e-05, + "loss": 0.1075, + "step": 14755 + }, + { + "epoch": 41.21787709497207, + "grad_norm": 0.4646933078765869, + "learning_rate": 2.784022346368715e-05, + "loss": 0.1006, + "step": 14756 + }, + { + "epoch": 41.220670391061454, + "grad_norm": 0.4136560261249542, + "learning_rate": 2.7839851024208564e-05, + "loss": 0.0946, + "step": 14757 + }, + { + "epoch": 41.22346368715084, + "grad_norm": 0.4261341989040375, + "learning_rate": 2.783947858472998e-05, + "loss": 0.0905, + "step": 14758 + }, + { + "epoch": 41.226256983240226, + "grad_norm": 0.606845498085022, + "learning_rate": 2.7839106145251397e-05, + "loss": 0.0805, + "step": 14759 + }, + { + "epoch": 41.229050279329606, + "grad_norm": 0.5032708644866943, + "learning_rate": 2.7838733705772814e-05, + "loss": 0.0913, + "step": 14760 + }, + { + "epoch": 41.23184357541899, + "grad_norm": 0.3250669836997986, + "learning_rate": 2.783836126629423e-05, + "loss": 0.0754, + "step": 14761 + }, + { + "epoch": 41.23463687150838, + "grad_norm": 0.47635143995285034, + "learning_rate": 2.7837988826815644e-05, + "loss": 0.0744, + "step": 14762 + }, + { + "epoch": 41.237430167597765, + "grad_norm": 0.3945865333080292, + "learning_rate": 2.7837616387337057e-05, + "loss": 0.067, + "step": 14763 + }, + { + "epoch": 41.24022346368715, + "grad_norm": 0.4206119775772095, + "learning_rate": 2.7837243947858473e-05, + "loss": 0.0672, + "step": 14764 + }, + { + "epoch": 41.24301675977654, + "grad_norm": 0.49129223823547363, + "learning_rate": 2.783687150837989e-05, + "loss": 0.0855, + "step": 14765 + }, + { + "epoch": 41.245810055865924, + "grad_norm": 0.4776935279369354, + "learning_rate": 2.7836499068901303e-05, + "loss": 0.0651, + "step": 14766 + }, + { + "epoch": 41.24860335195531, + "grad_norm": 0.47951287031173706, + "learning_rate": 2.783612662942272e-05, + "loss": 0.0638, + "step": 14767 + }, + { + "epoch": 41.25139664804469, + "grad_norm": 0.5277390480041504, + "learning_rate": 2.7835754189944136e-05, + "loss": 0.0512, + "step": 14768 + }, + { + "epoch": 41.254189944134076, + "grad_norm": 0.38485580682754517, + "learning_rate": 2.7835381750465553e-05, + "loss": 0.0575, + "step": 14769 + }, + { + "epoch": 41.25698324022346, + "grad_norm": 1.2995383739471436, + "learning_rate": 2.7835009310986966e-05, + "loss": 0.0551, + "step": 14770 + }, + { + "epoch": 41.25977653631285, + "grad_norm": 0.46281006932258606, + "learning_rate": 2.783463687150838e-05, + "loss": 0.0561, + "step": 14771 + }, + { + "epoch": 41.262569832402235, + "grad_norm": 0.513760507106781, + "learning_rate": 2.7834264432029795e-05, + "loss": 0.051, + "step": 14772 + }, + { + "epoch": 41.26536312849162, + "grad_norm": 0.5489118099212646, + "learning_rate": 2.7833891992551212e-05, + "loss": 0.0606, + "step": 14773 + }, + { + "epoch": 41.26815642458101, + "grad_norm": 0.5417560935020447, + "learning_rate": 2.783351955307263e-05, + "loss": 0.0486, + "step": 14774 + }, + { + "epoch": 41.270949720670394, + "grad_norm": 0.7627074122428894, + "learning_rate": 2.783314711359404e-05, + "loss": 0.0491, + "step": 14775 + }, + { + "epoch": 41.273743016759774, + "grad_norm": 0.6051290035247803, + "learning_rate": 2.7832774674115455e-05, + "loss": 0.0311, + "step": 14776 + }, + { + "epoch": 41.27653631284916, + "grad_norm": 0.6617611050605774, + "learning_rate": 2.783240223463687e-05, + "loss": 0.0432, + "step": 14777 + }, + { + "epoch": 41.279329608938546, + "grad_norm": 17.764041900634766, + "learning_rate": 2.7832029795158288e-05, + "loss": 0.0604, + "step": 14778 + }, + { + "epoch": 41.28212290502793, + "grad_norm": 0.5937397480010986, + "learning_rate": 2.7831657355679704e-05, + "loss": 0.1078, + "step": 14779 + }, + { + "epoch": 41.28491620111732, + "grad_norm": 0.6563060879707336, + "learning_rate": 2.7831284916201118e-05, + "loss": 0.0946, + "step": 14780 + }, + { + "epoch": 41.287709497206706, + "grad_norm": 0.4445837438106537, + "learning_rate": 2.7830912476722534e-05, + "loss": 0.098, + "step": 14781 + }, + { + "epoch": 41.29050279329609, + "grad_norm": 0.5120139122009277, + "learning_rate": 2.783054003724395e-05, + "loss": 0.0918, + "step": 14782 + }, + { + "epoch": 41.29329608938548, + "grad_norm": 0.6017833948135376, + "learning_rate": 2.7830167597765364e-05, + "loss": 0.0891, + "step": 14783 + }, + { + "epoch": 41.29608938547486, + "grad_norm": 0.767500102519989, + "learning_rate": 2.7829795158286777e-05, + "loss": 0.0868, + "step": 14784 + }, + { + "epoch": 41.298882681564244, + "grad_norm": 0.676042914390564, + "learning_rate": 2.7829422718808193e-05, + "loss": 0.0838, + "step": 14785 + }, + { + "epoch": 41.30167597765363, + "grad_norm": 0.5611730813980103, + "learning_rate": 2.782905027932961e-05, + "loss": 0.0785, + "step": 14786 + }, + { + "epoch": 41.30446927374302, + "grad_norm": 0.7556858062744141, + "learning_rate": 2.7828677839851027e-05, + "loss": 0.0829, + "step": 14787 + }, + { + "epoch": 41.3072625698324, + "grad_norm": 0.9348226189613342, + "learning_rate": 2.782830540037244e-05, + "loss": 0.0885, + "step": 14788 + }, + { + "epoch": 41.31005586592179, + "grad_norm": 0.41349339485168457, + "learning_rate": 2.7827932960893856e-05, + "loss": 0.0741, + "step": 14789 + }, + { + "epoch": 41.312849162011176, + "grad_norm": 0.4495737552642822, + "learning_rate": 2.782756052141527e-05, + "loss": 0.0621, + "step": 14790 + }, + { + "epoch": 41.315642458100555, + "grad_norm": 0.4021644592285156, + "learning_rate": 2.7827188081936686e-05, + "loss": 0.0598, + "step": 14791 + }, + { + "epoch": 41.31843575418994, + "grad_norm": 0.5408711433410645, + "learning_rate": 2.7826815642458102e-05, + "loss": 0.0687, + "step": 14792 + }, + { + "epoch": 41.32122905027933, + "grad_norm": 0.4309297800064087, + "learning_rate": 2.7826443202979516e-05, + "loss": 0.0673, + "step": 14793 + }, + { + "epoch": 41.324022346368714, + "grad_norm": 0.6148296594619751, + "learning_rate": 2.7826070763500932e-05, + "loss": 0.06, + "step": 14794 + }, + { + "epoch": 41.3268156424581, + "grad_norm": 0.8107187747955322, + "learning_rate": 2.782569832402235e-05, + "loss": 0.0495, + "step": 14795 + }, + { + "epoch": 41.32960893854749, + "grad_norm": 0.8252013921737671, + "learning_rate": 2.7825325884543762e-05, + "loss": 0.0426, + "step": 14796 + }, + { + "epoch": 41.332402234636874, + "grad_norm": 1.2006474733352661, + "learning_rate": 2.7824953445065175e-05, + "loss": 0.0603, + "step": 14797 + }, + { + "epoch": 41.33519553072626, + "grad_norm": 0.6341233253479004, + "learning_rate": 2.782458100558659e-05, + "loss": 0.049, + "step": 14798 + }, + { + "epoch": 41.33798882681564, + "grad_norm": 0.7733345627784729, + "learning_rate": 2.7824208566108008e-05, + "loss": 0.057, + "step": 14799 + }, + { + "epoch": 41.340782122905026, + "grad_norm": 0.502625048160553, + "learning_rate": 2.7823836126629425e-05, + "loss": 0.0552, + "step": 14800 + }, + { + "epoch": 41.34357541899441, + "grad_norm": 0.6361004710197449, + "learning_rate": 2.782346368715084e-05, + "loss": 0.0574, + "step": 14801 + }, + { + "epoch": 41.3463687150838, + "grad_norm": 1.6219550371170044, + "learning_rate": 2.7823091247672254e-05, + "loss": 0.0587, + "step": 14802 + }, + { + "epoch": 41.349162011173185, + "grad_norm": 0.9112234711647034, + "learning_rate": 2.7822718808193667e-05, + "loss": 0.0701, + "step": 14803 + }, + { + "epoch": 41.35195530726257, + "grad_norm": 0.5648272633552551, + "learning_rate": 2.7822346368715084e-05, + "loss": 0.1074, + "step": 14804 + }, + { + "epoch": 41.35474860335196, + "grad_norm": 0.5708702206611633, + "learning_rate": 2.78219739292365e-05, + "loss": 0.1054, + "step": 14805 + }, + { + "epoch": 41.357541899441344, + "grad_norm": 0.534580647945404, + "learning_rate": 2.7821601489757914e-05, + "loss": 0.1089, + "step": 14806 + }, + { + "epoch": 41.36033519553072, + "grad_norm": 0.7114267945289612, + "learning_rate": 2.782122905027933e-05, + "loss": 0.0812, + "step": 14807 + }, + { + "epoch": 41.36312849162011, + "grad_norm": 3.2281599044799805, + "learning_rate": 2.7820856610800747e-05, + "loss": 0.0887, + "step": 14808 + }, + { + "epoch": 41.365921787709496, + "grad_norm": 0.6568450927734375, + "learning_rate": 2.7820484171322163e-05, + "loss": 0.0921, + "step": 14809 + }, + { + "epoch": 41.36871508379888, + "grad_norm": 0.670698344707489, + "learning_rate": 2.7820111731843576e-05, + "loss": 0.0749, + "step": 14810 + }, + { + "epoch": 41.37150837988827, + "grad_norm": 0.5498164892196655, + "learning_rate": 2.781973929236499e-05, + "loss": 0.0751, + "step": 14811 + }, + { + "epoch": 41.374301675977655, + "grad_norm": 0.9663040637969971, + "learning_rate": 2.7819366852886406e-05, + "loss": 0.0692, + "step": 14812 + }, + { + "epoch": 41.37709497206704, + "grad_norm": 0.5570074319839478, + "learning_rate": 2.7818994413407823e-05, + "loss": 0.0634, + "step": 14813 + }, + { + "epoch": 41.37988826815643, + "grad_norm": 0.47200533747673035, + "learning_rate": 2.781862197392924e-05, + "loss": 0.0631, + "step": 14814 + }, + { + "epoch": 41.38268156424581, + "grad_norm": 0.373282253742218, + "learning_rate": 2.7818249534450652e-05, + "loss": 0.0688, + "step": 14815 + }, + { + "epoch": 41.385474860335194, + "grad_norm": 0.5124996304512024, + "learning_rate": 2.7817877094972065e-05, + "loss": 0.0659, + "step": 14816 + }, + { + "epoch": 41.38826815642458, + "grad_norm": 0.5339503288269043, + "learning_rate": 2.7817504655493482e-05, + "loss": 0.0911, + "step": 14817 + }, + { + "epoch": 41.391061452513966, + "grad_norm": 0.6565989255905151, + "learning_rate": 2.78171322160149e-05, + "loss": 0.0705, + "step": 14818 + }, + { + "epoch": 41.39385474860335, + "grad_norm": 1.3767865896224976, + "learning_rate": 2.7816759776536315e-05, + "loss": 0.0648, + "step": 14819 + }, + { + "epoch": 41.39664804469274, + "grad_norm": 0.5889513492584229, + "learning_rate": 2.7816387337057728e-05, + "loss": 0.0728, + "step": 14820 + }, + { + "epoch": 41.399441340782126, + "grad_norm": 0.7196378707885742, + "learning_rate": 2.7816014897579145e-05, + "loss": 0.0701, + "step": 14821 + }, + { + "epoch": 41.40223463687151, + "grad_norm": 0.7367963194847107, + "learning_rate": 2.781564245810056e-05, + "loss": 0.0539, + "step": 14822 + }, + { + "epoch": 41.40502793296089, + "grad_norm": 1.181085467338562, + "learning_rate": 2.7815270018621974e-05, + "loss": 0.0598, + "step": 14823 + }, + { + "epoch": 41.40782122905028, + "grad_norm": 0.6476032137870789, + "learning_rate": 2.7814897579143388e-05, + "loss": 0.0449, + "step": 14824 + }, + { + "epoch": 41.410614525139664, + "grad_norm": 1.217656135559082, + "learning_rate": 2.7814525139664804e-05, + "loss": 0.0573, + "step": 14825 + }, + { + "epoch": 41.41340782122905, + "grad_norm": 1.12455153465271, + "learning_rate": 2.781415270018622e-05, + "loss": 0.0625, + "step": 14826 + }, + { + "epoch": 41.41620111731844, + "grad_norm": 1.2016258239746094, + "learning_rate": 2.7813780260707637e-05, + "loss": 0.0712, + "step": 14827 + }, + { + "epoch": 41.41899441340782, + "grad_norm": 2.489788770675659, + "learning_rate": 2.7813407821229054e-05, + "loss": 0.0981, + "step": 14828 + }, + { + "epoch": 41.42178770949721, + "grad_norm": 0.45046940445899963, + "learning_rate": 2.7813035381750467e-05, + "loss": 0.0969, + "step": 14829 + }, + { + "epoch": 41.42458100558659, + "grad_norm": 0.43510663509368896, + "learning_rate": 2.781266294227188e-05, + "loss": 0.1228, + "step": 14830 + }, + { + "epoch": 41.427374301675975, + "grad_norm": 0.46358102560043335, + "learning_rate": 2.7812290502793296e-05, + "loss": 0.1066, + "step": 14831 + }, + { + "epoch": 41.43016759776536, + "grad_norm": 0.32895955443382263, + "learning_rate": 2.7811918063314713e-05, + "loss": 0.0881, + "step": 14832 + }, + { + "epoch": 41.43296089385475, + "grad_norm": 0.44561266899108887, + "learning_rate": 2.7811545623836126e-05, + "loss": 0.0996, + "step": 14833 + }, + { + "epoch": 41.435754189944134, + "grad_norm": 0.40712451934814453, + "learning_rate": 2.7811173184357543e-05, + "loss": 0.0719, + "step": 14834 + }, + { + "epoch": 41.43854748603352, + "grad_norm": 0.5657070875167847, + "learning_rate": 2.781080074487896e-05, + "loss": 0.0947, + "step": 14835 + }, + { + "epoch": 41.44134078212291, + "grad_norm": 0.4253842830657959, + "learning_rate": 2.7810428305400372e-05, + "loss": 0.0916, + "step": 14836 + }, + { + "epoch": 41.444134078212294, + "grad_norm": 0.4819422960281372, + "learning_rate": 2.781005586592179e-05, + "loss": 0.0757, + "step": 14837 + }, + { + "epoch": 41.44692737430167, + "grad_norm": 0.41262009739875793, + "learning_rate": 2.7809683426443202e-05, + "loss": 0.0773, + "step": 14838 + }, + { + "epoch": 41.44972067039106, + "grad_norm": 0.5748059153556824, + "learning_rate": 2.780931098696462e-05, + "loss": 0.0904, + "step": 14839 + }, + { + "epoch": 41.452513966480446, + "grad_norm": 0.49391263723373413, + "learning_rate": 2.7808938547486035e-05, + "loss": 0.0837, + "step": 14840 + }, + { + "epoch": 41.45530726256983, + "grad_norm": 0.8098183274269104, + "learning_rate": 2.780856610800745e-05, + "loss": 0.0783, + "step": 14841 + }, + { + "epoch": 41.45810055865922, + "grad_norm": 0.5164819359779358, + "learning_rate": 2.7808193668528865e-05, + "loss": 0.0655, + "step": 14842 + }, + { + "epoch": 41.460893854748605, + "grad_norm": 0.5091065168380737, + "learning_rate": 2.7807821229050278e-05, + "loss": 0.0486, + "step": 14843 + }, + { + "epoch": 41.46368715083799, + "grad_norm": 0.7388013601303101, + "learning_rate": 2.7807448789571695e-05, + "loss": 0.0584, + "step": 14844 + }, + { + "epoch": 41.46648044692738, + "grad_norm": 0.740967869758606, + "learning_rate": 2.780707635009311e-05, + "loss": 0.0557, + "step": 14845 + }, + { + "epoch": 41.46927374301676, + "grad_norm": 0.8581821918487549, + "learning_rate": 2.7806703910614528e-05, + "loss": 0.0628, + "step": 14846 + }, + { + "epoch": 41.47206703910614, + "grad_norm": 0.6697985529899597, + "learning_rate": 2.780633147113594e-05, + "loss": 0.0605, + "step": 14847 + }, + { + "epoch": 41.47486033519553, + "grad_norm": 0.6577515006065369, + "learning_rate": 2.7805959031657357e-05, + "loss": 0.0725, + "step": 14848 + }, + { + "epoch": 41.477653631284916, + "grad_norm": 1.3916298151016235, + "learning_rate": 2.7805586592178774e-05, + "loss": 0.0606, + "step": 14849 + }, + { + "epoch": 41.4804469273743, + "grad_norm": 1.206573724746704, + "learning_rate": 2.7805214152700187e-05, + "loss": 0.0622, + "step": 14850 + }, + { + "epoch": 41.48324022346369, + "grad_norm": 1.037818431854248, + "learning_rate": 2.78048417132216e-05, + "loss": 0.0629, + "step": 14851 + }, + { + "epoch": 41.486033519553075, + "grad_norm": 1.7236969470977783, + "learning_rate": 2.7804469273743017e-05, + "loss": 0.0515, + "step": 14852 + }, + { + "epoch": 41.48882681564246, + "grad_norm": 1.9304693937301636, + "learning_rate": 2.7804096834264433e-05, + "loss": 0.0646, + "step": 14853 + }, + { + "epoch": 41.49162011173184, + "grad_norm": 0.4429226517677307, + "learning_rate": 2.780372439478585e-05, + "loss": 0.1187, + "step": 14854 + }, + { + "epoch": 41.49441340782123, + "grad_norm": 0.4544937312602997, + "learning_rate": 2.7803351955307266e-05, + "loss": 0.1166, + "step": 14855 + }, + { + "epoch": 41.497206703910614, + "grad_norm": 0.6274206638336182, + "learning_rate": 2.7802979515828676e-05, + "loss": 0.0952, + "step": 14856 + }, + { + "epoch": 41.5, + "grad_norm": 1.326375961303711, + "learning_rate": 2.7802607076350093e-05, + "loss": 0.0816, + "step": 14857 + }, + { + "epoch": 41.502793296089386, + "grad_norm": 2.798722505569458, + "learning_rate": 2.780223463687151e-05, + "loss": 0.084, + "step": 14858 + }, + { + "epoch": 41.50558659217877, + "grad_norm": 0.43449631333351135, + "learning_rate": 2.7801862197392926e-05, + "loss": 0.0975, + "step": 14859 + }, + { + "epoch": 41.50837988826816, + "grad_norm": 0.45040830969810486, + "learning_rate": 2.780148975791434e-05, + "loss": 0.0959, + "step": 14860 + }, + { + "epoch": 41.51117318435754, + "grad_norm": 0.4898139536380768, + "learning_rate": 2.7801117318435755e-05, + "loss": 0.0901, + "step": 14861 + }, + { + "epoch": 41.513966480446925, + "grad_norm": 0.3513268828392029, + "learning_rate": 2.7800744878957172e-05, + "loss": 0.0625, + "step": 14862 + }, + { + "epoch": 41.51675977653631, + "grad_norm": 0.5029136538505554, + "learning_rate": 2.7800372439478585e-05, + "loss": 0.0867, + "step": 14863 + }, + { + "epoch": 41.5195530726257, + "grad_norm": 1.39842689037323, + "learning_rate": 2.78e-05, + "loss": 0.0974, + "step": 14864 + }, + { + "epoch": 41.522346368715084, + "grad_norm": 0.43698546290397644, + "learning_rate": 2.7799627560521415e-05, + "loss": 0.0692, + "step": 14865 + }, + { + "epoch": 41.52513966480447, + "grad_norm": 0.5350784659385681, + "learning_rate": 2.779925512104283e-05, + "loss": 0.0635, + "step": 14866 + }, + { + "epoch": 41.52793296089386, + "grad_norm": 0.5769796967506409, + "learning_rate": 2.7798882681564248e-05, + "loss": 0.0696, + "step": 14867 + }, + { + "epoch": 41.53072625698324, + "grad_norm": 0.4338921308517456, + "learning_rate": 2.7798510242085664e-05, + "loss": 0.0574, + "step": 14868 + }, + { + "epoch": 41.53351955307262, + "grad_norm": 0.5946993231773376, + "learning_rate": 2.7798137802607077e-05, + "loss": 0.0639, + "step": 14869 + }, + { + "epoch": 41.53631284916201, + "grad_norm": 0.45603877305984497, + "learning_rate": 2.779776536312849e-05, + "loss": 0.0486, + "step": 14870 + }, + { + "epoch": 41.539106145251395, + "grad_norm": 0.9038728475570679, + "learning_rate": 2.7797392923649907e-05, + "loss": 0.0651, + "step": 14871 + }, + { + "epoch": 41.54189944134078, + "grad_norm": 0.8404735922813416, + "learning_rate": 2.7797020484171324e-05, + "loss": 0.0551, + "step": 14872 + }, + { + "epoch": 41.54469273743017, + "grad_norm": 1.026926040649414, + "learning_rate": 2.779664804469274e-05, + "loss": 0.0555, + "step": 14873 + }, + { + "epoch": 41.547486033519554, + "grad_norm": 0.63999342918396, + "learning_rate": 2.7796275605214153e-05, + "loss": 0.0464, + "step": 14874 + }, + { + "epoch": 41.55027932960894, + "grad_norm": 0.5580688714981079, + "learning_rate": 2.779590316573557e-05, + "loss": 0.0572, + "step": 14875 + }, + { + "epoch": 41.55307262569833, + "grad_norm": 1.0276579856872559, + "learning_rate": 2.7795530726256983e-05, + "loss": 0.0554, + "step": 14876 + }, + { + "epoch": 41.555865921787706, + "grad_norm": 1.3307157754898071, + "learning_rate": 2.77951582867784e-05, + "loss": 0.0497, + "step": 14877 + }, + { + "epoch": 41.55865921787709, + "grad_norm": 1.9916534423828125, + "learning_rate": 2.7794785847299813e-05, + "loss": 0.0936, + "step": 14878 + }, + { + "epoch": 41.56145251396648, + "grad_norm": 0.42835739254951477, + "learning_rate": 2.779441340782123e-05, + "loss": 0.104, + "step": 14879 + }, + { + "epoch": 41.564245810055866, + "grad_norm": 0.5651381611824036, + "learning_rate": 2.7794040968342646e-05, + "loss": 0.0984, + "step": 14880 + }, + { + "epoch": 41.56703910614525, + "grad_norm": 0.7299178242683411, + "learning_rate": 2.7793668528864062e-05, + "loss": 0.0886, + "step": 14881 + }, + { + "epoch": 41.56983240223464, + "grad_norm": 0.4124048352241516, + "learning_rate": 2.7793296089385475e-05, + "loss": 0.0876, + "step": 14882 + }, + { + "epoch": 41.572625698324025, + "grad_norm": 0.7444042563438416, + "learning_rate": 2.779292364990689e-05, + "loss": 0.0938, + "step": 14883 + }, + { + "epoch": 41.57541899441341, + "grad_norm": 0.585882306098938, + "learning_rate": 2.7792551210428305e-05, + "loss": 0.0772, + "step": 14884 + }, + { + "epoch": 41.57821229050279, + "grad_norm": 0.5560806393623352, + "learning_rate": 2.779217877094972e-05, + "loss": 0.0995, + "step": 14885 + }, + { + "epoch": 41.58100558659218, + "grad_norm": 0.4742416441440582, + "learning_rate": 2.7791806331471138e-05, + "loss": 0.076, + "step": 14886 + }, + { + "epoch": 41.58379888268156, + "grad_norm": 0.38448366522789, + "learning_rate": 2.779143389199255e-05, + "loss": 0.0805, + "step": 14887 + }, + { + "epoch": 41.58659217877095, + "grad_norm": 0.4591265618801117, + "learning_rate": 2.7791061452513968e-05, + "loss": 0.0719, + "step": 14888 + }, + { + "epoch": 41.589385474860336, + "grad_norm": 0.6505791544914246, + "learning_rate": 2.7790689013035384e-05, + "loss": 0.0973, + "step": 14889 + }, + { + "epoch": 41.59217877094972, + "grad_norm": 0.5147098898887634, + "learning_rate": 2.7790316573556798e-05, + "loss": 0.0627, + "step": 14890 + }, + { + "epoch": 41.59497206703911, + "grad_norm": 0.4578930139541626, + "learning_rate": 2.778994413407821e-05, + "loss": 0.0657, + "step": 14891 + }, + { + "epoch": 41.59776536312849, + "grad_norm": 1.2492740154266357, + "learning_rate": 2.7789571694599627e-05, + "loss": 0.06, + "step": 14892 + }, + { + "epoch": 41.600558659217874, + "grad_norm": 0.8781992197036743, + "learning_rate": 2.7789199255121044e-05, + "loss": 0.0664, + "step": 14893 + }, + { + "epoch": 41.60335195530726, + "grad_norm": 0.39853760600090027, + "learning_rate": 2.778882681564246e-05, + "loss": 0.0638, + "step": 14894 + }, + { + "epoch": 41.60614525139665, + "grad_norm": 0.50192791223526, + "learning_rate": 2.7788454376163877e-05, + "loss": 0.0609, + "step": 14895 + }, + { + "epoch": 41.608938547486034, + "grad_norm": 0.5180484056472778, + "learning_rate": 2.7788081936685287e-05, + "loss": 0.0562, + "step": 14896 + }, + { + "epoch": 41.61173184357542, + "grad_norm": 0.6195234060287476, + "learning_rate": 2.7787709497206703e-05, + "loss": 0.0602, + "step": 14897 + }, + { + "epoch": 41.614525139664806, + "grad_norm": 0.7008521556854248, + "learning_rate": 2.778733705772812e-05, + "loss": 0.0479, + "step": 14898 + }, + { + "epoch": 41.61731843575419, + "grad_norm": 0.7038863897323608, + "learning_rate": 2.7786964618249536e-05, + "loss": 0.0495, + "step": 14899 + }, + { + "epoch": 41.62011173184357, + "grad_norm": 0.6439720988273621, + "learning_rate": 2.778659217877095e-05, + "loss": 0.0594, + "step": 14900 + }, + { + "epoch": 41.62290502793296, + "grad_norm": 2.1874327659606934, + "learning_rate": 2.7786219739292366e-05, + "loss": 0.0617, + "step": 14901 + }, + { + "epoch": 41.625698324022345, + "grad_norm": 1.0075485706329346, + "learning_rate": 2.7785847299813782e-05, + "loss": 0.0519, + "step": 14902 + }, + { + "epoch": 41.62849162011173, + "grad_norm": 1.585569143295288, + "learning_rate": 2.7785474860335196e-05, + "loss": 0.0748, + "step": 14903 + }, + { + "epoch": 41.63128491620112, + "grad_norm": 0.4286902844905853, + "learning_rate": 2.7785102420856612e-05, + "loss": 0.1217, + "step": 14904 + }, + { + "epoch": 41.634078212290504, + "grad_norm": 0.921953558921814, + "learning_rate": 2.7784729981378025e-05, + "loss": 0.1197, + "step": 14905 + }, + { + "epoch": 41.63687150837989, + "grad_norm": 0.896013617515564, + "learning_rate": 2.7784357541899442e-05, + "loss": 0.1022, + "step": 14906 + }, + { + "epoch": 41.63966480446928, + "grad_norm": 0.8218348026275635, + "learning_rate": 2.778398510242086e-05, + "loss": 0.0882, + "step": 14907 + }, + { + "epoch": 41.642458100558656, + "grad_norm": 0.34573882818222046, + "learning_rate": 2.7783612662942275e-05, + "loss": 0.0751, + "step": 14908 + }, + { + "epoch": 41.64525139664804, + "grad_norm": 0.4505094885826111, + "learning_rate": 2.7783240223463685e-05, + "loss": 0.0844, + "step": 14909 + }, + { + "epoch": 41.64804469273743, + "grad_norm": 0.5797669887542725, + "learning_rate": 2.77828677839851e-05, + "loss": 0.0974, + "step": 14910 + }, + { + "epoch": 41.650837988826815, + "grad_norm": 0.6740428805351257, + "learning_rate": 2.7782495344506518e-05, + "loss": 0.0799, + "step": 14911 + }, + { + "epoch": 41.6536312849162, + "grad_norm": 0.5576908588409424, + "learning_rate": 2.7782122905027934e-05, + "loss": 0.0757, + "step": 14912 + }, + { + "epoch": 41.65642458100559, + "grad_norm": 0.5252158045768738, + "learning_rate": 2.778175046554935e-05, + "loss": 0.0894, + "step": 14913 + }, + { + "epoch": 41.659217877094974, + "grad_norm": 0.4640348255634308, + "learning_rate": 2.7781378026070764e-05, + "loss": 0.0823, + "step": 14914 + }, + { + "epoch": 41.66201117318436, + "grad_norm": 0.5629348158836365, + "learning_rate": 2.778100558659218e-05, + "loss": 0.0887, + "step": 14915 + }, + { + "epoch": 41.66480446927374, + "grad_norm": 0.5911336541175842, + "learning_rate": 2.7780633147113594e-05, + "loss": 0.0654, + "step": 14916 + }, + { + "epoch": 41.667597765363126, + "grad_norm": 0.49098217487335205, + "learning_rate": 2.778026070763501e-05, + "loss": 0.0767, + "step": 14917 + }, + { + "epoch": 41.67039106145251, + "grad_norm": 1.0355116128921509, + "learning_rate": 2.7779888268156423e-05, + "loss": 0.0549, + "step": 14918 + }, + { + "epoch": 41.6731843575419, + "grad_norm": 0.4135434925556183, + "learning_rate": 2.777951582867784e-05, + "loss": 0.0657, + "step": 14919 + }, + { + "epoch": 41.675977653631286, + "grad_norm": 0.512968122959137, + "learning_rate": 2.7779143389199256e-05, + "loss": 0.0602, + "step": 14920 + }, + { + "epoch": 41.67877094972067, + "grad_norm": 0.9360942244529724, + "learning_rate": 2.7778770949720673e-05, + "loss": 0.0655, + "step": 14921 + }, + { + "epoch": 41.68156424581006, + "grad_norm": 1.2096656560897827, + "learning_rate": 2.777839851024209e-05, + "loss": 0.0515, + "step": 14922 + }, + { + "epoch": 41.684357541899445, + "grad_norm": 0.5980818271636963, + "learning_rate": 2.77780260707635e-05, + "loss": 0.064, + "step": 14923 + }, + { + "epoch": 41.687150837988824, + "grad_norm": 0.8037233352661133, + "learning_rate": 2.7777653631284916e-05, + "loss": 0.0575, + "step": 14924 + }, + { + "epoch": 41.68994413407821, + "grad_norm": 0.7582312226295471, + "learning_rate": 2.7777281191806332e-05, + "loss": 0.054, + "step": 14925 + }, + { + "epoch": 41.6927374301676, + "grad_norm": 1.0473291873931885, + "learning_rate": 2.777690875232775e-05, + "loss": 0.0623, + "step": 14926 + }, + { + "epoch": 41.69553072625698, + "grad_norm": 0.8502528071403503, + "learning_rate": 2.7776536312849162e-05, + "loss": 0.0643, + "step": 14927 + }, + { + "epoch": 41.69832402234637, + "grad_norm": 1.3158092498779297, + "learning_rate": 2.777616387337058e-05, + "loss": 0.0727, + "step": 14928 + }, + { + "epoch": 41.701117318435756, + "grad_norm": 0.4083884060382843, + "learning_rate": 2.777579143389199e-05, + "loss": 0.1053, + "step": 14929 + }, + { + "epoch": 41.70391061452514, + "grad_norm": 0.9422934651374817, + "learning_rate": 2.7775418994413408e-05, + "loss": 0.1117, + "step": 14930 + }, + { + "epoch": 41.70670391061452, + "grad_norm": 0.6116060614585876, + "learning_rate": 2.7775046554934825e-05, + "loss": 0.0788, + "step": 14931 + }, + { + "epoch": 41.70949720670391, + "grad_norm": 0.8707770109176636, + "learning_rate": 2.7774674115456238e-05, + "loss": 0.1155, + "step": 14932 + }, + { + "epoch": 41.712290502793294, + "grad_norm": 0.4858730137348175, + "learning_rate": 2.7774301675977654e-05, + "loss": 0.0845, + "step": 14933 + }, + { + "epoch": 41.71508379888268, + "grad_norm": 0.4707864224910736, + "learning_rate": 2.777392923649907e-05, + "loss": 0.0756, + "step": 14934 + }, + { + "epoch": 41.71787709497207, + "grad_norm": 1.030653953552246, + "learning_rate": 2.7773556797020487e-05, + "loss": 0.0825, + "step": 14935 + }, + { + "epoch": 41.720670391061454, + "grad_norm": 0.4262174665927887, + "learning_rate": 2.7773184357541897e-05, + "loss": 0.0935, + "step": 14936 + }, + { + "epoch": 41.72346368715084, + "grad_norm": 0.561479926109314, + "learning_rate": 2.7772811918063314e-05, + "loss": 0.0901, + "step": 14937 + }, + { + "epoch": 41.726256983240226, + "grad_norm": 0.46449539065361023, + "learning_rate": 2.777243947858473e-05, + "loss": 0.0991, + "step": 14938 + }, + { + "epoch": 41.729050279329606, + "grad_norm": 0.6700923442840576, + "learning_rate": 2.7772067039106147e-05, + "loss": 0.0619, + "step": 14939 + }, + { + "epoch": 41.73184357541899, + "grad_norm": 0.5495648980140686, + "learning_rate": 2.7771694599627563e-05, + "loss": 0.076, + "step": 14940 + }, + { + "epoch": 41.73463687150838, + "grad_norm": 0.5077648758888245, + "learning_rate": 2.7771322160148976e-05, + "loss": 0.0659, + "step": 14941 + }, + { + "epoch": 41.737430167597765, + "grad_norm": 0.5018128752708435, + "learning_rate": 2.7770949720670393e-05, + "loss": 0.0802, + "step": 14942 + }, + { + "epoch": 41.74022346368715, + "grad_norm": 0.58824622631073, + "learning_rate": 2.7770577281191806e-05, + "loss": 0.0853, + "step": 14943 + }, + { + "epoch": 41.74301675977654, + "grad_norm": 0.846481204032898, + "learning_rate": 2.7770204841713223e-05, + "loss": 0.0763, + "step": 14944 + }, + { + "epoch": 41.745810055865924, + "grad_norm": 0.5854530334472656, + "learning_rate": 2.7769832402234636e-05, + "loss": 0.0638, + "step": 14945 + }, + { + "epoch": 41.74860335195531, + "grad_norm": 0.594271719455719, + "learning_rate": 2.7769459962756052e-05, + "loss": 0.0714, + "step": 14946 + }, + { + "epoch": 41.75139664804469, + "grad_norm": 0.5914281606674194, + "learning_rate": 2.776908752327747e-05, + "loss": 0.0687, + "step": 14947 + }, + { + "epoch": 41.754189944134076, + "grad_norm": 0.5023000836372375, + "learning_rate": 2.7768715083798885e-05, + "loss": 0.0702, + "step": 14948 + }, + { + "epoch": 41.75698324022346, + "grad_norm": 0.6542255878448486, + "learning_rate": 2.77683426443203e-05, + "loss": 0.0472, + "step": 14949 + }, + { + "epoch": 41.75977653631285, + "grad_norm": 0.6411840319633484, + "learning_rate": 2.7767970204841712e-05, + "loss": 0.0658, + "step": 14950 + }, + { + "epoch": 41.762569832402235, + "grad_norm": 0.8899617195129395, + "learning_rate": 2.7767597765363128e-05, + "loss": 0.0483, + "step": 14951 + }, + { + "epoch": 41.76536312849162, + "grad_norm": 0.8953210115432739, + "learning_rate": 2.7767225325884545e-05, + "loss": 0.0631, + "step": 14952 + }, + { + "epoch": 41.76815642458101, + "grad_norm": 1.4072473049163818, + "learning_rate": 2.776685288640596e-05, + "loss": 0.0754, + "step": 14953 + }, + { + "epoch": 41.770949720670394, + "grad_norm": 0.8885411620140076, + "learning_rate": 2.7766480446927374e-05, + "loss": 0.108, + "step": 14954 + }, + { + "epoch": 41.773743016759774, + "grad_norm": 0.5091105699539185, + "learning_rate": 2.776610800744879e-05, + "loss": 0.1118, + "step": 14955 + }, + { + "epoch": 41.77653631284916, + "grad_norm": 0.4205143451690674, + "learning_rate": 2.7765735567970204e-05, + "loss": 0.0912, + "step": 14956 + }, + { + "epoch": 41.779329608938546, + "grad_norm": 0.3487218916416168, + "learning_rate": 2.776536312849162e-05, + "loss": 0.0888, + "step": 14957 + }, + { + "epoch": 41.78212290502793, + "grad_norm": 0.5427373647689819, + "learning_rate": 2.7764990689013037e-05, + "loss": 0.1258, + "step": 14958 + }, + { + "epoch": 41.78491620111732, + "grad_norm": 0.44418948888778687, + "learning_rate": 2.776461824953445e-05, + "loss": 0.0897, + "step": 14959 + }, + { + "epoch": 41.787709497206706, + "grad_norm": 0.5765896439552307, + "learning_rate": 2.7764245810055867e-05, + "loss": 0.0836, + "step": 14960 + }, + { + "epoch": 41.79050279329609, + "grad_norm": 0.5737472176551819, + "learning_rate": 2.7763873370577283e-05, + "loss": 0.0842, + "step": 14961 + }, + { + "epoch": 41.79329608938548, + "grad_norm": 0.6971997618675232, + "learning_rate": 2.77635009310987e-05, + "loss": 0.0928, + "step": 14962 + }, + { + "epoch": 41.79608938547486, + "grad_norm": 0.46742939949035645, + "learning_rate": 2.776312849162011e-05, + "loss": 0.0819, + "step": 14963 + }, + { + "epoch": 41.798882681564244, + "grad_norm": 0.4254261255264282, + "learning_rate": 2.7762756052141526e-05, + "loss": 0.0802, + "step": 14964 + }, + { + "epoch": 41.80167597765363, + "grad_norm": 0.6402746438980103, + "learning_rate": 2.7762383612662943e-05, + "loss": 0.0717, + "step": 14965 + }, + { + "epoch": 41.80446927374302, + "grad_norm": 0.45485788583755493, + "learning_rate": 2.776201117318436e-05, + "loss": 0.0758, + "step": 14966 + }, + { + "epoch": 41.8072625698324, + "grad_norm": 0.5302246809005737, + "learning_rate": 2.7761638733705776e-05, + "loss": 0.084, + "step": 14967 + }, + { + "epoch": 41.81005586592179, + "grad_norm": 0.6025182604789734, + "learning_rate": 2.776126629422719e-05, + "loss": 0.0726, + "step": 14968 + }, + { + "epoch": 41.812849162011176, + "grad_norm": 0.49360501766204834, + "learning_rate": 2.7760893854748602e-05, + "loss": 0.0507, + "step": 14969 + }, + { + "epoch": 41.815642458100555, + "grad_norm": 0.48614275455474854, + "learning_rate": 2.776052141527002e-05, + "loss": 0.0756, + "step": 14970 + }, + { + "epoch": 41.81843575418994, + "grad_norm": 0.6221330165863037, + "learning_rate": 2.7760148975791435e-05, + "loss": 0.069, + "step": 14971 + }, + { + "epoch": 41.82122905027933, + "grad_norm": 0.46178507804870605, + "learning_rate": 2.775977653631285e-05, + "loss": 0.0559, + "step": 14972 + }, + { + "epoch": 41.824022346368714, + "grad_norm": 0.5102673768997192, + "learning_rate": 2.7759404096834265e-05, + "loss": 0.062, + "step": 14973 + }, + { + "epoch": 41.8268156424581, + "grad_norm": 2.0573298931121826, + "learning_rate": 2.775903165735568e-05, + "loss": 0.0605, + "step": 14974 + }, + { + "epoch": 41.82960893854749, + "grad_norm": 1.4258012771606445, + "learning_rate": 2.7758659217877098e-05, + "loss": 0.0564, + "step": 14975 + }, + { + "epoch": 41.832402234636874, + "grad_norm": 0.6132301688194275, + "learning_rate": 2.775828677839851e-05, + "loss": 0.0617, + "step": 14976 + }, + { + "epoch": 41.83519553072626, + "grad_norm": 0.8928762078285217, + "learning_rate": 2.7757914338919924e-05, + "loss": 0.0537, + "step": 14977 + }, + { + "epoch": 41.83798882681564, + "grad_norm": 1.0122393369674683, + "learning_rate": 2.775754189944134e-05, + "loss": 0.0683, + "step": 14978 + }, + { + "epoch": 41.840782122905026, + "grad_norm": 0.462863564491272, + "learning_rate": 2.7757169459962757e-05, + "loss": 0.1168, + "step": 14979 + }, + { + "epoch": 41.84357541899441, + "grad_norm": 0.4146672487258911, + "learning_rate": 2.7756797020484174e-05, + "loss": 0.1099, + "step": 14980 + }, + { + "epoch": 41.8463687150838, + "grad_norm": 0.5577225685119629, + "learning_rate": 2.7756424581005587e-05, + "loss": 0.0986, + "step": 14981 + }, + { + "epoch": 41.849162011173185, + "grad_norm": 1.3911149501800537, + "learning_rate": 2.7756052141527004e-05, + "loss": 0.1019, + "step": 14982 + }, + { + "epoch": 41.85195530726257, + "grad_norm": 0.3628711402416229, + "learning_rate": 2.7755679702048417e-05, + "loss": 0.0995, + "step": 14983 + }, + { + "epoch": 41.85474860335196, + "grad_norm": 0.5253500938415527, + "learning_rate": 2.7755307262569833e-05, + "loss": 0.1018, + "step": 14984 + }, + { + "epoch": 41.857541899441344, + "grad_norm": 0.6663055419921875, + "learning_rate": 2.7754934823091246e-05, + "loss": 0.0891, + "step": 14985 + }, + { + "epoch": 41.86033519553072, + "grad_norm": 0.4310595393180847, + "learning_rate": 2.7754562383612663e-05, + "loss": 0.0844, + "step": 14986 + }, + { + "epoch": 41.86312849162011, + "grad_norm": 0.5669034123420715, + "learning_rate": 2.775418994413408e-05, + "loss": 0.0866, + "step": 14987 + }, + { + "epoch": 41.865921787709496, + "grad_norm": 0.470303475856781, + "learning_rate": 2.7753817504655496e-05, + "loss": 0.0993, + "step": 14988 + }, + { + "epoch": 41.86871508379888, + "grad_norm": 0.4092751145362854, + "learning_rate": 2.775344506517691e-05, + "loss": 0.0896, + "step": 14989 + }, + { + "epoch": 41.87150837988827, + "grad_norm": 0.5512726902961731, + "learning_rate": 2.7753072625698322e-05, + "loss": 0.0929, + "step": 14990 + }, + { + "epoch": 41.874301675977655, + "grad_norm": 0.462289422750473, + "learning_rate": 2.775270018621974e-05, + "loss": 0.066, + "step": 14991 + }, + { + "epoch": 41.87709497206704, + "grad_norm": 0.5267193913459778, + "learning_rate": 2.7752327746741155e-05, + "loss": 0.0633, + "step": 14992 + }, + { + "epoch": 41.87988826815643, + "grad_norm": 0.43307554721832275, + "learning_rate": 2.7751955307262572e-05, + "loss": 0.0457, + "step": 14993 + }, + { + "epoch": 41.88268156424581, + "grad_norm": 2.2835745811462402, + "learning_rate": 2.7751582867783985e-05, + "loss": 0.0529, + "step": 14994 + }, + { + "epoch": 41.885474860335194, + "grad_norm": 1.1386154890060425, + "learning_rate": 2.77512104283054e-05, + "loss": 0.0701, + "step": 14995 + }, + { + "epoch": 41.88826815642458, + "grad_norm": 0.6758648753166199, + "learning_rate": 2.7750837988826815e-05, + "loss": 0.0554, + "step": 14996 + }, + { + "epoch": 41.891061452513966, + "grad_norm": 3.4787709712982178, + "learning_rate": 2.775046554934823e-05, + "loss": 0.0591, + "step": 14997 + }, + { + "epoch": 41.89385474860335, + "grad_norm": 0.49814748764038086, + "learning_rate": 2.7750093109869648e-05, + "loss": 0.0546, + "step": 14998 + }, + { + "epoch": 41.89664804469274, + "grad_norm": 0.8538738489151001, + "learning_rate": 2.774972067039106e-05, + "loss": 0.0582, + "step": 14999 + }, + { + "epoch": 41.899441340782126, + "grad_norm": 0.7768288254737854, + "learning_rate": 2.7749348230912478e-05, + "loss": 0.0564, + "step": 15000 + }, + { + "epoch": 41.899441340782126, + "eval_cer": 0.19385617492062454, + "eval_loss": 0.3839842975139618, + "eval_runtime": 24.1013, + "eval_samples_per_second": 188.289, + "eval_steps_per_second": 0.622, + "eval_wer": 0.35409625429457275, + "step": 15000 + }, + { + "epoch": 41.899441340782126, + "step": 15000, + "total_flos": 4.631234010246996e+20, + "train_loss": 0.3056531836404155, + "train_runtime": 30334.2374, + "train_samples_per_second": 944.049, + "train_steps_per_second": 2.95 + } + ], + "logging_steps": 1.0, + "max_steps": 89500, + "num_input_tokens_seen": 0, + "num_train_epochs": 250, + "save_steps": 1000, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 5, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 5 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.631234010246996e+20, + "train_batch_size": 160, + "trial_name": null, + "trial_params": null +}