diff --git "a/checkpoint-700/trainer_state.json" "b/checkpoint-700/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-700/trainer_state.json" @@ -0,0 +1,4934 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.2950819672131146, + "eval_steps": 500, + "global_step": 700, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.003278688524590164, + "grad_norm": 58.999969482421875, + "learning_rate": 0.0, + "loss": 3.8125, + "step": 1 + }, + { + "epoch": 0.006557377049180328, + "grad_norm": 56.5999870300293, + "learning_rate": 3.5714285714285716e-07, + "loss": 3.8984, + "step": 2 + }, + { + "epoch": 0.009836065573770493, + "grad_norm": 58.442317962646484, + "learning_rate": 7.142857142857143e-07, + "loss": 3.8359, + "step": 3 + }, + { + "epoch": 0.013114754098360656, + "grad_norm": 59.13441467285156, + "learning_rate": 1.0714285714285714e-06, + "loss": 3.7422, + "step": 4 + }, + { + "epoch": 0.01639344262295082, + "grad_norm": 59.55280685424805, + "learning_rate": 1.4285714285714286e-06, + "loss": 3.7383, + "step": 5 + }, + { + "epoch": 0.019672131147540985, + "grad_norm": 57.113956451416016, + "learning_rate": 1.7857142857142859e-06, + "loss": 3.8281, + "step": 6 + }, + { + "epoch": 0.022950819672131147, + "grad_norm": 44.69753646850586, + "learning_rate": 2.1428571428571427e-06, + "loss": 3.6562, + "step": 7 + }, + { + "epoch": 0.02622950819672131, + "grad_norm": 42.74599075317383, + "learning_rate": 2.5e-06, + "loss": 3.668, + "step": 8 + }, + { + "epoch": 0.029508196721311476, + "grad_norm": 27.557815551757812, + "learning_rate": 2.8571428571428573e-06, + "loss": 3.3438, + "step": 9 + }, + { + "epoch": 0.03278688524590164, + "grad_norm": 25.87570571899414, + "learning_rate": 3.2142857142857147e-06, + "loss": 3.3633, + "step": 10 + }, + { + "epoch": 0.036065573770491806, + "grad_norm": 25.204158782958984, + "learning_rate": 3.5714285714285718e-06, + "loss": 3.2852, + "step": 11 + }, + { + "epoch": 0.03934426229508197, + "grad_norm": 25.082277297973633, + "learning_rate": 3.928571428571429e-06, + "loss": 3.2188, + "step": 12 + }, + { + "epoch": 0.04262295081967213, + "grad_norm": 21.52399253845215, + "learning_rate": 4.2857142857142855e-06, + "loss": 3.1016, + "step": 13 + }, + { + "epoch": 0.04590163934426229, + "grad_norm": 21.302213668823242, + "learning_rate": 4.642857142857144e-06, + "loss": 3.0938, + "step": 14 + }, + { + "epoch": 0.04918032786885246, + "grad_norm": 17.48920249938965, + "learning_rate": 5e-06, + "loss": 3.0703, + "step": 15 + }, + { + "epoch": 0.05245901639344262, + "grad_norm": 14.723689079284668, + "learning_rate": 5.357142857142857e-06, + "loss": 2.7812, + "step": 16 + }, + { + "epoch": 0.05573770491803279, + "grad_norm": 13.425204277038574, + "learning_rate": 5.7142857142857145e-06, + "loss": 2.8398, + "step": 17 + }, + { + "epoch": 0.05901639344262295, + "grad_norm": 12.121846199035645, + "learning_rate": 6.071428571428571e-06, + "loss": 2.7539, + "step": 18 + }, + { + "epoch": 0.06229508196721312, + "grad_norm": 12.97847843170166, + "learning_rate": 6.4285714285714295e-06, + "loss": 2.6719, + "step": 19 + }, + { + "epoch": 0.06557377049180328, + "grad_norm": 12.588919639587402, + "learning_rate": 6.785714285714287e-06, + "loss": 2.6758, + "step": 20 + }, + { + "epoch": 0.06885245901639345, + "grad_norm": 10.492659568786621, + "learning_rate": 7.1428571428571436e-06, + "loss": 2.6562, + "step": 21 + }, + { + "epoch": 0.07213114754098361, + "grad_norm": 9.676924705505371, + "learning_rate": 7.500000000000001e-06, + "loss": 2.6094, + "step": 22 + }, + { + "epoch": 0.07540983606557378, + "grad_norm": 10.222829818725586, + "learning_rate": 7.857142857142858e-06, + "loss": 2.6211, + "step": 23 + }, + { + "epoch": 0.07868852459016394, + "grad_norm": 10.020594596862793, + "learning_rate": 8.214285714285714e-06, + "loss": 2.5273, + "step": 24 + }, + { + "epoch": 0.08196721311475409, + "grad_norm": 9.404995918273926, + "learning_rate": 8.571428571428571e-06, + "loss": 2.5352, + "step": 25 + }, + { + "epoch": 0.08524590163934426, + "grad_norm": 10.194731712341309, + "learning_rate": 8.92857142857143e-06, + "loss": 2.543, + "step": 26 + }, + { + "epoch": 0.08852459016393442, + "grad_norm": 9.045639991760254, + "learning_rate": 9.285714285714288e-06, + "loss": 2.4961, + "step": 27 + }, + { + "epoch": 0.09180327868852459, + "grad_norm": 9.108716011047363, + "learning_rate": 9.642857142857144e-06, + "loss": 2.4336, + "step": 28 + }, + { + "epoch": 0.09508196721311475, + "grad_norm": 9.336297988891602, + "learning_rate": 1e-05, + "loss": 2.457, + "step": 29 + }, + { + "epoch": 0.09836065573770492, + "grad_norm": 8.170738220214844, + "learning_rate": 9.999968638844173e-06, + "loss": 2.4688, + "step": 30 + }, + { + "epoch": 0.10163934426229508, + "grad_norm": 8.507989883422852, + "learning_rate": 9.999874555770099e-06, + "loss": 2.4062, + "step": 31 + }, + { + "epoch": 0.10491803278688525, + "grad_norm": 8.38835334777832, + "learning_rate": 9.999717751958002e-06, + "loss": 2.3711, + "step": 32 + }, + { + "epoch": 0.10819672131147541, + "grad_norm": 8.36460018157959, + "learning_rate": 9.9994982293749e-06, + "loss": 2.4805, + "step": 33 + }, + { + "epoch": 0.11147540983606558, + "grad_norm": 8.547989845275879, + "learning_rate": 9.999215990774584e-06, + "loss": 2.3906, + "step": 34 + }, + { + "epoch": 0.11475409836065574, + "grad_norm": 8.938812255859375, + "learning_rate": 9.998871039697589e-06, + "loss": 2.3906, + "step": 35 + }, + { + "epoch": 0.1180327868852459, + "grad_norm": 8.230469703674316, + "learning_rate": 9.998463380471138e-06, + "loss": 2.3164, + "step": 36 + }, + { + "epoch": 0.12131147540983607, + "grad_norm": 7.883755683898926, + "learning_rate": 9.997993018209098e-06, + "loss": 2.3633, + "step": 37 + }, + { + "epoch": 0.12459016393442623, + "grad_norm": 8.220748901367188, + "learning_rate": 9.997459958811911e-06, + "loss": 2.3516, + "step": 38 + }, + { + "epoch": 0.12786885245901639, + "grad_norm": 8.738327980041504, + "learning_rate": 9.996864208966522e-06, + "loss": 2.3398, + "step": 39 + }, + { + "epoch": 0.13114754098360656, + "grad_norm": 7.944454669952393, + "learning_rate": 9.996205776146288e-06, + "loss": 2.3867, + "step": 40 + }, + { + "epoch": 0.13442622950819672, + "grad_norm": 7.631208419799805, + "learning_rate": 9.995484668610897e-06, + "loss": 2.3242, + "step": 41 + }, + { + "epoch": 0.1377049180327869, + "grad_norm": 9.301526069641113, + "learning_rate": 9.994700895406258e-06, + "loss": 2.3203, + "step": 42 + }, + { + "epoch": 0.14098360655737704, + "grad_norm": 7.625834941864014, + "learning_rate": 9.99385446636438e-06, + "loss": 2.3398, + "step": 43 + }, + { + "epoch": 0.14426229508196722, + "grad_norm": 7.908705234527588, + "learning_rate": 9.992945392103264e-06, + "loss": 2.3477, + "step": 44 + }, + { + "epoch": 0.14754098360655737, + "grad_norm": 8.002193450927734, + "learning_rate": 9.991973684026755e-06, + "loss": 2.2852, + "step": 45 + }, + { + "epoch": 0.15081967213114755, + "grad_norm": 8.058524131774902, + "learning_rate": 9.99093935432441e-06, + "loss": 2.3711, + "step": 46 + }, + { + "epoch": 0.1540983606557377, + "grad_norm": 8.803773880004883, + "learning_rate": 9.989842415971338e-06, + "loss": 2.2344, + "step": 47 + }, + { + "epoch": 0.15737704918032788, + "grad_norm": 7.851852893829346, + "learning_rate": 9.988682882728043e-06, + "loss": 2.3047, + "step": 48 + }, + { + "epoch": 0.16065573770491803, + "grad_norm": 8.29660415649414, + "learning_rate": 9.987460769140242e-06, + "loss": 2.3047, + "step": 49 + }, + { + "epoch": 0.16393442622950818, + "grad_norm": 7.779457092285156, + "learning_rate": 9.986176090538697e-06, + "loss": 2.1992, + "step": 50 + }, + { + "epoch": 0.16721311475409836, + "grad_norm": 8.602660179138184, + "learning_rate": 9.984828863039008e-06, + "loss": 2.25, + "step": 51 + }, + { + "epoch": 0.17049180327868851, + "grad_norm": 8.079516410827637, + "learning_rate": 9.983419103541421e-06, + "loss": 2.3203, + "step": 52 + }, + { + "epoch": 0.1737704918032787, + "grad_norm": 8.443635940551758, + "learning_rate": 9.981946829730611e-06, + "loss": 2.3242, + "step": 53 + }, + { + "epoch": 0.17704918032786884, + "grad_norm": 7.873276710510254, + "learning_rate": 9.980412060075459e-06, + "loss": 2.2656, + "step": 54 + }, + { + "epoch": 0.18032786885245902, + "grad_norm": 8.633429527282715, + "learning_rate": 9.978814813828827e-06, + "loss": 2.2227, + "step": 55 + }, + { + "epoch": 0.18360655737704917, + "grad_norm": 9.135279655456543, + "learning_rate": 9.97715511102731e-06, + "loss": 2.2148, + "step": 56 + }, + { + "epoch": 0.18688524590163935, + "grad_norm": 7.584745407104492, + "learning_rate": 9.975432972490985e-06, + "loss": 2.2773, + "step": 57 + }, + { + "epoch": 0.1901639344262295, + "grad_norm": 9.086216926574707, + "learning_rate": 9.973648419823161e-06, + "loss": 2.2656, + "step": 58 + }, + { + "epoch": 0.19344262295081968, + "grad_norm": 8.223559379577637, + "learning_rate": 9.971801475410084e-06, + "loss": 2.2773, + "step": 59 + }, + { + "epoch": 0.19672131147540983, + "grad_norm": 8.726387023925781, + "learning_rate": 9.969892162420682e-06, + "loss": 2.3125, + "step": 60 + }, + { + "epoch": 0.2, + "grad_norm": 8.304938316345215, + "learning_rate": 9.96792050480626e-06, + "loss": 2.168, + "step": 61 + }, + { + "epoch": 0.20327868852459016, + "grad_norm": 8.20266342163086, + "learning_rate": 9.965886527300201e-06, + "loss": 2.25, + "step": 62 + }, + { + "epoch": 0.20655737704918034, + "grad_norm": 7.833676815032959, + "learning_rate": 9.963790255417663e-06, + "loss": 2.2422, + "step": 63 + }, + { + "epoch": 0.2098360655737705, + "grad_norm": 7.891449928283691, + "learning_rate": 9.961631715455245e-06, + "loss": 2.1758, + "step": 64 + }, + { + "epoch": 0.21311475409836064, + "grad_norm": 8.999512672424316, + "learning_rate": 9.959410934490673e-06, + "loss": 2.1484, + "step": 65 + }, + { + "epoch": 0.21639344262295082, + "grad_norm": 7.718410015106201, + "learning_rate": 9.95712794038245e-06, + "loss": 2.1172, + "step": 66 + }, + { + "epoch": 0.21967213114754097, + "grad_norm": 8.272541999816895, + "learning_rate": 9.954782761769509e-06, + "loss": 2.25, + "step": 67 + }, + { + "epoch": 0.22295081967213115, + "grad_norm": 9.607975006103516, + "learning_rate": 9.952375428070853e-06, + "loss": 2.2812, + "step": 68 + }, + { + "epoch": 0.2262295081967213, + "grad_norm": 8.297865867614746, + "learning_rate": 9.949905969485192e-06, + "loss": 2.1406, + "step": 69 + }, + { + "epoch": 0.22950819672131148, + "grad_norm": 8.192915916442871, + "learning_rate": 9.947374416990554e-06, + "loss": 2.2109, + "step": 70 + }, + { + "epoch": 0.23278688524590163, + "grad_norm": 8.489333152770996, + "learning_rate": 9.944780802343906e-06, + "loss": 2.1914, + "step": 71 + }, + { + "epoch": 0.2360655737704918, + "grad_norm": 8.244175910949707, + "learning_rate": 9.942125158080747e-06, + "loss": 2.1797, + "step": 72 + }, + { + "epoch": 0.23934426229508196, + "grad_norm": 8.009477615356445, + "learning_rate": 9.939407517514709e-06, + "loss": 2.2578, + "step": 73 + }, + { + "epoch": 0.24262295081967214, + "grad_norm": 8.284704208374023, + "learning_rate": 9.936627914737129e-06, + "loss": 2.2305, + "step": 74 + }, + { + "epoch": 0.2459016393442623, + "grad_norm": 9.479103088378906, + "learning_rate": 9.933786384616631e-06, + "loss": 2.1406, + "step": 75 + }, + { + "epoch": 0.24918032786885247, + "grad_norm": 8.558723449707031, + "learning_rate": 9.930882962798683e-06, + "loss": 2.2812, + "step": 76 + }, + { + "epoch": 0.25245901639344265, + "grad_norm": 8.608718872070312, + "learning_rate": 9.927917685705148e-06, + "loss": 2.2461, + "step": 77 + }, + { + "epoch": 0.25573770491803277, + "grad_norm": 7.863804817199707, + "learning_rate": 9.924890590533837e-06, + "loss": 2.1914, + "step": 78 + }, + { + "epoch": 0.25901639344262295, + "grad_norm": 8.842366218566895, + "learning_rate": 9.921801715258027e-06, + "loss": 2.0996, + "step": 79 + }, + { + "epoch": 0.26229508196721313, + "grad_norm": 8.273311614990234, + "learning_rate": 9.918651098626e-06, + "loss": 2.1641, + "step": 80 + }, + { + "epoch": 0.26557377049180325, + "grad_norm": 8.011392593383789, + "learning_rate": 9.915438780160547e-06, + "loss": 2.168, + "step": 81 + }, + { + "epoch": 0.26885245901639343, + "grad_norm": 7.946706771850586, + "learning_rate": 9.912164800158474e-06, + "loss": 2.2227, + "step": 82 + }, + { + "epoch": 0.2721311475409836, + "grad_norm": 9.032291412353516, + "learning_rate": 9.908829199690106e-06, + "loss": 2.2891, + "step": 83 + }, + { + "epoch": 0.2754098360655738, + "grad_norm": 8.33333969116211, + "learning_rate": 9.905432020598751e-06, + "loss": 2.2344, + "step": 84 + }, + { + "epoch": 0.2786885245901639, + "grad_norm": 7.967901229858398, + "learning_rate": 9.901973305500197e-06, + "loss": 2.1172, + "step": 85 + }, + { + "epoch": 0.2819672131147541, + "grad_norm": 8.24918270111084, + "learning_rate": 9.898453097782164e-06, + "loss": 2.1602, + "step": 86 + }, + { + "epoch": 0.28524590163934427, + "grad_norm": 8.877111434936523, + "learning_rate": 9.894871441603766e-06, + "loss": 2.2383, + "step": 87 + }, + { + "epoch": 0.28852459016393445, + "grad_norm": 8.32304573059082, + "learning_rate": 9.891228381894954e-06, + "loss": 2.1719, + "step": 88 + }, + { + "epoch": 0.29180327868852457, + "grad_norm": 9.091806411743164, + "learning_rate": 9.887523964355953e-06, + "loss": 2.168, + "step": 89 + }, + { + "epoch": 0.29508196721311475, + "grad_norm": 8.54736614227295, + "learning_rate": 9.88375823545669e-06, + "loss": 2.1211, + "step": 90 + }, + { + "epoch": 0.2983606557377049, + "grad_norm": 9.3049955368042, + "learning_rate": 9.879931242436208e-06, + "loss": 2.1719, + "step": 91 + }, + { + "epoch": 0.3016393442622951, + "grad_norm": 8.523479461669922, + "learning_rate": 9.876043033302079e-06, + "loss": 2.1719, + "step": 92 + }, + { + "epoch": 0.30491803278688523, + "grad_norm": 7.741846561431885, + "learning_rate": 9.872093656829792e-06, + "loss": 2.0391, + "step": 93 + }, + { + "epoch": 0.3081967213114754, + "grad_norm": 8.36119270324707, + "learning_rate": 9.868083162562155e-06, + "loss": 2.0938, + "step": 94 + }, + { + "epoch": 0.3114754098360656, + "grad_norm": 8.163023948669434, + "learning_rate": 9.864011600808663e-06, + "loss": 2.0938, + "step": 95 + }, + { + "epoch": 0.31475409836065577, + "grad_norm": 8.827863693237305, + "learning_rate": 9.859879022644865e-06, + "loss": 2.1406, + "step": 96 + }, + { + "epoch": 0.3180327868852459, + "grad_norm": 8.434041023254395, + "learning_rate": 9.855685479911736e-06, + "loss": 2.168, + "step": 97 + }, + { + "epoch": 0.32131147540983607, + "grad_norm": 8.144036293029785, + "learning_rate": 9.851431025215012e-06, + "loss": 2.0742, + "step": 98 + }, + { + "epoch": 0.32459016393442625, + "grad_norm": 8.843683242797852, + "learning_rate": 9.847115711924542e-06, + "loss": 2.1016, + "step": 99 + }, + { + "epoch": 0.32786885245901637, + "grad_norm": 8.300407409667969, + "learning_rate": 9.842739594173607e-06, + "loss": 2.1953, + "step": 100 + }, + { + "epoch": 0.33114754098360655, + "grad_norm": 8.825713157653809, + "learning_rate": 9.838302726858258e-06, + "loss": 2.1406, + "step": 101 + }, + { + "epoch": 0.3344262295081967, + "grad_norm": 8.763128280639648, + "learning_rate": 9.833805165636603e-06, + "loss": 2.1523, + "step": 102 + }, + { + "epoch": 0.3377049180327869, + "grad_norm": 8.53940486907959, + "learning_rate": 9.829246966928135e-06, + "loss": 2.0703, + "step": 103 + }, + { + "epoch": 0.34098360655737703, + "grad_norm": 8.830872535705566, + "learning_rate": 9.824628187913001e-06, + "loss": 2.1602, + "step": 104 + }, + { + "epoch": 0.3442622950819672, + "grad_norm": 8.062509536743164, + "learning_rate": 9.819948886531305e-06, + "loss": 2.1211, + "step": 105 + }, + { + "epoch": 0.3475409836065574, + "grad_norm": 8.8229398727417, + "learning_rate": 9.815209121482363e-06, + "loss": 2.1172, + "step": 106 + }, + { + "epoch": 0.35081967213114756, + "grad_norm": 9.78320598602295, + "learning_rate": 9.810408952223982e-06, + "loss": 2.0586, + "step": 107 + }, + { + "epoch": 0.3540983606557377, + "grad_norm": 8.267208099365234, + "learning_rate": 9.805548438971702e-06, + "loss": 2.1406, + "step": 108 + }, + { + "epoch": 0.35737704918032787, + "grad_norm": 10.9806489944458, + "learning_rate": 9.80062764269805e-06, + "loss": 2.1035, + "step": 109 + }, + { + "epoch": 0.36065573770491804, + "grad_norm": 9.699831008911133, + "learning_rate": 9.795646625131771e-06, + "loss": 2.1797, + "step": 110 + }, + { + "epoch": 0.3639344262295082, + "grad_norm": 10.373594284057617, + "learning_rate": 9.790605448757049e-06, + "loss": 2.1445, + "step": 111 + }, + { + "epoch": 0.36721311475409835, + "grad_norm": 8.51201343536377, + "learning_rate": 9.785504176812733e-06, + "loss": 2.1875, + "step": 112 + }, + { + "epoch": 0.3704918032786885, + "grad_norm": 10.090012550354004, + "learning_rate": 9.780342873291535e-06, + "loss": 2.2109, + "step": 113 + }, + { + "epoch": 0.3737704918032787, + "grad_norm": 8.939981460571289, + "learning_rate": 9.775121602939234e-06, + "loss": 2.0273, + "step": 114 + }, + { + "epoch": 0.3770491803278688, + "grad_norm": 8.556815147399902, + "learning_rate": 9.76984043125386e-06, + "loss": 2.002, + "step": 115 + }, + { + "epoch": 0.380327868852459, + "grad_norm": 9.271822929382324, + "learning_rate": 9.764499424484873e-06, + "loss": 2.0742, + "step": 116 + }, + { + "epoch": 0.3836065573770492, + "grad_norm": 9.027301788330078, + "learning_rate": 9.759098649632326e-06, + "loss": 2.1133, + "step": 117 + }, + { + "epoch": 0.38688524590163936, + "grad_norm": 9.81689453125, + "learning_rate": 9.753638174446042e-06, + "loss": 2.0645, + "step": 118 + }, + { + "epoch": 0.3901639344262295, + "grad_norm": 9.42857551574707, + "learning_rate": 9.748118067424742e-06, + "loss": 2.1172, + "step": 119 + }, + { + "epoch": 0.39344262295081966, + "grad_norm": 8.636916160583496, + "learning_rate": 9.742538397815203e-06, + "loss": 2.0859, + "step": 120 + }, + { + "epoch": 0.39672131147540984, + "grad_norm": 9.01211166381836, + "learning_rate": 9.73689923561138e-06, + "loss": 2.0605, + "step": 121 + }, + { + "epoch": 0.4, + "grad_norm": 8.320965766906738, + "learning_rate": 9.73120065155353e-06, + "loss": 2.0352, + "step": 122 + }, + { + "epoch": 0.40327868852459015, + "grad_norm": 8.969964027404785, + "learning_rate": 9.725442717127324e-06, + "loss": 2.1055, + "step": 123 + }, + { + "epoch": 0.4065573770491803, + "grad_norm": 7.825037002563477, + "learning_rate": 9.719625504562959e-06, + "loss": 2.1133, + "step": 124 + }, + { + "epoch": 0.4098360655737705, + "grad_norm": 9.594252586364746, + "learning_rate": 9.713749086834234e-06, + "loss": 2.1367, + "step": 125 + }, + { + "epoch": 0.4131147540983607, + "grad_norm": 9.286437034606934, + "learning_rate": 9.707813537657652e-06, + "loss": 2.0293, + "step": 126 + }, + { + "epoch": 0.4163934426229508, + "grad_norm": 11.219290733337402, + "learning_rate": 9.701818931491485e-06, + "loss": 2.1992, + "step": 127 + }, + { + "epoch": 0.419672131147541, + "grad_norm": 8.125706672668457, + "learning_rate": 9.695765343534846e-06, + "loss": 2.0273, + "step": 128 + }, + { + "epoch": 0.42295081967213116, + "grad_norm": 9.294842720031738, + "learning_rate": 9.689652849726739e-06, + "loss": 2.082, + "step": 129 + }, + { + "epoch": 0.4262295081967213, + "grad_norm": 8.297966957092285, + "learning_rate": 9.683481526745114e-06, + "loss": 2.084, + "step": 130 + }, + { + "epoch": 0.42950819672131146, + "grad_norm": 9.68375015258789, + "learning_rate": 9.677251452005897e-06, + "loss": 2.0996, + "step": 131 + }, + { + "epoch": 0.43278688524590164, + "grad_norm": 8.27182388305664, + "learning_rate": 9.670962703662032e-06, + "loss": 2.0312, + "step": 132 + }, + { + "epoch": 0.4360655737704918, + "grad_norm": 9.681833267211914, + "learning_rate": 9.664615360602478e-06, + "loss": 2.0547, + "step": 133 + }, + { + "epoch": 0.43934426229508194, + "grad_norm": 8.213553428649902, + "learning_rate": 9.658209502451245e-06, + "loss": 1.9746, + "step": 134 + }, + { + "epoch": 0.4426229508196721, + "grad_norm": 9.513401985168457, + "learning_rate": 9.651745209566379e-06, + "loss": 2.0625, + "step": 135 + }, + { + "epoch": 0.4459016393442623, + "grad_norm": 9.364225387573242, + "learning_rate": 9.645222563038959e-06, + "loss": 2.0078, + "step": 136 + }, + { + "epoch": 0.4491803278688525, + "grad_norm": 8.659600257873535, + "learning_rate": 9.638641644692077e-06, + "loss": 2.0469, + "step": 137 + }, + { + "epoch": 0.4524590163934426, + "grad_norm": 9.638226509094238, + "learning_rate": 9.632002537079816e-06, + "loss": 2.0391, + "step": 138 + }, + { + "epoch": 0.4557377049180328, + "grad_norm": 9.299017906188965, + "learning_rate": 9.625305323486214e-06, + "loss": 2.0508, + "step": 139 + }, + { + "epoch": 0.45901639344262296, + "grad_norm": 9.940561294555664, + "learning_rate": 9.61855008792421e-06, + "loss": 2.082, + "step": 140 + }, + { + "epoch": 0.46229508196721314, + "grad_norm": 9.618109703063965, + "learning_rate": 9.611736915134605e-06, + "loss": 1.998, + "step": 141 + }, + { + "epoch": 0.46557377049180326, + "grad_norm": 8.827125549316406, + "learning_rate": 9.604865890584987e-06, + "loss": 2.1055, + "step": 142 + }, + { + "epoch": 0.46885245901639344, + "grad_norm": 10.079780578613281, + "learning_rate": 9.597937100468668e-06, + "loss": 2.0859, + "step": 143 + }, + { + "epoch": 0.4721311475409836, + "grad_norm": 9.007691383361816, + "learning_rate": 9.590950631703589e-06, + "loss": 2.125, + "step": 144 + }, + { + "epoch": 0.47540983606557374, + "grad_norm": 9.498886108398438, + "learning_rate": 9.583906571931248e-06, + "loss": 2.1172, + "step": 145 + }, + { + "epoch": 0.4786885245901639, + "grad_norm": 10.918264389038086, + "learning_rate": 9.576805009515588e-06, + "loss": 2.0508, + "step": 146 + }, + { + "epoch": 0.4819672131147541, + "grad_norm": 8.792062759399414, + "learning_rate": 9.569646033541888e-06, + "loss": 2.1484, + "step": 147 + }, + { + "epoch": 0.4852459016393443, + "grad_norm": 9.490158081054688, + "learning_rate": 9.562429733815655e-06, + "loss": 2.1133, + "step": 148 + }, + { + "epoch": 0.4885245901639344, + "grad_norm": 9.344221115112305, + "learning_rate": 9.55515620086149e-06, + "loss": 2.0684, + "step": 149 + }, + { + "epoch": 0.4918032786885246, + "grad_norm": 10.885297775268555, + "learning_rate": 9.54782552592195e-06, + "loss": 2.0957, + "step": 150 + }, + { + "epoch": 0.49508196721311476, + "grad_norm": 8.593646049499512, + "learning_rate": 9.540437800956412e-06, + "loss": 2.0625, + "step": 151 + }, + { + "epoch": 0.49836065573770494, + "grad_norm": 8.966227531433105, + "learning_rate": 9.532993118639915e-06, + "loss": 2.0, + "step": 152 + }, + { + "epoch": 0.5016393442622951, + "grad_norm": 10.044122695922852, + "learning_rate": 9.525491572361995e-06, + "loss": 2.0176, + "step": 153 + }, + { + "epoch": 0.5049180327868853, + "grad_norm": 8.53621768951416, + "learning_rate": 9.517933256225516e-06, + "loss": 2.1094, + "step": 154 + }, + { + "epoch": 0.5081967213114754, + "grad_norm": 10.106084823608398, + "learning_rate": 9.510318265045488e-06, + "loss": 2.0195, + "step": 155 + }, + { + "epoch": 0.5114754098360655, + "grad_norm": 9.872212409973145, + "learning_rate": 9.502646694347886e-06, + "loss": 2.1055, + "step": 156 + }, + { + "epoch": 0.5147540983606558, + "grad_norm": 8.695504188537598, + "learning_rate": 9.494918640368435e-06, + "loss": 1.9961, + "step": 157 + }, + { + "epoch": 0.5180327868852459, + "grad_norm": 9.583698272705078, + "learning_rate": 9.487134200051422e-06, + "loss": 2.0801, + "step": 158 + }, + { + "epoch": 0.521311475409836, + "grad_norm": 9.81774616241455, + "learning_rate": 9.479293471048461e-06, + "loss": 2.043, + "step": 159 + }, + { + "epoch": 0.5245901639344263, + "grad_norm": 8.542972564697266, + "learning_rate": 9.471396551717283e-06, + "loss": 1.9922, + "step": 160 + }, + { + "epoch": 0.5278688524590164, + "grad_norm": 9.894759178161621, + "learning_rate": 9.463443541120496e-06, + "loss": 2.0469, + "step": 161 + }, + { + "epoch": 0.5311475409836065, + "grad_norm": 9.396486282348633, + "learning_rate": 9.45543453902434e-06, + "loss": 1.9688, + "step": 162 + }, + { + "epoch": 0.5344262295081967, + "grad_norm": 9.620316505432129, + "learning_rate": 9.447369645897441e-06, + "loss": 2.0391, + "step": 163 + }, + { + "epoch": 0.5377049180327869, + "grad_norm": 10.236593246459961, + "learning_rate": 9.439248962909547e-06, + "loss": 1.957, + "step": 164 + }, + { + "epoch": 0.5409836065573771, + "grad_norm": 11.980149269104004, + "learning_rate": 9.431072591930261e-06, + "loss": 2.0293, + "step": 165 + }, + { + "epoch": 0.5442622950819672, + "grad_norm": 14.74476146697998, + "learning_rate": 9.42284063552776e-06, + "loss": 1.9844, + "step": 166 + }, + { + "epoch": 0.5475409836065573, + "grad_norm": 9.34021282196045, + "learning_rate": 9.414553196967508e-06, + "loss": 2.0703, + "step": 167 + }, + { + "epoch": 0.5508196721311476, + "grad_norm": 12.805119514465332, + "learning_rate": 9.406210380210971e-06, + "loss": 2.0117, + "step": 168 + }, + { + "epoch": 0.5540983606557377, + "grad_norm": 9.839883804321289, + "learning_rate": 9.397812289914297e-06, + "loss": 1.9727, + "step": 169 + }, + { + "epoch": 0.5573770491803278, + "grad_norm": 14.596024513244629, + "learning_rate": 9.389359031427013e-06, + "loss": 2.0742, + "step": 170 + }, + { + "epoch": 0.5606557377049181, + "grad_norm": 13.544656753540039, + "learning_rate": 9.380850710790701e-06, + "loss": 2.0742, + "step": 171 + }, + { + "epoch": 0.5639344262295082, + "grad_norm": 9.45289421081543, + "learning_rate": 9.372287434737673e-06, + "loss": 2.0586, + "step": 172 + }, + { + "epoch": 0.5672131147540984, + "grad_norm": 10.174077987670898, + "learning_rate": 9.363669310689617e-06, + "loss": 2.1367, + "step": 173 + }, + { + "epoch": 0.5704918032786885, + "grad_norm": 12.690667152404785, + "learning_rate": 9.354996446756268e-06, + "loss": 1.9395, + "step": 174 + }, + { + "epoch": 0.5737704918032787, + "grad_norm": 10.625876426696777, + "learning_rate": 9.346268951734042e-06, + "loss": 2.0215, + "step": 175 + }, + { + "epoch": 0.5770491803278689, + "grad_norm": 11.856694221496582, + "learning_rate": 9.337486935104672e-06, + "loss": 1.9766, + "step": 176 + }, + { + "epoch": 0.580327868852459, + "grad_norm": 10.2452974319458, + "learning_rate": 9.328650507033832e-06, + "loss": 2.0586, + "step": 177 + }, + { + "epoch": 0.5836065573770491, + "grad_norm": 12.020367622375488, + "learning_rate": 9.319759778369764e-06, + "loss": 2.0352, + "step": 178 + }, + { + "epoch": 0.5868852459016394, + "grad_norm": 11.364824295043945, + "learning_rate": 9.310814860641874e-06, + "loss": 2.0352, + "step": 179 + }, + { + "epoch": 0.5901639344262295, + "grad_norm": 12.439066886901855, + "learning_rate": 9.301815866059353e-06, + "loss": 2.0703, + "step": 180 + }, + { + "epoch": 0.5934426229508196, + "grad_norm": 12.66345500946045, + "learning_rate": 9.292762907509742e-06, + "loss": 1.9922, + "step": 181 + }, + { + "epoch": 0.5967213114754099, + "grad_norm": 11.868645668029785, + "learning_rate": 9.283656098557545e-06, + "loss": 1.9863, + "step": 182 + }, + { + "epoch": 0.6, + "grad_norm": 11.235689163208008, + "learning_rate": 9.27449555344278e-06, + "loss": 2.0195, + "step": 183 + }, + { + "epoch": 0.6032786885245902, + "grad_norm": 10.719315528869629, + "learning_rate": 9.26528138707956e-06, + "loss": 1.9883, + "step": 184 + }, + { + "epoch": 0.6065573770491803, + "grad_norm": 11.960482597351074, + "learning_rate": 9.256013715054648e-06, + "loss": 1.9082, + "step": 185 + }, + { + "epoch": 0.6098360655737705, + "grad_norm": 9.54014778137207, + "learning_rate": 9.24669265362601e-06, + "loss": 2.0156, + "step": 186 + }, + { + "epoch": 0.6131147540983607, + "grad_norm": 10.753348350524902, + "learning_rate": 9.237318319721346e-06, + "loss": 1.9746, + "step": 187 + }, + { + "epoch": 0.6163934426229508, + "grad_norm": 9.75783634185791, + "learning_rate": 9.227890830936634e-06, + "loss": 2.0195, + "step": 188 + }, + { + "epoch": 0.6196721311475409, + "grad_norm": 15.201295852661133, + "learning_rate": 9.218410305534657e-06, + "loss": 2.0117, + "step": 189 + }, + { + "epoch": 0.6229508196721312, + "grad_norm": 10.927349090576172, + "learning_rate": 9.208876862443504e-06, + "loss": 1.9316, + "step": 190 + }, + { + "epoch": 0.6262295081967213, + "grad_norm": 13.822660446166992, + "learning_rate": 9.199290621255095e-06, + "loss": 2.0371, + "step": 191 + }, + { + "epoch": 0.6295081967213115, + "grad_norm": 13.426775932312012, + "learning_rate": 9.189651702223672e-06, + "loss": 2.0273, + "step": 192 + }, + { + "epoch": 0.6327868852459017, + "grad_norm": 9.90354061126709, + "learning_rate": 9.179960226264288e-06, + "loss": 1.957, + "step": 193 + }, + { + "epoch": 0.6360655737704918, + "grad_norm": 20.340970993041992, + "learning_rate": 9.170216314951301e-06, + "loss": 2.0488, + "step": 194 + }, + { + "epoch": 0.639344262295082, + "grad_norm": 9.366877555847168, + "learning_rate": 9.16042009051684e-06, + "loss": 1.9688, + "step": 195 + }, + { + "epoch": 0.6426229508196721, + "grad_norm": 14.084966659545898, + "learning_rate": 9.150571675849273e-06, + "loss": 2.0938, + "step": 196 + }, + { + "epoch": 0.6459016393442623, + "grad_norm": 16.520889282226562, + "learning_rate": 9.140671194491666e-06, + "loss": 2.0508, + "step": 197 + }, + { + "epoch": 0.6491803278688525, + "grad_norm": 13.539576530456543, + "learning_rate": 9.130718770640235e-06, + "loss": 1.9922, + "step": 198 + }, + { + "epoch": 0.6524590163934426, + "grad_norm": 14.922486305236816, + "learning_rate": 9.120714529142784e-06, + "loss": 2.0371, + "step": 199 + }, + { + "epoch": 0.6557377049180327, + "grad_norm": 12.108205795288086, + "learning_rate": 9.110658595497148e-06, + "loss": 1.9902, + "step": 200 + }, + { + "epoch": 0.659016393442623, + "grad_norm": 10.225913047790527, + "learning_rate": 9.100551095849603e-06, + "loss": 1.8574, + "step": 201 + }, + { + "epoch": 0.6622950819672131, + "grad_norm": 10.973288536071777, + "learning_rate": 9.090392156993301e-06, + "loss": 1.9902, + "step": 202 + }, + { + "epoch": 0.6655737704918033, + "grad_norm": 12.73608684539795, + "learning_rate": 9.080181906366668e-06, + "loss": 1.9551, + "step": 203 + }, + { + "epoch": 0.6688524590163935, + "grad_norm": 11.886429786682129, + "learning_rate": 9.069920472051804e-06, + "loss": 1.957, + "step": 204 + }, + { + "epoch": 0.6721311475409836, + "grad_norm": 13.453667640686035, + "learning_rate": 9.05960798277289e-06, + "loss": 1.9512, + "step": 205 + }, + { + "epoch": 0.6754098360655738, + "grad_norm": 19.135602951049805, + "learning_rate": 9.04924456789456e-06, + "loss": 1.9688, + "step": 206 + }, + { + "epoch": 0.6786885245901639, + "grad_norm": 16.149311065673828, + "learning_rate": 9.038830357420276e-06, + "loss": 1.9688, + "step": 207 + }, + { + "epoch": 0.6819672131147541, + "grad_norm": 14.444845199584961, + "learning_rate": 9.028365481990713e-06, + "loss": 2.0625, + "step": 208 + }, + { + "epoch": 0.6852459016393443, + "grad_norm": 12.08676528930664, + "learning_rate": 9.017850072882107e-06, + "loss": 1.9121, + "step": 209 + }, + { + "epoch": 0.6885245901639344, + "grad_norm": 11.756418228149414, + "learning_rate": 9.007284262004608e-06, + "loss": 1.9102, + "step": 210 + }, + { + "epoch": 0.6918032786885245, + "grad_norm": 12.56296443939209, + "learning_rate": 8.996668181900636e-06, + "loss": 2.0312, + "step": 211 + }, + { + "epoch": 0.6950819672131148, + "grad_norm": 10.218338966369629, + "learning_rate": 8.986001965743205e-06, + "loss": 1.9629, + "step": 212 + }, + { + "epoch": 0.6983606557377049, + "grad_norm": 9.9441556930542, + "learning_rate": 8.975285747334267e-06, + "loss": 1.9844, + "step": 213 + }, + { + "epoch": 0.7016393442622951, + "grad_norm": 10.612693786621094, + "learning_rate": 8.964519661103014e-06, + "loss": 1.9746, + "step": 214 + }, + { + "epoch": 0.7049180327868853, + "grad_norm": 10.07176399230957, + "learning_rate": 8.953703842104212e-06, + "loss": 1.9531, + "step": 215 + }, + { + "epoch": 0.7081967213114754, + "grad_norm": 9.502864837646484, + "learning_rate": 8.942838426016496e-06, + "loss": 1.9492, + "step": 216 + }, + { + "epoch": 0.7114754098360656, + "grad_norm": 9.91391372680664, + "learning_rate": 8.931923549140668e-06, + "loss": 1.9844, + "step": 217 + }, + { + "epoch": 0.7147540983606557, + "grad_norm": 9.780139923095703, + "learning_rate": 8.920959348397988e-06, + "loss": 1.9473, + "step": 218 + }, + { + "epoch": 0.7180327868852459, + "grad_norm": 15.030618667602539, + "learning_rate": 8.909945961328462e-06, + "loss": 1.9922, + "step": 219 + }, + { + "epoch": 0.7213114754098361, + "grad_norm": 10.924345016479492, + "learning_rate": 8.898883526089108e-06, + "loss": 2.0195, + "step": 220 + }, + { + "epoch": 0.7245901639344262, + "grad_norm": 10.839616775512695, + "learning_rate": 8.887772181452228e-06, + "loss": 2.0234, + "step": 221 + }, + { + "epoch": 0.7278688524590164, + "grad_norm": 11.212786674499512, + "learning_rate": 8.876612066803666e-06, + "loss": 1.9727, + "step": 222 + }, + { + "epoch": 0.7311475409836066, + "grad_norm": 9.742143630981445, + "learning_rate": 8.86540332214106e-06, + "loss": 1.9844, + "step": 223 + }, + { + "epoch": 0.7344262295081967, + "grad_norm": 12.224595069885254, + "learning_rate": 8.854146088072087e-06, + "loss": 2.0566, + "step": 224 + }, + { + "epoch": 0.7377049180327869, + "grad_norm": 9.907976150512695, + "learning_rate": 8.842840505812691e-06, + "loss": 2.0215, + "step": 225 + }, + { + "epoch": 0.740983606557377, + "grad_norm": 10.605637550354004, + "learning_rate": 8.83148671718533e-06, + "loss": 1.959, + "step": 226 + }, + { + "epoch": 0.7442622950819672, + "grad_norm": 10.371489524841309, + "learning_rate": 8.82008486461717e-06, + "loss": 1.9492, + "step": 227 + }, + { + "epoch": 0.7475409836065574, + "grad_norm": 13.147061347961426, + "learning_rate": 8.808635091138327e-06, + "loss": 1.9375, + "step": 228 + }, + { + "epoch": 0.7508196721311475, + "grad_norm": 13.442245483398438, + "learning_rate": 8.79713754038005e-06, + "loss": 1.9219, + "step": 229 + }, + { + "epoch": 0.7540983606557377, + "grad_norm": 10.87454891204834, + "learning_rate": 8.785592356572931e-06, + "loss": 1.9609, + "step": 230 + }, + { + "epoch": 0.7573770491803279, + "grad_norm": 12.752360343933105, + "learning_rate": 8.773999684545096e-06, + "loss": 1.9414, + "step": 231 + }, + { + "epoch": 0.760655737704918, + "grad_norm": 10.660215377807617, + "learning_rate": 8.76235966972038e-06, + "loss": 1.9395, + "step": 232 + }, + { + "epoch": 0.7639344262295082, + "grad_norm": 13.825894355773926, + "learning_rate": 8.750672458116513e-06, + "loss": 1.9102, + "step": 233 + }, + { + "epoch": 0.7672131147540984, + "grad_norm": 11.399506568908691, + "learning_rate": 8.738938196343278e-06, + "loss": 2.002, + "step": 234 + }, + { + "epoch": 0.7704918032786885, + "grad_norm": 10.732869148254395, + "learning_rate": 8.72715703160068e-06, + "loss": 1.8887, + "step": 235 + }, + { + "epoch": 0.7737704918032787, + "grad_norm": 13.854918479919434, + "learning_rate": 8.715329111677102e-06, + "loss": 1.9922, + "step": 236 + }, + { + "epoch": 0.7770491803278688, + "grad_norm": 12.330432891845703, + "learning_rate": 8.703454584947433e-06, + "loss": 1.9395, + "step": 237 + }, + { + "epoch": 0.780327868852459, + "grad_norm": 9.50846004486084, + "learning_rate": 8.691533600371227e-06, + "loss": 1.8965, + "step": 238 + }, + { + "epoch": 0.7836065573770492, + "grad_norm": 10.240988731384277, + "learning_rate": 8.67956630749083e-06, + "loss": 1.9121, + "step": 239 + }, + { + "epoch": 0.7868852459016393, + "grad_norm": 10.392232894897461, + "learning_rate": 8.667552856429494e-06, + "loss": 1.8809, + "step": 240 + }, + { + "epoch": 0.7901639344262295, + "grad_norm": 10.917051315307617, + "learning_rate": 8.655493397889503e-06, + "loss": 1.9805, + "step": 241 + }, + { + "epoch": 0.7934426229508197, + "grad_norm": 10.163880348205566, + "learning_rate": 8.64338808315028e-06, + "loss": 1.8711, + "step": 242 + }, + { + "epoch": 0.7967213114754098, + "grad_norm": 11.161616325378418, + "learning_rate": 8.631237064066491e-06, + "loss": 1.9102, + "step": 243 + }, + { + "epoch": 0.8, + "grad_norm": 11.414627075195312, + "learning_rate": 8.61904049306614e-06, + "loss": 1.9727, + "step": 244 + }, + { + "epoch": 0.8032786885245902, + "grad_norm": 10.282430648803711, + "learning_rate": 8.606798523148648e-06, + "loss": 1.9414, + "step": 245 + }, + { + "epoch": 0.8065573770491803, + "grad_norm": 12.212276458740234, + "learning_rate": 8.594511307882945e-06, + "loss": 1.8496, + "step": 246 + }, + { + "epoch": 0.8098360655737705, + "grad_norm": 14.458395957946777, + "learning_rate": 8.582179001405544e-06, + "loss": 1.9414, + "step": 247 + }, + { + "epoch": 0.8131147540983606, + "grad_norm": 10.135568618774414, + "learning_rate": 8.569801758418596e-06, + "loss": 1.8691, + "step": 248 + }, + { + "epoch": 0.8163934426229508, + "grad_norm": 14.193105697631836, + "learning_rate": 8.55737973418796e-06, + "loss": 1.834, + "step": 249 + }, + { + "epoch": 0.819672131147541, + "grad_norm": 10.561907768249512, + "learning_rate": 8.544913084541253e-06, + "loss": 2.043, + "step": 250 + }, + { + "epoch": 0.8229508196721311, + "grad_norm": 12.753805160522461, + "learning_rate": 8.532401965865889e-06, + "loss": 1.9141, + "step": 251 + }, + { + "epoch": 0.8262295081967214, + "grad_norm": 11.008598327636719, + "learning_rate": 8.519846535107129e-06, + "loss": 1.9258, + "step": 252 + }, + { + "epoch": 0.8295081967213115, + "grad_norm": 10.186516761779785, + "learning_rate": 8.507246949766095e-06, + "loss": 1.9062, + "step": 253 + }, + { + "epoch": 0.8327868852459016, + "grad_norm": 12.843867301940918, + "learning_rate": 8.494603367897817e-06, + "loss": 1.8516, + "step": 254 + }, + { + "epoch": 0.8360655737704918, + "grad_norm": 11.588765144348145, + "learning_rate": 8.481915948109226e-06, + "loss": 1.9121, + "step": 255 + }, + { + "epoch": 0.839344262295082, + "grad_norm": 12.169210433959961, + "learning_rate": 8.469184849557188e-06, + "loss": 1.9238, + "step": 256 + }, + { + "epoch": 0.8426229508196721, + "grad_norm": 17.143991470336914, + "learning_rate": 8.456410231946484e-06, + "loss": 1.8887, + "step": 257 + }, + { + "epoch": 0.8459016393442623, + "grad_norm": 9.951960563659668, + "learning_rate": 8.443592255527824e-06, + "loss": 1.9355, + "step": 258 + }, + { + "epoch": 0.8491803278688524, + "grad_norm": 11.594623565673828, + "learning_rate": 8.430731081095832e-06, + "loss": 1.9512, + "step": 259 + }, + { + "epoch": 0.8524590163934426, + "grad_norm": 10.37341022491455, + "learning_rate": 8.417826869987026e-06, + "loss": 1.9883, + "step": 260 + }, + { + "epoch": 0.8557377049180328, + "grad_norm": 13.815895080566406, + "learning_rate": 8.404879784077793e-06, + "loss": 1.9746, + "step": 261 + }, + { + "epoch": 0.8590163934426229, + "grad_norm": 10.756362915039062, + "learning_rate": 8.39188998578237e-06, + "loss": 1.9023, + "step": 262 + }, + { + "epoch": 0.8622950819672132, + "grad_norm": 11.125969886779785, + "learning_rate": 8.37885763805079e-06, + "loss": 1.873, + "step": 263 + }, + { + "epoch": 0.8655737704918033, + "grad_norm": 10.486440658569336, + "learning_rate": 8.365782904366845e-06, + "loss": 1.8789, + "step": 264 + }, + { + "epoch": 0.8688524590163934, + "grad_norm": 10.452739715576172, + "learning_rate": 8.352665948746044e-06, + "loss": 1.9453, + "step": 265 + }, + { + "epoch": 0.8721311475409836, + "grad_norm": 10.094730377197266, + "learning_rate": 8.33950693573354e-06, + "loss": 1.8516, + "step": 266 + }, + { + "epoch": 0.8754098360655738, + "grad_norm": 11.559701919555664, + "learning_rate": 8.326306030402076e-06, + "loss": 1.9844, + "step": 267 + }, + { + "epoch": 0.8786885245901639, + "grad_norm": 11.006271362304688, + "learning_rate": 8.313063398349914e-06, + "loss": 1.8887, + "step": 268 + }, + { + "epoch": 0.8819672131147541, + "grad_norm": 12.529583930969238, + "learning_rate": 8.29977920569875e-06, + "loss": 1.877, + "step": 269 + }, + { + "epoch": 0.8852459016393442, + "grad_norm": 10.77204704284668, + "learning_rate": 8.28645361909164e-06, + "loss": 1.998, + "step": 270 + }, + { + "epoch": 0.8885245901639345, + "grad_norm": 13.109503746032715, + "learning_rate": 8.273086805690903e-06, + "loss": 1.8711, + "step": 271 + }, + { + "epoch": 0.8918032786885246, + "grad_norm": 15.495031356811523, + "learning_rate": 8.259678933176025e-06, + "loss": 1.916, + "step": 272 + }, + { + "epoch": 0.8950819672131147, + "grad_norm": 11.081769943237305, + "learning_rate": 8.24623016974156e-06, + "loss": 1.8809, + "step": 273 + }, + { + "epoch": 0.898360655737705, + "grad_norm": 10.445962905883789, + "learning_rate": 8.232740684095012e-06, + "loss": 1.8301, + "step": 274 + }, + { + "epoch": 0.9016393442622951, + "grad_norm": 11.152101516723633, + "learning_rate": 8.219210645454727e-06, + "loss": 1.9258, + "step": 275 + }, + { + "epoch": 0.9049180327868852, + "grad_norm": 10.69294261932373, + "learning_rate": 8.205640223547766e-06, + "loss": 1.8242, + "step": 276 + }, + { + "epoch": 0.9081967213114754, + "grad_norm": 13.176959991455078, + "learning_rate": 8.192029588607772e-06, + "loss": 1.9453, + "step": 277 + }, + { + "epoch": 0.9114754098360656, + "grad_norm": 12.394086837768555, + "learning_rate": 8.178378911372846e-06, + "loss": 1.9434, + "step": 278 + }, + { + "epoch": 0.9147540983606557, + "grad_norm": 13.449362754821777, + "learning_rate": 8.16468836308339e-06, + "loss": 1.9961, + "step": 279 + }, + { + "epoch": 0.9180327868852459, + "grad_norm": 10.922250747680664, + "learning_rate": 8.150958115479976e-06, + "loss": 1.9531, + "step": 280 + }, + { + "epoch": 0.921311475409836, + "grad_norm": 14.1903657913208, + "learning_rate": 8.137188340801177e-06, + "loss": 1.9336, + "step": 281 + }, + { + "epoch": 0.9245901639344263, + "grad_norm": 13.266737937927246, + "learning_rate": 8.12337921178141e-06, + "loss": 1.9688, + "step": 282 + }, + { + "epoch": 0.9278688524590164, + "grad_norm": 12.7962064743042, + "learning_rate": 8.109530901648776e-06, + "loss": 1.875, + "step": 283 + }, + { + "epoch": 0.9311475409836065, + "grad_norm": 12.458824157714844, + "learning_rate": 8.095643584122879e-06, + "loss": 1.9141, + "step": 284 + }, + { + "epoch": 0.9344262295081968, + "grad_norm": 10.26793384552002, + "learning_rate": 8.08171743341265e-06, + "loss": 1.8633, + "step": 285 + }, + { + "epoch": 0.9377049180327869, + "grad_norm": 12.272446632385254, + "learning_rate": 8.067752624214164e-06, + "loss": 1.8945, + "step": 286 + }, + { + "epoch": 0.940983606557377, + "grad_norm": 11.974457740783691, + "learning_rate": 8.053749331708444e-06, + "loss": 1.877, + "step": 287 + }, + { + "epoch": 0.9442622950819672, + "grad_norm": 10.955528259277344, + "learning_rate": 8.039707731559262e-06, + "loss": 1.8867, + "step": 288 + }, + { + "epoch": 0.9475409836065574, + "grad_norm": 12.32547378540039, + "learning_rate": 8.025627999910946e-06, + "loss": 1.9629, + "step": 289 + }, + { + "epoch": 0.9508196721311475, + "grad_norm": 12.567078590393066, + "learning_rate": 8.011510313386156e-06, + "loss": 1.877, + "step": 290 + }, + { + "epoch": 0.9540983606557377, + "grad_norm": 10.757590293884277, + "learning_rate": 7.997354849083682e-06, + "loss": 1.8926, + "step": 291 + }, + { + "epoch": 0.9573770491803278, + "grad_norm": 12.629182815551758, + "learning_rate": 7.98316178457621e-06, + "loss": 1.9199, + "step": 292 + }, + { + "epoch": 0.9606557377049181, + "grad_norm": 11.572930335998535, + "learning_rate": 7.968931297908105e-06, + "loss": 1.9492, + "step": 293 + }, + { + "epoch": 0.9639344262295082, + "grad_norm": 13.813660621643066, + "learning_rate": 7.954663567593169e-06, + "loss": 1.793, + "step": 294 + }, + { + "epoch": 0.9672131147540983, + "grad_norm": 10.30826473236084, + "learning_rate": 7.940358772612408e-06, + "loss": 1.8809, + "step": 295 + }, + { + "epoch": 0.9704918032786886, + "grad_norm": 11.040604591369629, + "learning_rate": 7.926017092411786e-06, + "loss": 1.8438, + "step": 296 + }, + { + "epoch": 0.9737704918032787, + "grad_norm": 11.625853538513184, + "learning_rate": 7.911638706899968e-06, + "loss": 1.8652, + "step": 297 + }, + { + "epoch": 0.9770491803278688, + "grad_norm": 12.791728973388672, + "learning_rate": 7.89722379644607e-06, + "loss": 1.8555, + "step": 298 + }, + { + "epoch": 0.980327868852459, + "grad_norm": 15.870569229125977, + "learning_rate": 7.882772541877392e-06, + "loss": 1.9043, + "step": 299 + }, + { + "epoch": 0.9836065573770492, + "grad_norm": 11.703765869140625, + "learning_rate": 7.868285124477156e-06, + "loss": 1.8613, + "step": 300 + }, + { + "epoch": 0.9868852459016394, + "grad_norm": 11.708532333374023, + "learning_rate": 7.853761725982221e-06, + "loss": 1.8574, + "step": 301 + }, + { + "epoch": 0.9901639344262295, + "grad_norm": 10.677489280700684, + "learning_rate": 7.839202528580812e-06, + "loss": 1.9004, + "step": 302 + }, + { + "epoch": 0.9934426229508196, + "grad_norm": 11.307282447814941, + "learning_rate": 7.824607714910233e-06, + "loss": 1.7969, + "step": 303 + }, + { + "epoch": 0.9967213114754099, + "grad_norm": 14.72847843170166, + "learning_rate": 7.809977468054574e-06, + "loss": 1.8672, + "step": 304 + }, + { + "epoch": 1.0, + "grad_norm": 14.625303268432617, + "learning_rate": 7.795311971542418e-06, + "loss": 1.8945, + "step": 305 + }, + { + "epoch": 1.0032786885245901, + "grad_norm": 10.49122142791748, + "learning_rate": 7.78061140934453e-06, + "loss": 1.709, + "step": 306 + }, + { + "epoch": 1.0065573770491802, + "grad_norm": 11.381379127502441, + "learning_rate": 7.765875965871561e-06, + "loss": 1.6719, + "step": 307 + }, + { + "epoch": 1.0098360655737706, + "grad_norm": 13.590991973876953, + "learning_rate": 7.751105825971727e-06, + "loss": 1.7441, + "step": 308 + }, + { + "epoch": 1.0131147540983607, + "grad_norm": 10.184123992919922, + "learning_rate": 7.736301174928489e-06, + "loss": 1.7246, + "step": 309 + }, + { + "epoch": 1.0163934426229508, + "grad_norm": 11.995527267456055, + "learning_rate": 7.721462198458235e-06, + "loss": 1.6797, + "step": 310 + }, + { + "epoch": 1.019672131147541, + "grad_norm": 17.523670196533203, + "learning_rate": 7.706589082707951e-06, + "loss": 1.6953, + "step": 311 + }, + { + "epoch": 1.022950819672131, + "grad_norm": 11.37983226776123, + "learning_rate": 7.69168201425287e-06, + "loss": 1.6387, + "step": 312 + }, + { + "epoch": 1.0262295081967212, + "grad_norm": 16.715530395507812, + "learning_rate": 7.676741180094156e-06, + "loss": 1.6797, + "step": 313 + }, + { + "epoch": 1.0295081967213116, + "grad_norm": 13.554855346679688, + "learning_rate": 7.661766767656538e-06, + "loss": 1.6738, + "step": 314 + }, + { + "epoch": 1.0327868852459017, + "grad_norm": 20.694581985473633, + "learning_rate": 7.64675896478597e-06, + "loss": 1.6992, + "step": 315 + }, + { + "epoch": 1.0360655737704918, + "grad_norm": 16.622581481933594, + "learning_rate": 7.631717959747269e-06, + "loss": 1.7305, + "step": 316 + }, + { + "epoch": 1.039344262295082, + "grad_norm": 19.643783569335938, + "learning_rate": 7.616643941221755e-06, + "loss": 1.752, + "step": 317 + }, + { + "epoch": 1.042622950819672, + "grad_norm": 14.100711822509766, + "learning_rate": 7.6015370983048875e-06, + "loss": 1.7285, + "step": 318 + }, + { + "epoch": 1.0459016393442624, + "grad_norm": 19.153236389160156, + "learning_rate": 7.586397620503887e-06, + "loss": 1.6953, + "step": 319 + }, + { + "epoch": 1.0491803278688525, + "grad_norm": 11.576682090759277, + "learning_rate": 7.571225697735365e-06, + "loss": 1.668, + "step": 320 + }, + { + "epoch": 1.0524590163934426, + "grad_norm": 11.654809951782227, + "learning_rate": 7.556021520322932e-06, + "loss": 1.7539, + "step": 321 + }, + { + "epoch": 1.0557377049180328, + "grad_norm": 14.915346145629883, + "learning_rate": 7.54078527899482e-06, + "loss": 1.6797, + "step": 322 + }, + { + "epoch": 1.0590163934426229, + "grad_norm": 16.05927276611328, + "learning_rate": 7.525517164881486e-06, + "loss": 1.5918, + "step": 323 + }, + { + "epoch": 1.0622950819672132, + "grad_norm": 12.948592185974121, + "learning_rate": 7.51021736951321e-06, + "loss": 1.5879, + "step": 324 + }, + { + "epoch": 1.0655737704918034, + "grad_norm": 14.317193031311035, + "learning_rate": 7.4948860848176984e-06, + "loss": 1.709, + "step": 325 + }, + { + "epoch": 1.0688524590163935, + "grad_norm": 14.903904914855957, + "learning_rate": 7.479523503117676e-06, + "loss": 1.6738, + "step": 326 + }, + { + "epoch": 1.0721311475409836, + "grad_norm": 19.48590087890625, + "learning_rate": 7.4641298171284715e-06, + "loss": 1.6777, + "step": 327 + }, + { + "epoch": 1.0754098360655737, + "grad_norm": 12.375781059265137, + "learning_rate": 7.448705219955596e-06, + "loss": 1.6719, + "step": 328 + }, + { + "epoch": 1.0786885245901638, + "grad_norm": 13.3538179397583, + "learning_rate": 7.433249905092331e-06, + "loss": 1.707, + "step": 329 + }, + { + "epoch": 1.0819672131147542, + "grad_norm": 14.281739234924316, + "learning_rate": 7.417764066417289e-06, + "loss": 1.7148, + "step": 330 + }, + { + "epoch": 1.0852459016393443, + "grad_norm": 12.570676803588867, + "learning_rate": 7.4022478981919884e-06, + "loss": 1.6445, + "step": 331 + }, + { + "epoch": 1.0885245901639344, + "grad_norm": 14.846644401550293, + "learning_rate": 7.386701595058422e-06, + "loss": 1.6309, + "step": 332 + }, + { + "epoch": 1.0918032786885246, + "grad_norm": 12.106386184692383, + "learning_rate": 7.3711253520365995e-06, + "loss": 1.6641, + "step": 333 + }, + { + "epoch": 1.0950819672131147, + "grad_norm": 14.234429359436035, + "learning_rate": 7.355519364522117e-06, + "loss": 1.6738, + "step": 334 + }, + { + "epoch": 1.098360655737705, + "grad_norm": 13.406672477722168, + "learning_rate": 7.339883828283696e-06, + "loss": 1.6035, + "step": 335 + }, + { + "epoch": 1.1016393442622952, + "grad_norm": 12.718591690063477, + "learning_rate": 7.324218939460733e-06, + "loss": 1.6172, + "step": 336 + }, + { + "epoch": 1.1049180327868853, + "grad_norm": 12.047405242919922, + "learning_rate": 7.308524894560835e-06, + "loss": 1.6328, + "step": 337 + }, + { + "epoch": 1.1081967213114754, + "grad_norm": 10.482436180114746, + "learning_rate": 7.292801890457357e-06, + "loss": 1.5742, + "step": 338 + }, + { + "epoch": 1.1114754098360655, + "grad_norm": 13.475817680358887, + "learning_rate": 7.2770501243869315e-06, + "loss": 1.7402, + "step": 339 + }, + { + "epoch": 1.1147540983606556, + "grad_norm": 12.393983840942383, + "learning_rate": 7.261269793946994e-06, + "loss": 1.6562, + "step": 340 + }, + { + "epoch": 1.118032786885246, + "grad_norm": 12.331396102905273, + "learning_rate": 7.245461097093307e-06, + "loss": 1.6758, + "step": 341 + }, + { + "epoch": 1.1213114754098361, + "grad_norm": 14.81454086303711, + "learning_rate": 7.229624232137472e-06, + "loss": 1.6152, + "step": 342 + }, + { + "epoch": 1.1245901639344262, + "grad_norm": 15.760224342346191, + "learning_rate": 7.213759397744446e-06, + "loss": 1.6465, + "step": 343 + }, + { + "epoch": 1.1278688524590164, + "grad_norm": 12.658391952514648, + "learning_rate": 7.197866792930044e-06, + "loss": 1.6309, + "step": 344 + }, + { + "epoch": 1.1311475409836065, + "grad_norm": 13.215618133544922, + "learning_rate": 7.1819466170584485e-06, + "loss": 1.582, + "step": 345 + }, + { + "epoch": 1.1344262295081968, + "grad_norm": 17.812692642211914, + "learning_rate": 7.165999069839708e-06, + "loss": 1.6035, + "step": 346 + }, + { + "epoch": 1.137704918032787, + "grad_norm": 12.584436416625977, + "learning_rate": 7.1500243513272275e-06, + "loss": 1.6875, + "step": 347 + }, + { + "epoch": 1.140983606557377, + "grad_norm": 18.275434494018555, + "learning_rate": 7.1340226619152595e-06, + "loss": 1.6914, + "step": 348 + }, + { + "epoch": 1.1442622950819672, + "grad_norm": 13.600272178649902, + "learning_rate": 7.117994202336398e-06, + "loss": 1.6367, + "step": 349 + }, + { + "epoch": 1.1475409836065573, + "grad_norm": 16.630714416503906, + "learning_rate": 7.101939173659045e-06, + "loss": 1.6191, + "step": 350 + }, + { + "epoch": 1.1508196721311474, + "grad_norm": 15.959205627441406, + "learning_rate": 7.085857777284908e-06, + "loss": 1.7305, + "step": 351 + }, + { + "epoch": 1.1540983606557378, + "grad_norm": 13.133319854736328, + "learning_rate": 7.069750214946458e-06, + "loss": 1.6641, + "step": 352 + }, + { + "epoch": 1.157377049180328, + "grad_norm": 20.936429977416992, + "learning_rate": 7.053616688704399e-06, + "loss": 1.6738, + "step": 353 + }, + { + "epoch": 1.160655737704918, + "grad_norm": 12.605098724365234, + "learning_rate": 7.037457400945147e-06, + "loss": 1.6953, + "step": 354 + }, + { + "epoch": 1.1639344262295082, + "grad_norm": 14.746296882629395, + "learning_rate": 7.021272554378279e-06, + "loss": 1.75, + "step": 355 + }, + { + "epoch": 1.1672131147540983, + "grad_norm": 14.727060317993164, + "learning_rate": 7.005062352033991e-06, + "loss": 1.6406, + "step": 356 + }, + { + "epoch": 1.1704918032786886, + "grad_norm": 11.865753173828125, + "learning_rate": 6.988826997260556e-06, + "loss": 1.584, + "step": 357 + }, + { + "epoch": 1.1737704918032787, + "grad_norm": 12.896044731140137, + "learning_rate": 6.972566693721772e-06, + "loss": 1.6797, + "step": 358 + }, + { + "epoch": 1.1770491803278689, + "grad_norm": 12.144120216369629, + "learning_rate": 6.956281645394404e-06, + "loss": 1.6055, + "step": 359 + }, + { + "epoch": 1.180327868852459, + "grad_norm": 12.519413948059082, + "learning_rate": 6.939972056565626e-06, + "loss": 1.6895, + "step": 360 + }, + { + "epoch": 1.1836065573770491, + "grad_norm": 14.610289573669434, + "learning_rate": 6.923638131830461e-06, + "loss": 1.6582, + "step": 361 + }, + { + "epoch": 1.1868852459016392, + "grad_norm": 11.816085815429688, + "learning_rate": 6.907280076089213e-06, + "loss": 1.5996, + "step": 362 + }, + { + "epoch": 1.1901639344262296, + "grad_norm": 13.615991592407227, + "learning_rate": 6.890898094544894e-06, + "loss": 1.6504, + "step": 363 + }, + { + "epoch": 1.1934426229508197, + "grad_norm": 16.203214645385742, + "learning_rate": 6.874492392700659e-06, + "loss": 1.6758, + "step": 364 + }, + { + "epoch": 1.1967213114754098, + "grad_norm": 12.530990600585938, + "learning_rate": 6.858063176357214e-06, + "loss": 1.6738, + "step": 365 + }, + { + "epoch": 1.2, + "grad_norm": 13.96877384185791, + "learning_rate": 6.841610651610243e-06, + "loss": 1.6953, + "step": 366 + }, + { + "epoch": 1.20327868852459, + "grad_norm": 13.824159622192383, + "learning_rate": 6.825135024847825e-06, + "loss": 1.6543, + "step": 367 + }, + { + "epoch": 1.2065573770491804, + "grad_norm": 12.192745208740234, + "learning_rate": 6.808636502747841e-06, + "loss": 1.6445, + "step": 368 + }, + { + "epoch": 1.2098360655737705, + "grad_norm": 12.605790138244629, + "learning_rate": 6.792115292275374e-06, + "loss": 1.668, + "step": 369 + }, + { + "epoch": 1.2131147540983607, + "grad_norm": 15.496692657470703, + "learning_rate": 6.775571600680133e-06, + "loss": 1.6426, + "step": 370 + }, + { + "epoch": 1.2163934426229508, + "grad_norm": 13.692412376403809, + "learning_rate": 6.759005635493831e-06, + "loss": 1.7207, + "step": 371 + }, + { + "epoch": 1.219672131147541, + "grad_norm": 17.969615936279297, + "learning_rate": 6.742417604527594e-06, + "loss": 1.5938, + "step": 372 + }, + { + "epoch": 1.222950819672131, + "grad_norm": 18.373065948486328, + "learning_rate": 6.725807715869352e-06, + "loss": 1.6875, + "step": 373 + }, + { + "epoch": 1.2262295081967214, + "grad_norm": 13.26612663269043, + "learning_rate": 6.709176177881228e-06, + "loss": 1.6602, + "step": 374 + }, + { + "epoch": 1.2295081967213115, + "grad_norm": 21.26793098449707, + "learning_rate": 6.692523199196926e-06, + "loss": 1.7227, + "step": 375 + }, + { + "epoch": 1.2327868852459016, + "grad_norm": 12.17033576965332, + "learning_rate": 6.675848988719102e-06, + "loss": 1.6465, + "step": 376 + }, + { + "epoch": 1.2360655737704918, + "grad_norm": 21.347976684570312, + "learning_rate": 6.659153755616771e-06, + "loss": 1.6289, + "step": 377 + }, + { + "epoch": 1.2393442622950819, + "grad_norm": 19.131080627441406, + "learning_rate": 6.64243770932265e-06, + "loss": 1.6406, + "step": 378 + }, + { + "epoch": 1.2426229508196722, + "grad_norm": 15.474773406982422, + "learning_rate": 6.625701059530552e-06, + "loss": 1.668, + "step": 379 + }, + { + "epoch": 1.2459016393442623, + "grad_norm": 12.6391019821167, + "learning_rate": 6.608944016192754e-06, + "loss": 1.623, + "step": 380 + }, + { + "epoch": 1.2491803278688525, + "grad_norm": 12.770636558532715, + "learning_rate": 6.5921667895173515e-06, + "loss": 1.6523, + "step": 381 + }, + { + "epoch": 1.2524590163934426, + "grad_norm": 18.565126419067383, + "learning_rate": 6.575369589965632e-06, + "loss": 1.6641, + "step": 382 + }, + { + "epoch": 1.2557377049180327, + "grad_norm": 13.343095779418945, + "learning_rate": 6.558552628249435e-06, + "loss": 1.6426, + "step": 383 + }, + { + "epoch": 1.2590163934426228, + "grad_norm": 12.588839530944824, + "learning_rate": 6.541716115328503e-06, + "loss": 1.6172, + "step": 384 + }, + { + "epoch": 1.2622950819672132, + "grad_norm": 13.01188850402832, + "learning_rate": 6.524860262407835e-06, + "loss": 1.5957, + "step": 385 + }, + { + "epoch": 1.2655737704918033, + "grad_norm": 11.886181831359863, + "learning_rate": 6.507985280935047e-06, + "loss": 1.6445, + "step": 386 + }, + { + "epoch": 1.2688524590163934, + "grad_norm": 18.22747039794922, + "learning_rate": 6.491091382597705e-06, + "loss": 1.5781, + "step": 387 + }, + { + "epoch": 1.2721311475409836, + "grad_norm": 11.928472518920898, + "learning_rate": 6.4741787793206824e-06, + "loss": 1.5645, + "step": 388 + }, + { + "epoch": 1.275409836065574, + "grad_norm": 15.313958168029785, + "learning_rate": 6.457247683263494e-06, + "loss": 1.6289, + "step": 389 + }, + { + "epoch": 1.278688524590164, + "grad_norm": 14.239816665649414, + "learning_rate": 6.440298306817637e-06, + "loss": 1.6094, + "step": 390 + }, + { + "epoch": 1.2819672131147541, + "grad_norm": 21.43814468383789, + "learning_rate": 6.423330862603924e-06, + "loss": 1.6113, + "step": 391 + }, + { + "epoch": 1.2852459016393443, + "grad_norm": 14.304780006408691, + "learning_rate": 6.406345563469819e-06, + "loss": 1.625, + "step": 392 + }, + { + "epoch": 1.2885245901639344, + "grad_norm": 15.105829238891602, + "learning_rate": 6.38934262248677e-06, + "loss": 1.5078, + "step": 393 + }, + { + "epoch": 1.2918032786885245, + "grad_norm": 13.64173412322998, + "learning_rate": 6.372322252947529e-06, + "loss": 1.6094, + "step": 394 + }, + { + "epoch": 1.2950819672131146, + "grad_norm": 12.446194648742676, + "learning_rate": 6.355284668363479e-06, + "loss": 1.6172, + "step": 395 + }, + { + "epoch": 1.298360655737705, + "grad_norm": 12.818573951721191, + "learning_rate": 6.3382300824619584e-06, + "loss": 1.623, + "step": 396 + }, + { + "epoch": 1.301639344262295, + "grad_norm": 16.565000534057617, + "learning_rate": 6.321158709183579e-06, + "loss": 1.6016, + "step": 397 + }, + { + "epoch": 1.3049180327868852, + "grad_norm": 12.29269027709961, + "learning_rate": 6.30407076267954e-06, + "loss": 1.666, + "step": 398 + }, + { + "epoch": 1.3081967213114754, + "grad_norm": 13.712056159973145, + "learning_rate": 6.286966457308941e-06, + "loss": 1.6289, + "step": 399 + }, + { + "epoch": 1.3114754098360657, + "grad_norm": 12.996943473815918, + "learning_rate": 6.2698460076360966e-06, + "loss": 1.6035, + "step": 400 + }, + { + "epoch": 1.3147540983606558, + "grad_norm": 12.406416893005371, + "learning_rate": 6.252709628427842e-06, + "loss": 1.6953, + "step": 401 + }, + { + "epoch": 1.318032786885246, + "grad_norm": 14.069875717163086, + "learning_rate": 6.235557534650844e-06, + "loss": 1.6367, + "step": 402 + }, + { + "epoch": 1.321311475409836, + "grad_norm": 12.035994529724121, + "learning_rate": 6.218389941468894e-06, + "loss": 1.7168, + "step": 403 + }, + { + "epoch": 1.3245901639344262, + "grad_norm": 13.915684700012207, + "learning_rate": 6.20120706424022e-06, + "loss": 1.6953, + "step": 404 + }, + { + "epoch": 1.3278688524590163, + "grad_norm": 13.748454093933105, + "learning_rate": 6.184009118514776e-06, + "loss": 1.7012, + "step": 405 + }, + { + "epoch": 1.3311475409836064, + "grad_norm": 12.660347938537598, + "learning_rate": 6.166796320031545e-06, + "loss": 1.5762, + "step": 406 + }, + { + "epoch": 1.3344262295081968, + "grad_norm": 13.897893905639648, + "learning_rate": 6.149568884715831e-06, + "loss": 1.6816, + "step": 407 + }, + { + "epoch": 1.337704918032787, + "grad_norm": 15.359755516052246, + "learning_rate": 6.132327028676544e-06, + "loss": 1.5742, + "step": 408 + }, + { + "epoch": 1.340983606557377, + "grad_norm": 15.530616760253906, + "learning_rate": 6.1150709682034995e-06, + "loss": 1.6094, + "step": 409 + }, + { + "epoch": 1.3442622950819672, + "grad_norm": 13.77632999420166, + "learning_rate": 6.097800919764698e-06, + "loss": 1.6895, + "step": 410 + }, + { + "epoch": 1.3475409836065575, + "grad_norm": 14.026497840881348, + "learning_rate": 6.080517100003611e-06, + "loss": 1.6523, + "step": 411 + }, + { + "epoch": 1.3508196721311476, + "grad_norm": 11.663541793823242, + "learning_rate": 6.063219725736468e-06, + "loss": 1.6328, + "step": 412 + }, + { + "epoch": 1.3540983606557377, + "grad_norm": 13.371345520019531, + "learning_rate": 6.0459090139495235e-06, + "loss": 1.7012, + "step": 413 + }, + { + "epoch": 1.3573770491803279, + "grad_norm": 14.87950325012207, + "learning_rate": 6.028585181796351e-06, + "loss": 1.6211, + "step": 414 + }, + { + "epoch": 1.360655737704918, + "grad_norm": 13.860067367553711, + "learning_rate": 6.011248446595113e-06, + "loss": 1.6816, + "step": 415 + }, + { + "epoch": 1.3639344262295081, + "grad_norm": 20.906126022338867, + "learning_rate": 5.99389902582583e-06, + "loss": 1.6582, + "step": 416 + }, + { + "epoch": 1.3672131147540982, + "grad_norm": 13.586402893066406, + "learning_rate": 5.9765371371276544e-06, + "loss": 1.625, + "step": 417 + }, + { + "epoch": 1.3704918032786886, + "grad_norm": 19.022512435913086, + "learning_rate": 5.959162998296149e-06, + "loss": 1.6875, + "step": 418 + }, + { + "epoch": 1.3737704918032787, + "grad_norm": 15.78264045715332, + "learning_rate": 5.941776827280544e-06, + "loss": 1.5957, + "step": 419 + }, + { + "epoch": 1.3770491803278688, + "grad_norm": 13.40628433227539, + "learning_rate": 5.924378842181002e-06, + "loss": 1.6582, + "step": 420 + }, + { + "epoch": 1.380327868852459, + "grad_norm": 16.419567108154297, + "learning_rate": 5.906969261245896e-06, + "loss": 1.6055, + "step": 421 + }, + { + "epoch": 1.3836065573770493, + "grad_norm": 14.798409461975098, + "learning_rate": 5.889548302869057e-06, + "loss": 1.5508, + "step": 422 + }, + { + "epoch": 1.3868852459016394, + "grad_norm": 11.23767375946045, + "learning_rate": 5.87211618558704e-06, + "loss": 1.5664, + "step": 423 + }, + { + "epoch": 1.3901639344262295, + "grad_norm": 14.773558616638184, + "learning_rate": 5.8546731280763846e-06, + "loss": 1.6973, + "step": 424 + }, + { + "epoch": 1.3934426229508197, + "grad_norm": 12.42011833190918, + "learning_rate": 5.83721934915087e-06, + "loss": 1.6016, + "step": 425 + }, + { + "epoch": 1.3967213114754098, + "grad_norm": 13.847085952758789, + "learning_rate": 5.819755067758767e-06, + "loss": 1.7383, + "step": 426 + }, + { + "epoch": 1.4, + "grad_norm": 14.425787925720215, + "learning_rate": 5.802280502980095e-06, + "loss": 1.6406, + "step": 427 + }, + { + "epoch": 1.40327868852459, + "grad_norm": 17.782608032226562, + "learning_rate": 5.784795874023876e-06, + "loss": 1.6113, + "step": 428 + }, + { + "epoch": 1.4065573770491804, + "grad_norm": 12.806195259094238, + "learning_rate": 5.767301400225378e-06, + "loss": 1.6465, + "step": 429 + }, + { + "epoch": 1.4098360655737705, + "grad_norm": 12.222467422485352, + "learning_rate": 5.749797301043366e-06, + "loss": 1.6016, + "step": 430 + }, + { + "epoch": 1.4131147540983606, + "grad_norm": 23.587997436523438, + "learning_rate": 5.732283796057356e-06, + "loss": 1.6426, + "step": 431 + }, + { + "epoch": 1.4163934426229507, + "grad_norm": 15.535636901855469, + "learning_rate": 5.714761104964852e-06, + "loss": 1.6348, + "step": 432 + }, + { + "epoch": 1.419672131147541, + "grad_norm": 24.00181770324707, + "learning_rate": 5.69722944757859e-06, + "loss": 1.6445, + "step": 433 + }, + { + "epoch": 1.4229508196721312, + "grad_norm": 22.185400009155273, + "learning_rate": 5.679689043823787e-06, + "loss": 1.6016, + "step": 434 + }, + { + "epoch": 1.4262295081967213, + "grad_norm": 15.497023582458496, + "learning_rate": 5.6621401137353784e-06, + "loss": 1.6172, + "step": 435 + }, + { + "epoch": 1.4295081967213115, + "grad_norm": 19.018062591552734, + "learning_rate": 5.644582877455254e-06, + "loss": 1.5918, + "step": 436 + }, + { + "epoch": 1.4327868852459016, + "grad_norm": 16.216062545776367, + "learning_rate": 5.6270175552295035e-06, + "loss": 1.6445, + "step": 437 + }, + { + "epoch": 1.4360655737704917, + "grad_norm": 17.771896362304688, + "learning_rate": 5.609444367405652e-06, + "loss": 1.6133, + "step": 438 + }, + { + "epoch": 1.4393442622950818, + "grad_norm": 15.940698623657227, + "learning_rate": 5.5918635344298885e-06, + "loss": 1.6914, + "step": 439 + }, + { + "epoch": 1.4426229508196722, + "grad_norm": 13.144735336303711, + "learning_rate": 5.574275276844315e-06, + "loss": 1.6055, + "step": 440 + }, + { + "epoch": 1.4459016393442623, + "grad_norm": 22.014144897460938, + "learning_rate": 5.556679815284162e-06, + "loss": 1.5879, + "step": 441 + }, + { + "epoch": 1.4491803278688524, + "grad_norm": 15.573799133300781, + "learning_rate": 5.5390773704750355e-06, + "loss": 1.6191, + "step": 442 + }, + { + "epoch": 1.4524590163934425, + "grad_norm": 14.217083930969238, + "learning_rate": 5.5214681632301425e-06, + "loss": 1.5469, + "step": 443 + }, + { + "epoch": 1.455737704918033, + "grad_norm": 15.9507417678833, + "learning_rate": 5.503852414447519e-06, + "loss": 1.6602, + "step": 444 + }, + { + "epoch": 1.459016393442623, + "grad_norm": 12.370546340942383, + "learning_rate": 5.486230345107262e-06, + "loss": 1.6348, + "step": 445 + }, + { + "epoch": 1.4622950819672131, + "grad_norm": 18.62546730041504, + "learning_rate": 5.468602176268756e-06, + "loss": 1.5859, + "step": 446 + }, + { + "epoch": 1.4655737704918033, + "grad_norm": 13.223871231079102, + "learning_rate": 5.4509681290679014e-06, + "loss": 1.5605, + "step": 447 + }, + { + "epoch": 1.4688524590163934, + "grad_norm": 14.363565444946289, + "learning_rate": 5.43332842471434e-06, + "loss": 1.6504, + "step": 448 + }, + { + "epoch": 1.4721311475409835, + "grad_norm": 13.501849174499512, + "learning_rate": 5.4156832844886775e-06, + "loss": 1.6016, + "step": 449 + }, + { + "epoch": 1.4754098360655736, + "grad_norm": 13.610998153686523, + "learning_rate": 5.398032929739712e-06, + "loss": 1.6133, + "step": 450 + }, + { + "epoch": 1.478688524590164, + "grad_norm": 18.596263885498047, + "learning_rate": 5.380377581881653e-06, + "loss": 1.5664, + "step": 451 + }, + { + "epoch": 1.481967213114754, + "grad_norm": 15.4996919631958, + "learning_rate": 5.362717462391345e-06, + "loss": 1.5176, + "step": 452 + }, + { + "epoch": 1.4852459016393442, + "grad_norm": 11.770246505737305, + "learning_rate": 5.345052792805494e-06, + "loss": 1.4805, + "step": 453 + }, + { + "epoch": 1.4885245901639343, + "grad_norm": 18.300512313842773, + "learning_rate": 5.327383794717881e-06, + "loss": 1.6074, + "step": 454 + }, + { + "epoch": 1.4918032786885247, + "grad_norm": 18.962173461914062, + "learning_rate": 5.309710689776587e-06, + "loss": 1.5898, + "step": 455 + }, + { + "epoch": 1.4950819672131148, + "grad_norm": 12.80952262878418, + "learning_rate": 5.2920336996812125e-06, + "loss": 1.5762, + "step": 456 + }, + { + "epoch": 1.498360655737705, + "grad_norm": 22.957740783691406, + "learning_rate": 5.274353046180093e-06, + "loss": 1.5918, + "step": 457 + }, + { + "epoch": 1.501639344262295, + "grad_norm": 13.40424633026123, + "learning_rate": 5.256668951067521e-06, + "loss": 1.6406, + "step": 458 + }, + { + "epoch": 1.5049180327868852, + "grad_norm": 13.17849349975586, + "learning_rate": 5.23898163618096e-06, + "loss": 1.7012, + "step": 459 + }, + { + "epoch": 1.5081967213114753, + "grad_norm": 17.233837127685547, + "learning_rate": 5.221291323398267e-06, + "loss": 1.6543, + "step": 460 + }, + { + "epoch": 1.5114754098360654, + "grad_norm": 12.22879695892334, + "learning_rate": 5.203598234634903e-06, + "loss": 1.6133, + "step": 461 + }, + { + "epoch": 1.5147540983606558, + "grad_norm": 22.00145721435547, + "learning_rate": 5.185902591841152e-06, + "loss": 1.6191, + "step": 462 + }, + { + "epoch": 1.518032786885246, + "grad_norm": 15.629786491394043, + "learning_rate": 5.168204616999342e-06, + "loss": 1.5605, + "step": 463 + }, + { + "epoch": 1.521311475409836, + "grad_norm": 14.949792861938477, + "learning_rate": 5.15050453212105e-06, + "loss": 1.6094, + "step": 464 + }, + { + "epoch": 1.5245901639344264, + "grad_norm": 18.31121826171875, + "learning_rate": 5.132802559244326e-06, + "loss": 1.5098, + "step": 465 + }, + { + "epoch": 1.5278688524590165, + "grad_norm": 12.61305046081543, + "learning_rate": 5.115098920430897e-06, + "loss": 1.5137, + "step": 466 + }, + { + "epoch": 1.5311475409836066, + "grad_norm": 15.578408241271973, + "learning_rate": 5.097393837763398e-06, + "loss": 1.6055, + "step": 467 + }, + { + "epoch": 1.5344262295081967, + "grad_norm": 18.71062469482422, + "learning_rate": 5.079687533342569e-06, + "loss": 1.6172, + "step": 468 + }, + { + "epoch": 1.5377049180327869, + "grad_norm": 13.096431732177734, + "learning_rate": 5.061980229284481e-06, + "loss": 1.6094, + "step": 469 + }, + { + "epoch": 1.540983606557377, + "grad_norm": 18.373483657836914, + "learning_rate": 5.04427214771774e-06, + "loss": 1.6484, + "step": 470 + }, + { + "epoch": 1.544262295081967, + "grad_norm": 13.364453315734863, + "learning_rate": 5.02656351078071e-06, + "loss": 1.6035, + "step": 471 + }, + { + "epoch": 1.5475409836065572, + "grad_norm": 17.02009391784668, + "learning_rate": 5.008854540618721e-06, + "loss": 1.6191, + "step": 472 + }, + { + "epoch": 1.5508196721311476, + "grad_norm": 18.287324905395508, + "learning_rate": 4.991145459381279e-06, + "loss": 1.6445, + "step": 473 + }, + { + "epoch": 1.5540983606557377, + "grad_norm": 13.834267616271973, + "learning_rate": 4.97343648921929e-06, + "loss": 1.6582, + "step": 474 + }, + { + "epoch": 1.5573770491803278, + "grad_norm": 15.57867431640625, + "learning_rate": 4.955727852282261e-06, + "loss": 1.6699, + "step": 475 + }, + { + "epoch": 1.5606557377049182, + "grad_norm": 15.531145095825195, + "learning_rate": 4.938019770715521e-06, + "loss": 1.4805, + "step": 476 + }, + { + "epoch": 1.5639344262295083, + "grad_norm": 20.70842742919922, + "learning_rate": 4.9203124666574325e-06, + "loss": 1.5957, + "step": 477 + }, + { + "epoch": 1.5672131147540984, + "grad_norm": 16.70767593383789, + "learning_rate": 4.902606162236605e-06, + "loss": 1.5566, + "step": 478 + }, + { + "epoch": 1.5704918032786885, + "grad_norm": 12.975936889648438, + "learning_rate": 4.8849010795691035e-06, + "loss": 1.6016, + "step": 479 + }, + { + "epoch": 1.5737704918032787, + "grad_norm": 16.035783767700195, + "learning_rate": 4.867197440755677e-06, + "loss": 1.5273, + "step": 480 + }, + { + "epoch": 1.5770491803278688, + "grad_norm": 25.241321563720703, + "learning_rate": 4.849495467878951e-06, + "loss": 1.6172, + "step": 481 + }, + { + "epoch": 1.580327868852459, + "grad_norm": 13.46953296661377, + "learning_rate": 4.831795383000659e-06, + "loss": 1.5547, + "step": 482 + }, + { + "epoch": 1.583606557377049, + "grad_norm": 24.528270721435547, + "learning_rate": 4.81409740815885e-06, + "loss": 1.5859, + "step": 483 + }, + { + "epoch": 1.5868852459016394, + "grad_norm": 24.17436981201172, + "learning_rate": 4.796401765365101e-06, + "loss": 1.5664, + "step": 484 + }, + { + "epoch": 1.5901639344262295, + "grad_norm": 13.636819839477539, + "learning_rate": 4.778708676601736e-06, + "loss": 1.6641, + "step": 485 + }, + { + "epoch": 1.5934426229508196, + "grad_norm": 17.118165969848633, + "learning_rate": 4.761018363819041e-06, + "loss": 1.6348, + "step": 486 + }, + { + "epoch": 1.59672131147541, + "grad_norm": 21.050155639648438, + "learning_rate": 4.743331048932481e-06, + "loss": 1.5957, + "step": 487 + }, + { + "epoch": 1.6, + "grad_norm": 14.285764694213867, + "learning_rate": 4.725646953819908e-06, + "loss": 1.6621, + "step": 488 + }, + { + "epoch": 1.6032786885245902, + "grad_norm": 19.979021072387695, + "learning_rate": 4.707966300318788e-06, + "loss": 1.6328, + "step": 489 + }, + { + "epoch": 1.6065573770491803, + "grad_norm": 24.40378189086914, + "learning_rate": 4.6902893102234145e-06, + "loss": 1.6113, + "step": 490 + }, + { + "epoch": 1.6098360655737705, + "grad_norm": 17.49124526977539, + "learning_rate": 4.6726162052821214e-06, + "loss": 1.6816, + "step": 491 + }, + { + "epoch": 1.6131147540983606, + "grad_norm": 18.845081329345703, + "learning_rate": 4.654947207194507e-06, + "loss": 1.6562, + "step": 492 + }, + { + "epoch": 1.6163934426229507, + "grad_norm": 24.073442459106445, + "learning_rate": 4.637282537608656e-06, + "loss": 1.6934, + "step": 493 + }, + { + "epoch": 1.6196721311475408, + "grad_norm": 13.014028549194336, + "learning_rate": 4.619622418118349e-06, + "loss": 1.5996, + "step": 494 + }, + { + "epoch": 1.6229508196721312, + "grad_norm": 15.70720100402832, + "learning_rate": 4.601967070260289e-06, + "loss": 1.5918, + "step": 495 + }, + { + "epoch": 1.6262295081967213, + "grad_norm": 20.249507904052734, + "learning_rate": 4.584316715511324e-06, + "loss": 1.6152, + "step": 496 + }, + { + "epoch": 1.6295081967213116, + "grad_norm": 13.594206809997559, + "learning_rate": 4.5666715752856624e-06, + "loss": 1.582, + "step": 497 + }, + { + "epoch": 1.6327868852459018, + "grad_norm": 14.592798233032227, + "learning_rate": 4.549031870932101e-06, + "loss": 1.5449, + "step": 498 + }, + { + "epoch": 1.6360655737704919, + "grad_norm": 16.72430419921875, + "learning_rate": 4.531397823731245e-06, + "loss": 1.5566, + "step": 499 + }, + { + "epoch": 1.639344262295082, + "grad_norm": 13.189786911010742, + "learning_rate": 4.51376965489274e-06, + "loss": 1.5977, + "step": 500 + }, + { + "epoch": 1.6426229508196721, + "grad_norm": 17.594057083129883, + "learning_rate": 4.4961475855524825e-06, + "loss": 1.6211, + "step": 501 + }, + { + "epoch": 1.6459016393442623, + "grad_norm": 14.489039421081543, + "learning_rate": 4.478531836769859e-06, + "loss": 1.5957, + "step": 502 + }, + { + "epoch": 1.6491803278688524, + "grad_norm": 13.191361427307129, + "learning_rate": 4.460922629524966e-06, + "loss": 1.5879, + "step": 503 + }, + { + "epoch": 1.6524590163934425, + "grad_norm": 17.483108520507812, + "learning_rate": 4.44332018471584e-06, + "loss": 1.5723, + "step": 504 + }, + { + "epoch": 1.6557377049180326, + "grad_norm": 14.9766206741333, + "learning_rate": 4.425724723155686e-06, + "loss": 1.6211, + "step": 505 + }, + { + "epoch": 1.659016393442623, + "grad_norm": 18.884735107421875, + "learning_rate": 4.408136465570112e-06, + "loss": 1.5801, + "step": 506 + }, + { + "epoch": 1.662295081967213, + "grad_norm": 18.028215408325195, + "learning_rate": 4.39055563259435e-06, + "loss": 1.6094, + "step": 507 + }, + { + "epoch": 1.6655737704918034, + "grad_norm": 12.109418869018555, + "learning_rate": 4.372982444770498e-06, + "loss": 1.6016, + "step": 508 + }, + { + "epoch": 1.6688524590163936, + "grad_norm": 15.456534385681152, + "learning_rate": 4.355417122544748e-06, + "loss": 1.5996, + "step": 509 + }, + { + "epoch": 1.6721311475409837, + "grad_norm": 17.789567947387695, + "learning_rate": 4.337859886264624e-06, + "loss": 1.6367, + "step": 510 + }, + { + "epoch": 1.6754098360655738, + "grad_norm": 12.670686721801758, + "learning_rate": 4.320310956176214e-06, + "loss": 1.5723, + "step": 511 + }, + { + "epoch": 1.678688524590164, + "grad_norm": 20.38010597229004, + "learning_rate": 4.302770552421411e-06, + "loss": 1.5449, + "step": 512 + }, + { + "epoch": 1.681967213114754, + "grad_norm": 16.769620895385742, + "learning_rate": 4.2852388950351496e-06, + "loss": 1.6328, + "step": 513 + }, + { + "epoch": 1.6852459016393442, + "grad_norm": 12.062372207641602, + "learning_rate": 4.2677162039426454e-06, + "loss": 1.6055, + "step": 514 + }, + { + "epoch": 1.6885245901639343, + "grad_norm": 16.456920623779297, + "learning_rate": 4.2502026989566354e-06, + "loss": 1.5801, + "step": 515 + }, + { + "epoch": 1.6918032786885244, + "grad_norm": 13.964550018310547, + "learning_rate": 4.232698599774625e-06, + "loss": 1.5762, + "step": 516 + }, + { + "epoch": 1.6950819672131148, + "grad_norm": 13.01318645477295, + "learning_rate": 4.215204125976126e-06, + "loss": 1.5859, + "step": 517 + }, + { + "epoch": 1.698360655737705, + "grad_norm": 16.137088775634766, + "learning_rate": 4.197719497019906e-06, + "loss": 1.6953, + "step": 518 + }, + { + "epoch": 1.7016393442622952, + "grad_norm": 13.629316329956055, + "learning_rate": 4.180244932241234e-06, + "loss": 1.582, + "step": 519 + }, + { + "epoch": 1.7049180327868854, + "grad_norm": 12.793038368225098, + "learning_rate": 4.162780650849131e-06, + "loss": 1.5332, + "step": 520 + }, + { + "epoch": 1.7081967213114755, + "grad_norm": 12.091232299804688, + "learning_rate": 4.145326871923616e-06, + "loss": 1.5898, + "step": 521 + }, + { + "epoch": 1.7114754098360656, + "grad_norm": 14.179364204406738, + "learning_rate": 4.1278838144129614e-06, + "loss": 1.623, + "step": 522 + }, + { + "epoch": 1.7147540983606557, + "grad_norm": 13.386200904846191, + "learning_rate": 4.110451697130946e-06, + "loss": 1.541, + "step": 523 + }, + { + "epoch": 1.7180327868852459, + "grad_norm": 13.438315391540527, + "learning_rate": 4.093030738754105e-06, + "loss": 1.5527, + "step": 524 + }, + { + "epoch": 1.721311475409836, + "grad_norm": 14.931065559387207, + "learning_rate": 4.0756211578189995e-06, + "loss": 1.6602, + "step": 525 + }, + { + "epoch": 1.724590163934426, + "grad_norm": 12.362815856933594, + "learning_rate": 4.058223172719459e-06, + "loss": 1.5801, + "step": 526 + }, + { + "epoch": 1.7278688524590164, + "grad_norm": 14.543976783752441, + "learning_rate": 4.0408370017038515e-06, + "loss": 1.5, + "step": 527 + }, + { + "epoch": 1.7311475409836066, + "grad_norm": 16.341663360595703, + "learning_rate": 4.023462862872346e-06, + "loss": 1.5957, + "step": 528 + }, + { + "epoch": 1.7344262295081967, + "grad_norm": 13.109537124633789, + "learning_rate": 4.006100974174173e-06, + "loss": 1.6426, + "step": 529 + }, + { + "epoch": 1.737704918032787, + "grad_norm": 18.410266876220703, + "learning_rate": 3.9887515534048866e-06, + "loss": 1.5898, + "step": 530 + }, + { + "epoch": 1.7409836065573772, + "grad_norm": 12.593286514282227, + "learning_rate": 3.971414818203648e-06, + "loss": 1.625, + "step": 531 + }, + { + "epoch": 1.7442622950819673, + "grad_norm": 14.592744827270508, + "learning_rate": 3.954090986050477e-06, + "loss": 1.5918, + "step": 532 + }, + { + "epoch": 1.7475409836065574, + "grad_norm": 13.23892593383789, + "learning_rate": 3.9367802742635335e-06, + "loss": 1.6504, + "step": 533 + }, + { + "epoch": 1.7508196721311475, + "grad_norm": 13.610597610473633, + "learning_rate": 3.9194828999963905e-06, + "loss": 1.5742, + "step": 534 + }, + { + "epoch": 1.7540983606557377, + "grad_norm": 12.243042945861816, + "learning_rate": 3.902199080235305e-06, + "loss": 1.5625, + "step": 535 + }, + { + "epoch": 1.7573770491803278, + "grad_norm": 13.913399696350098, + "learning_rate": 3.884929031796504e-06, + "loss": 1.5547, + "step": 536 + }, + { + "epoch": 1.760655737704918, + "grad_norm": 12.08027458190918, + "learning_rate": 3.867672971323457e-06, + "loss": 1.543, + "step": 537 + }, + { + "epoch": 1.7639344262295082, + "grad_norm": 13.890872955322266, + "learning_rate": 3.850431115284171e-06, + "loss": 1.5391, + "step": 538 + }, + { + "epoch": 1.7672131147540984, + "grad_norm": 13.448533058166504, + "learning_rate": 3.8332036799684554e-06, + "loss": 1.5547, + "step": 539 + }, + { + "epoch": 1.7704918032786885, + "grad_norm": 12.864797592163086, + "learning_rate": 3.815990881485225e-06, + "loss": 1.5625, + "step": 540 + }, + { + "epoch": 1.7737704918032788, + "grad_norm": 13.268163681030273, + "learning_rate": 3.798792935759782e-06, + "loss": 1.5684, + "step": 541 + }, + { + "epoch": 1.777049180327869, + "grad_norm": 12.976286888122559, + "learning_rate": 3.781610058531107e-06, + "loss": 1.5996, + "step": 542 + }, + { + "epoch": 1.780327868852459, + "grad_norm": 14.119148254394531, + "learning_rate": 3.7644424653491564e-06, + "loss": 1.5918, + "step": 543 + }, + { + "epoch": 1.7836065573770492, + "grad_norm": 14.292675018310547, + "learning_rate": 3.7472903715721576e-06, + "loss": 1.5098, + "step": 544 + }, + { + "epoch": 1.7868852459016393, + "grad_norm": 13.508952140808105, + "learning_rate": 3.730153992363905e-06, + "loss": 1.5176, + "step": 545 + }, + { + "epoch": 1.7901639344262295, + "grad_norm": 13.02145767211914, + "learning_rate": 3.713033542691061e-06, + "loss": 1.6113, + "step": 546 + }, + { + "epoch": 1.7934426229508196, + "grad_norm": 12.737781524658203, + "learning_rate": 3.6959292373204615e-06, + "loss": 1.623, + "step": 547 + }, + { + "epoch": 1.7967213114754097, + "grad_norm": 12.331480026245117, + "learning_rate": 3.678841290816422e-06, + "loss": 1.4766, + "step": 548 + }, + { + "epoch": 1.8, + "grad_norm": 12.874061584472656, + "learning_rate": 3.661769917538043e-06, + "loss": 1.6211, + "step": 549 + }, + { + "epoch": 1.8032786885245902, + "grad_norm": 12.349608421325684, + "learning_rate": 3.6447153316365226e-06, + "loss": 1.5898, + "step": 550 + }, + { + "epoch": 1.8065573770491803, + "grad_norm": 13.023985862731934, + "learning_rate": 3.6276777470524727e-06, + "loss": 1.5469, + "step": 551 + }, + { + "epoch": 1.8098360655737706, + "grad_norm": 15.1737699508667, + "learning_rate": 3.6106573775132315e-06, + "loss": 1.5918, + "step": 552 + }, + { + "epoch": 1.8131147540983608, + "grad_norm": 14.11626148223877, + "learning_rate": 3.5936544365301824e-06, + "loss": 1.5684, + "step": 553 + }, + { + "epoch": 1.8163934426229509, + "grad_norm": 15.703447341918945, + "learning_rate": 3.5766691373960787e-06, + "loss": 1.6035, + "step": 554 + }, + { + "epoch": 1.819672131147541, + "grad_norm": 13.144828796386719, + "learning_rate": 3.559701693182366e-06, + "loss": 1.5723, + "step": 555 + }, + { + "epoch": 1.8229508196721311, + "grad_norm": 16.379074096679688, + "learning_rate": 3.542752316736506e-06, + "loss": 1.6133, + "step": 556 + }, + { + "epoch": 1.8262295081967213, + "grad_norm": 18.03300666809082, + "learning_rate": 3.5258212206793184e-06, + "loss": 1.582, + "step": 557 + }, + { + "epoch": 1.8295081967213114, + "grad_norm": 12.72996997833252, + "learning_rate": 3.5089086174022964e-06, + "loss": 1.6328, + "step": 558 + }, + { + "epoch": 1.8327868852459015, + "grad_norm": 17.85942840576172, + "learning_rate": 3.4920147190649555e-06, + "loss": 1.6523, + "step": 559 + }, + { + "epoch": 1.8360655737704918, + "grad_norm": 21.671913146972656, + "learning_rate": 3.475139737592167e-06, + "loss": 1.5801, + "step": 560 + }, + { + "epoch": 1.839344262295082, + "grad_norm": 13.127819061279297, + "learning_rate": 3.4582838846715e-06, + "loss": 1.5469, + "step": 561 + }, + { + "epoch": 1.842622950819672, + "grad_norm": 18.55519676208496, + "learning_rate": 3.441447371750566e-06, + "loss": 1.5176, + "step": 562 + }, + { + "epoch": 1.8459016393442624, + "grad_norm": 21.33066177368164, + "learning_rate": 3.4246304100343686e-06, + "loss": 1.5566, + "step": 563 + }, + { + "epoch": 1.8491803278688526, + "grad_norm": 16.495615005493164, + "learning_rate": 3.4078332104826506e-06, + "loss": 1.543, + "step": 564 + }, + { + "epoch": 1.8524590163934427, + "grad_norm": 13.102204322814941, + "learning_rate": 3.3910559838072475e-06, + "loss": 1.6289, + "step": 565 + }, + { + "epoch": 1.8557377049180328, + "grad_norm": 21.321346282958984, + "learning_rate": 3.374298940469448e-06, + "loss": 1.6191, + "step": 566 + }, + { + "epoch": 1.859016393442623, + "grad_norm": 21.263696670532227, + "learning_rate": 3.357562290677352e-06, + "loss": 1.5312, + "step": 567 + }, + { + "epoch": 1.862295081967213, + "grad_norm": 14.131853103637695, + "learning_rate": 3.340846244383231e-06, + "loss": 1.5762, + "step": 568 + }, + { + "epoch": 1.8655737704918032, + "grad_norm": 17.478233337402344, + "learning_rate": 3.324151011280897e-06, + "loss": 1.543, + "step": 569 + }, + { + "epoch": 1.8688524590163933, + "grad_norm": 19.541519165039062, + "learning_rate": 3.307476800803077e-06, + "loss": 1.5312, + "step": 570 + }, + { + "epoch": 1.8721311475409836, + "grad_norm": 18.075885772705078, + "learning_rate": 3.2908238221187736e-06, + "loss": 1.6348, + "step": 571 + }, + { + "epoch": 1.8754098360655738, + "grad_norm": 12.494379043579102, + "learning_rate": 3.27419228413065e-06, + "loss": 1.5527, + "step": 572 + }, + { + "epoch": 1.8786885245901639, + "grad_norm": 20.536605834960938, + "learning_rate": 3.257582395472408e-06, + "loss": 1.5508, + "step": 573 + }, + { + "epoch": 1.8819672131147542, + "grad_norm": 16.939430236816406, + "learning_rate": 3.240994364506172e-06, + "loss": 1.5449, + "step": 574 + }, + { + "epoch": 1.8852459016393444, + "grad_norm": 14.016660690307617, + "learning_rate": 3.224428399319868e-06, + "loss": 1.5352, + "step": 575 + }, + { + "epoch": 1.8885245901639345, + "grad_norm": 13.903589248657227, + "learning_rate": 3.207884707724626e-06, + "loss": 1.4844, + "step": 576 + }, + { + "epoch": 1.8918032786885246, + "grad_norm": 20.456378936767578, + "learning_rate": 3.1913634972521614e-06, + "loss": 1.6328, + "step": 577 + }, + { + "epoch": 1.8950819672131147, + "grad_norm": 22.15937042236328, + "learning_rate": 3.1748649751521754e-06, + "loss": 1.5977, + "step": 578 + }, + { + "epoch": 1.8983606557377048, + "grad_norm": 13.084799766540527, + "learning_rate": 3.158389348389758e-06, + "loss": 1.5273, + "step": 579 + }, + { + "epoch": 1.901639344262295, + "grad_norm": 21.793209075927734, + "learning_rate": 3.1419368236427894e-06, + "loss": 1.6094, + "step": 580 + }, + { + "epoch": 1.904918032786885, + "grad_norm": 24.526473999023438, + "learning_rate": 3.125507607299341e-06, + "loss": 1.5508, + "step": 581 + }, + { + "epoch": 1.9081967213114754, + "grad_norm": 18.045026779174805, + "learning_rate": 3.1091019054551053e-06, + "loss": 1.5508, + "step": 582 + }, + { + "epoch": 1.9114754098360656, + "grad_norm": 16.46772575378418, + "learning_rate": 3.0927199239107887e-06, + "loss": 1.5957, + "step": 583 + }, + { + "epoch": 1.9147540983606557, + "grad_norm": 19.109159469604492, + "learning_rate": 3.0763618681695405e-06, + "loss": 1.5059, + "step": 584 + }, + { + "epoch": 1.918032786885246, + "grad_norm": 18.3557186126709, + "learning_rate": 3.060027943434376e-06, + "loss": 1.5645, + "step": 585 + }, + { + "epoch": 1.9213114754098362, + "grad_norm": 13.355094909667969, + "learning_rate": 3.0437183546055977e-06, + "loss": 1.5508, + "step": 586 + }, + { + "epoch": 1.9245901639344263, + "grad_norm": 20.06283950805664, + "learning_rate": 3.027433306278229e-06, + "loss": 1.6016, + "step": 587 + }, + { + "epoch": 1.9278688524590164, + "grad_norm": 24.55561065673828, + "learning_rate": 3.0111730027394443e-06, + "loss": 1.5703, + "step": 588 + }, + { + "epoch": 1.9311475409836065, + "grad_norm": 16.579423904418945, + "learning_rate": 2.9949376479660104e-06, + "loss": 1.4688, + "step": 589 + }, + { + "epoch": 1.9344262295081966, + "grad_norm": 12.7395658493042, + "learning_rate": 2.9787274456217225e-06, + "loss": 1.5918, + "step": 590 + }, + { + "epoch": 1.9377049180327868, + "grad_norm": 13.82606029510498, + "learning_rate": 2.962542599054854e-06, + "loss": 1.5332, + "step": 591 + }, + { + "epoch": 1.940983606557377, + "grad_norm": 23.461950302124023, + "learning_rate": 2.9463833112956026e-06, + "loss": 1.5234, + "step": 592 + }, + { + "epoch": 1.9442622950819672, + "grad_norm": 14.912766456604004, + "learning_rate": 2.9302497850535447e-06, + "loss": 1.6074, + "step": 593 + }, + { + "epoch": 1.9475409836065574, + "grad_norm": 12.350762367248535, + "learning_rate": 2.9141422227150917e-06, + "loss": 1.541, + "step": 594 + }, + { + "epoch": 1.9508196721311475, + "grad_norm": 21.709842681884766, + "learning_rate": 2.898060826340954e-06, + "loss": 1.5742, + "step": 595 + }, + { + "epoch": 1.9540983606557378, + "grad_norm": 21.633665084838867, + "learning_rate": 2.882005797663604e-06, + "loss": 1.5781, + "step": 596 + }, + { + "epoch": 1.957377049180328, + "grad_norm": 17.24868392944336, + "learning_rate": 2.86597733808474e-06, + "loss": 1.5273, + "step": 597 + }, + { + "epoch": 1.960655737704918, + "grad_norm": 16.522829055786133, + "learning_rate": 2.849975648672775e-06, + "loss": 1.5645, + "step": 598 + }, + { + "epoch": 1.9639344262295082, + "grad_norm": 20.824695587158203, + "learning_rate": 2.8340009301602943e-06, + "loss": 1.5938, + "step": 599 + }, + { + "epoch": 1.9672131147540983, + "grad_norm": 24.554397583007812, + "learning_rate": 2.8180533829415536e-06, + "loss": 1.5195, + "step": 600 + }, + { + "epoch": 1.9704918032786884, + "grad_norm": 15.339919090270996, + "learning_rate": 2.8021332070699582e-06, + "loss": 1.5918, + "step": 601 + }, + { + "epoch": 1.9737704918032786, + "grad_norm": 16.299833297729492, + "learning_rate": 2.7862406022555547e-06, + "loss": 1.5469, + "step": 602 + }, + { + "epoch": 1.9770491803278687, + "grad_norm": 21.7740535736084, + "learning_rate": 2.7703757678625286e-06, + "loss": 1.6309, + "step": 603 + }, + { + "epoch": 1.980327868852459, + "grad_norm": 15.119240760803223, + "learning_rate": 2.754538902906693e-06, + "loss": 1.4824, + "step": 604 + }, + { + "epoch": 1.9836065573770492, + "grad_norm": 12.31226921081543, + "learning_rate": 2.7387302060530076e-06, + "loss": 1.5566, + "step": 605 + }, + { + "epoch": 1.9868852459016395, + "grad_norm": 14.131988525390625, + "learning_rate": 2.72294987561307e-06, + "loss": 1.6523, + "step": 606 + }, + { + "epoch": 1.9901639344262296, + "grad_norm": 16.699298858642578, + "learning_rate": 2.7071981095426435e-06, + "loss": 1.5371, + "step": 607 + }, + { + "epoch": 1.9934426229508198, + "grad_norm": 12.761926651000977, + "learning_rate": 2.6914751054391665e-06, + "loss": 1.584, + "step": 608 + }, + { + "epoch": 1.9967213114754099, + "grad_norm": 12.398103713989258, + "learning_rate": 2.6757810605392675e-06, + "loss": 1.543, + "step": 609 + }, + { + "epoch": 2.0, + "grad_norm": 14.358308792114258, + "learning_rate": 2.660116171716305e-06, + "loss": 1.5586, + "step": 610 + }, + { + "epoch": 2.00327868852459, + "grad_norm": 13.359660148620605, + "learning_rate": 2.6444806354778833e-06, + "loss": 1.4277, + "step": 611 + }, + { + "epoch": 2.0065573770491802, + "grad_norm": 12.226639747619629, + "learning_rate": 2.628874647963402e-06, + "loss": 1.418, + "step": 612 + }, + { + "epoch": 2.0098360655737704, + "grad_norm": 14.138394355773926, + "learning_rate": 2.6132984049415798e-06, + "loss": 1.4336, + "step": 613 + }, + { + "epoch": 2.0131147540983605, + "grad_norm": 13.725980758666992, + "learning_rate": 2.5977521018080115e-06, + "loss": 1.4082, + "step": 614 + }, + { + "epoch": 2.0163934426229506, + "grad_norm": 14.25224494934082, + "learning_rate": 2.582235933582714e-06, + "loss": 1.3633, + "step": 615 + }, + { + "epoch": 2.019672131147541, + "grad_norm": 13.436676979064941, + "learning_rate": 2.5667500949076705e-06, + "loss": 1.3887, + "step": 616 + }, + { + "epoch": 2.0229508196721313, + "grad_norm": 12.966423034667969, + "learning_rate": 2.5512947800444054e-06, + "loss": 1.3633, + "step": 617 + }, + { + "epoch": 2.0262295081967214, + "grad_norm": 14.51209545135498, + "learning_rate": 2.5358701828715298e-06, + "loss": 1.4883, + "step": 618 + }, + { + "epoch": 2.0295081967213116, + "grad_norm": 13.7439546585083, + "learning_rate": 2.520476496882326e-06, + "loss": 1.4238, + "step": 619 + }, + { + "epoch": 2.0327868852459017, + "grad_norm": 14.126566886901855, + "learning_rate": 2.5051139151823032e-06, + "loss": 1.4219, + "step": 620 + }, + { + "epoch": 2.036065573770492, + "grad_norm": 13.732039451599121, + "learning_rate": 2.4897826304867917e-06, + "loss": 1.4238, + "step": 621 + }, + { + "epoch": 2.039344262295082, + "grad_norm": 13.921870231628418, + "learning_rate": 2.4744828351185167e-06, + "loss": 1.4434, + "step": 622 + }, + { + "epoch": 2.042622950819672, + "grad_norm": 17.297060012817383, + "learning_rate": 2.459214721005181e-06, + "loss": 1.3398, + "step": 623 + }, + { + "epoch": 2.045901639344262, + "grad_norm": 13.58006763458252, + "learning_rate": 2.4439784796770705e-06, + "loss": 1.4238, + "step": 624 + }, + { + "epoch": 2.0491803278688523, + "grad_norm": 14.021245956420898, + "learning_rate": 2.428774302264637e-06, + "loss": 1.3906, + "step": 625 + }, + { + "epoch": 2.0524590163934424, + "grad_norm": 14.446585655212402, + "learning_rate": 2.4136023794961126e-06, + "loss": 1.4121, + "step": 626 + }, + { + "epoch": 2.055737704918033, + "grad_norm": 16.137575149536133, + "learning_rate": 2.3984629016951146e-06, + "loss": 1.457, + "step": 627 + }, + { + "epoch": 2.059016393442623, + "grad_norm": 13.775222778320312, + "learning_rate": 2.3833560587782462e-06, + "loss": 1.3477, + "step": 628 + }, + { + "epoch": 2.0622950819672132, + "grad_norm": 13.22794246673584, + "learning_rate": 2.3682820402527342e-06, + "loss": 1.4043, + "step": 629 + }, + { + "epoch": 2.0655737704918034, + "grad_norm": 14.258461952209473, + "learning_rate": 2.3532410352140317e-06, + "loss": 1.4277, + "step": 630 + }, + { + "epoch": 2.0688524590163935, + "grad_norm": 14.627102851867676, + "learning_rate": 2.3382332323434644e-06, + "loss": 1.4727, + "step": 631 + }, + { + "epoch": 2.0721311475409836, + "grad_norm": 13.571576118469238, + "learning_rate": 2.3232588199058455e-06, + "loss": 1.4453, + "step": 632 + }, + { + "epoch": 2.0754098360655737, + "grad_norm": 13.201940536499023, + "learning_rate": 2.3083179857471304e-06, + "loss": 1.4805, + "step": 633 + }, + { + "epoch": 2.078688524590164, + "grad_norm": 15.000838279724121, + "learning_rate": 2.29341091729205e-06, + "loss": 1.3867, + "step": 634 + }, + { + "epoch": 2.081967213114754, + "grad_norm": 15.08394718170166, + "learning_rate": 2.278537801541765e-06, + "loss": 1.3535, + "step": 635 + }, + { + "epoch": 2.085245901639344, + "grad_norm": 17.611570358276367, + "learning_rate": 2.2636988250715136e-06, + "loss": 1.4238, + "step": 636 + }, + { + "epoch": 2.088524590163934, + "grad_norm": 13.167271614074707, + "learning_rate": 2.2488941740282757e-06, + "loss": 1.377, + "step": 637 + }, + { + "epoch": 2.091803278688525, + "grad_norm": 16.23381996154785, + "learning_rate": 2.2341240341284416e-06, + "loss": 1.4219, + "step": 638 + }, + { + "epoch": 2.095081967213115, + "grad_norm": 12.979382514953613, + "learning_rate": 2.2193885906554694e-06, + "loss": 1.3652, + "step": 639 + }, + { + "epoch": 2.098360655737705, + "grad_norm": 13.941397666931152, + "learning_rate": 2.2046880284575833e-06, + "loss": 1.3711, + "step": 640 + }, + { + "epoch": 2.101639344262295, + "grad_norm": 13.71773624420166, + "learning_rate": 2.1900225319454254e-06, + "loss": 1.4043, + "step": 641 + }, + { + "epoch": 2.1049180327868853, + "grad_norm": 12.830635070800781, + "learning_rate": 2.1753922850897684e-06, + "loss": 1.3809, + "step": 642 + }, + { + "epoch": 2.1081967213114754, + "grad_norm": 15.367294311523438, + "learning_rate": 2.1607974714191894e-06, + "loss": 1.4609, + "step": 643 + }, + { + "epoch": 2.1114754098360655, + "grad_norm": 12.984966278076172, + "learning_rate": 2.1462382740177814e-06, + "loss": 1.4219, + "step": 644 + }, + { + "epoch": 2.1147540983606556, + "grad_norm": 14.171280860900879, + "learning_rate": 2.131714875522845e-06, + "loss": 1.3906, + "step": 645 + }, + { + "epoch": 2.1180327868852458, + "grad_norm": 13.716917037963867, + "learning_rate": 2.1172274581226064e-06, + "loss": 1.3574, + "step": 646 + }, + { + "epoch": 2.121311475409836, + "grad_norm": 13.324223518371582, + "learning_rate": 2.1027762035539314e-06, + "loss": 1.3457, + "step": 647 + }, + { + "epoch": 2.1245901639344265, + "grad_norm": 14.234905242919922, + "learning_rate": 2.0883612931000323e-06, + "loss": 1.3887, + "step": 648 + }, + { + "epoch": 2.1278688524590166, + "grad_norm": 14.057714462280273, + "learning_rate": 2.073982907588215e-06, + "loss": 1.4199, + "step": 649 + }, + { + "epoch": 2.1311475409836067, + "grad_norm": 13.39286994934082, + "learning_rate": 2.059641227387592e-06, + "loss": 1.4238, + "step": 650 + }, + { + "epoch": 2.134426229508197, + "grad_norm": 13.911747932434082, + "learning_rate": 2.0453364324068335e-06, + "loss": 1.4434, + "step": 651 + }, + { + "epoch": 2.137704918032787, + "grad_norm": 14.115652084350586, + "learning_rate": 2.0310687020918972e-06, + "loss": 1.3516, + "step": 652 + }, + { + "epoch": 2.140983606557377, + "grad_norm": 13.558361053466797, + "learning_rate": 2.016838215423791e-06, + "loss": 1.4375, + "step": 653 + }, + { + "epoch": 2.144262295081967, + "grad_norm": 13.805608749389648, + "learning_rate": 2.0026451509163204e-06, + "loss": 1.4199, + "step": 654 + }, + { + "epoch": 2.1475409836065573, + "grad_norm": 14.977523803710938, + "learning_rate": 1.988489686613845e-06, + "loss": 1.4492, + "step": 655 + }, + { + "epoch": 2.1508196721311474, + "grad_norm": 14.626441955566406, + "learning_rate": 1.9743720000890564e-06, + "loss": 1.3691, + "step": 656 + }, + { + "epoch": 2.1540983606557376, + "grad_norm": 14.720403671264648, + "learning_rate": 1.9602922684407388e-06, + "loss": 1.4023, + "step": 657 + }, + { + "epoch": 2.1573770491803277, + "grad_norm": 14.904003143310547, + "learning_rate": 1.9462506682915573e-06, + "loss": 1.4004, + "step": 658 + }, + { + "epoch": 2.160655737704918, + "grad_norm": 14.525262832641602, + "learning_rate": 1.932247375785837e-06, + "loss": 1.3984, + "step": 659 + }, + { + "epoch": 2.1639344262295084, + "grad_norm": 13.78215503692627, + "learning_rate": 1.91828256658735e-06, + "loss": 1.4082, + "step": 660 + }, + { + "epoch": 2.1672131147540985, + "grad_norm": 14.337031364440918, + "learning_rate": 1.9043564158771226e-06, + "loss": 1.4062, + "step": 661 + }, + { + "epoch": 2.1704918032786886, + "grad_norm": 14.619343757629395, + "learning_rate": 1.8904690983512253e-06, + "loss": 1.3965, + "step": 662 + }, + { + "epoch": 2.1737704918032787, + "grad_norm": 14.823307991027832, + "learning_rate": 1.876620788218592e-06, + "loss": 1.3223, + "step": 663 + }, + { + "epoch": 2.177049180327869, + "grad_norm": 14.898371696472168, + "learning_rate": 1.8628116591988232e-06, + "loss": 1.4004, + "step": 664 + }, + { + "epoch": 2.180327868852459, + "grad_norm": 17.55912208557129, + "learning_rate": 1.8490418845200241e-06, + "loss": 1.373, + "step": 665 + }, + { + "epoch": 2.183606557377049, + "grad_norm": 16.01456069946289, + "learning_rate": 1.8353116369166107e-06, + "loss": 1.4727, + "step": 666 + }, + { + "epoch": 2.1868852459016392, + "grad_norm": 13.750204086303711, + "learning_rate": 1.8216210886271562e-06, + "loss": 1.3711, + "step": 667 + }, + { + "epoch": 2.1901639344262294, + "grad_norm": 15.304807662963867, + "learning_rate": 1.8079704113922303e-06, + "loss": 1.4102, + "step": 668 + }, + { + "epoch": 2.1934426229508195, + "grad_norm": 14.734195709228516, + "learning_rate": 1.7943597764522363e-06, + "loss": 1.4199, + "step": 669 + }, + { + "epoch": 2.19672131147541, + "grad_norm": 13.935043334960938, + "learning_rate": 1.7807893545452748e-06, + "loss": 1.4258, + "step": 670 + }, + { + "epoch": 2.2, + "grad_norm": 13.949748039245605, + "learning_rate": 1.7672593159049877e-06, + "loss": 1.4141, + "step": 671 + }, + { + "epoch": 2.2032786885245903, + "grad_norm": 13.482948303222656, + "learning_rate": 1.753769830258441e-06, + "loss": 1.4199, + "step": 672 + }, + { + "epoch": 2.2065573770491804, + "grad_norm": 19.490720748901367, + "learning_rate": 1.740321066823975e-06, + "loss": 1.3574, + "step": 673 + }, + { + "epoch": 2.2098360655737705, + "grad_norm": 14.87624740600586, + "learning_rate": 1.7269131943090984e-06, + "loss": 1.4414, + "step": 674 + }, + { + "epoch": 2.2131147540983607, + "grad_norm": 14.58738899230957, + "learning_rate": 1.7135463809083625e-06, + "loss": 1.3867, + "step": 675 + }, + { + "epoch": 2.216393442622951, + "grad_norm": 14.250307083129883, + "learning_rate": 1.7002207943012516e-06, + "loss": 1.4277, + "step": 676 + }, + { + "epoch": 2.219672131147541, + "grad_norm": 14.0363187789917, + "learning_rate": 1.6869366016500872e-06, + "loss": 1.4062, + "step": 677 + }, + { + "epoch": 2.222950819672131, + "grad_norm": 13.977777481079102, + "learning_rate": 1.6736939695979237e-06, + "loss": 1.4102, + "step": 678 + }, + { + "epoch": 2.226229508196721, + "grad_norm": 13.99720573425293, + "learning_rate": 1.6604930642664614e-06, + "loss": 1.3828, + "step": 679 + }, + { + "epoch": 2.2295081967213113, + "grad_norm": 14.72916030883789, + "learning_rate": 1.6473340512539564e-06, + "loss": 1.3418, + "step": 680 + }, + { + "epoch": 2.2327868852459014, + "grad_norm": 14.218852043151855, + "learning_rate": 1.634217095633156e-06, + "loss": 1.4043, + "step": 681 + }, + { + "epoch": 2.236065573770492, + "grad_norm": 14.683760643005371, + "learning_rate": 1.6211423619492117e-06, + "loss": 1.3711, + "step": 682 + }, + { + "epoch": 2.239344262295082, + "grad_norm": 14.292176246643066, + "learning_rate": 1.6081100142176314e-06, + "loss": 1.416, + "step": 683 + }, + { + "epoch": 2.2426229508196722, + "grad_norm": 14.603547096252441, + "learning_rate": 1.5951202159222073e-06, + "loss": 1.3789, + "step": 684 + }, + { + "epoch": 2.2459016393442623, + "grad_norm": 14.563721656799316, + "learning_rate": 1.5821731300129755e-06, + "loss": 1.4062, + "step": 685 + }, + { + "epoch": 2.2491803278688525, + "grad_norm": 14.297794342041016, + "learning_rate": 1.5692689189041698e-06, + "loss": 1.4434, + "step": 686 + }, + { + "epoch": 2.2524590163934426, + "grad_norm": 16.67906379699707, + "learning_rate": 1.556407744472177e-06, + "loss": 1.4043, + "step": 687 + }, + { + "epoch": 2.2557377049180327, + "grad_norm": 14.280274391174316, + "learning_rate": 1.5435897680535184e-06, + "loss": 1.4023, + "step": 688 + }, + { + "epoch": 2.259016393442623, + "grad_norm": 14.600772857666016, + "learning_rate": 1.530815150442813e-06, + "loss": 1.416, + "step": 689 + }, + { + "epoch": 2.262295081967213, + "grad_norm": 13.637638092041016, + "learning_rate": 1.518084051890773e-06, + "loss": 1.4238, + "step": 690 + }, + { + "epoch": 2.265573770491803, + "grad_norm": 15.59545612335205, + "learning_rate": 1.5053966321021847e-06, + "loss": 1.3535, + "step": 691 + }, + { + "epoch": 2.2688524590163937, + "grad_norm": 15.58894157409668, + "learning_rate": 1.4927530502339048e-06, + "loss": 1.4199, + "step": 692 + }, + { + "epoch": 2.2721311475409838, + "grad_norm": 13.924785614013672, + "learning_rate": 1.4801534648928733e-06, + "loss": 1.2949, + "step": 693 + }, + { + "epoch": 2.275409836065574, + "grad_norm": 13.374207496643066, + "learning_rate": 1.4675980341341112e-06, + "loss": 1.4219, + "step": 694 + }, + { + "epoch": 2.278688524590164, + "grad_norm": 13.448588371276855, + "learning_rate": 1.4550869154587487e-06, + "loss": 1.3379, + "step": 695 + }, + { + "epoch": 2.281967213114754, + "grad_norm": 15.436904907226562, + "learning_rate": 1.4426202658120408e-06, + "loss": 1.3945, + "step": 696 + }, + { + "epoch": 2.2852459016393443, + "grad_norm": 13.138762474060059, + "learning_rate": 1.4301982415814046e-06, + "loss": 1.4316, + "step": 697 + }, + { + "epoch": 2.2885245901639344, + "grad_norm": 13.476716995239258, + "learning_rate": 1.417820998594458e-06, + "loss": 1.3887, + "step": 698 + }, + { + "epoch": 2.2918032786885245, + "grad_norm": 14.63200569152832, + "learning_rate": 1.4054886921170557e-06, + "loss": 1.4043, + "step": 699 + }, + { + "epoch": 2.2950819672131146, + "grad_norm": 14.927804946899414, + "learning_rate": 1.3932014768513553e-06, + "loss": 1.4395, + "step": 700 + } + ], + "logging_steps": 1, + "max_steps": 915, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 8.212482959920333e+16, + "train_batch_size": 512, + "trial_name": null, + "trial_params": null +}