| { | |
| "best_metric": 0.9798994974874372, | |
| "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-biopsy/checkpoint-1134", | |
| "epoch": 50.0, | |
| "eval_steps": 500, | |
| "global_step": 2100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 1.9525200128555298, | |
| "learning_rate": 2.3809523809523808e-06, | |
| "loss": 1.4367, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.9581096172332764, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 1.383, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.039808511734009, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 1.2902, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 1.1895326375961304, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 1.1862, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5611390284757118, | |
| "eval_loss": 1.1166571378707886, | |
| "eval_runtime": 104.4408, | |
| "eval_samples_per_second": 5.716, | |
| "eval_steps_per_second": 0.182, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.9201633930206299, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 1.0861, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.0600992441177368, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.9457, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.0473607778549194, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.8483, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 1.1092392206192017, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.7235, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8542713567839196, | |
| "eval_loss": 0.602853536605835, | |
| "eval_runtime": 104.148, | |
| "eval_samples_per_second": 5.732, | |
| "eval_steps_per_second": 0.182, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 1.8511453866958618, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.6091, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 2.7441365718841553, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.5427, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.619047619047619, | |
| "grad_norm": 4.292701721191406, | |
| "learning_rate": 2.6190476190476192e-05, | |
| "loss": 0.4896, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 1.5600124597549438, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.4286, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9279731993299832, | |
| "eval_loss": 0.3451606035232544, | |
| "eval_runtime": 103.7235, | |
| "eval_samples_per_second": 5.756, | |
| "eval_steps_per_second": 0.183, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.0952380952380953, | |
| "grad_norm": 2.245466470718384, | |
| "learning_rate": 3.095238095238095e-05, | |
| "loss": 0.383, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 2.2018802165985107, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.3709, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 1.7421743869781494, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.328, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.8095238095238093, | |
| "grad_norm": 4.41985559463501, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.3612, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8944723618090452, | |
| "eval_loss": 0.34848451614379883, | |
| "eval_runtime": 102.4218, | |
| "eval_samples_per_second": 5.829, | |
| "eval_steps_per_second": 0.186, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 4.0476190476190474, | |
| "grad_norm": 3.4240400791168213, | |
| "learning_rate": 4.047619047619048e-05, | |
| "loss": 0.3198, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 3.2982099056243896, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.3305, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.523809523809524, | |
| "grad_norm": 1.3844854831695557, | |
| "learning_rate": 4.523809523809524e-05, | |
| "loss": 0.295, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.761904761904762, | |
| "grad_norm": 1.4766120910644531, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.2811, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.445556879043579, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3015, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9296482412060302, | |
| "eval_loss": 0.2589903771877289, | |
| "eval_runtime": 102.9927, | |
| "eval_samples_per_second": 5.797, | |
| "eval_steps_per_second": 0.184, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.238095238095238, | |
| "grad_norm": 2.2521181106567383, | |
| "learning_rate": 4.973544973544973e-05, | |
| "loss": 0.2306, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 5.476190476190476, | |
| "grad_norm": 1.8446944952011108, | |
| "learning_rate": 4.9470899470899475e-05, | |
| "loss": 0.2521, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 1.3707401752471924, | |
| "learning_rate": 4.9206349206349204e-05, | |
| "loss": 0.2368, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.9523809523809526, | |
| "grad_norm": 5.455785751342773, | |
| "learning_rate": 4.894179894179895e-05, | |
| "loss": 0.2917, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9413735343383585, | |
| "eval_loss": 0.22188201546669006, | |
| "eval_runtime": 102.3843, | |
| "eval_samples_per_second": 5.831, | |
| "eval_steps_per_second": 0.186, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 6.190476190476191, | |
| "grad_norm": 4.942572593688965, | |
| "learning_rate": 4.8677248677248676e-05, | |
| "loss": 0.26, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 2.746673107147217, | |
| "learning_rate": 4.841269841269841e-05, | |
| "loss": 0.2444, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 1.1653099060058594, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.2231, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 6.904761904761905, | |
| "grad_norm": 2.306840419769287, | |
| "learning_rate": 4.7883597883597884e-05, | |
| "loss": 0.2312, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9279731993299832, | |
| "eval_loss": 0.24004340171813965, | |
| "eval_runtime": 103.0964, | |
| "eval_samples_per_second": 5.791, | |
| "eval_steps_per_second": 0.184, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 1.45888090133667, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.2121, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.380952380952381, | |
| "grad_norm": 2.0533416271209717, | |
| "learning_rate": 4.7354497354497356e-05, | |
| "loss": 0.1907, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 7.619047619047619, | |
| "grad_norm": 3.498002767562866, | |
| "learning_rate": 4.708994708994709e-05, | |
| "loss": 0.1825, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 3.9980995655059814, | |
| "learning_rate": 4.682539682539683e-05, | |
| "loss": 0.1708, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9413735343383585, | |
| "eval_loss": 0.21199914813041687, | |
| "eval_runtime": 102.2389, | |
| "eval_samples_per_second": 5.839, | |
| "eval_steps_per_second": 0.186, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 8.095238095238095, | |
| "grad_norm": 2.3464365005493164, | |
| "learning_rate": 4.656084656084656e-05, | |
| "loss": 0.2199, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 5.443611145019531, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.2172, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 3.144594430923462, | |
| "learning_rate": 4.603174603174603e-05, | |
| "loss": 0.2218, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 8.80952380952381, | |
| "grad_norm": 2.5637638568878174, | |
| "learning_rate": 4.576719576719577e-05, | |
| "loss": 0.1806, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9514237855946399, | |
| "eval_loss": 0.17835275828838348, | |
| "eval_runtime": 101.6107, | |
| "eval_samples_per_second": 5.875, | |
| "eval_steps_per_second": 0.187, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 9.047619047619047, | |
| "grad_norm": 2.3099260330200195, | |
| "learning_rate": 4.55026455026455e-05, | |
| "loss": 0.1635, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 1.4256964921951294, | |
| "learning_rate": 4.523809523809524e-05, | |
| "loss": 0.1678, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 9.523809523809524, | |
| "grad_norm": 3.0002613067626953, | |
| "learning_rate": 4.4973544973544974e-05, | |
| "loss": 0.142, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.761904761904763, | |
| "grad_norm": 1.4789687395095825, | |
| "learning_rate": 4.470899470899471e-05, | |
| "loss": 0.1607, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.7972970008850098, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.1703, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.948073701842546, | |
| "eval_loss": 0.157081738114357, | |
| "eval_runtime": 100.824, | |
| "eval_samples_per_second": 5.921, | |
| "eval_steps_per_second": 0.188, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 10.238095238095237, | |
| "grad_norm": 1.5848140716552734, | |
| "learning_rate": 4.417989417989418e-05, | |
| "loss": 0.1397, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 10.476190476190476, | |
| "grad_norm": 2.9363107681274414, | |
| "learning_rate": 4.391534391534391e-05, | |
| "loss": 0.1672, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "grad_norm": 2.2171630859375, | |
| "learning_rate": 4.3650793650793655e-05, | |
| "loss": 0.1389, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 10.952380952380953, | |
| "grad_norm": 1.635779619216919, | |
| "learning_rate": 4.3386243386243384e-05, | |
| "loss": 0.139, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.15439380705356598, | |
| "eval_runtime": 102.2671, | |
| "eval_samples_per_second": 5.838, | |
| "eval_steps_per_second": 0.186, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 11.19047619047619, | |
| "grad_norm": 1.8428142070770264, | |
| "learning_rate": 4.312169312169313e-05, | |
| "loss": 0.1591, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 2.849888801574707, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.1257, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 11.666666666666666, | |
| "grad_norm": 1.4126508235931396, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.1369, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 11.904761904761905, | |
| "grad_norm": 1.5928796529769897, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 0.1301, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9597989949748744, | |
| "eval_loss": 0.143105611205101, | |
| "eval_runtime": 101.7618, | |
| "eval_samples_per_second": 5.867, | |
| "eval_steps_per_second": 0.187, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 12.142857142857142, | |
| "grad_norm": 1.9849919080734253, | |
| "learning_rate": 4.2063492063492065e-05, | |
| "loss": 0.1381, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 12.380952380952381, | |
| "grad_norm": 1.4877374172210693, | |
| "learning_rate": 4.17989417989418e-05, | |
| "loss": 0.1538, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 12.619047619047619, | |
| "grad_norm": 1.9960867166519165, | |
| "learning_rate": 4.153439153439154e-05, | |
| "loss": 0.1269, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "grad_norm": 2.2734787464141846, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 0.122, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.12966834008693695, | |
| "eval_runtime": 102.3139, | |
| "eval_samples_per_second": 5.835, | |
| "eval_steps_per_second": 0.186, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 13.095238095238095, | |
| "grad_norm": 3.210118055343628, | |
| "learning_rate": 4.100529100529101e-05, | |
| "loss": 0.1153, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 1.9536188840866089, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.1429, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 13.571428571428571, | |
| "grad_norm": 1.364351749420166, | |
| "learning_rate": 4.047619047619048e-05, | |
| "loss": 0.1315, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 13.80952380952381, | |
| "grad_norm": 1.046488881111145, | |
| "learning_rate": 4.021164021164021e-05, | |
| "loss": 0.1104, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9597989949748744, | |
| "eval_loss": 0.1400885283946991, | |
| "eval_runtime": 102.2729, | |
| "eval_samples_per_second": 5.837, | |
| "eval_steps_per_second": 0.186, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 14.047619047619047, | |
| "grad_norm": 1.5574325323104858, | |
| "learning_rate": 3.9947089947089946e-05, | |
| "loss": 0.1154, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 2.0739004611968994, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 0.1451, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 14.523809523809524, | |
| "grad_norm": 2.3485310077667236, | |
| "learning_rate": 3.941798941798942e-05, | |
| "loss": 0.0912, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 14.761904761904763, | |
| "grad_norm": 2.368014097213745, | |
| "learning_rate": 3.9153439153439155e-05, | |
| "loss": 0.1088, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.749390721321106, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.1075, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.11995427310466766, | |
| "eval_runtime": 102.833, | |
| "eval_samples_per_second": 5.806, | |
| "eval_steps_per_second": 0.185, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 15.238095238095237, | |
| "grad_norm": 1.4212672710418701, | |
| "learning_rate": 3.862433862433863e-05, | |
| "loss": 0.1171, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 15.476190476190476, | |
| "grad_norm": 1.671434998512268, | |
| "learning_rate": 3.835978835978836e-05, | |
| "loss": 0.1011, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 15.714285714285714, | |
| "grad_norm": 2.932440757751465, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.115, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 15.952380952380953, | |
| "grad_norm": 0.7532219290733337, | |
| "learning_rate": 3.7830687830687835e-05, | |
| "loss": 0.0986, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.16654644906520844, | |
| "eval_runtime": 107.3167, | |
| "eval_samples_per_second": 5.563, | |
| "eval_steps_per_second": 0.177, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 16.19047619047619, | |
| "grad_norm": 1.4447543621063232, | |
| "learning_rate": 3.7566137566137564e-05, | |
| "loss": 0.1057, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 16.428571428571427, | |
| "grad_norm": 0.7543544173240662, | |
| "learning_rate": 3.730158730158731e-05, | |
| "loss": 0.0892, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 2.0222179889678955, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.1124, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 16.904761904761905, | |
| "grad_norm": 0.4199140965938568, | |
| "learning_rate": 3.677248677248677e-05, | |
| "loss": 0.092, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9530988274706867, | |
| "eval_loss": 0.13986486196517944, | |
| "eval_runtime": 102.7548, | |
| "eval_samples_per_second": 5.81, | |
| "eval_steps_per_second": 0.185, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 3.0130505561828613, | |
| "learning_rate": 3.650793650793651e-05, | |
| "loss": 0.1337, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 17.38095238095238, | |
| "grad_norm": 1.515160083770752, | |
| "learning_rate": 3.6243386243386245e-05, | |
| "loss": 0.0903, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 17.61904761904762, | |
| "grad_norm": 3.3534371852874756, | |
| "learning_rate": 3.597883597883598e-05, | |
| "loss": 0.0981, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 17.857142857142858, | |
| "grad_norm": 1.0088682174682617, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.1123, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9698492462311558, | |
| "eval_loss": 0.11222123354673386, | |
| "eval_runtime": 102.0899, | |
| "eval_samples_per_second": 5.848, | |
| "eval_steps_per_second": 0.186, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 18.095238095238095, | |
| "grad_norm": 1.7243051528930664, | |
| "learning_rate": 3.5449735449735446e-05, | |
| "loss": 0.0753, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 18.333333333333332, | |
| "grad_norm": 1.9596099853515625, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.0944, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 18.571428571428573, | |
| "grad_norm": 1.1566182374954224, | |
| "learning_rate": 3.492063492063492e-05, | |
| "loss": 0.0995, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 18.80952380952381, | |
| "grad_norm": 1.8758305311203003, | |
| "learning_rate": 3.465608465608466e-05, | |
| "loss": 0.0766, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9564489112227805, | |
| "eval_loss": 0.13372252881526947, | |
| "eval_runtime": 101.4158, | |
| "eval_samples_per_second": 5.887, | |
| "eval_steps_per_second": 0.187, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 19.047619047619047, | |
| "grad_norm": 1.219541072845459, | |
| "learning_rate": 3.439153439153439e-05, | |
| "loss": 0.0855, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 19.285714285714285, | |
| "grad_norm": 3.616736650466919, | |
| "learning_rate": 3.412698412698413e-05, | |
| "loss": 0.1017, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 19.523809523809526, | |
| "grad_norm": 1.773658037185669, | |
| "learning_rate": 3.386243386243386e-05, | |
| "loss": 0.0868, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 19.761904761904763, | |
| "grad_norm": 1.0607361793518066, | |
| "learning_rate": 3.35978835978836e-05, | |
| "loss": 0.0853, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.2637231647968292, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0762, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9731993299832495, | |
| "eval_loss": 0.09744315594434738, | |
| "eval_runtime": 102.366, | |
| "eval_samples_per_second": 5.832, | |
| "eval_steps_per_second": 0.186, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 20.238095238095237, | |
| "grad_norm": 1.8454608917236328, | |
| "learning_rate": 3.306878306878307e-05, | |
| "loss": 0.0822, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 20.476190476190474, | |
| "grad_norm": 3.0975053310394287, | |
| "learning_rate": 3.280423280423281e-05, | |
| "loss": 0.1028, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 20.714285714285715, | |
| "grad_norm": 1.2053009271621704, | |
| "learning_rate": 3.253968253968254e-05, | |
| "loss": 0.072, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 20.952380952380953, | |
| "grad_norm": 1.6915127038955688, | |
| "learning_rate": 3.227513227513227e-05, | |
| "loss": 0.0994, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.9698492462311558, | |
| "eval_loss": 0.10233575850725174, | |
| "eval_runtime": 103.9281, | |
| "eval_samples_per_second": 5.744, | |
| "eval_steps_per_second": 0.183, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 21.19047619047619, | |
| "grad_norm": 2.1695563793182373, | |
| "learning_rate": 3.2010582010582015e-05, | |
| "loss": 0.0996, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "grad_norm": 1.6543734073638916, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 0.0795, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 21.666666666666668, | |
| "grad_norm": 1.129447102546692, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.0579, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 21.904761904761905, | |
| "grad_norm": 0.9281991720199585, | |
| "learning_rate": 3.121693121693122e-05, | |
| "loss": 0.0687, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.9748743718592965, | |
| "eval_loss": 0.0976216122508049, | |
| "eval_runtime": 104.1214, | |
| "eval_samples_per_second": 5.734, | |
| "eval_steps_per_second": 0.182, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 22.142857142857142, | |
| "grad_norm": 0.8745301961898804, | |
| "learning_rate": 3.095238095238095e-05, | |
| "loss": 0.0729, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 22.38095238095238, | |
| "grad_norm": 0.5599234700202942, | |
| "learning_rate": 3.068783068783069e-05, | |
| "loss": 0.0509, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 22.61904761904762, | |
| "grad_norm": 2.2772858142852783, | |
| "learning_rate": 3.0423280423280425e-05, | |
| "loss": 0.0569, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 1.4757581949234009, | |
| "learning_rate": 3.0158730158730158e-05, | |
| "loss": 0.0767, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.9765494137353434, | |
| "eval_loss": 0.09517333656549454, | |
| "eval_runtime": 102.0605, | |
| "eval_samples_per_second": 5.849, | |
| "eval_steps_per_second": 0.186, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 23.095238095238095, | |
| "grad_norm": 1.75167715549469, | |
| "learning_rate": 2.9894179894179897e-05, | |
| "loss": 0.0743, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 23.333333333333332, | |
| "grad_norm": 2.2827818393707275, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0545, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 23.571428571428573, | |
| "grad_norm": 1.0787564516067505, | |
| "learning_rate": 2.9365079365079366e-05, | |
| "loss": 0.0751, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 23.80952380952381, | |
| "grad_norm": 1.3239110708236694, | |
| "learning_rate": 2.91005291005291e-05, | |
| "loss": 0.0581, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.10956421494483948, | |
| "eval_runtime": 100.8901, | |
| "eval_samples_per_second": 5.917, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 24.047619047619047, | |
| "grad_norm": 1.6604154109954834, | |
| "learning_rate": 2.8835978835978838e-05, | |
| "loss": 0.0605, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 24.285714285714285, | |
| "grad_norm": 1.335884928703308, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0818, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 24.523809523809526, | |
| "grad_norm": 1.7045536041259766, | |
| "learning_rate": 2.830687830687831e-05, | |
| "loss": 0.0617, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 24.761904761904763, | |
| "grad_norm": 0.8292408585548401, | |
| "learning_rate": 2.8042328042328043e-05, | |
| "loss": 0.0695, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.2549328804016113, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0544, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.9715242881072027, | |
| "eval_loss": 0.11226008832454681, | |
| "eval_runtime": 102.7141, | |
| "eval_samples_per_second": 5.812, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 25.238095238095237, | |
| "grad_norm": 2.550447463989258, | |
| "learning_rate": 2.7513227513227512e-05, | |
| "loss": 0.0653, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 25.476190476190474, | |
| "grad_norm": 1.7058266401290894, | |
| "learning_rate": 2.724867724867725e-05, | |
| "loss": 0.0545, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 0.9403615593910217, | |
| "learning_rate": 2.6984126984126984e-05, | |
| "loss": 0.0721, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 25.952380952380953, | |
| "grad_norm": 1.375019907951355, | |
| "learning_rate": 2.6719576719576723e-05, | |
| "loss": 0.079, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.9681742043551089, | |
| "eval_loss": 0.10403668135404587, | |
| "eval_runtime": 104.8805, | |
| "eval_samples_per_second": 5.692, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 26.19047619047619, | |
| "grad_norm": 2.5057268142700195, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 0.064, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 26.428571428571427, | |
| "grad_norm": 1.9105467796325684, | |
| "learning_rate": 2.6190476190476192e-05, | |
| "loss": 0.0646, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 0.8713690638542175, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.0612, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 26.904761904761905, | |
| "grad_norm": 0.41031908988952637, | |
| "learning_rate": 2.5661375661375664e-05, | |
| "loss": 0.0661, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.9798994974874372, | |
| "eval_loss": 0.08376247435808182, | |
| "eval_runtime": 102.2339, | |
| "eval_samples_per_second": 5.84, | |
| "eval_steps_per_second": 0.186, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 27.142857142857142, | |
| "grad_norm": 0.7565813064575195, | |
| "learning_rate": 2.5396825396825397e-05, | |
| "loss": 0.0521, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 27.38095238095238, | |
| "grad_norm": 1.0543544292449951, | |
| "learning_rate": 2.5132275132275137e-05, | |
| "loss": 0.0414, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 27.61904761904762, | |
| "grad_norm": 0.8543218970298767, | |
| "learning_rate": 2.4867724867724866e-05, | |
| "loss": 0.0363, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 27.857142857142858, | |
| "grad_norm": 0.8904032707214355, | |
| "learning_rate": 2.4603174603174602e-05, | |
| "loss": 0.068, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.9715242881072027, | |
| "eval_loss": 0.11693750321865082, | |
| "eval_runtime": 101.9232, | |
| "eval_samples_per_second": 5.857, | |
| "eval_steps_per_second": 0.186, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 28.095238095238095, | |
| "grad_norm": 1.8682641983032227, | |
| "learning_rate": 2.4338624338624338e-05, | |
| "loss": 0.0506, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 28.333333333333332, | |
| "grad_norm": 1.2170352935791016, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.0447, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 1.7995257377624512, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.06, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 28.80952380952381, | |
| "grad_norm": 1.7307757139205933, | |
| "learning_rate": 2.3544973544973546e-05, | |
| "loss": 0.0722, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.9731993299832495, | |
| "eval_loss": 0.08973471820354462, | |
| "eval_runtime": 102.7807, | |
| "eval_samples_per_second": 5.808, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 29.047619047619047, | |
| "grad_norm": 1.751111626625061, | |
| "learning_rate": 2.328042328042328e-05, | |
| "loss": 0.0593, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 29.285714285714285, | |
| "grad_norm": 1.2809324264526367, | |
| "learning_rate": 2.3015873015873015e-05, | |
| "loss": 0.0782, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 29.523809523809526, | |
| "grad_norm": 1.383988618850708, | |
| "learning_rate": 2.275132275132275e-05, | |
| "loss": 0.0424, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 29.761904761904763, | |
| "grad_norm": 0.8208743929862976, | |
| "learning_rate": 2.2486772486772487e-05, | |
| "loss": 0.0445, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 2.3345565795898438, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.048, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.9731993299832495, | |
| "eval_loss": 0.08638785034418106, | |
| "eval_runtime": 102.6141, | |
| "eval_samples_per_second": 5.818, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 30.238095238095237, | |
| "grad_norm": 0.1527481973171234, | |
| "learning_rate": 2.1957671957671956e-05, | |
| "loss": 0.047, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 30.476190476190474, | |
| "grad_norm": 2.870774745941162, | |
| "learning_rate": 2.1693121693121692e-05, | |
| "loss": 0.0439, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 30.714285714285715, | |
| "grad_norm": 1.335099697113037, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0481, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 30.952380952380953, | |
| "grad_norm": 0.7925770878791809, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 0.0509, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.9748743718592965, | |
| "eval_loss": 0.08578111976385117, | |
| "eval_runtime": 102.7903, | |
| "eval_samples_per_second": 5.808, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 31.19047619047619, | |
| "grad_norm": 0.871812641620636, | |
| "learning_rate": 2.08994708994709e-05, | |
| "loss": 0.0634, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 1.0023894309997559, | |
| "learning_rate": 2.0634920634920636e-05, | |
| "loss": 0.0453, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 31.666666666666668, | |
| "grad_norm": 2.041550636291504, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0664, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 31.904761904761905, | |
| "grad_norm": 0.7777080535888672, | |
| "learning_rate": 2.0105820105820105e-05, | |
| "loss": 0.047, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.0800652801990509, | |
| "eval_runtime": 105.2901, | |
| "eval_samples_per_second": 5.67, | |
| "eval_steps_per_second": 0.18, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 32.142857142857146, | |
| "grad_norm": 1.1469173431396484, | |
| "learning_rate": 1.984126984126984e-05, | |
| "loss": 0.0482, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 32.38095238095238, | |
| "grad_norm": 1.5965325832366943, | |
| "learning_rate": 1.9576719576719577e-05, | |
| "loss": 0.0491, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 32.61904761904762, | |
| "grad_norm": 1.1459414958953857, | |
| "learning_rate": 1.9312169312169313e-05, | |
| "loss": 0.0489, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "grad_norm": 0.28279682993888855, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.0411, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.12206684798002243, | |
| "eval_runtime": 104.9484, | |
| "eval_samples_per_second": 5.689, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 33.095238095238095, | |
| "grad_norm": 3.9715614318847656, | |
| "learning_rate": 1.8783068783068782e-05, | |
| "loss": 0.0511, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "grad_norm": 0.9649419188499451, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.0399, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 33.57142857142857, | |
| "grad_norm": 0.7443895936012268, | |
| "learning_rate": 1.8253968253968254e-05, | |
| "loss": 0.0543, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 33.80952380952381, | |
| "grad_norm": 0.8323255777359009, | |
| "learning_rate": 1.798941798941799e-05, | |
| "loss": 0.0378, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.10112123191356659, | |
| "eval_runtime": 106.0234, | |
| "eval_samples_per_second": 5.631, | |
| "eval_steps_per_second": 0.179, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 34.04761904761905, | |
| "grad_norm": 0.8554814457893372, | |
| "learning_rate": 1.7724867724867723e-05, | |
| "loss": 0.0401, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 0.10374557971954346, | |
| "learning_rate": 1.746031746031746e-05, | |
| "loss": 0.0473, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 34.523809523809526, | |
| "grad_norm": 1.8587803840637207, | |
| "learning_rate": 1.7195767195767195e-05, | |
| "loss": 0.0484, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 34.76190476190476, | |
| "grad_norm": 3.2876217365264893, | |
| "learning_rate": 1.693121693121693e-05, | |
| "loss": 0.0574, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 1.5491383075714111, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0358, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.9798994974874372, | |
| "eval_loss": 0.08340263366699219, | |
| "eval_runtime": 105.9899, | |
| "eval_samples_per_second": 5.633, | |
| "eval_steps_per_second": 0.179, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 35.23809523809524, | |
| "grad_norm": 1.2243473529815674, | |
| "learning_rate": 1.6402116402116404e-05, | |
| "loss": 0.034, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 35.476190476190474, | |
| "grad_norm": 1.0530498027801514, | |
| "learning_rate": 1.6137566137566136e-05, | |
| "loss": 0.0446, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 35.714285714285715, | |
| "grad_norm": 0.9862790107727051, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 0.0515, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 35.95238095238095, | |
| "grad_norm": 0.3827548325061798, | |
| "learning_rate": 1.560846560846561e-05, | |
| "loss": 0.0347, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.9715242881072027, | |
| "eval_loss": 0.0993322879076004, | |
| "eval_runtime": 106.124, | |
| "eval_samples_per_second": 5.625, | |
| "eval_steps_per_second": 0.179, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 36.19047619047619, | |
| "grad_norm": 0.9916484355926514, | |
| "learning_rate": 1.5343915343915344e-05, | |
| "loss": 0.0312, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 36.42857142857143, | |
| "grad_norm": 1.0411128997802734, | |
| "learning_rate": 1.5079365079365079e-05, | |
| "loss": 0.0454, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 36.666666666666664, | |
| "grad_norm": 1.4398282766342163, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.0382, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 36.904761904761905, | |
| "grad_norm": 1.5881459712982178, | |
| "learning_rate": 1.455026455026455e-05, | |
| "loss": 0.0434, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.9731993299832495, | |
| "eval_loss": 0.09377118200063705, | |
| "eval_runtime": 103.2924, | |
| "eval_samples_per_second": 5.78, | |
| "eval_steps_per_second": 0.184, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 37.142857142857146, | |
| "grad_norm": 0.7819347381591797, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0471, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 37.38095238095238, | |
| "grad_norm": 2.019038200378418, | |
| "learning_rate": 1.4021164021164022e-05, | |
| "loss": 0.0467, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 37.61904761904762, | |
| "grad_norm": 2.1753976345062256, | |
| "learning_rate": 1.3756613756613756e-05, | |
| "loss": 0.0491, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 37.857142857142854, | |
| "grad_norm": 1.6137406826019287, | |
| "learning_rate": 1.3492063492063492e-05, | |
| "loss": 0.0507, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.08741559088230133, | |
| "eval_runtime": 106.3602, | |
| "eval_samples_per_second": 5.613, | |
| "eval_steps_per_second": 0.179, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 38.095238095238095, | |
| "grad_norm": 1.4368014335632324, | |
| "learning_rate": 1.3227513227513228e-05, | |
| "loss": 0.048, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 38.333333333333336, | |
| "grad_norm": 0.7738118171691895, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.0476, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 38.57142857142857, | |
| "grad_norm": 1.3715893030166626, | |
| "learning_rate": 1.2698412698412699e-05, | |
| "loss": 0.0453, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 38.80952380952381, | |
| "grad_norm": 0.9568972587585449, | |
| "learning_rate": 1.2433862433862433e-05, | |
| "loss": 0.0466, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.9765494137353434, | |
| "eval_loss": 0.09318649768829346, | |
| "eval_runtime": 103.2637, | |
| "eval_samples_per_second": 5.781, | |
| "eval_steps_per_second": 0.184, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 39.04761904761905, | |
| "grad_norm": 0.7897378206253052, | |
| "learning_rate": 1.2169312169312169e-05, | |
| "loss": 0.0506, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 39.285714285714285, | |
| "grad_norm": 0.580004096031189, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 0.035, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 39.523809523809526, | |
| "grad_norm": 0.05374649167060852, | |
| "learning_rate": 1.164021164021164e-05, | |
| "loss": 0.0332, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 39.76190476190476, | |
| "grad_norm": 1.3745430707931519, | |
| "learning_rate": 1.1375661375661376e-05, | |
| "loss": 0.0349, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.9970341324806213, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0502, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.9698492462311558, | |
| "eval_loss": 0.101199209690094, | |
| "eval_runtime": 104.9435, | |
| "eval_samples_per_second": 5.689, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 40.23809523809524, | |
| "grad_norm": 1.5024595260620117, | |
| "learning_rate": 1.0846560846560846e-05, | |
| "loss": 0.0519, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 40.476190476190474, | |
| "grad_norm": 0.6853582859039307, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 0.0278, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 40.714285714285715, | |
| "grad_norm": 1.2199345827102661, | |
| "learning_rate": 1.0317460317460318e-05, | |
| "loss": 0.051, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 40.95238095238095, | |
| "grad_norm": 1.440384030342102, | |
| "learning_rate": 1.0052910052910053e-05, | |
| "loss": 0.0289, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_accuracy": 0.9715242881072027, | |
| "eval_loss": 0.08410957455635071, | |
| "eval_runtime": 103.4275, | |
| "eval_samples_per_second": 5.772, | |
| "eval_steps_per_second": 0.184, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 41.19047619047619, | |
| "grad_norm": 1.2516354322433472, | |
| "learning_rate": 9.788359788359789e-06, | |
| "loss": 0.0571, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 41.42857142857143, | |
| "grad_norm": 1.3431121110916138, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.0347, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 41.666666666666664, | |
| "grad_norm": 1.746293544769287, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.0288, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 41.904761904761905, | |
| "grad_norm": 0.5595345497131348, | |
| "learning_rate": 8.994708994708995e-06, | |
| "loss": 0.0274, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.9681742043551089, | |
| "eval_loss": 0.08830881863832474, | |
| "eval_runtime": 104.6396, | |
| "eval_samples_per_second": 5.705, | |
| "eval_steps_per_second": 0.182, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 42.142857142857146, | |
| "grad_norm": 0.7204629778862, | |
| "learning_rate": 8.73015873015873e-06, | |
| "loss": 0.0334, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 42.38095238095238, | |
| "grad_norm": 2.038654327392578, | |
| "learning_rate": 8.465608465608466e-06, | |
| "loss": 0.0307, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 42.61904761904762, | |
| "grad_norm": 0.7497786283493042, | |
| "learning_rate": 8.201058201058202e-06, | |
| "loss": 0.0307, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 0.8559117913246155, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 0.0251, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.08430066704750061, | |
| "eval_runtime": 104.6379, | |
| "eval_samples_per_second": 5.705, | |
| "eval_steps_per_second": 0.182, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 43.095238095238095, | |
| "grad_norm": 0.4223964810371399, | |
| "learning_rate": 7.671957671957672e-06, | |
| "loss": 0.0246, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 43.333333333333336, | |
| "grad_norm": 1.1295182704925537, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0377, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 43.57142857142857, | |
| "grad_norm": 0.18845248222351074, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0318, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 43.80952380952381, | |
| "grad_norm": 0.5233401656150818, | |
| "learning_rate": 6.878306878306878e-06, | |
| "loss": 0.0343, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.08117161691188812, | |
| "eval_runtime": 104.6378, | |
| "eval_samples_per_second": 5.705, | |
| "eval_steps_per_second": 0.182, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 44.04761904761905, | |
| "grad_norm": 0.24697835743427277, | |
| "learning_rate": 6.613756613756614e-06, | |
| "loss": 0.0275, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 44.285714285714285, | |
| "grad_norm": 0.4542170763015747, | |
| "learning_rate": 6.349206349206349e-06, | |
| "loss": 0.0253, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 44.523809523809526, | |
| "grad_norm": 0.6393638849258423, | |
| "learning_rate": 6.0846560846560845e-06, | |
| "loss": 0.0294, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 44.76190476190476, | |
| "grad_norm": 1.173812747001648, | |
| "learning_rate": 5.82010582010582e-06, | |
| "loss": 0.0315, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 0.3166700005531311, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0289, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.08049461245536804, | |
| "eval_runtime": 104.7685, | |
| "eval_samples_per_second": 5.698, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 45.23809523809524, | |
| "grad_norm": 2.791168689727783, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 0.0426, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 45.476190476190474, | |
| "grad_norm": 0.08095259964466095, | |
| "learning_rate": 5.026455026455026e-06, | |
| "loss": 0.0142, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 45.714285714285715, | |
| "grad_norm": 1.3561227321624756, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 0.0349, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 45.95238095238095, | |
| "grad_norm": 0.7681119441986084, | |
| "learning_rate": 4.497354497354498e-06, | |
| "loss": 0.0277, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.9698492462311558, | |
| "eval_loss": 0.09433051943778992, | |
| "eval_runtime": 103.7424, | |
| "eval_samples_per_second": 5.755, | |
| "eval_steps_per_second": 0.183, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 46.19047619047619, | |
| "grad_norm": 0.1386665552854538, | |
| "learning_rate": 4.232804232804233e-06, | |
| "loss": 0.0299, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 46.42857142857143, | |
| "grad_norm": 1.531664490699768, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 0.0406, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 46.666666666666664, | |
| "grad_norm": 0.6857447624206543, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0256, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 46.904761904761905, | |
| "grad_norm": 2.32024884223938, | |
| "learning_rate": 3.439153439153439e-06, | |
| "loss": 0.0332, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "eval_accuracy": 0.9765494137353434, | |
| "eval_loss": 0.08073202520608902, | |
| "eval_runtime": 104.425, | |
| "eval_samples_per_second": 5.717, | |
| "eval_steps_per_second": 0.182, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 47.142857142857146, | |
| "grad_norm": 0.38239943981170654, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 0.0263, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 47.38095238095238, | |
| "grad_norm": 2.018458843231201, | |
| "learning_rate": 2.91005291005291e-06, | |
| "loss": 0.0304, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 47.61904761904762, | |
| "grad_norm": 0.7840559482574463, | |
| "learning_rate": 2.6455026455026455e-06, | |
| "loss": 0.0304, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 47.857142857142854, | |
| "grad_norm": 0.3612171709537506, | |
| "learning_rate": 2.3809523809523808e-06, | |
| "loss": 0.0328, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.9748743718592965, | |
| "eval_loss": 0.08264525234699249, | |
| "eval_runtime": 104.0038, | |
| "eval_samples_per_second": 5.74, | |
| "eval_steps_per_second": 0.183, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 48.095238095238095, | |
| "grad_norm": 0.5553507804870605, | |
| "learning_rate": 2.1164021164021164e-06, | |
| "loss": 0.0365, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 48.333333333333336, | |
| "grad_norm": 0.475077360868454, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0237, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 48.57142857142857, | |
| "grad_norm": 0.0878726914525032, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.0319, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 48.80952380952381, | |
| "grad_norm": 0.4963679909706116, | |
| "learning_rate": 1.3227513227513228e-06, | |
| "loss": 0.0257, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "eval_accuracy": 0.9748743718592965, | |
| "eval_loss": 0.08518210053443909, | |
| "eval_runtime": 103.3836, | |
| "eval_samples_per_second": 5.775, | |
| "eval_steps_per_second": 0.184, | |
| "step": 2058 | |
| }, | |
| { | |
| "epoch": 49.04761904761905, | |
| "grad_norm": 1.105210304260254, | |
| "learning_rate": 1.0582010582010582e-06, | |
| "loss": 0.0293, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 49.285714285714285, | |
| "grad_norm": 0.16146498918533325, | |
| "learning_rate": 7.936507936507937e-07, | |
| "loss": 0.0244, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 49.523809523809526, | |
| "grad_norm": 1.7809348106384277, | |
| "learning_rate": 5.291005291005291e-07, | |
| "loss": 0.0424, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 49.76190476190476, | |
| "grad_norm": 0.6645485758781433, | |
| "learning_rate": 2.6455026455026455e-07, | |
| "loss": 0.0223, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 1.225805401802063, | |
| "learning_rate": 0.0, | |
| "loss": 0.0287, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.9782244556113903, | |
| "eval_loss": 0.08483675867319107, | |
| "eval_runtime": 105.2572, | |
| "eval_samples_per_second": 5.672, | |
| "eval_steps_per_second": 0.181, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 2100, | |
| "total_flos": 2.0803097508518707e+19, | |
| "train_loss": 0.13851347144160953, | |
| "train_runtime": 52719.3565, | |
| "train_samples_per_second": 5.092, | |
| "train_steps_per_second": 0.04 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0803097508518707e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |