| { | |
| "best_metric": 0.9681742043551089, | |
| "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-biopsy/checkpoint-1596", | |
| "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": 1.7908698320388794, | |
| "learning_rate": 2.3809523809523808e-06, | |
| "loss": 1.3741, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.7948147058486938, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 1.3269, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.6937719583511353, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 1.2328, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.9957559108734131, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 1.1332, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5494137353433836, | |
| "eval_loss": 1.071189522743225, | |
| "eval_runtime": 102.6887, | |
| "eval_samples_per_second": 5.814, | |
| "eval_steps_per_second": 0.185, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.9576915502548218, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 1.0191, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.0786559581756592, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.9104, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.0609694719314575, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.7667, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 1.6637555360794067, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.6556, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8492462311557789, | |
| "eval_loss": 0.5741624236106873, | |
| "eval_runtime": 102.8217, | |
| "eval_samples_per_second": 5.806, | |
| "eval_steps_per_second": 0.185, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 2.9285330772399902, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.5688, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 1.4198558330535889, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.4944, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.619047619047619, | |
| "grad_norm": 1.1169885396957397, | |
| "learning_rate": 2.6190476190476192e-05, | |
| "loss": 0.4229, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 1.9193533658981323, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.3987, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8894472361809045, | |
| "eval_loss": 0.3950396776199341, | |
| "eval_runtime": 102.779, | |
| "eval_samples_per_second": 5.809, | |
| "eval_steps_per_second": 0.185, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.0952380952380953, | |
| "grad_norm": 4.524629592895508, | |
| "learning_rate": 3.095238095238095e-05, | |
| "loss": 0.3845, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 2.838951587677002, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.3616, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 2.4979946613311768, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.3479, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.8095238095238093, | |
| "grad_norm": 4.321163177490234, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.2825, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8777219430485762, | |
| "eval_loss": 0.39236369729042053, | |
| "eval_runtime": 102.4215, | |
| "eval_samples_per_second": 5.829, | |
| "eval_steps_per_second": 0.186, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 4.0476190476190474, | |
| "grad_norm": 2.3643758296966553, | |
| "learning_rate": 4.047619047619048e-05, | |
| "loss": 0.3009, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 6.605370998382568, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.3091, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.523809523809524, | |
| "grad_norm": 4.201854228973389, | |
| "learning_rate": 4.523809523809524e-05, | |
| "loss": 0.277, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.761904761904762, | |
| "grad_norm": 1.754778265953064, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.2548, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.1804463863372803, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3662, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8860971524288107, | |
| "eval_loss": 0.36218151450157166, | |
| "eval_runtime": 101.769, | |
| "eval_samples_per_second": 5.866, | |
| "eval_steps_per_second": 0.187, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.238095238095238, | |
| "grad_norm": 1.8210253715515137, | |
| "learning_rate": 4.973544973544973e-05, | |
| "loss": 0.3025, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 5.476190476190476, | |
| "grad_norm": 2.0568735599517822, | |
| "learning_rate": 4.9470899470899475e-05, | |
| "loss": 0.2431, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 5.020086765289307, | |
| "learning_rate": 4.9206349206349204e-05, | |
| "loss": 0.2462, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.9523809523809526, | |
| "grad_norm": 2.455803632736206, | |
| "learning_rate": 4.894179894179895e-05, | |
| "loss": 0.2218, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9246231155778895, | |
| "eval_loss": 0.2705938518047333, | |
| "eval_runtime": 102.0179, | |
| "eval_samples_per_second": 5.852, | |
| "eval_steps_per_second": 0.186, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 6.190476190476191, | |
| "grad_norm": 1.1127841472625732, | |
| "learning_rate": 4.8677248677248676e-05, | |
| "loss": 0.2424, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 5.003824234008789, | |
| "learning_rate": 4.841269841269841e-05, | |
| "loss": 0.2726, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 1.7662922143936157, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.2482, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 6.904761904761905, | |
| "grad_norm": 4.009079933166504, | |
| "learning_rate": 4.7883597883597884e-05, | |
| "loss": 0.2236, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9346733668341709, | |
| "eval_loss": 0.2283419966697693, | |
| "eval_runtime": 101.6184, | |
| "eval_samples_per_second": 5.875, | |
| "eval_steps_per_second": 0.187, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 1.4519894123077393, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.2111, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.380952380952381, | |
| "grad_norm": 2.688704490661621, | |
| "learning_rate": 4.7354497354497356e-05, | |
| "loss": 0.2187, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 7.619047619047619, | |
| "grad_norm": 2.2603533267974854, | |
| "learning_rate": 4.708994708994709e-05, | |
| "loss": 0.1934, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 2.341301679611206, | |
| "learning_rate": 4.682539682539683e-05, | |
| "loss": 0.2224, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.931323283082077, | |
| "eval_loss": 0.23672878742218018, | |
| "eval_runtime": 100.0543, | |
| "eval_samples_per_second": 5.967, | |
| "eval_steps_per_second": 0.19, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 8.095238095238095, | |
| "grad_norm": 1.5976060628890991, | |
| "learning_rate": 4.656084656084656e-05, | |
| "loss": 0.1954, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 1.360371708869934, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.1607, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 1.7058426141738892, | |
| "learning_rate": 4.603174603174603e-05, | |
| "loss": 0.1804, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 8.80952380952381, | |
| "grad_norm": 2.388390302658081, | |
| "learning_rate": 4.576719576719577e-05, | |
| "loss": 0.1754, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9296482412060302, | |
| "eval_loss": 0.21387061476707458, | |
| "eval_runtime": 101.5101, | |
| "eval_samples_per_second": 5.881, | |
| "eval_steps_per_second": 0.187, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 9.047619047619047, | |
| "grad_norm": 2.2065589427948, | |
| "learning_rate": 4.55026455026455e-05, | |
| "loss": 0.1695, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 1.8705061674118042, | |
| "learning_rate": 4.523809523809524e-05, | |
| "loss": 0.1655, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 9.523809523809524, | |
| "grad_norm": 1.9261928796768188, | |
| "learning_rate": 4.4973544973544974e-05, | |
| "loss": 0.1584, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.761904761904763, | |
| "grad_norm": 2.372283697128296, | |
| "learning_rate": 4.470899470899471e-05, | |
| "loss": 0.1711, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.0945327281951904, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.1707, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.949748743718593, | |
| "eval_loss": 0.18289239704608917, | |
| "eval_runtime": 100.379, | |
| "eval_samples_per_second": 5.947, | |
| "eval_steps_per_second": 0.189, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 10.238095238095237, | |
| "grad_norm": 2.2974770069122314, | |
| "learning_rate": 4.417989417989418e-05, | |
| "loss": 0.1446, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 10.476190476190476, | |
| "grad_norm": 3.294903039932251, | |
| "learning_rate": 4.391534391534391e-05, | |
| "loss": 0.1848, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "grad_norm": 2.059147834777832, | |
| "learning_rate": 4.3650793650793655e-05, | |
| "loss": 0.1923, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 10.952380952380953, | |
| "grad_norm": 1.4008002281188965, | |
| "learning_rate": 4.3386243386243384e-05, | |
| "loss": 0.1619, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9463986599664992, | |
| "eval_loss": 0.21719853579998016, | |
| "eval_runtime": 100.686, | |
| "eval_samples_per_second": 5.929, | |
| "eval_steps_per_second": 0.189, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 11.19047619047619, | |
| "grad_norm": 3.9857630729675293, | |
| "learning_rate": 4.312169312169313e-05, | |
| "loss": 0.1479, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 1.8493316173553467, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.1399, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 11.666666666666666, | |
| "grad_norm": 3.7601022720336914, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.1693, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 11.904761904761905, | |
| "grad_norm": 2.0924296379089355, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 0.1547, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9380234505862647, | |
| "eval_loss": 0.19601453840732574, | |
| "eval_runtime": 100.959, | |
| "eval_samples_per_second": 5.913, | |
| "eval_steps_per_second": 0.188, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 12.142857142857142, | |
| "grad_norm": 0.876983106136322, | |
| "learning_rate": 4.2063492063492065e-05, | |
| "loss": 0.1467, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 12.380952380952381, | |
| "grad_norm": 1.4364930391311646, | |
| "learning_rate": 4.17989417989418e-05, | |
| "loss": 0.1293, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 12.619047619047619, | |
| "grad_norm": 2.0503087043762207, | |
| "learning_rate": 4.153439153439154e-05, | |
| "loss": 0.123, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "grad_norm": 4.358797550201416, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 0.1213, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.14836081862449646, | |
| "eval_runtime": 100.2262, | |
| "eval_samples_per_second": 5.957, | |
| "eval_steps_per_second": 0.19, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 13.095238095238095, | |
| "grad_norm": 5.3521199226379395, | |
| "learning_rate": 4.100529100529101e-05, | |
| "loss": 0.1213, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 1.1555789709091187, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.1353, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 13.571428571428571, | |
| "grad_norm": 2.3314497470855713, | |
| "learning_rate": 4.047619047619048e-05, | |
| "loss": 0.1392, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 13.80952380952381, | |
| "grad_norm": 2.4166135787963867, | |
| "learning_rate": 4.021164021164021e-05, | |
| "loss": 0.1388, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.16891160607337952, | |
| "eval_runtime": 101.124, | |
| "eval_samples_per_second": 5.904, | |
| "eval_steps_per_second": 0.188, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 14.047619047619047, | |
| "grad_norm": 1.2457020282745361, | |
| "learning_rate": 3.9947089947089946e-05, | |
| "loss": 0.1075, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 1.428454041481018, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 0.1149, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 14.523809523809524, | |
| "grad_norm": 1.9853533506393433, | |
| "learning_rate": 3.941798941798942e-05, | |
| "loss": 0.1232, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 14.761904761904763, | |
| "grad_norm": 2.842386484146118, | |
| "learning_rate": 3.9153439153439155e-05, | |
| "loss": 0.1123, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.4970998764038086, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.1009, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.1493997424840927, | |
| "eval_runtime": 103.3785, | |
| "eval_samples_per_second": 5.775, | |
| "eval_steps_per_second": 0.184, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 15.238095238095237, | |
| "grad_norm": 2.815284490585327, | |
| "learning_rate": 3.862433862433863e-05, | |
| "loss": 0.1301, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 15.476190476190476, | |
| "grad_norm": 2.5713961124420166, | |
| "learning_rate": 3.835978835978836e-05, | |
| "loss": 0.0979, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 15.714285714285714, | |
| "grad_norm": 1.771735668182373, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.12, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 15.952380952380953, | |
| "grad_norm": 2.1721739768981934, | |
| "learning_rate": 3.7830687830687835e-05, | |
| "loss": 0.124, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.15640056133270264, | |
| "eval_runtime": 101.1931, | |
| "eval_samples_per_second": 5.9, | |
| "eval_steps_per_second": 0.188, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 16.19047619047619, | |
| "grad_norm": 3.04827880859375, | |
| "learning_rate": 3.7566137566137564e-05, | |
| "loss": 0.1047, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 16.428571428571427, | |
| "grad_norm": 1.7996746301651, | |
| "learning_rate": 3.730158730158731e-05, | |
| "loss": 0.0987, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 2.0100674629211426, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.1132, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 16.904761904761905, | |
| "grad_norm": 3.321587562561035, | |
| "learning_rate": 3.677248677248677e-05, | |
| "loss": 0.1078, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9514237855946399, | |
| "eval_loss": 0.17282074689865112, | |
| "eval_runtime": 103.2232, | |
| "eval_samples_per_second": 5.784, | |
| "eval_steps_per_second": 0.184, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 1.6693847179412842, | |
| "learning_rate": 3.650793650793651e-05, | |
| "loss": 0.0762, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 17.38095238095238, | |
| "grad_norm": 1.9719501733779907, | |
| "learning_rate": 3.6243386243386245e-05, | |
| "loss": 0.0961, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 17.61904761904762, | |
| "grad_norm": 1.4737845659255981, | |
| "learning_rate": 3.597883597883598e-05, | |
| "loss": 0.0777, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 17.857142857142858, | |
| "grad_norm": 5.473972797393799, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.102, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9447236180904522, | |
| "eval_loss": 0.16686658561229706, | |
| "eval_runtime": 100.8047, | |
| "eval_samples_per_second": 5.922, | |
| "eval_steps_per_second": 0.188, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 18.095238095238095, | |
| "grad_norm": 2.001469612121582, | |
| "learning_rate": 3.5449735449735446e-05, | |
| "loss": 0.0987, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 18.333333333333332, | |
| "grad_norm": 1.7356058359146118, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.0933, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 18.571428571428573, | |
| "grad_norm": 2.967357635498047, | |
| "learning_rate": 3.492063492063492e-05, | |
| "loss": 0.122, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 18.80952380952381, | |
| "grad_norm": 1.3743668794631958, | |
| "learning_rate": 3.465608465608466e-05, | |
| "loss": 0.1006, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9581239530988275, | |
| "eval_loss": 0.1404878944158554, | |
| "eval_runtime": 101.9768, | |
| "eval_samples_per_second": 5.854, | |
| "eval_steps_per_second": 0.186, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 19.047619047619047, | |
| "grad_norm": 1.777043342590332, | |
| "learning_rate": 3.439153439153439e-05, | |
| "loss": 0.0782, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 19.285714285714285, | |
| "grad_norm": 0.9696865677833557, | |
| "learning_rate": 3.412698412698413e-05, | |
| "loss": 0.08, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 19.523809523809526, | |
| "grad_norm": 2.0073580741882324, | |
| "learning_rate": 3.386243386243386e-05, | |
| "loss": 0.095, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 19.761904761904763, | |
| "grad_norm": 2.404144048690796, | |
| "learning_rate": 3.35978835978836e-05, | |
| "loss": 0.116, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.8298864960670471, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0791, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.11790105700492859, | |
| "eval_runtime": 102.1963, | |
| "eval_samples_per_second": 5.842, | |
| "eval_steps_per_second": 0.186, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 20.238095238095237, | |
| "grad_norm": 1.1204876899719238, | |
| "learning_rate": 3.306878306878307e-05, | |
| "loss": 0.0827, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 20.476190476190474, | |
| "grad_norm": 1.7101030349731445, | |
| "learning_rate": 3.280423280423281e-05, | |
| "loss": 0.0809, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 20.714285714285715, | |
| "grad_norm": 1.100553035736084, | |
| "learning_rate": 3.253968253968254e-05, | |
| "loss": 0.0768, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 20.952380952380953, | |
| "grad_norm": 1.7138772010803223, | |
| "learning_rate": 3.227513227513227e-05, | |
| "loss": 0.0694, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.14235423505306244, | |
| "eval_runtime": 100.0469, | |
| "eval_samples_per_second": 5.967, | |
| "eval_steps_per_second": 0.19, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 21.19047619047619, | |
| "grad_norm": 2.3028171062469482, | |
| "learning_rate": 3.2010582010582015e-05, | |
| "loss": 0.0691, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "grad_norm": 1.7097909450531006, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 0.0748, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 21.666666666666668, | |
| "grad_norm": 1.1676493883132935, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.0991, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 21.904761904761905, | |
| "grad_norm": 1.6388928890228271, | |
| "learning_rate": 3.121693121693122e-05, | |
| "loss": 0.0627, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.12243743985891342, | |
| "eval_runtime": 101.3465, | |
| "eval_samples_per_second": 5.891, | |
| "eval_steps_per_second": 0.187, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 22.142857142857142, | |
| "grad_norm": 6.947789192199707, | |
| "learning_rate": 3.095238095238095e-05, | |
| "loss": 0.0952, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 22.38095238095238, | |
| "grad_norm": 2.1520142555236816, | |
| "learning_rate": 3.068783068783069e-05, | |
| "loss": 0.0697, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 22.61904761904762, | |
| "grad_norm": 2.1891136169433594, | |
| "learning_rate": 3.0423280423280425e-05, | |
| "loss": 0.0727, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 1.557528018951416, | |
| "learning_rate": 3.0158730158730158e-05, | |
| "loss": 0.0883, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.9447236180904522, | |
| "eval_loss": 0.16018958389759064, | |
| "eval_runtime": 101.009, | |
| "eval_samples_per_second": 5.91, | |
| "eval_steps_per_second": 0.188, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 23.095238095238095, | |
| "grad_norm": 0.4970981478691101, | |
| "learning_rate": 2.9894179894179897e-05, | |
| "loss": 0.0727, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 23.333333333333332, | |
| "grad_norm": 0.9330291748046875, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0776, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 23.571428571428573, | |
| "grad_norm": 0.5385302305221558, | |
| "learning_rate": 2.9365079365079366e-05, | |
| "loss": 0.068, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 23.80952380952381, | |
| "grad_norm": 1.3941255807876587, | |
| "learning_rate": 2.91005291005291e-05, | |
| "loss": 0.074, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.9614740368509213, | |
| "eval_loss": 0.13146798312664032, | |
| "eval_runtime": 101.4893, | |
| "eval_samples_per_second": 5.882, | |
| "eval_steps_per_second": 0.187, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 24.047619047619047, | |
| "grad_norm": 1.8402540683746338, | |
| "learning_rate": 2.8835978835978838e-05, | |
| "loss": 0.0529, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 24.285714285714285, | |
| "grad_norm": 1.1238616704940796, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0645, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 24.523809523809526, | |
| "grad_norm": 2.4385814666748047, | |
| "learning_rate": 2.830687830687831e-05, | |
| "loss": 0.0891, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 24.761904761904763, | |
| "grad_norm": 0.7520690560340881, | |
| "learning_rate": 2.8042328042328043e-05, | |
| "loss": 0.0856, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.3286396265029907, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0708, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.1330818086862564, | |
| "eval_runtime": 101.2634, | |
| "eval_samples_per_second": 5.896, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 25.238095238095237, | |
| "grad_norm": 1.7304377555847168, | |
| "learning_rate": 2.7513227513227512e-05, | |
| "loss": 0.0685, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 25.476190476190474, | |
| "grad_norm": 3.330090284347534, | |
| "learning_rate": 2.724867724867725e-05, | |
| "loss": 0.0664, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 0.801770806312561, | |
| "learning_rate": 2.6984126984126984e-05, | |
| "loss": 0.0641, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 25.952380952380953, | |
| "grad_norm": 1.6211001873016357, | |
| "learning_rate": 2.6719576719576723e-05, | |
| "loss": 0.06, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.11908143013715744, | |
| "eval_runtime": 100.9095, | |
| "eval_samples_per_second": 5.916, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 26.19047619047619, | |
| "grad_norm": 1.944045066833496, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 0.0636, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 26.428571428571427, | |
| "grad_norm": 1.0481789112091064, | |
| "learning_rate": 2.6190476190476192e-05, | |
| "loss": 0.0612, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 2.2019481658935547, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.0637, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 26.904761904761905, | |
| "grad_norm": 3.6049156188964844, | |
| "learning_rate": 2.5661375661375664e-05, | |
| "loss": 0.083, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.9530988274706867, | |
| "eval_loss": 0.1582762449979782, | |
| "eval_runtime": 100.0407, | |
| "eval_samples_per_second": 5.968, | |
| "eval_steps_per_second": 0.19, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 27.142857142857142, | |
| "grad_norm": 2.5039260387420654, | |
| "learning_rate": 2.5396825396825397e-05, | |
| "loss": 0.0606, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 27.38095238095238, | |
| "grad_norm": 1.956338882446289, | |
| "learning_rate": 2.5132275132275137e-05, | |
| "loss": 0.077, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 27.61904761904762, | |
| "grad_norm": 1.0798221826553345, | |
| "learning_rate": 2.4867724867724866e-05, | |
| "loss": 0.0427, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 27.857142857142858, | |
| "grad_norm": 3.189793348312378, | |
| "learning_rate": 2.4603174603174602e-05, | |
| "loss": 0.0584, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.9564489112227805, | |
| "eval_loss": 0.13481402397155762, | |
| "eval_runtime": 100.6312, | |
| "eval_samples_per_second": 5.933, | |
| "eval_steps_per_second": 0.189, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 28.095238095238095, | |
| "grad_norm": 1.0723867416381836, | |
| "learning_rate": 2.4338624338624338e-05, | |
| "loss": 0.0432, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 28.333333333333332, | |
| "grad_norm": 0.41823190450668335, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.0747, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 1.465440034866333, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.0451, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 28.80952380952381, | |
| "grad_norm": 1.9235646724700928, | |
| "learning_rate": 2.3544973544973546e-05, | |
| "loss": 0.0627, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.9564489112227805, | |
| "eval_loss": 0.12703652679920197, | |
| "eval_runtime": 101.0473, | |
| "eval_samples_per_second": 5.908, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 29.047619047619047, | |
| "grad_norm": 1.0633718967437744, | |
| "learning_rate": 2.328042328042328e-05, | |
| "loss": 0.0453, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 29.285714285714285, | |
| "grad_norm": 1.1838030815124512, | |
| "learning_rate": 2.3015873015873015e-05, | |
| "loss": 0.0549, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 29.523809523809526, | |
| "grad_norm": 1.2892875671386719, | |
| "learning_rate": 2.275132275132275e-05, | |
| "loss": 0.0506, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 29.761904761904763, | |
| "grad_norm": 0.46934711933135986, | |
| "learning_rate": 2.2486772486772487e-05, | |
| "loss": 0.0586, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.5484461784362793, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0627, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.9564489112227805, | |
| "eval_loss": 0.14105232059955597, | |
| "eval_runtime": 100.5476, | |
| "eval_samples_per_second": 5.937, | |
| "eval_steps_per_second": 0.189, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 30.238095238095237, | |
| "grad_norm": 1.0722055435180664, | |
| "learning_rate": 2.1957671957671956e-05, | |
| "loss": 0.0436, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 30.476190476190474, | |
| "grad_norm": 3.5860719680786133, | |
| "learning_rate": 2.1693121693121692e-05, | |
| "loss": 0.0477, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 30.714285714285715, | |
| "grad_norm": 0.9268338680267334, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0464, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 30.952380952380953, | |
| "grad_norm": 0.41101688146591187, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 0.038, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.12079473584890366, | |
| "eval_runtime": 101.8569, | |
| "eval_samples_per_second": 5.861, | |
| "eval_steps_per_second": 0.187, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 31.19047619047619, | |
| "grad_norm": 1.7808339595794678, | |
| "learning_rate": 2.08994708994709e-05, | |
| "loss": 0.0485, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 2.029599666595459, | |
| "learning_rate": 2.0634920634920636e-05, | |
| "loss": 0.0391, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 31.666666666666668, | |
| "grad_norm": 0.1751292645931244, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0393, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 31.904761904761905, | |
| "grad_norm": 1.3681291341781616, | |
| "learning_rate": 2.0105820105820105e-05, | |
| "loss": 0.0569, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.9514237855946399, | |
| "eval_loss": 0.15873192250728607, | |
| "eval_runtime": 101.2347, | |
| "eval_samples_per_second": 5.897, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 32.142857142857146, | |
| "grad_norm": 1.053713321685791, | |
| "learning_rate": 1.984126984126984e-05, | |
| "loss": 0.0385, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 32.38095238095238, | |
| "grad_norm": 1.0178366899490356, | |
| "learning_rate": 1.9576719576719577e-05, | |
| "loss": 0.0561, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 32.61904761904762, | |
| "grad_norm": 1.472675085067749, | |
| "learning_rate": 1.9312169312169313e-05, | |
| "loss": 0.0434, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "grad_norm": 1.232201099395752, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.0502, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.949748743718593, | |
| "eval_loss": 0.1500607430934906, | |
| "eval_runtime": 102.5738, | |
| "eval_samples_per_second": 5.82, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 33.095238095238095, | |
| "grad_norm": 4.194485664367676, | |
| "learning_rate": 1.8783068783068782e-05, | |
| "loss": 0.0589, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "grad_norm": 0.24997162818908691, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.0389, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 33.57142857142857, | |
| "grad_norm": 1.9876259565353394, | |
| "learning_rate": 1.8253968253968254e-05, | |
| "loss": 0.039, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 33.80952380952381, | |
| "grad_norm": 1.257222056388855, | |
| "learning_rate": 1.798941798941799e-05, | |
| "loss": 0.0464, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.9614740368509213, | |
| "eval_loss": 0.15079863369464874, | |
| "eval_runtime": 103.3413, | |
| "eval_samples_per_second": 5.777, | |
| "eval_steps_per_second": 0.184, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 34.04761904761905, | |
| "grad_norm": 1.1036677360534668, | |
| "learning_rate": 1.7724867724867723e-05, | |
| "loss": 0.0542, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 1.4602792263031006, | |
| "learning_rate": 1.746031746031746e-05, | |
| "loss": 0.0599, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 34.523809523809526, | |
| "grad_norm": 1.127734899520874, | |
| "learning_rate": 1.7195767195767195e-05, | |
| "loss": 0.0456, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 34.76190476190476, | |
| "grad_norm": 2.489420175552368, | |
| "learning_rate": 1.693121693121693e-05, | |
| "loss": 0.034, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 1.582891821861267, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0317, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.13086940348148346, | |
| "eval_runtime": 102.9849, | |
| "eval_samples_per_second": 5.797, | |
| "eval_steps_per_second": 0.184, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 35.23809523809524, | |
| "grad_norm": 1.932721495628357, | |
| "learning_rate": 1.6402116402116404e-05, | |
| "loss": 0.0387, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 35.476190476190474, | |
| "grad_norm": 3.2499399185180664, | |
| "learning_rate": 1.6137566137566136e-05, | |
| "loss": 0.0556, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 35.714285714285715, | |
| "grad_norm": 1.335181713104248, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 0.0514, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 35.95238095238095, | |
| "grad_norm": 0.5096970200538635, | |
| "learning_rate": 1.560846560846561e-05, | |
| "loss": 0.0552, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.9597989949748744, | |
| "eval_loss": 0.13715249300003052, | |
| "eval_runtime": 101.916, | |
| "eval_samples_per_second": 5.858, | |
| "eval_steps_per_second": 0.186, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 36.19047619047619, | |
| "grad_norm": 0.6248478293418884, | |
| "learning_rate": 1.5343915343915344e-05, | |
| "loss": 0.0634, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 36.42857142857143, | |
| "grad_norm": 0.9641430377960205, | |
| "learning_rate": 1.5079365079365079e-05, | |
| "loss": 0.0369, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 36.666666666666664, | |
| "grad_norm": 0.7052914500236511, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.0429, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 36.904761904761905, | |
| "grad_norm": 0.45987239480018616, | |
| "learning_rate": 1.455026455026455e-05, | |
| "loss": 0.031, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.9597989949748744, | |
| "eval_loss": 0.12577290832996368, | |
| "eval_runtime": 102.0267, | |
| "eval_samples_per_second": 5.851, | |
| "eval_steps_per_second": 0.186, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 37.142857142857146, | |
| "grad_norm": 1.7245473861694336, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0519, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 37.38095238095238, | |
| "grad_norm": 1.6640981435775757, | |
| "learning_rate": 1.4021164021164022e-05, | |
| "loss": 0.0448, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 37.61904761904762, | |
| "grad_norm": 0.358934611082077, | |
| "learning_rate": 1.3756613756613756e-05, | |
| "loss": 0.0315, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 37.857142857142854, | |
| "grad_norm": 3.198369026184082, | |
| "learning_rate": 1.3492063492063492e-05, | |
| "loss": 0.0383, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.9681742043551089, | |
| "eval_loss": 0.12486154586076736, | |
| "eval_runtime": 101.5464, | |
| "eval_samples_per_second": 5.879, | |
| "eval_steps_per_second": 0.187, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 38.095238095238095, | |
| "grad_norm": 0.6694173216819763, | |
| "learning_rate": 1.3227513227513228e-05, | |
| "loss": 0.0289, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 38.333333333333336, | |
| "grad_norm": 2.153737783432007, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.038, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 38.57142857142857, | |
| "grad_norm": 0.9305407404899597, | |
| "learning_rate": 1.2698412698412699e-05, | |
| "loss": 0.0467, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 38.80952380952381, | |
| "grad_norm": 1.1094582080841064, | |
| "learning_rate": 1.2433862433862433e-05, | |
| "loss": 0.036, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.13118697702884674, | |
| "eval_runtime": 101.001, | |
| "eval_samples_per_second": 5.911, | |
| "eval_steps_per_second": 0.188, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 39.04761904761905, | |
| "grad_norm": 0.5792038440704346, | |
| "learning_rate": 1.2169312169312169e-05, | |
| "loss": 0.0499, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 39.285714285714285, | |
| "grad_norm": 0.9286515116691589, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 0.0366, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 39.523809523809526, | |
| "grad_norm": 0.1906491369009018, | |
| "learning_rate": 1.164021164021164e-05, | |
| "loss": 0.0479, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 39.76190476190476, | |
| "grad_norm": 0.3992154896259308, | |
| "learning_rate": 1.1375661375661376e-05, | |
| "loss": 0.037, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.3215673863887787, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0405, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.1207132637500763, | |
| "eval_runtime": 101.9503, | |
| "eval_samples_per_second": 5.856, | |
| "eval_steps_per_second": 0.186, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 40.23809523809524, | |
| "grad_norm": 0.8763807415962219, | |
| "learning_rate": 1.0846560846560846e-05, | |
| "loss": 0.0258, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 40.476190476190474, | |
| "grad_norm": 1.2383830547332764, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 0.0373, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 40.714285714285715, | |
| "grad_norm": 1.2714569568634033, | |
| "learning_rate": 1.0317460317460318e-05, | |
| "loss": 0.0345, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 40.95238095238095, | |
| "grad_norm": 1.3528138399124146, | |
| "learning_rate": 1.0052910052910053e-05, | |
| "loss": 0.0343, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.12328309565782547, | |
| "eval_runtime": 102.5699, | |
| "eval_samples_per_second": 5.82, | |
| "eval_steps_per_second": 0.185, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 41.19047619047619, | |
| "grad_norm": 0.28864073753356934, | |
| "learning_rate": 9.788359788359789e-06, | |
| "loss": 0.0306, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 41.42857142857143, | |
| "grad_norm": 0.915897011756897, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.0335, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 41.666666666666664, | |
| "grad_norm": 1.0583347082138062, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.036, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 41.904761904761905, | |
| "grad_norm": 1.7799439430236816, | |
| "learning_rate": 8.994708994708995e-06, | |
| "loss": 0.0325, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.12857390940189362, | |
| "eval_runtime": 104.7251, | |
| "eval_samples_per_second": 5.701, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 42.142857142857146, | |
| "grad_norm": 1.5153943300247192, | |
| "learning_rate": 8.73015873015873e-06, | |
| "loss": 0.0305, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 42.38095238095238, | |
| "grad_norm": 1.5341339111328125, | |
| "learning_rate": 8.465608465608466e-06, | |
| "loss": 0.0422, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 42.61904761904762, | |
| "grad_norm": 0.7007244825363159, | |
| "learning_rate": 8.201058201058202e-06, | |
| "loss": 0.0462, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 1.075047254562378, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 0.0293, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "eval_accuracy": 0.9681742043551089, | |
| "eval_loss": 0.11350709199905396, | |
| "eval_runtime": 105.2069, | |
| "eval_samples_per_second": 5.675, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 43.095238095238095, | |
| "grad_norm": 0.27737095952033997, | |
| "learning_rate": 7.671957671957672e-06, | |
| "loss": 0.0308, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 43.333333333333336, | |
| "grad_norm": 0.4645295739173889, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0296, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 43.57142857142857, | |
| "grad_norm": 0.7459697723388672, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.039, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 43.80952380952381, | |
| "grad_norm": 1.326521396636963, | |
| "learning_rate": 6.878306878306878e-06, | |
| "loss": 0.0306, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "eval_accuracy": 0.9614740368509213, | |
| "eval_loss": 0.12575660645961761, | |
| "eval_runtime": 104.9936, | |
| "eval_samples_per_second": 5.686, | |
| "eval_steps_per_second": 0.181, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 44.04761904761905, | |
| "grad_norm": 0.11777963489294052, | |
| "learning_rate": 6.613756613756614e-06, | |
| "loss": 0.0355, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 44.285714285714285, | |
| "grad_norm": 0.9748639464378357, | |
| "learning_rate": 6.349206349206349e-06, | |
| "loss": 0.0281, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 44.523809523809526, | |
| "grad_norm": 1.6148064136505127, | |
| "learning_rate": 6.0846560846560845e-06, | |
| "loss": 0.0251, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 44.76190476190476, | |
| "grad_norm": 0.1060701385140419, | |
| "learning_rate": 5.82010582010582e-06, | |
| "loss": 0.0242, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 1.808620572090149, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0267, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.12606783211231232, | |
| "eval_runtime": 106.1357, | |
| "eval_samples_per_second": 5.625, | |
| "eval_steps_per_second": 0.179, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 45.23809523809524, | |
| "grad_norm": 1.064568042755127, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 0.0295, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 45.476190476190474, | |
| "grad_norm": 0.4456542730331421, | |
| "learning_rate": 5.026455026455026e-06, | |
| "loss": 0.0341, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 45.714285714285715, | |
| "grad_norm": 1.5678577423095703, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 0.0323, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 45.95238095238095, | |
| "grad_norm": 1.6082887649536133, | |
| "learning_rate": 4.497354497354498e-06, | |
| "loss": 0.0338, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.12089279294013977, | |
| "eval_runtime": 105.3726, | |
| "eval_samples_per_second": 5.666, | |
| "eval_steps_per_second": 0.18, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 46.19047619047619, | |
| "grad_norm": 0.628757894039154, | |
| "learning_rate": 4.232804232804233e-06, | |
| "loss": 0.039, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 46.42857142857143, | |
| "grad_norm": 0.13868111371994019, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 0.0264, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 46.666666666666664, | |
| "grad_norm": 1.5409092903137207, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0312, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 46.904761904761905, | |
| "grad_norm": 1.158074975013733, | |
| "learning_rate": 3.439153439153439e-06, | |
| "loss": 0.0213, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.11565249413251877, | |
| "eval_runtime": 104.0226, | |
| "eval_samples_per_second": 5.739, | |
| "eval_steps_per_second": 0.183, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 47.142857142857146, | |
| "grad_norm": 1.6575576066970825, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 0.0262, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 47.38095238095238, | |
| "grad_norm": 1.0997730493545532, | |
| "learning_rate": 2.91005291005291e-06, | |
| "loss": 0.0263, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 47.61904761904762, | |
| "grad_norm": 0.41330787539482117, | |
| "learning_rate": 2.6455026455026455e-06, | |
| "loss": 0.031, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 47.857142857142854, | |
| "grad_norm": 0.409661203622818, | |
| "learning_rate": 2.3809523809523808e-06, | |
| "loss": 0.0285, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "eval_accuracy": 0.9631490787269682, | |
| "eval_loss": 0.12032893300056458, | |
| "eval_runtime": 103.9176, | |
| "eval_samples_per_second": 5.745, | |
| "eval_steps_per_second": 0.183, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 48.095238095238095, | |
| "grad_norm": 1.2258131504058838, | |
| "learning_rate": 2.1164021164021164e-06, | |
| "loss": 0.0379, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 48.333333333333336, | |
| "grad_norm": 1.0911734104156494, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0267, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 48.57142857142857, | |
| "grad_norm": 1.2295212745666504, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.0194, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 48.80952380952381, | |
| "grad_norm": 0.645757257938385, | |
| "learning_rate": 1.3227513227513228e-06, | |
| "loss": 0.0287, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "eval_accuracy": 0.964824120603015, | |
| "eval_loss": 0.12397059053182602, | |
| "eval_runtime": 104.3655, | |
| "eval_samples_per_second": 5.72, | |
| "eval_steps_per_second": 0.182, | |
| "step": 2058 | |
| }, | |
| { | |
| "epoch": 49.04761904761905, | |
| "grad_norm": 1.8496387004852295, | |
| "learning_rate": 1.0582010582010582e-06, | |
| "loss": 0.0338, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 49.285714285714285, | |
| "grad_norm": 0.7916781306266785, | |
| "learning_rate": 7.936507936507937e-07, | |
| "loss": 0.0271, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 49.523809523809526, | |
| "grad_norm": 1.60358464717865, | |
| "learning_rate": 5.291005291005291e-07, | |
| "loss": 0.0329, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 49.76190476190476, | |
| "grad_norm": 1.0516901016235352, | |
| "learning_rate": 2.6455026455026455e-07, | |
| "loss": 0.0373, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 1.148287296295166, | |
| "learning_rate": 0.0, | |
| "loss": 0.0183, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_accuracy": 0.966499162479062, | |
| "eval_loss": 0.12244007736444473, | |
| "eval_runtime": 105.8827, | |
| "eval_samples_per_second": 5.638, | |
| "eval_steps_per_second": 0.179, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 2100, | |
| "total_flos": 2.0803097508518707e+19, | |
| "train_loss": 0.1363242975941726, | |
| "train_runtime": 52918.3087, | |
| "train_samples_per_second": 5.073, | |
| "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 | |
| } | |