| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 39.95179987797437, | |
| "eval_steps": 100.0, | |
| "global_step": 32760, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6101281269066504, | |
| "grad_norm": 55.342838287353516, | |
| "learning_rate": 1.1904e-06, | |
| "loss": 29.232, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 1.1283783783783783, | |
| "eval_loss": 11.350740432739258, | |
| "eval_runtime": 78.3217, | |
| "eval_samples_per_second": 87.332, | |
| "eval_steps_per_second": 10.917, | |
| "eval_wer": 1.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2196461256863942, | |
| "grad_norm": 47.82231903076172, | |
| "learning_rate": 2.3880000000000003e-06, | |
| "loss": 11.653, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.8297742525930445, | |
| "grad_norm": 40.30710220336914, | |
| "learning_rate": 3.588e-06, | |
| "loss": 9.4623, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 1.1284080132764343, | |
| "eval_loss": 7.82855224609375, | |
| "eval_runtime": 75.8772, | |
| "eval_samples_per_second": 90.146, | |
| "eval_steps_per_second": 11.268, | |
| "eval_wer": 1.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.4392922513727884, | |
| "grad_norm": 20.09738540649414, | |
| "learning_rate": 4.788e-06, | |
| "loss": 7.1817, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 1.1284080132764343, | |
| "eval_loss": 4.291374683380127, | |
| "eval_runtime": 70.7133, | |
| "eval_samples_per_second": 96.729, | |
| "eval_steps_per_second": 12.091, | |
| "eval_wer": 1.0, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.048810250152532, | |
| "grad_norm": 11.131582260131836, | |
| "learning_rate": 5.988e-06, | |
| "loss": 4.9943, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.6589383770591826, | |
| "grad_norm": 3.2968924045562744, | |
| "learning_rate": 7.1880000000000005e-06, | |
| "loss": 3.6746, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 1.1284080132764343, | |
| "eval_loss": 2.9958813190460205, | |
| "eval_runtime": 70.8535, | |
| "eval_samples_per_second": 96.537, | |
| "eval_steps_per_second": 12.067, | |
| "eval_wer": 1.0, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 4.268456375838926, | |
| "grad_norm": 1.5761828422546387, | |
| "learning_rate": 8.388e-06, | |
| "loss": 3.0354, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.878584502745577, | |
| "grad_norm": 2.042921543121338, | |
| "learning_rate": 9.588e-06, | |
| "loss": 2.6344, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 1.1284376481744902, | |
| "eval_loss": 2.379577159881592, | |
| "eval_runtime": 71.7131, | |
| "eval_samples_per_second": 95.38, | |
| "eval_steps_per_second": 11.923, | |
| "eval_wer": 1.0, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 5.48810250152532, | |
| "grad_norm": 4.554098606109619, | |
| "learning_rate": 1.0787999999999999e-05, | |
| "loss": 2.224, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 1.0735538169748695, | |
| "eval_loss": 1.1762758493423462, | |
| "eval_runtime": 74.8637, | |
| "eval_samples_per_second": 91.366, | |
| "eval_steps_per_second": 11.421, | |
| "eval_wer": 1.0, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 6.097620500305064, | |
| "grad_norm": 3.2407827377319336, | |
| "learning_rate": 1.1988000000000001e-05, | |
| "loss": 1.5426, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.707748627211714, | |
| "grad_norm": 8.200541496276855, | |
| "learning_rate": 1.3188e-05, | |
| "loss": 0.8482, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.9927098150782361, | |
| "eval_loss": 0.46800941228866577, | |
| "eval_runtime": 71.6177, | |
| "eval_samples_per_second": 95.507, | |
| "eval_steps_per_second": 11.938, | |
| "eval_wer": 0.9998538011695907, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 7.317266625991458, | |
| "grad_norm": 3.0007307529449463, | |
| "learning_rate": 1.4388000000000002e-05, | |
| "loss": 0.5869, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.927394752898109, | |
| "grad_norm": 4.457747459411621, | |
| "learning_rate": 1.5588e-05, | |
| "loss": 0.4357, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.9734471313418682, | |
| "eval_loss": 0.27095767855644226, | |
| "eval_runtime": 76.0314, | |
| "eval_samples_per_second": 89.963, | |
| "eval_steps_per_second": 11.245, | |
| "eval_wer": 0.9992690058479532, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 8.536912751677852, | |
| "grad_norm": 6.115583419799805, | |
| "learning_rate": 1.6788e-05, | |
| "loss": 0.3568, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 1.0188477951635846, | |
| "eval_loss": 0.1256408840417862, | |
| "eval_runtime": 71.254, | |
| "eval_samples_per_second": 95.995, | |
| "eval_steps_per_second": 11.999, | |
| "eval_wer": 0.999561403508772, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 9.146430750457595, | |
| "grad_norm": 4.245608806610107, | |
| "learning_rate": 1.7988e-05, | |
| "loss": 0.2776, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 9.756558877364247, | |
| "grad_norm": 5.325539588928223, | |
| "learning_rate": 1.9188e-05, | |
| "loss": 0.2374, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 1.0219001896633475, | |
| "eval_loss": 0.14763644337654114, | |
| "eval_runtime": 73.2557, | |
| "eval_samples_per_second": 93.372, | |
| "eval_steps_per_second": 11.671, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 10.36607687614399, | |
| "grad_norm": 5.118485450744629, | |
| "learning_rate": 2.0388e-05, | |
| "loss": 0.2199, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 10.97620500305064, | |
| "grad_norm": 7.805136203765869, | |
| "learning_rate": 2.1588e-05, | |
| "loss": 0.1941, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 1.018669985775249, | |
| "eval_loss": 0.059657976031303406, | |
| "eval_runtime": 72.2722, | |
| "eval_samples_per_second": 94.642, | |
| "eval_steps_per_second": 11.83, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 11.585723001830385, | |
| "grad_norm": 8.909073829650879, | |
| "learning_rate": 2.2788000000000003e-05, | |
| "loss": 0.1829, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 1.0188477951635846, | |
| "eval_loss": 0.046978700906038284, | |
| "eval_runtime": 70.6222, | |
| "eval_samples_per_second": 96.853, | |
| "eval_steps_per_second": 12.107, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 12.195241000610128, | |
| "grad_norm": 2.5103747844696045, | |
| "learning_rate": 2.3988e-05, | |
| "loss": 0.1802, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 12.805369127516778, | |
| "grad_norm": 3.005159378051758, | |
| "learning_rate": 2.5188e-05, | |
| "loss": 0.1701, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 1.018136557610242, | |
| "eval_loss": 0.05100085958838463, | |
| "eval_runtime": 70.6009, | |
| "eval_samples_per_second": 96.883, | |
| "eval_steps_per_second": 12.11, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 13.414887126296522, | |
| "grad_norm": 8.558382987976074, | |
| "learning_rate": 2.6388000000000002e-05, | |
| "loss": 0.1603, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 1.0192626837363679, | |
| "eval_loss": 0.037522342056035995, | |
| "eval_runtime": 70.9174, | |
| "eval_samples_per_second": 96.45, | |
| "eval_steps_per_second": 12.056, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 11480 | |
| }, | |
| { | |
| "epoch": 14.024405125076266, | |
| "grad_norm": 0.5282774567604065, | |
| "learning_rate": 2.7588e-05, | |
| "loss": 0.1579, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 14.634533251982916, | |
| "grad_norm": 6.527963161468506, | |
| "learning_rate": 2.8788e-05, | |
| "loss": 0.1564, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 1.0064011379800855, | |
| "eval_loss": 0.0865325778722763, | |
| "eval_runtime": 72.0213, | |
| "eval_samples_per_second": 94.972, | |
| "eval_steps_per_second": 11.871, | |
| "eval_wer": 0.9991228070175439, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 15.24405125076266, | |
| "grad_norm": 3.7430195808410645, | |
| "learning_rate": 2.9988e-05, | |
| "loss": 0.1632, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 15.854179377669311, | |
| "grad_norm": 3.735182762145996, | |
| "learning_rate": 2.9956013070043084e-05, | |
| "loss": 0.1555, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 1.0171882408724513, | |
| "eval_loss": 0.0239888783544302, | |
| "eval_runtime": 79.7324, | |
| "eval_samples_per_second": 85.787, | |
| "eval_steps_per_second": 10.723, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 13120 | |
| }, | |
| { | |
| "epoch": 16.463697376449055, | |
| "grad_norm": 2.248210906982422, | |
| "learning_rate": 2.9822174136311704e-05, | |
| "loss": 0.1446, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 1.0184329065908013, | |
| "eval_loss": 0.03877296671271324, | |
| "eval_runtime": 70.4641, | |
| "eval_samples_per_second": 97.071, | |
| "eval_steps_per_second": 12.134, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 13940 | |
| }, | |
| { | |
| "epoch": 17.0732153752288, | |
| "grad_norm": 1.9990071058273315, | |
| "learning_rate": 2.9599280835811145e-05, | |
| "loss": 0.1432, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 17.683343502135447, | |
| "grad_norm": 1.6087830066680908, | |
| "learning_rate": 2.9289379955813937e-05, | |
| "loss": 0.1374, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 1.0229077761972498, | |
| "eval_loss": 0.07236260920763016, | |
| "eval_runtime": 72.8414, | |
| "eval_samples_per_second": 93.903, | |
| "eval_steps_per_second": 11.738, | |
| "eval_wer": 0.9991228070175439, | |
| "step": 14760 | |
| }, | |
| { | |
| "epoch": 18.29286150091519, | |
| "grad_norm": 1.2661323547363281, | |
| "learning_rate": 2.8893091974003682e-05, | |
| "loss": 0.1271, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 18.902989627821842, | |
| "grad_norm": 2.638475179672241, | |
| "learning_rate": 2.841333172308954e-05, | |
| "loss": 0.1358, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 1.016802987197724, | |
| "eval_loss": 0.029252657666802406, | |
| "eval_runtime": 79.4655, | |
| "eval_samples_per_second": 86.075, | |
| "eval_steps_per_second": 10.759, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 15580 | |
| }, | |
| { | |
| "epoch": 19.512507626601586, | |
| "grad_norm": 3.171802043914795, | |
| "learning_rate": 2.785298169149414e-05, | |
| "loss": 0.126, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 1.0173364153627311, | |
| "eval_loss": 0.02503075823187828, | |
| "eval_runtime": 82.6447, | |
| "eval_samples_per_second": 82.764, | |
| "eval_steps_per_second": 10.345, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 20.12202562538133, | |
| "grad_norm": 2.895329236984253, | |
| "learning_rate": 2.7215408565964914e-05, | |
| "loss": 0.1259, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 20.73215375228798, | |
| "grad_norm": 8.28530502319336, | |
| "learning_rate": 2.650444300389672e-05, | |
| "loss": 0.1238, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_cer": 1.017366050260787, | |
| "eval_loss": 0.03371915966272354, | |
| "eval_runtime": 80.9247, | |
| "eval_samples_per_second": 84.523, | |
| "eval_steps_per_second": 10.565, | |
| "eval_wer": 0.9991228070175439, | |
| "step": 17220 | |
| }, | |
| { | |
| "epoch": 21.341671751067725, | |
| "grad_norm": 10.950956344604492, | |
| "learning_rate": 2.5724356618032884e-05, | |
| "loss": 0.1193, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 21.951799877974373, | |
| "grad_norm": 2.780343770980835, | |
| "learning_rate": 2.4879836311824927e-05, | |
| "loss": 0.1106, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_cer": 1.0159139402560455, | |
| "eval_loss": 0.018149759620428085, | |
| "eval_runtime": 71.3853, | |
| "eval_samples_per_second": 95.818, | |
| "eval_steps_per_second": 11.977, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 18040 | |
| }, | |
| { | |
| "epoch": 22.561317876754117, | |
| "grad_norm": 1.4264533519744873, | |
| "learning_rate": 2.397595611964874e-05, | |
| "loss": 0.1124, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_cer": 1.0165066382171646, | |
| "eval_loss": 0.019898999482393265, | |
| "eval_runtime": 81.6329, | |
| "eval_samples_per_second": 83.79, | |
| "eval_steps_per_second": 10.474, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 18860 | |
| }, | |
| { | |
| "epoch": 23.17083587553386, | |
| "grad_norm": 3.757054328918457, | |
| "learning_rate": 2.302011238680703e-05, | |
| "loss": 0.1079, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 23.780964002440513, | |
| "grad_norm": 6.899264812469482, | |
| "learning_rate": 2.2014218885552525e-05, | |
| "loss": 0.104, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_cer": 1.015054528212423, | |
| "eval_loss": 0.01506368163973093, | |
| "eval_runtime": 72.1117, | |
| "eval_samples_per_second": 94.853, | |
| "eval_steps_per_second": 11.857, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 19680 | |
| }, | |
| { | |
| "epoch": 24.390482001220256, | |
| "grad_norm": 1.1912193298339844, | |
| "learning_rate": 2.096618265844089e-05, | |
| "loss": 0.1025, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 2.540283679962158, | |
| "learning_rate": 1.988230050118496e-05, | |
| "loss": 0.1039, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 1.0149359886201992, | |
| "eval_loss": 0.013627970591187477, | |
| "eval_runtime": 73.5874, | |
| "eval_samples_per_second": 92.951, | |
| "eval_steps_per_second": 11.619, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 25.610128126906652, | |
| "grad_norm": 1.817094326019287, | |
| "learning_rate": 1.876908457848333e-05, | |
| "loss": 0.094, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_cer": 1.0147878141299194, | |
| "eval_loss": 0.013174526393413544, | |
| "eval_runtime": 78.6671, | |
| "eval_samples_per_second": 86.949, | |
| "eval_steps_per_second": 10.869, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 21320 | |
| }, | |
| { | |
| "epoch": 26.219646125686396, | |
| "grad_norm": 1.6206278800964355, | |
| "learning_rate": 1.7633223297728993e-05, | |
| "loss": 0.0919, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 26.829774252593044, | |
| "grad_norm": 0.5614987015724182, | |
| "learning_rate": 1.6481541123819273e-05, | |
| "loss": 0.0921, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_cer": 1.0157954006638217, | |
| "eval_loss": 0.017084894701838493, | |
| "eval_runtime": 76.9298, | |
| "eval_samples_per_second": 88.912, | |
| "eval_steps_per_second": 11.114, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 22140 | |
| }, | |
| { | |
| "epoch": 27.439292251372787, | |
| "grad_norm": 1.8238394260406494, | |
| "learning_rate": 1.532095757650705e-05, | |
| "loss": 0.0832, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_cer": 1.0146396396396395, | |
| "eval_loss": 0.012186683714389801, | |
| "eval_runtime": 72.328, | |
| "eval_samples_per_second": 94.569, | |
| "eval_steps_per_second": 11.821, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 22960 | |
| }, | |
| { | |
| "epoch": 28.04881025015253, | |
| "grad_norm": 3.1914637088775635, | |
| "learning_rate": 1.4160767960306099e-05, | |
| "loss": 0.0899, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 28.658938377059183, | |
| "grad_norm": 1.6626567840576172, | |
| "learning_rate": 1.300329518493389e-05, | |
| "loss": 0.08, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_cer": 1.015054528212423, | |
| "eval_loss": 0.014703178778290749, | |
| "eval_runtime": 69.3866, | |
| "eval_samples_per_second": 98.578, | |
| "eval_steps_per_second": 12.322, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 23780 | |
| }, | |
| { | |
| "epoch": 29.268456375838927, | |
| "grad_norm": 1.1348814964294434, | |
| "learning_rate": 1.1857818981811845e-05, | |
| "loss": 0.0822, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 29.878584502745575, | |
| "grad_norm": 0.9157238602638245, | |
| "learning_rate": 1.0731221584071209e-05, | |
| "loss": 0.0797, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_cer": 1.0146692745376955, | |
| "eval_loss": 0.01217756699770689, | |
| "eval_runtime": 70.4102, | |
| "eval_samples_per_second": 97.145, | |
| "eval_steps_per_second": 12.143, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 30.48810250152532, | |
| "grad_norm": 2.62127947807312, | |
| "learning_rate": 9.630271797484814e-06, | |
| "loss": 0.0775, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_cer": 1.015054528212423, | |
| "eval_loss": 0.013129099272191525, | |
| "eval_runtime": 69.5638, | |
| "eval_samples_per_second": 98.327, | |
| "eval_steps_per_second": 12.291, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 25420 | |
| }, | |
| { | |
| "epoch": 31.097620500305062, | |
| "grad_norm": 1.8141522407531738, | |
| "learning_rate": 8.561584332228596e-06, | |
| "loss": 0.071, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 31.707748627211714, | |
| "grad_norm": 0.32249507308006287, | |
| "learning_rate": 7.533597307465705e-06, | |
| "loss": 0.0675, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_cer": 1.0148174490279753, | |
| "eval_loss": 0.01305685006082058, | |
| "eval_runtime": 70.307, | |
| "eval_samples_per_second": 97.288, | |
| "eval_steps_per_second": 12.161, | |
| "eval_wer": 0.9991228070175439, | |
| "step": 26240 | |
| }, | |
| { | |
| "epoch": 32.31726662599146, | |
| "grad_norm": 0.06660401076078415, | |
| "learning_rate": 6.550290643366546e-06, | |
| "loss": 0.07, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 32.92739475289811, | |
| "grad_norm": 0.08903522789478302, | |
| "learning_rate": 5.617548632301114e-06, | |
| "loss": 0.0676, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_cer": 1.0146100047415836, | |
| "eval_loss": 0.017076797783374786, | |
| "eval_runtime": 71.0362, | |
| "eval_samples_per_second": 96.289, | |
| "eval_steps_per_second": 12.036, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 27060 | |
| }, | |
| { | |
| "epoch": 33.53691275167785, | |
| "grad_norm": 4.330328464508057, | |
| "learning_rate": 4.739256219207167e-06, | |
| "loss": 0.069, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_cer": 1.0145803698435278, | |
| "eval_loss": 0.014736202545464039, | |
| "eval_runtime": 73.5978, | |
| "eval_samples_per_second": 92.938, | |
| "eval_steps_per_second": 11.617, | |
| "eval_wer": 0.9991228070175439, | |
| "step": 27880 | |
| }, | |
| { | |
| "epoch": 34.1464307504576, | |
| "grad_norm": 0.1510840207338333, | |
| "learning_rate": 3.922612254686006e-06, | |
| "loss": 0.0629, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 34.75655887736425, | |
| "grad_norm": 0.51495361328125, | |
| "learning_rate": 3.1725232868909293e-06, | |
| "loss": 0.0588, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_cer": 1.0146396396396395, | |
| "eval_loss": 0.014013656415045261, | |
| "eval_runtime": 73.9594, | |
| "eval_samples_per_second": 92.483, | |
| "eval_steps_per_second": 11.56, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 35.36607687614399, | |
| "grad_norm": 2.207364559173584, | |
| "learning_rate": 2.493495989231198e-06, | |
| "loss": 0.0623, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 35.97620500305064, | |
| "grad_norm": 1.6189491748809814, | |
| "learning_rate": 1.8896100834437107e-06, | |
| "loss": 0.0608, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_cer": 1.0145803698435278, | |
| "eval_loss": 0.015181286260485649, | |
| "eval_runtime": 73.2042, | |
| "eval_samples_per_second": 93.437, | |
| "eval_steps_per_second": 11.68, | |
| "eval_wer": 0.9988304093567252, | |
| "step": 29520 | |
| }, | |
| { | |
| "epoch": 36.58572300183038, | |
| "grad_norm": 3.607675790786743, | |
| "learning_rate": 1.3644938278693997e-06, | |
| "loss": 0.0598, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_cer": 1.0146396396396395, | |
| "eval_loss": 0.015759674832224846, | |
| "eval_runtime": 73.8898, | |
| "eval_samples_per_second": 92.57, | |
| "eval_steps_per_second": 11.571, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 30340 | |
| }, | |
| { | |
| "epoch": 37.195241000610125, | |
| "grad_norm": 1.4858845472335815, | |
| "learning_rate": 9.213022182052699e-07, | |
| "loss": 0.0579, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 37.80536912751678, | |
| "grad_norm": 0.11195004731416702, | |
| "learning_rate": 5.626980317060648e-07, | |
| "loss": 0.0578, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_cer": 1.0146989094357515, | |
| "eval_loss": 0.014966564252972603, | |
| "eval_runtime": 73.4635, | |
| "eval_samples_per_second": 93.107, | |
| "eval_steps_per_second": 11.638, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 31160 | |
| }, | |
| { | |
| "epoch": 38.41488712629652, | |
| "grad_norm": 1.9088082313537598, | |
| "learning_rate": 2.912918111057888e-07, | |
| "loss": 0.0563, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_cer": 1.0147285443338074, | |
| "eval_loss": 0.015355916693806648, | |
| "eval_runtime": 63.935, | |
| "eval_samples_per_second": 106.984, | |
| "eval_steps_per_second": 13.373, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 31980 | |
| }, | |
| { | |
| "epoch": 39.024405125076264, | |
| "grad_norm": 3.6777658462524414, | |
| "learning_rate": 1.0762696080869105e-07, | |
| "loss": 0.059, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 39.634533251982916, | |
| "grad_norm": 1.0012741088867188, | |
| "learning_rate": 1.3438245287707985e-08, | |
| "loss": 0.0591, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 39.95179987797437, | |
| "eval_cer": 1.0146692745376955, | |
| "eval_loss": 0.014642004854977131, | |
| "eval_runtime": 66.4954, | |
| "eval_samples_per_second": 102.864, | |
| "eval_steps_per_second": 12.858, | |
| "eval_wer": 0.9989766081871345, | |
| "step": 32760 | |
| }, | |
| { | |
| "epoch": 39.95179987797437, | |
| "step": 32760, | |
| "total_flos": 1.7574609702583173e+19, | |
| "train_loss": 1.277195220610743, | |
| "train_runtime": 29323.9298, | |
| "train_samples_per_second": 71.532, | |
| "train_steps_per_second": 1.117 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32760, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 40, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7574609702583173e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |