| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 8200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.4058000000000002e-05, | |
| "loss": 8.5696, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 2.8258000000000003e-05, | |
| "loss": 3.3881, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "eval_loss": 2.9710121154785156, | |
| "eval_runtime": 37.4848, | |
| "eval_samples_per_second": 31.826, | |
| "eval_steps_per_second": 1.014, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "learning_rate": 4.2458e-05, | |
| "loss": 2.9695, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "learning_rate": 5.6658000000000005e-05, | |
| "loss": 2.8538, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 7.0858e-05, | |
| "loss": 2.6401, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "eval_loss": 1.7677034139633179, | |
| "eval_runtime": 36.6194, | |
| "eval_samples_per_second": 32.578, | |
| "eval_steps_per_second": 1.038, | |
| "eval_wer": 0.9733567559654516, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "learning_rate": 6.904750000000001e-05, | |
| "loss": 1.81, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "learning_rate": 6.707527777777778e-05, | |
| "loss": 1.5152, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 18.29, | |
| "eval_loss": 0.5564321875572205, | |
| "eval_runtime": 36.9716, | |
| "eval_samples_per_second": 32.268, | |
| "eval_steps_per_second": 1.028, | |
| "eval_wer": 0.6010832967354707, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "learning_rate": 6.510305555555556e-05, | |
| "loss": 1.386, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 21.95, | |
| "learning_rate": 6.313083333333334e-05, | |
| "loss": 1.3129, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 24.39, | |
| "learning_rate": 6.115861111111111e-05, | |
| "loss": 1.2191, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 24.39, | |
| "eval_loss": 0.43192508816719055, | |
| "eval_runtime": 36.8038, | |
| "eval_samples_per_second": 32.415, | |
| "eval_steps_per_second": 1.033, | |
| "eval_wer": 0.4390279607670912, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 26.83, | |
| "learning_rate": 5.918638888888889e-05, | |
| "loss": 1.1031, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 29.27, | |
| "learning_rate": 5.7214166666666665e-05, | |
| "loss": 1.0237, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 30.49, | |
| "eval_loss": 0.3140852749347687, | |
| "eval_runtime": 36.7964, | |
| "eval_samples_per_second": 32.422, | |
| "eval_steps_per_second": 1.033, | |
| "eval_wer": 0.31752305665349145, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 31.71, | |
| "learning_rate": 5.5241944444444446e-05, | |
| "loss": 0.9707, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 34.15, | |
| "learning_rate": 5.326972222222223e-05, | |
| "loss": 0.9189, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 36.59, | |
| "learning_rate": 5.129750000000001e-05, | |
| "loss": 0.8892, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 36.59, | |
| "eval_loss": 0.27478542923927307, | |
| "eval_runtime": 37.4013, | |
| "eval_samples_per_second": 31.897, | |
| "eval_steps_per_second": 1.016, | |
| "eval_wer": 0.2689210950080515, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "learning_rate": 4.932527777777778e-05, | |
| "loss": 0.8573, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 41.46, | |
| "learning_rate": 4.735305555555556e-05, | |
| "loss": 0.8296, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 42.68, | |
| "eval_loss": 0.26798102259635925, | |
| "eval_runtime": 37.6187, | |
| "eval_samples_per_second": 31.713, | |
| "eval_steps_per_second": 1.01, | |
| "eval_wer": 0.2534036012296882, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 43.9, | |
| "learning_rate": 4.538083333333334e-05, | |
| "loss": 0.8047, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 46.34, | |
| "learning_rate": 4.340861111111111e-05, | |
| "loss": 0.7829, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 48.78, | |
| "learning_rate": 4.1436388888888886e-05, | |
| "loss": 0.7602, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 48.78, | |
| "eval_loss": 0.28198665380477905, | |
| "eval_runtime": 36.6815, | |
| "eval_samples_per_second": 32.523, | |
| "eval_steps_per_second": 1.036, | |
| "eval_wer": 0.25062216366564194, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 51.22, | |
| "learning_rate": 3.946416666666667e-05, | |
| "loss": 0.7536, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 53.66, | |
| "learning_rate": 3.7501805555555555e-05, | |
| "loss": 0.7186, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 54.88, | |
| "eval_loss": 0.2672194242477417, | |
| "eval_runtime": 36.55, | |
| "eval_samples_per_second": 32.64, | |
| "eval_steps_per_second": 1.04, | |
| "eval_wer": 0.23978919631093545, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 56.1, | |
| "learning_rate": 3.552958333333333e-05, | |
| "loss": 0.7166, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 58.54, | |
| "learning_rate": 3.355736111111111e-05, | |
| "loss": 0.706, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 60.98, | |
| "learning_rate": 3.1595000000000005e-05, | |
| "loss": 0.6887, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 60.98, | |
| "eval_loss": 0.27292799949645996, | |
| "eval_runtime": 36.0168, | |
| "eval_samples_per_second": 33.123, | |
| "eval_steps_per_second": 1.055, | |
| "eval_wer": 0.24022837066315328, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 63.41, | |
| "learning_rate": 2.962277777777778e-05, | |
| "loss": 0.6664, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 65.85, | |
| "learning_rate": 2.7650555555555557e-05, | |
| "loss": 0.6507, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 67.07, | |
| "eval_loss": 0.27668464183807373, | |
| "eval_runtime": 37.7209, | |
| "eval_samples_per_second": 31.627, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.23612941004245352, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 68.29, | |
| "learning_rate": 2.5678333333333338e-05, | |
| "loss": 0.6418, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 70.73, | |
| "learning_rate": 2.3706111111111112e-05, | |
| "loss": 0.6432, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 73.17, | |
| "learning_rate": 2.173388888888889e-05, | |
| "loss": 0.6226, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 73.17, | |
| "eval_loss": 0.2817266285419464, | |
| "eval_runtime": 36.1511, | |
| "eval_samples_per_second": 33.0, | |
| "eval_steps_per_second": 1.051, | |
| "eval_wer": 0.233201581027668, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 75.61, | |
| "learning_rate": 1.9761666666666667e-05, | |
| "loss": 0.6183, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 78.05, | |
| "learning_rate": 1.7789444444444445e-05, | |
| "loss": 0.6024, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 79.27, | |
| "eval_loss": 0.26792478561401367, | |
| "eval_runtime": 36.4629, | |
| "eval_samples_per_second": 32.718, | |
| "eval_steps_per_second": 1.042, | |
| "eval_wer": 0.22793148880105402, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 80.49, | |
| "learning_rate": 1.5817222222222223e-05, | |
| "loss": 0.6028, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 82.93, | |
| "learning_rate": 1.3845000000000002e-05, | |
| "loss": 0.5855, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 85.37, | |
| "learning_rate": 1.1872777777777778e-05, | |
| "loss": 0.5787, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 85.37, | |
| "eval_loss": 0.28374168276786804, | |
| "eval_runtime": 36.0785, | |
| "eval_samples_per_second": 33.067, | |
| "eval_steps_per_second": 1.053, | |
| "eval_wer": 0.23159127506953595, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 87.8, | |
| "learning_rate": 9.910416666666668e-06, | |
| "loss": 0.5824, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 90.24, | |
| "learning_rate": 7.938194444444445e-06, | |
| "loss": 0.5744, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 91.46, | |
| "eval_loss": 0.28380560874938965, | |
| "eval_runtime": 35.876, | |
| "eval_samples_per_second": 33.253, | |
| "eval_steps_per_second": 1.059, | |
| "eval_wer": 0.22837066315327184, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 92.68, | |
| "learning_rate": 5.965972222222223e-06, | |
| "loss": 0.569, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 95.12, | |
| "learning_rate": 3.9937500000000005e-06, | |
| "loss": 0.561, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 97.56, | |
| "learning_rate": 2.0215277777777777e-06, | |
| "loss": 0.5556, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 97.56, | |
| "eval_loss": 0.2762868106365204, | |
| "eval_runtime": 36.6971, | |
| "eval_samples_per_second": 32.509, | |
| "eval_steps_per_second": 1.036, | |
| "eval_wer": 0.2280778802517933, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 4.930555555555556e-08, | |
| "loss": 0.558, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 8200, | |
| "total_flos": 2.609429292408082e+19, | |
| "train_loss": 1.2146642443028892, | |
| "train_runtime": 11060.029, | |
| "train_samples_per_second": 23.562, | |
| "train_steps_per_second": 0.741 | |
| } | |
| ], | |
| "max_steps": 8200, | |
| "num_train_epochs": 100, | |
| "total_flos": 2.609429292408082e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |