| { | |
| "best_metric": 57.99031476997578, | |
| "best_model_checkpoint": "./checkpoint-400", | |
| "epoch": 14.814814814814815, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 2.1183, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "eval_loss": 1.3170489072799683, | |
| "eval_runtime": 435.5063, | |
| "eval_samples_per_second": 1.176, | |
| "eval_steps_per_second": 0.147, | |
| "eval_wer": 76.95217917675545, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.8565, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "eval_loss": 0.9366902709007263, | |
| "eval_runtime": 473.9252, | |
| "eval_samples_per_second": 1.08, | |
| "eval_steps_per_second": 0.135, | |
| "eval_wer": 61.99303874092009, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 4.12e-06, | |
| "loss": 0.5824, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 4.32e-06, | |
| "loss": 0.5195, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.52, | |
| "learning_rate": 4.520000000000001e-06, | |
| "loss": 0.4472, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 4.7200000000000005e-06, | |
| "loss": 0.4615, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.4068, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "learning_rate": 5.12e-06, | |
| "loss": 0.3408, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5.320000000000001e-06, | |
| "loss": 0.3718, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 10.37, | |
| "learning_rate": 5.5200000000000005e-06, | |
| "loss": 0.2749, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 10.74, | |
| "learning_rate": 5.72e-06, | |
| "loss": 0.2625, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.2246, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "eval_loss": 0.9642460942268372, | |
| "eval_runtime": 439.7231, | |
| "eval_samples_per_second": 1.164, | |
| "eval_steps_per_second": 0.146, | |
| "eval_wer": 58.830205811138015, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.48, | |
| "learning_rate": 6.120000000000001e-06, | |
| "loss": 0.1991, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 11.85, | |
| "learning_rate": 6.3200000000000005e-06, | |
| "loss": 0.1798, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "learning_rate": 6.520000000000001e-06, | |
| "loss": 0.1426, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 12.59, | |
| "learning_rate": 6.720000000000001e-06, | |
| "loss": 0.125, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 12.96, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.1177, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 7.1200000000000004e-06, | |
| "loss": 0.0874, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "learning_rate": 7.32e-06, | |
| "loss": 0.0852, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 14.07, | |
| "learning_rate": 7.520000000000001e-06, | |
| "loss": 0.0666, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "learning_rate": 7.72e-06, | |
| "loss": 0.0531, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.054, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "eval_loss": 1.0876343250274658, | |
| "eval_runtime": 428.9567, | |
| "eval_samples_per_second": 1.194, | |
| "eval_steps_per_second": 0.149, | |
| "eval_wer": 57.99031476997578, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "step": 400, | |
| "total_flos": 1.84290636644352e+18, | |
| "train_loss": 0.027764302641153336, | |
| "train_runtime": 533.3671, | |
| "train_samples_per_second": 11.999, | |
| "train_steps_per_second": 0.75 | |
| } | |
| ], | |
| "max_steps": 400, | |
| "num_train_epochs": 15, | |
| "total_flos": 1.84290636644352e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |