| { | |
| "best_metric": 30.521861777150917, | |
| "best_model_checkpoint": "./checkpoint-500", | |
| "epoch": 49.02, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.0003, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.0001, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0004, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0014, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0034, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.0053, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0033, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.0031, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "learning_rate": 9.387500000000001e-06, | |
| "loss": 0.0041, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "eval_loss": 0.4539676606655121, | |
| "eval_runtime": 80.2542, | |
| "eval_samples_per_second": 2.741, | |
| "eval_steps_per_second": 0.087, | |
| "eval_wer": 31.509167842031026, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "learning_rate": 9.075e-06, | |
| "loss": 0.0024, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 8.762500000000001e-06, | |
| "loss": 0.0021, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 8.45e-06, | |
| "loss": 0.0014, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "learning_rate": 8.1375e-06, | |
| "loss": 0.0015, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "learning_rate": 7.825e-06, | |
| "loss": 0.0014, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "learning_rate": 7.5125000000000005e-06, | |
| "loss": 0.001, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.0012, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 22.01, | |
| "learning_rate": 6.8875000000000005e-06, | |
| "loss": 0.001, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "learning_rate": 6.5750000000000006e-06, | |
| "loss": 0.0008, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "learning_rate": 6.262500000000001e-06, | |
| "loss": 0.0011, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "eval_loss": 0.48038455843925476, | |
| "eval_runtime": 86.9882, | |
| "eval_samples_per_second": 2.529, | |
| "eval_steps_per_second": 0.08, | |
| "eval_wer": 30.521861777150917, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "learning_rate": 5.950000000000001e-06, | |
| "loss": 0.0004, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 27.01, | |
| "learning_rate": 5.6375e-06, | |
| "loss": 0.0004, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "learning_rate": 5.325e-06, | |
| "loss": 0.0003, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "learning_rate": 5.0125e-06, | |
| "loss": 0.0005, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.0004, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 32.01, | |
| "learning_rate": 4.3875e-06, | |
| "loss": 0.0005, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "learning_rate": 4.075e-06, | |
| "loss": 0.0001, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "learning_rate": 3.7625e-06, | |
| "loss": 0.0002, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 36.01, | |
| "learning_rate": 3.45e-06, | |
| "loss": 0.0002, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 37.01, | |
| "learning_rate": 3.1375e-06, | |
| "loss": 0.0004, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 37.01, | |
| "eval_loss": 0.5306031703948975, | |
| "eval_runtime": 70.8611, | |
| "eval_samples_per_second": 3.105, | |
| "eval_steps_per_second": 0.099, | |
| "eval_wer": 31.48095909732017, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 38.02, | |
| "learning_rate": 2.825e-06, | |
| "loss": 0.0001, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "learning_rate": 2.5125e-06, | |
| "loss": 0.0001, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 41.01, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 42.01, | |
| "learning_rate": 1.8875000000000001e-06, | |
| "loss": 0.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 43.02, | |
| "learning_rate": 1.5750000000000002e-06, | |
| "loss": 0.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 44.02, | |
| "learning_rate": 1.2625000000000002e-06, | |
| "loss": 0.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 46.01, | |
| "learning_rate": 9.500000000000001e-07, | |
| "loss": 0.0, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 47.01, | |
| "learning_rate": 6.375e-07, | |
| "loss": 0.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 48.02, | |
| "learning_rate": 3.25e-07, | |
| "loss": 0.0, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "learning_rate": 1.2500000000000001e-08, | |
| "loss": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "eval_loss": 0.5505905747413635, | |
| "eval_runtime": 68.5701, | |
| "eval_samples_per_second": 3.208, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 32.77856135401975, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "step": 1000, | |
| "total_flos": 8.917288869888e+18, | |
| "train_loss": 0.0009783178368234075, | |
| "train_runtime": 10298.1352, | |
| "train_samples_per_second": 3.107, | |
| "train_steps_per_second": 0.097 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 8.917288869888e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |