| { | |
| "best_metric": 45.29159303206261, | |
| "best_model_checkpoint": "./output/checkpoint-1000", | |
| "epoch": 1.0, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.6000000000000004e-07, | |
| "loss": 3.7453, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 3.4563, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.46e-06, | |
| "loss": 3.1538, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 2.6311, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.46e-06, | |
| "loss": 2.1979, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.96e-06, | |
| "loss": 1.9003, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.46e-06, | |
| "loss": 1.7659, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3.96e-06, | |
| "loss": 1.5471, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 1.3988, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 1.4343, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.4900375604629517, | |
| "eval_runtime": 696.9939, | |
| "eval_samples_per_second": 0.717, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 77.9348649330977, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 1.2074, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 1.3247, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 1.2725, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 6.96e-06, | |
| "loss": 1.15, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 1.0425, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 1.0967, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 8.46e-06, | |
| "loss": 1.0666, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.9832, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.6704, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.7722, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.8731086850166321, | |
| "eval_runtime": 702.8815, | |
| "eval_samples_per_second": 0.711, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 60.28780610956829, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 9.54e-06, | |
| "loss": 1.0086, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.040000000000002e-06, | |
| "loss": 0.7804, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.540000000000001e-06, | |
| "loss": 0.7918, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.040000000000001e-06, | |
| "loss": 0.832, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 7.540000000000001e-06, | |
| "loss": 0.7615, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 7.04e-06, | |
| "loss": 0.6704, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 6.540000000000001e-06, | |
| "loss": 0.6398, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 6.040000000000001e-06, | |
| "loss": 0.7214, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5.540000000000001e-06, | |
| "loss": 0.731, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5.04e-06, | |
| "loss": 0.7181, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.6534215211868286, | |
| "eval_runtime": 681.508, | |
| "eval_samples_per_second": 0.734, | |
| "eval_steps_per_second": 0.047, | |
| "eval_wer": 46.50340823024489, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.540000000000001e-06, | |
| "loss": 0.9827, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.04e-06, | |
| "loss": 0.8043, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.54e-06, | |
| "loss": 0.6616, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.04e-06, | |
| "loss": 0.6631, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.5400000000000002e-06, | |
| "loss": 0.6304, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2.04e-06, | |
| "loss": 0.6005, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.54e-06, | |
| "loss": 0.6095, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.04e-06, | |
| "loss": 0.6342, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5.4e-07, | |
| "loss": 0.5842, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4e-08, | |
| "loss": 0.5733, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6313478946685791, | |
| "eval_runtime": 708.6015, | |
| "eval_samples_per_second": 0.706, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 45.29159303206261, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1000, | |
| "total_flos": 1.57560864768e+18, | |
| "train_loss": 1.2053918581008911, | |
| "train_runtime": 78451.7893, | |
| "train_samples_per_second": 0.816, | |
| "train_steps_per_second": 0.013 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.57560864768e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |