| { | |
| "best_metric": 53.125, | |
| "best_model_checkpoint": "./whsiper-tiny/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": 3.3664, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 3.1517, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.46e-06, | |
| "loss": 2.7882, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 2.452, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.46e-06, | |
| "loss": 2.2192, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.96e-06, | |
| "loss": 2.0116, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.46e-06, | |
| "loss": 1.8317, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3.96e-06, | |
| "loss": 1.6729, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 1.5381, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 1.4531, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.5444352626800537, | |
| "eval_runtime": 21.3508, | |
| "eval_samples_per_second": 4.684, | |
| "eval_steps_per_second": 0.328, | |
| "eval_wer": 88.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 1.3681, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 1.2935, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 1.2256, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 6.96e-06, | |
| "loss": 1.1241, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 1.1015, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 1.052, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 8.46e-06, | |
| "loss": 1.0188, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.981, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.9204, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.8932, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 1.007778286933899, | |
| "eval_runtime": 19.7185, | |
| "eval_samples_per_second": 5.071, | |
| "eval_steps_per_second": 0.355, | |
| "eval_wer": 62.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 9.54e-06, | |
| "loss": 0.8921, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.040000000000002e-06, | |
| "loss": 0.8553, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.540000000000001e-06, | |
| "loss": 0.8444, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.040000000000001e-06, | |
| "loss": 0.829, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 7.540000000000001e-06, | |
| "loss": 0.7873, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 7.04e-06, | |
| "loss": 0.7691, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 6.540000000000001e-06, | |
| "loss": 0.7523, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 6.040000000000001e-06, | |
| "loss": 0.7421, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5.540000000000001e-06, | |
| "loss": 0.7464, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5.04e-06, | |
| "loss": 0.7225, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.8465978503227234, | |
| "eval_runtime": 19.7413, | |
| "eval_samples_per_second": 5.066, | |
| "eval_steps_per_second": 0.355, | |
| "eval_wer": 53.625, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.540000000000001e-06, | |
| "loss": 0.7234, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.04e-06, | |
| "loss": 0.7127, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.54e-06, | |
| "loss": 0.7057, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.04e-06, | |
| "loss": 0.6736, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.5400000000000002e-06, | |
| "loss": 0.7037, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2.04e-06, | |
| "loss": 0.7098, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.54e-06, | |
| "loss": 0.6867, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.04e-06, | |
| "loss": 0.6803, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5.4e-07, | |
| "loss": 0.6766, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4e-08, | |
| "loss": 0.6828, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.8060594201087952, | |
| "eval_runtime": 36.6359, | |
| "eval_samples_per_second": 2.73, | |
| "eval_steps_per_second": 0.191, | |
| "eval_wer": 53.125, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1000, | |
| "total_flos": 1.57560864768e+18, | |
| "train_loss": 1.2089761180877685, | |
| "train_runtime": 16872.5239, | |
| "train_samples_per_second": 3.793, | |
| "train_steps_per_second": 0.059 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.57560864768e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |