| { | |
| "best_metric": 0.7814126394052044, | |
| "best_model_checkpoint": "./whisper-top-3/checkpoint-854", | |
| "epoch": 8.0, | |
| "global_step": 976, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.200000000000001e-07, | |
| "loss": 4.5543, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 4.1278, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.92e-06, | |
| "loss": 3.7088, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 3.2758, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.8959147930145264, | |
| "eval_runtime": 20.5606, | |
| "eval_samples_per_second": 10.554, | |
| "eval_steps_per_second": 1.362, | |
| "eval_wer": 4.915241635687733, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 4.92e-06, | |
| "loss": 2.9945, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 5.92e-06, | |
| "loss": 2.701, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 6.92e-06, | |
| "loss": 2.4989, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 7.92e-06, | |
| "loss": 2.453, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 2.2775, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.221217632293701, | |
| "eval_runtime": 17.0795, | |
| "eval_samples_per_second": 12.705, | |
| "eval_steps_per_second": 1.639, | |
| "eval_wer": 2.5866171003717473, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 2.155, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 1.0920000000000002e-05, | |
| "loss": 1.9709, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 1.1920000000000001e-05, | |
| "loss": 1.9147, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 1.2920000000000002e-05, | |
| "loss": 1.817, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 1.392e-05, | |
| "loss": 1.7023, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.8355822563171387, | |
| "eval_runtime": 15.5981, | |
| "eval_samples_per_second": 13.912, | |
| "eval_steps_per_second": 1.795, | |
| "eval_wer": 1.420074349442379, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 1.4920000000000001e-05, | |
| "loss": 1.5954, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 1.5920000000000003e-05, | |
| "loss": 1.3409, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 1.692e-05, | |
| "loss": 1.3539, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 1.792e-05, | |
| "loss": 1.2949, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 1.8920000000000002e-05, | |
| "loss": 1.2397, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.5363762378692627, | |
| "eval_runtime": 13.9831, | |
| "eval_samples_per_second": 15.519, | |
| "eval_steps_per_second": 2.002, | |
| "eval_wer": 0.8698884758364313, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 1.9920000000000002e-05, | |
| "loss": 1.1056, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "learning_rate": 1.9033613445378152e-05, | |
| "loss": 0.9068, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "learning_rate": 1.7983193277310925e-05, | |
| "loss": 0.8977, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "learning_rate": 1.69327731092437e-05, | |
| "loss": 0.8429, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 1.5882352941176473e-05, | |
| "loss": 0.8297, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.3302078247070312, | |
| "eval_runtime": 13.9868, | |
| "eval_samples_per_second": 15.515, | |
| "eval_steps_per_second": 2.002, | |
| "eval_wer": 0.8200743494423792, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "learning_rate": 1.4831932773109245e-05, | |
| "loss": 0.663, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 1.3781512605042017e-05, | |
| "loss": 0.5803, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "learning_rate": 1.2731092436974791e-05, | |
| "loss": 0.545, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "learning_rate": 1.1680672268907565e-05, | |
| "loss": 0.53, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "learning_rate": 1.0630252100840337e-05, | |
| "loss": 0.4978, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.2111059427261353, | |
| "eval_runtime": 14.3784, | |
| "eval_samples_per_second": 15.092, | |
| "eval_steps_per_second": 1.947, | |
| "eval_wer": 0.9405204460966543, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 9.579831932773111e-06, | |
| "loss": 0.4161, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "learning_rate": 8.529411764705883e-06, | |
| "loss": 0.342, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 7.478991596638656e-06, | |
| "loss": 0.3587, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.76, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.3542, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 5.378151260504202e-06, | |
| "loss": 0.3281, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 1.1617300510406494, | |
| "eval_runtime": 13.8478, | |
| "eval_samples_per_second": 15.67, | |
| "eval_steps_per_second": 2.022, | |
| "eval_wer": 0.7814126394052044, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.17, | |
| "learning_rate": 4.327731092436976e-06, | |
| "loss": 0.2595, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 7.38, | |
| "learning_rate": 3.2773109243697483e-06, | |
| "loss": 0.2469, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "learning_rate": 2.2268907563025214e-06, | |
| "loss": 0.2455, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "learning_rate": 1.1764705882352942e-06, | |
| "loss": 0.2667, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 7.99, | |
| "learning_rate": 1.2605042016806723e-07, | |
| "loss": 0.2414, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.144942045211792, | |
| "eval_runtime": 14.0429, | |
| "eval_samples_per_second": 15.453, | |
| "eval_steps_per_second": 1.994, | |
| "eval_wer": 0.7851301115241636, | |
| "step": 976 | |
| } | |
| ], | |
| "max_steps": 976, | |
| "num_train_epochs": 8, | |
| "total_flos": 3.8385765679104e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |