| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "global_step": 2180, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.92, | |
| "eval_loss": 3.6019930839538574, | |
| "eval_runtime": 69.4999, | |
| "eval_samples_per_second": 23.698, | |
| "eval_steps_per_second": 2.964, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_loss": 2.9971394538879395, | |
| "eval_runtime": 69.2726, | |
| "eval_samples_per_second": 23.776, | |
| "eval_steps_per_second": 2.974, | |
| "eval_wer": 0.9998978653865795, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_loss": 0.9173915386199951, | |
| "eval_runtime": 69.6556, | |
| "eval_samples_per_second": 23.645, | |
| "eval_steps_per_second": 2.957, | |
| "eval_wer": 0.7772444081299152, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "eval_loss": 0.5667781829833984, | |
| "eval_runtime": 68.9782, | |
| "eval_samples_per_second": 23.877, | |
| "eval_steps_per_second": 2.986, | |
| "eval_wer": 0.6355836993156981, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 0.0002988, | |
| "loss": 3.1619, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "eval_loss": 0.4949137270450592, | |
| "eval_runtime": 69.6168, | |
| "eval_samples_per_second": 23.658, | |
| "eval_steps_per_second": 2.959, | |
| "eval_wer": 0.5255847206618323, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "eval_loss": 0.451631098985672, | |
| "eval_runtime": 69.6073, | |
| "eval_samples_per_second": 23.661, | |
| "eval_steps_per_second": 2.959, | |
| "eval_wer": 0.4744152793381677, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.42, | |
| "eval_loss": 0.42906084656715393, | |
| "eval_runtime": 68.1794, | |
| "eval_samples_per_second": 24.157, | |
| "eval_steps_per_second": 3.021, | |
| "eval_wer": 0.45746093351036665, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.34, | |
| "eval_loss": 0.4329916536808014, | |
| "eval_runtime": 69.7025, | |
| "eval_samples_per_second": 23.629, | |
| "eval_steps_per_second": 2.955, | |
| "eval_wer": 0.42733122255132266, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "eval_loss": 0.401591032743454, | |
| "eval_runtime": 69.0593, | |
| "eval_samples_per_second": 23.849, | |
| "eval_steps_per_second": 2.983, | |
| "eval_wer": 0.41446226126034114, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "learning_rate": 0.00021107142857142858, | |
| "loss": 0.2261, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "eval_loss": 0.4213583171367645, | |
| "eval_runtime": 69.3263, | |
| "eval_samples_per_second": 23.757, | |
| "eval_steps_per_second": 2.971, | |
| "eval_wer": 0.40046981922173425, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.09, | |
| "eval_loss": 0.4093494117259979, | |
| "eval_runtime": 69.6468, | |
| "eval_samples_per_second": 23.648, | |
| "eval_steps_per_second": 2.958, | |
| "eval_wer": 0.3946481462567664, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "eval_loss": 0.40514957904815674, | |
| "eval_runtime": 68.9496, | |
| "eval_samples_per_second": 23.887, | |
| "eval_steps_per_second": 2.988, | |
| "eval_wer": 0.3916862424675723, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.93, | |
| "eval_loss": 0.3908359408378601, | |
| "eval_runtime": 68.6938, | |
| "eval_samples_per_second": 23.976, | |
| "eval_steps_per_second": 2.999, | |
| "eval_wer": 0.37187212746399756, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.84, | |
| "eval_loss": 0.38503608107566833, | |
| "eval_runtime": 68.9973, | |
| "eval_samples_per_second": 23.871, | |
| "eval_steps_per_second": 2.986, | |
| "eval_wer": 0.36033091614748236, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "learning_rate": 0.00012178571428571428, | |
| "loss": 0.1119, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "eval_loss": 0.3967060148715973, | |
| "eval_runtime": 69.4393, | |
| "eval_samples_per_second": 23.719, | |
| "eval_steps_per_second": 2.967, | |
| "eval_wer": 0.3645184352977224, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.68, | |
| "eval_loss": 0.3821273744106293, | |
| "eval_runtime": 69.0903, | |
| "eval_samples_per_second": 23.838, | |
| "eval_steps_per_second": 2.982, | |
| "eval_wer": 0.3525686855275253, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "eval_loss": 0.39191436767578125, | |
| "eval_runtime": 69.7445, | |
| "eval_samples_per_second": 23.615, | |
| "eval_steps_per_second": 2.954, | |
| "eval_wer": 0.3518537432335819, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 16.51, | |
| "eval_loss": 0.37627777457237244, | |
| "eval_runtime": 70.1892, | |
| "eval_samples_per_second": 23.465, | |
| "eval_steps_per_second": 2.935, | |
| "eval_wer": 0.3366356858339291, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 17.43, | |
| "eval_loss": 0.3681693971157074, | |
| "eval_runtime": 69.0258, | |
| "eval_samples_per_second": 23.861, | |
| "eval_steps_per_second": 2.984, | |
| "eval_wer": 0.33489939740578084, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 18.35, | |
| "learning_rate": 3.25e-05, | |
| "loss": 0.074, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.35, | |
| "eval_loss": 0.37525057792663574, | |
| "eval_runtime": 69.5184, | |
| "eval_samples_per_second": 23.692, | |
| "eval_steps_per_second": 2.963, | |
| "eval_wer": 0.33234603207026864, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 19.27, | |
| "eval_loss": 0.3752918243408203, | |
| "eval_runtime": 68.5179, | |
| "eval_samples_per_second": 24.038, | |
| "eval_steps_per_second": 3.007, | |
| "eval_wer": 0.3267286283321418, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2180, | |
| "total_flos": 8.620382802507182e+18, | |
| "train_loss": 0.8245810762457891, | |
| "train_runtime": 5142.237, | |
| "train_samples_per_second": 13.527, | |
| "train_steps_per_second": 0.424 | |
| } | |
| ], | |
| "max_steps": 2180, | |
| "num_train_epochs": 20, | |
| "total_flos": 8.620382802507182e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |