| { | |
| "best_metric": 0.9848484848484849, | |
| "best_model_checkpoint": "audio/train/checkpoint/whisper-small/superb_ks_42/checkpoint-11179", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 15970, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6261740763932373, | |
| "grad_norm": 8.361194610595703, | |
| "learning_rate": 3.1308703819661866e-05, | |
| "loss": 0.3599, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9707266843189173, | |
| "eval_loss": 0.15458986163139343, | |
| "eval_runtime": 143.1918, | |
| "eval_samples_per_second": 47.475, | |
| "eval_steps_per_second": 0.189, | |
| "step": 1597 | |
| }, | |
| { | |
| "epoch": 1.2523481527864746, | |
| "grad_norm": 0.06838619709014893, | |
| "learning_rate": 4.859806581785292e-05, | |
| "loss": 0.0958, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.878522229179712, | |
| "grad_norm": 6.442350387573242, | |
| "learning_rate": 4.51193209490016e-05, | |
| "loss": 0.0819, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9761694616063548, | |
| "eval_loss": 0.09978443384170532, | |
| "eval_runtime": 138.5727, | |
| "eval_samples_per_second": 49.057, | |
| "eval_steps_per_second": 0.195, | |
| "step": 3194 | |
| }, | |
| { | |
| "epoch": 2.504696305572949, | |
| "grad_norm": 0.008366833440959454, | |
| "learning_rate": 4.1640576080150286e-05, | |
| "loss": 0.0635, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.979994115916446, | |
| "eval_loss": 0.10489170253276825, | |
| "eval_runtime": 138.2853, | |
| "eval_samples_per_second": 49.159, | |
| "eval_steps_per_second": 0.195, | |
| "step": 4791 | |
| }, | |
| { | |
| "epoch": 3.1308703819661865, | |
| "grad_norm": 0.06020672991871834, | |
| "learning_rate": 3.8161831211298965e-05, | |
| "loss": 0.0569, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.757044458359424, | |
| "grad_norm": 0.005801007151603699, | |
| "learning_rate": 3.468308634244764e-05, | |
| "loss": 0.0437, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9796999117387467, | |
| "eval_loss": 0.0905165895819664, | |
| "eval_runtime": 138.7617, | |
| "eval_samples_per_second": 48.99, | |
| "eval_steps_per_second": 0.195, | |
| "step": 6388 | |
| }, | |
| { | |
| "epoch": 4.383218534752661, | |
| "grad_norm": 1.9717774391174316, | |
| "learning_rate": 3.120434147359633e-05, | |
| "loss": 0.0411, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.980876728449544, | |
| "eval_loss": 0.08979170769453049, | |
| "eval_runtime": 138.2656, | |
| "eval_samples_per_second": 49.166, | |
| "eval_steps_per_second": 0.195, | |
| "step": 7985 | |
| }, | |
| { | |
| "epoch": 5.009392611145898, | |
| "grad_norm": 0.09022530913352966, | |
| "learning_rate": 2.7725596604745007e-05, | |
| "loss": 0.0386, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.635566687539136, | |
| "grad_norm": 0.003507192712277174, | |
| "learning_rate": 2.4246851735893692e-05, | |
| "loss": 0.0283, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9811709326272433, | |
| "eval_loss": 0.10055801272392273, | |
| "eval_runtime": 139.0437, | |
| "eval_samples_per_second": 48.891, | |
| "eval_steps_per_second": 0.194, | |
| "step": 9582 | |
| }, | |
| { | |
| "epoch": 6.261740763932373, | |
| "grad_norm": 0.6481895446777344, | |
| "learning_rate": 2.076810686704237e-05, | |
| "loss": 0.0322, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.88791484032561, | |
| "grad_norm": 15.71690559387207, | |
| "learning_rate": 1.7289361998191053e-05, | |
| "loss": 0.0229, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9848484848484849, | |
| "eval_loss": 0.09762121737003326, | |
| "eval_runtime": 138.2722, | |
| "eval_samples_per_second": 49.164, | |
| "eval_steps_per_second": 0.195, | |
| "step": 11179 | |
| }, | |
| { | |
| "epoch": 7.514088916718848, | |
| "grad_norm": 0.0020187441259622574, | |
| "learning_rate": 1.3810617129339735e-05, | |
| "loss": 0.0186, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9824948514268903, | |
| "eval_loss": 0.11428509652614594, | |
| "eval_runtime": 139.2486, | |
| "eval_samples_per_second": 48.819, | |
| "eval_steps_per_second": 0.194, | |
| "step": 12776 | |
| }, | |
| { | |
| "epoch": 8.140262993112085, | |
| "grad_norm": 0.6445852518081665, | |
| "learning_rate": 1.0331872260488417e-05, | |
| "loss": 0.0159, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.766437069505322, | |
| "grad_norm": 0.003636428155004978, | |
| "learning_rate": 6.853127391637097e-06, | |
| "loss": 0.0094, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9835245660488379, | |
| "eval_loss": 0.11361410468816757, | |
| "eval_runtime": 139.9155, | |
| "eval_samples_per_second": 48.586, | |
| "eval_steps_per_second": 0.193, | |
| "step": 14373 | |
| }, | |
| { | |
| "epoch": 9.39261114589856, | |
| "grad_norm": 0.0017493839841336012, | |
| "learning_rate": 3.3743825227857787e-06, | |
| "loss": 0.0066, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9833774639599883, | |
| "eval_loss": 0.11724763363599777, | |
| "eval_runtime": 138.7939, | |
| "eval_samples_per_second": 48.979, | |
| "eval_steps_per_second": 0.195, | |
| "step": 15970 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 15970, | |
| "total_flos": 6.415922957241599e+19, | |
| "train_loss": 0.057567716346506634, | |
| "train_runtime": 27132.3161, | |
| "train_samples_per_second": 18.831, | |
| "train_steps_per_second": 0.589 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 15970, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 6.415922957241599e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |