| { | |
| "best_metric": 0.9872021182700794, | |
| "best_model_checkpoint": "audio/train/checkpoint/hubert-xlarge/superb_ks_42/checkpoint-12776", | |
| "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": 4.316092491149902, | |
| "learning_rate": 3.1308703819661866e-05, | |
| "loss": 1.5639, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9794057075610474, | |
| "eval_loss": 0.10121691972017288, | |
| "eval_runtime": 36.0525, | |
| "eval_samples_per_second": 188.558, | |
| "eval_steps_per_second": 0.749, | |
| "step": 1597 | |
| }, | |
| { | |
| "epoch": 1.2523481527864746, | |
| "grad_norm": 1.4483940601348877, | |
| "learning_rate": 4.859806581785292e-05, | |
| "loss": 0.3591, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.878522229179712, | |
| "grad_norm": 0.9219438433647156, | |
| "learning_rate": 4.51193209490016e-05, | |
| "loss": 0.205, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9841129744042365, | |
| "eval_loss": 0.07202675193548203, | |
| "eval_runtime": 35.6833, | |
| "eval_samples_per_second": 190.509, | |
| "eval_steps_per_second": 0.757, | |
| "step": 3194 | |
| }, | |
| { | |
| "epoch": 2.504696305572949, | |
| "grad_norm": 1.6926661729812622, | |
| "learning_rate": 4.1640576080150286e-05, | |
| "loss": 0.1774, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9855839952927331, | |
| "eval_loss": 0.06639665365219116, | |
| "eval_runtime": 35.3875, | |
| "eval_samples_per_second": 192.102, | |
| "eval_steps_per_second": 0.763, | |
| "step": 4791 | |
| }, | |
| { | |
| "epoch": 3.1308703819661865, | |
| "grad_norm": 1.2424182891845703, | |
| "learning_rate": 3.8161831211298965e-05, | |
| "loss": 0.159, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.757044458359424, | |
| "grad_norm": 0.05382406339049339, | |
| "learning_rate": 3.468308634244764e-05, | |
| "loss": 0.1549, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9855839952927331, | |
| "eval_loss": 0.06212027370929718, | |
| "eval_runtime": 35.7277, | |
| "eval_samples_per_second": 190.273, | |
| "eval_steps_per_second": 0.756, | |
| "step": 6388 | |
| }, | |
| { | |
| "epoch": 4.383218534752661, | |
| "grad_norm": 0.4753202199935913, | |
| "learning_rate": 3.120434147359633e-05, | |
| "loss": 0.1416, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9858781994704324, | |
| "eval_loss": 0.06196163594722748, | |
| "eval_runtime": 35.6118, | |
| "eval_samples_per_second": 190.892, | |
| "eval_steps_per_second": 0.758, | |
| "step": 7985 | |
| }, | |
| { | |
| "epoch": 5.009392611145898, | |
| "grad_norm": 1.2926490306854248, | |
| "learning_rate": 2.7725596604745007e-05, | |
| "loss": 0.1378, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.635566687539136, | |
| "grad_norm": 0.03198292478919029, | |
| "learning_rate": 2.4246851735893692e-05, | |
| "loss": 0.1328, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9864666078258312, | |
| "eval_loss": 0.061353616416454315, | |
| "eval_runtime": 35.411, | |
| "eval_samples_per_second": 191.974, | |
| "eval_steps_per_second": 0.762, | |
| "step": 9582 | |
| }, | |
| { | |
| "epoch": 6.261740763932373, | |
| "grad_norm": 1.4529321193695068, | |
| "learning_rate": 2.076810686704237e-05, | |
| "loss": 0.1244, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.88791484032561, | |
| "grad_norm": 1.9151487350463867, | |
| "learning_rate": 1.7289361998191053e-05, | |
| "loss": 0.1159, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9864666078258312, | |
| "eval_loss": 0.06154850497841835, | |
| "eval_runtime": 35.249, | |
| "eval_samples_per_second": 192.857, | |
| "eval_steps_per_second": 0.766, | |
| "step": 11179 | |
| }, | |
| { | |
| "epoch": 7.514088916718848, | |
| "grad_norm": 4.778970241546631, | |
| "learning_rate": 1.3810617129339735e-05, | |
| "loss": 0.1211, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9872021182700794, | |
| "eval_loss": 0.057863999158144, | |
| "eval_runtime": 35.7046, | |
| "eval_samples_per_second": 190.396, | |
| "eval_steps_per_second": 0.756, | |
| "step": 12776 | |
| }, | |
| { | |
| "epoch": 8.140262993112085, | |
| "grad_norm": 0.4232797622680664, | |
| "learning_rate": 1.0331872260488417e-05, | |
| "loss": 0.114, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.766437069505322, | |
| "grad_norm": 1.9394092559814453, | |
| "learning_rate": 6.853127391637097e-06, | |
| "loss": 0.108, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9863195057369815, | |
| "eval_loss": 0.05663929507136345, | |
| "eval_runtime": 35.5366, | |
| "eval_samples_per_second": 191.296, | |
| "eval_steps_per_second": 0.76, | |
| "step": 14373 | |
| }, | |
| { | |
| "epoch": 9.39261114589856, | |
| "grad_norm": 3.1410608291625977, | |
| "learning_rate": 3.3743825227857787e-06, | |
| "loss": 0.1088, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9869079140923801, | |
| "eval_loss": 0.05716182291507721, | |
| "eval_runtime": 35.7353, | |
| "eval_samples_per_second": 190.232, | |
| "eval_steps_per_second": 0.756, | |
| "step": 15970 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 15970, | |
| "total_flos": 4.722696547222273e+19, | |
| "train_loss": 0.23973732257382005, | |
| "train_runtime": 8088.8257, | |
| "train_samples_per_second": 63.166, | |
| "train_steps_per_second": 1.974 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 15970, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 4.722696547222273e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |