| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.277904328018223, | |
| "eval_steps": 200, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22779043280182232, | |
| "eval_loss": 2.7455391883850098, | |
| "eval_runtime": 186.8417, | |
| "eval_samples_per_second": 37.583, | |
| "eval_steps_per_second": 4.699, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "eval_loss": 1.1437467336654663, | |
| "eval_runtime": 180.9385, | |
| "eval_samples_per_second": 38.809, | |
| "eval_steps_per_second": 4.852, | |
| "eval_wer": 0.7405758779680836, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 1.8708776235580444, | |
| "learning_rate": 0.0002508333333333333, | |
| "loss": 2.6988, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "eval_loss": 0.7285588979721069, | |
| "eval_runtime": 182.6087, | |
| "eval_samples_per_second": 38.454, | |
| "eval_steps_per_second": 4.808, | |
| "eval_wer": 0.5377164192986885, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "eval_loss": 0.6604043245315552, | |
| "eval_runtime": 181.72, | |
| "eval_samples_per_second": 38.642, | |
| "eval_steps_per_second": 4.832, | |
| "eval_wer": 0.4896557222392006, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 0.762070894241333, | |
| "learning_rate": 0.00016849999999999998, | |
| "loss": 0.8304, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "eval_loss": 0.5619407892227173, | |
| "eval_runtime": 181.9456, | |
| "eval_samples_per_second": 38.594, | |
| "eval_steps_per_second": 4.826, | |
| "eval_wer": 0.4338617308228916, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "eval_loss": 0.5691641569137573, | |
| "eval_runtime": 182.4948, | |
| "eval_samples_per_second": 38.478, | |
| "eval_steps_per_second": 4.811, | |
| "eval_wer": 0.4244675055475093, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "eval_loss": 0.563896656036377, | |
| "eval_runtime": 182.6815, | |
| "eval_samples_per_second": 38.438, | |
| "eval_steps_per_second": 4.806, | |
| "eval_wer": 0.4210260566842504, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 4.044058322906494, | |
| "learning_rate": 8.566666666666666e-05, | |
| "loss": 0.71, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "eval_loss": 0.5502046346664429, | |
| "eval_runtime": 183.2379, | |
| "eval_samples_per_second": 38.322, | |
| "eval_steps_per_second": 4.792, | |
| "eval_wer": 0.4186476036088707, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "eval_loss": 0.5441319942474365, | |
| "eval_runtime": 182.7119, | |
| "eval_samples_per_second": 38.432, | |
| "eval_steps_per_second": 4.805, | |
| "eval_wer": 0.4073266984679573, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 9.03835678100586, | |
| "learning_rate": 2.4999999999999998e-06, | |
| "loss": 0.6605, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.5418316721916199, | |
| "eval_runtime": 182.4712, | |
| "eval_samples_per_second": 38.483, | |
| "eval_steps_per_second": 4.812, | |
| "eval_wer": 0.3990751936645451, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "step": 2000, | |
| "total_flos": 7.50335828249239e+18, | |
| "train_loss": 1.2249342193603516, | |
| "train_runtime": 3791.7318, | |
| "train_samples_per_second": 16.879, | |
| "train_steps_per_second": 0.527 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.50335828249239e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |