| { | |
| "best_metric": 6.446480382145557, | |
| "best_model_checkpoint": "/opt/dlami/nvme/AAA_V2/checkpoint-1584", | |
| "epoch": 11.0, | |
| "eval_steps": 500, | |
| "global_step": 2178, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.28139322996139526, | |
| "learning_rate": 0.0007714285714285713, | |
| "loss": 0.0358, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 6.260606562914056, | |
| "eval_loss": 0.0037530818954110146, | |
| "eval_runtime": 73.4886, | |
| "eval_samples_per_second": 10.736, | |
| "eval_steps_per_second": 0.68, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.16617226600646973, | |
| "learning_rate": 0.0006428571428571428, | |
| "loss": 0.0052, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 6.308825370321139, | |
| "eval_loss": 0.001781024388037622, | |
| "eval_runtime": 73.278, | |
| "eval_samples_per_second": 10.767, | |
| "eval_steps_per_second": 0.682, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.1758955419063568, | |
| "learning_rate": 0.0005142857142857142, | |
| "loss": 0.0025, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 6.368129059471027, | |
| "eval_loss": 0.0011983831645920873, | |
| "eval_runtime": 73.2664, | |
| "eval_samples_per_second": 10.769, | |
| "eval_steps_per_second": 0.682, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0046499622985720634, | |
| "learning_rate": 0.00038571428571428567, | |
| "loss": 0.0015, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 6.385440196858329, | |
| "eval_loss": 0.0010980970691889524, | |
| "eval_runtime": 73.1472, | |
| "eval_samples_per_second": 10.786, | |
| "eval_steps_per_second": 0.684, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.0024423596914857626, | |
| "learning_rate": 0.0002571428571428571, | |
| "loss": 0.0008, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 6.41579040812938, | |
| "eval_loss": 0.0010172681650146842, | |
| "eval_runtime": 73.3825, | |
| "eval_samples_per_second": 10.752, | |
| "eval_steps_per_second": 0.681, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.00022464522044174373, | |
| "learning_rate": 0.00012857142857142855, | |
| "loss": 0.0006, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 6.392992917575001, | |
| "eval_loss": 0.0009776438819244504, | |
| "eval_runtime": 73.1926, | |
| "eval_samples_per_second": 10.78, | |
| "eval_steps_per_second": 0.683, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.0007703190785832703, | |
| "learning_rate": 0.0, | |
| "loss": 0.0004, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 6.4192986039574595, | |
| "eval_loss": 0.0009626392857171595, | |
| "eval_runtime": 73.2421, | |
| "eval_samples_per_second": 10.772, | |
| "eval_steps_per_second": 0.683, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.616127989720553e-05, | |
| "learning_rate": 0.0003, | |
| "loss": 0.0006, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_bleu": 6.446480382145557, | |
| "eval_loss": 0.0012350629549473524, | |
| "eval_runtime": 73.8593, | |
| "eval_samples_per_second": 10.682, | |
| "eval_steps_per_second": 0.677, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.939168113400228e-05, | |
| "learning_rate": 0.000225, | |
| "loss": 0.0008, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_bleu": 6.435039176248687, | |
| "eval_loss": 0.0010175537317991257, | |
| "eval_runtime": 73.5317, | |
| "eval_samples_per_second": 10.73, | |
| "eval_steps_per_second": 0.68, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.800251604639925e-05, | |
| "learning_rate": 0.00015, | |
| "loss": 0.0006, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_bleu": 6.399756503138505, | |
| "eval_loss": 0.0010247103637084365, | |
| "eval_runtime": 73.5802, | |
| "eval_samples_per_second": 10.723, | |
| "eval_steps_per_second": 0.68, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.0001712931552901864, | |
| "learning_rate": 7.5e-05, | |
| "loss": 0.0004, | |
| "step": 2178 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_bleu": 6.409543063646119, | |
| "eval_loss": 0.0009960704483091831, | |
| "eval_runtime": 73.5452, | |
| "eval_samples_per_second": 10.728, | |
| "eval_steps_per_second": 0.68, | |
| "step": 2178 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2376, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.112048651829248e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |