| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4470, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.573519229888916, | |
| "learning_rate": 0.0008, | |
| "loss": 0.827, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.517066240310669, | |
| "eval_rouge1": 0.6517, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6531, | |
| "eval_rougeLsum": 0.6496, | |
| "eval_runtime": 37.7451, | |
| "eval_samples_per_second": 19.791, | |
| "eval_steps_per_second": 0.636, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5823910236358643, | |
| "learning_rate": 0.0006, | |
| "loss": 0.6257, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.4943862855434418, | |
| "eval_rouge1": 0.6472, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6472, | |
| "eval_rougeLsum": 0.6469, | |
| "eval_runtime": 37.6794, | |
| "eval_samples_per_second": 19.825, | |
| "eval_steps_per_second": 0.637, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9307901859283447, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5832, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.48476719856262207, | |
| "eval_rouge1": 0.6329, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6335, | |
| "eval_rougeLsum": 0.634, | |
| "eval_runtime": 37.5561, | |
| "eval_samples_per_second": 19.89, | |
| "eval_steps_per_second": 0.639, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5210844874382019, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5581, | |
| "step": 3576 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.48243632912635803, | |
| "eval_rouge1": 0.6575, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6562, | |
| "eval_rougeLsum": 0.6576, | |
| "eval_runtime": 37.1819, | |
| "eval_samples_per_second": 20.09, | |
| "eval_steps_per_second": 0.645, | |
| "step": 3576 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.741267740726471, | |
| "learning_rate": 0.0, | |
| "loss": 0.5411, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.47984424233436584, | |
| "eval_rouge1": 0.6568, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6564, | |
| "eval_rougeLsum": 0.6586, | |
| "eval_runtime": 37.5756, | |
| "eval_samples_per_second": 19.88, | |
| "eval_steps_per_second": 0.639, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4470, | |
| "total_flos": 4.9120107429888e+16, | |
| "train_loss": 0.6270396870521358, | |
| "train_runtime": 3918.4765, | |
| "train_samples_per_second": 18.234, | |
| "train_steps_per_second": 1.141 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4470, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.9120107429888e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |