| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4460, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8804861903190613, | |
| "learning_rate": 0.0008, | |
| "loss": 0.7745, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5559388995170593, | |
| "eval_rouge1": 0.7822, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7815, | |
| "eval_rougeLsum": 0.7779, | |
| "eval_runtime": 36.4825, | |
| "eval_samples_per_second": 20.53, | |
| "eval_steps_per_second": 0.658, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7505515217781067, | |
| "learning_rate": 0.0006, | |
| "loss": 0.604, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5499536395072937, | |
| "eval_rouge1": 0.7671, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7671, | |
| "eval_rougeLsum": 0.7636, | |
| "eval_runtime": 35.8374, | |
| "eval_samples_per_second": 20.9, | |
| "eval_steps_per_second": 0.67, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8223933577537537, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5554, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5185900330543518, | |
| "eval_rouge1": 0.7816, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7796, | |
| "eval_rougeLsum": 0.7794, | |
| "eval_runtime": 36.64, | |
| "eval_samples_per_second": 20.442, | |
| "eval_steps_per_second": 0.655, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.8598939776420593, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5189, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5018280744552612, | |
| "eval_rouge1": 0.7767, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7758, | |
| "eval_rougeLsum": 0.7718, | |
| "eval_runtime": 36.4159, | |
| "eval_samples_per_second": 20.568, | |
| "eval_steps_per_second": 0.659, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.9939805269241333, | |
| "learning_rate": 0.0, | |
| "loss": 0.4887, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.4995948076248169, | |
| "eval_rouge1": 0.756, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7531, | |
| "eval_rougeLsum": 0.7521, | |
| "eval_runtime": 35.7601, | |
| "eval_samples_per_second": 20.945, | |
| "eval_steps_per_second": 0.671, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.92254164058112e+16, | |
| "train_loss": 0.5882825569187045, | |
| "train_runtime": 3728.6347, | |
| "train_samples_per_second": 19.126, | |
| "train_steps_per_second": 1.196 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.92254164058112e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |