| { | |
| "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": 1.0061594247817993, | |
| "learning_rate": 0.0008, | |
| "loss": 3.3224, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.384333372116089, | |
| "eval_rouge1": 0.6887, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6879, | |
| "eval_rougeLsum": 0.6876, | |
| "eval_runtime": 58.8999, | |
| "eval_samples_per_second": 12.733, | |
| "eval_steps_per_second": 0.407, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.9041750431060791, | |
| "learning_rate": 0.0006, | |
| "loss": 2.913, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.155667781829834, | |
| "eval_rouge1": 0.7307, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7301, | |
| "eval_rougeLsum": 0.7315, | |
| "eval_runtime": 52.5159, | |
| "eval_samples_per_second": 14.281, | |
| "eval_steps_per_second": 0.457, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0807160139083862, | |
| "learning_rate": 0.0004, | |
| "loss": 2.7028, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.9775158166885376, | |
| "eval_rouge1": 0.7149, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7155, | |
| "eval_rougeLsum": 0.7105, | |
| "eval_runtime": 53.7344, | |
| "eval_samples_per_second": 13.958, | |
| "eval_steps_per_second": 0.447, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9261811971664429, | |
| "learning_rate": 0.0002, | |
| "loss": 2.5387, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.8682197332382202, | |
| "eval_rouge1": 0.7337, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.731, | |
| "eval_rougeLsum": 0.7295, | |
| "eval_runtime": 53.1542, | |
| "eval_samples_per_second": 14.11, | |
| "eval_steps_per_second": 0.452, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.018937110900879, | |
| "learning_rate": 0.0, | |
| "loss": 2.4084, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.780213713645935, | |
| "eval_rouge1": 0.696, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6965, | |
| "eval_rougeLsum": 0.6913, | |
| "eval_runtime": 51.9124, | |
| "eval_samples_per_second": 14.447, | |
| "eval_steps_per_second": 0.462, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_loss": 2.7770426797225336, | |
| "train_runtime": 5662.7265, | |
| "train_samples_per_second": 12.593, | |
| "train_steps_per_second": 0.788 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |