| { | |
| "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": 0.7220906615257263, | |
| "learning_rate": 0.0008, | |
| "loss": 0.8166, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.541189432144165, | |
| "eval_rouge1": 0.6958, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6933, | |
| "eval_rougeLsum": 0.6944, | |
| "eval_runtime": 38.2682, | |
| "eval_samples_per_second": 19.599, | |
| "eval_steps_per_second": 0.627, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7783016562461853, | |
| "learning_rate": 0.0006, | |
| "loss": 0.6181, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5291976928710938, | |
| "eval_rouge1": 0.6767, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6749, | |
| "eval_rougeLsum": 0.6733, | |
| "eval_runtime": 38.0243, | |
| "eval_samples_per_second": 19.724, | |
| "eval_steps_per_second": 0.631, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.717491865158081, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5768, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.4997425377368927, | |
| "eval_rouge1": 0.6835, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6819, | |
| "eval_rougeLsum": 0.6808, | |
| "eval_runtime": 37.8683, | |
| "eval_samples_per_second": 19.805, | |
| "eval_steps_per_second": 0.634, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5613729953765869, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5516, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5060973763465881, | |
| "eval_rouge1": 0.6866, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6872, | |
| "eval_rougeLsum": 0.6829, | |
| "eval_runtime": 37.6793, | |
| "eval_samples_per_second": 19.905, | |
| "eval_steps_per_second": 0.637, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.717795193195343, | |
| "learning_rate": 0.0, | |
| "loss": 0.5337, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.49971678853034973, | |
| "eval_rouge1": 0.7062, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7054, | |
| "eval_rougeLsum": 0.7026, | |
| "eval_runtime": 37.7942, | |
| "eval_samples_per_second": 19.844, | |
| "eval_steps_per_second": 0.635, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.90238608932864e+16, | |
| "train_loss": 0.6193654355446854, | |
| "train_runtime": 3915.3347, | |
| "train_samples_per_second": 18.213, | |
| "train_steps_per_second": 1.139 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.90238608932864e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |