| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 8920, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.1375385522842407, | |
| "learning_rate": 4e-05, | |
| "loss": 1.2253, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6576545834541321, | |
| "eval_rouge1": 0.4006, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.3969, | |
| "eval_rougeLsum": 0.4007, | |
| "eval_runtime": 113.3266, | |
| "eval_samples_per_second": 6.618, | |
| "eval_steps_per_second": 0.212, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6233590841293335, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7893, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6133397817611694, | |
| "eval_rouge1": 0.4188, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4167, | |
| "eval_rougeLsum": 0.4194, | |
| "eval_runtime": 112.0833, | |
| "eval_samples_per_second": 6.691, | |
| "eval_steps_per_second": 0.214, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.4971694946289062, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7351, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6062405109405518, | |
| "eval_rouge1": 0.3925, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.3898, | |
| "eval_rougeLsum": 0.3908, | |
| "eval_runtime": 112.7718, | |
| "eval_samples_per_second": 6.651, | |
| "eval_steps_per_second": 0.213, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.2644437551498413, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7092, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.599039614200592, | |
| "eval_rouge1": 0.3834, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.381, | |
| "eval_rougeLsum": 0.3826, | |
| "eval_runtime": 112.5655, | |
| "eval_samples_per_second": 6.663, | |
| "eval_steps_per_second": 0.213, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.0209782123565674, | |
| "learning_rate": 0.0, | |
| "loss": 0.6978, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5916664600372314, | |
| "eval_rouge1": 0.3832, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.3816, | |
| "eval_rougeLsum": 0.3824, | |
| "eval_runtime": 112.9978, | |
| "eval_samples_per_second": 6.637, | |
| "eval_steps_per_second": 0.212, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 8920, | |
| "total_flos": 7.534082864399155e+16, | |
| "train_loss": 0.8313294962382637, | |
| "train_runtime": 5163.645, | |
| "train_samples_per_second": 13.82, | |
| "train_steps_per_second": 1.727 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 7.534082864399155e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |