| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "global_step": 129888, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.6666695181310364e-05, | |
| "loss": 0.6085, | |
| "step": 32472 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 18.2164, | |
| "eval_loss": 0.16991940140724182, | |
| "eval_rouge1": 17.6159, | |
| "eval_rouge2": 10.0939, | |
| "eval_rougeL": 17.5541, | |
| "eval_rougeLsum": 17.5459, | |
| "eval_runtime": 1185.7292, | |
| "eval_samples_per_second": 70.419, | |
| "eval_steps_per_second": 1.957, | |
| "step": 32472 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.1111130120873575e-05, | |
| "loss": 0.178, | |
| "step": 64944 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 18.2056, | |
| "eval_loss": 0.14107763767242432, | |
| "eval_rouge1": 17.7228, | |
| "eval_rouge2": 10.1174, | |
| "eval_rougeL": 17.6623, | |
| "eval_rougeLsum": 17.6545, | |
| "eval_runtime": 1187.4291, | |
| "eval_samples_per_second": 70.318, | |
| "eval_steps_per_second": 1.954, | |
| "step": 64944 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 5.5555650604367875e-06, | |
| "loss": 0.1504, | |
| "step": 97416 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 18.2497, | |
| "eval_loss": 0.13206903636455536, | |
| "eval_rouge1": 17.9419, | |
| "eval_rouge2": 10.3186, | |
| "eval_rougeL": 17.8803, | |
| "eval_rougeLsum": 17.869, | |
| "eval_runtime": 1186.9555, | |
| "eval_samples_per_second": 70.346, | |
| "eval_steps_per_second": 1.955, | |
| "step": 97416 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.1376, | |
| "step": 129888 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 18.253, | |
| "eval_loss": 0.12918101251125336, | |
| "eval_rouge1": 17.9188, | |
| "eval_rouge2": 10.2616, | |
| "eval_rougeL": 17.863, | |
| "eval_rougeLsum": 17.85, | |
| "eval_runtime": 1184.3447, | |
| "eval_samples_per_second": 70.501, | |
| "eval_steps_per_second": 1.959, | |
| "step": 129888 | |
| } | |
| ], | |
| "max_steps": 129888, | |
| "num_train_epochs": 4, | |
| "total_flos": 1.4647320279945216e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |