| { | |
| "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": 0.9127490520477295, | |
| "learning_rate": 0.0008, | |
| "loss": 3.0347, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.0001704692840576, | |
| "eval_rouge1": 0.6717, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6744, | |
| "eval_rougeLsum": 0.6696, | |
| "eval_runtime": 53.7591, | |
| "eval_samples_per_second": 13.895, | |
| "eval_steps_per_second": 0.446, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7276727557182312, | |
| "learning_rate": 0.0006, | |
| "loss": 2.4012, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.7295931577682495, | |
| "eval_rouge1": 0.7117, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7105, | |
| "eval_rougeLsum": 0.7111, | |
| "eval_runtime": 55.151, | |
| "eval_samples_per_second": 13.545, | |
| "eval_steps_per_second": 0.435, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8275261521339417, | |
| "learning_rate": 0.0004, | |
| "loss": 2.1461, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.478960633277893, | |
| "eval_rouge1": 0.6953, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6922, | |
| "eval_rougeLsum": 0.6902, | |
| "eval_runtime": 52.4262, | |
| "eval_samples_per_second": 14.249, | |
| "eval_steps_per_second": 0.458, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9719261527061462, | |
| "learning_rate": 0.0002, | |
| "loss": 1.952, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.3588956594467163, | |
| "eval_rouge1": 0.6758, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6753, | |
| "eval_rougeLsum": 0.6736, | |
| "eval_runtime": 54.0573, | |
| "eval_samples_per_second": 13.819, | |
| "eval_steps_per_second": 0.444, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.8105780482292175, | |
| "learning_rate": 0.0, | |
| "loss": 1.8052, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.2773828506469727, | |
| "eval_rouge1": 0.6685, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6677, | |
| "eval_rougeLsum": 0.6651, | |
| "eval_runtime": 55.3406, | |
| "eval_samples_per_second": 13.498, | |
| "eval_steps_per_second": 0.434, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.531847691862016e+16, | |
| "train_loss": 2.267836908588495, | |
| "train_runtime": 5835.4771, | |
| "train_samples_per_second": 12.224, | |
| "train_steps_per_second": 0.764 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.531847691862016e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |