| { | |
| "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": 1.6540483236312866, | |
| "learning_rate": 0.0008, | |
| "loss": 2.4494, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.2649613618850708, | |
| "eval_rouge1": 0.2566, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.2573, | |
| "eval_rougeLsum": 0.2579, | |
| "eval_runtime": 60.8639, | |
| "eval_samples_per_second": 12.273, | |
| "eval_steps_per_second": 0.394, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.7608369588851929, | |
| "learning_rate": 0.0006, | |
| "loss": 1.5203, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.9467156529426575, | |
| "eval_rouge1": 0.3425, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.3452, | |
| "eval_rougeLsum": 0.3442, | |
| "eval_runtime": 61.4729, | |
| "eval_samples_per_second": 12.152, | |
| "eval_steps_per_second": 0.39, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.6110718250274658, | |
| "learning_rate": 0.0004, | |
| "loss": 1.2188, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7931520342826843, | |
| "eval_rouge1": 0.3497, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.3495, | |
| "eval_rougeLsum": 0.3497, | |
| "eval_runtime": 64.3107, | |
| "eval_samples_per_second": 11.615, | |
| "eval_steps_per_second": 0.373, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.5266708135604858, | |
| "learning_rate": 0.0002, | |
| "loss": 1.045, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7192575335502625, | |
| "eval_rouge1": 0.4033, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4016, | |
| "eval_rougeLsum": 0.4056, | |
| "eval_runtime": 61.2273, | |
| "eval_samples_per_second": 12.2, | |
| "eval_steps_per_second": 0.392, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.427673578262329, | |
| "learning_rate": 0.0, | |
| "loss": 0.9373, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6836755871772766, | |
| "eval_rouge1": 0.7788, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7826, | |
| "eval_rougeLsum": 0.7777, | |
| "eval_runtime": 63.3918, | |
| "eval_samples_per_second": 11.784, | |
| "eval_steps_per_second": 0.379, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.592661664034816e+16, | |
| "train_loss": 1.4341732418590596, | |
| "train_runtime": 6521.1507, | |
| "train_samples_per_second": 10.938, | |
| "train_steps_per_second": 0.684 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.592661664034816e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |