| { | |
| "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": 1.0901676416397095, | |
| "learning_rate": 0.0008, | |
| "loss": 3.0294, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.9919179677963257, | |
| "eval_rouge1": 0.672, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6729, | |
| "eval_rougeLsum": 0.6683, | |
| "eval_runtime": 51.0214, | |
| "eval_samples_per_second": 14.7, | |
| "eval_steps_per_second": 0.47, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0671581029891968, | |
| "learning_rate": 0.0006, | |
| "loss": 2.4162, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.7555127143859863, | |
| "eval_rouge1": 0.7269, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7253, | |
| "eval_rougeLsum": 0.727, | |
| "eval_runtime": 52.3423, | |
| "eval_samples_per_second": 14.329, | |
| "eval_steps_per_second": 0.459, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.3621200323104858, | |
| "learning_rate": 0.0004, | |
| "loss": 2.1684, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.5470179319381714, | |
| "eval_rouge1": 0.693, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6917, | |
| "eval_rougeLsum": 0.6933, | |
| "eval_runtime": 51.2029, | |
| "eval_samples_per_second": 14.648, | |
| "eval_steps_per_second": 0.469, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.170156717300415, | |
| "learning_rate": 0.0002, | |
| "loss": 1.9893, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.430000901222229, | |
| "eval_rouge1": 0.6992, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6959, | |
| "eval_rougeLsum": 0.6968, | |
| "eval_runtime": 50.7411, | |
| "eval_samples_per_second": 14.781, | |
| "eval_steps_per_second": 0.473, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.9514837265014648, | |
| "learning_rate": 0.0, | |
| "loss": 1.8466, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.3436046838760376, | |
| "eval_rouge1": 0.6802, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6802, | |
| "eval_rougeLsum": 0.6778, | |
| "eval_runtime": 53.0932, | |
| "eval_samples_per_second": 14.126, | |
| "eval_steps_per_second": 0.452, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_loss": 2.2899972684714824, | |
| "train_runtime": 5739.3575, | |
| "train_samples_per_second": 12.425, | |
| "train_steps_per_second": 0.777 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |