| { | |
| "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": 1.1466261148452759, | |
| "learning_rate": 0.0008, | |
| "loss": 1.1936, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7944247126579285, | |
| "eval_rouge1": 0.7769, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7776, | |
| "eval_rougeLsum": 0.7776, | |
| "eval_runtime": 34.2455, | |
| "eval_samples_per_second": 21.871, | |
| "eval_steps_per_second": 0.701, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6689993739128113, | |
| "learning_rate": 0.0006, | |
| "loss": 0.678, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6952247023582458, | |
| "eval_rouge1": 0.7516, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7464, | |
| "eval_rougeLsum": 0.7488, | |
| "eval_runtime": 34.3119, | |
| "eval_samples_per_second": 21.829, | |
| "eval_steps_per_second": 0.699, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.6889210939407349, | |
| "learning_rate": 0.0004, | |
| "loss": 0.4823, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6426491141319275, | |
| "eval_rouge1": 0.752, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7509, | |
| "eval_rougeLsum": 0.7499, | |
| "eval_runtime": 35.0834, | |
| "eval_samples_per_second": 21.349, | |
| "eval_steps_per_second": 0.684, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6657813191413879, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3348, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.665139377117157, | |
| "eval_rouge1": 0.7449, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7455, | |
| "eval_rougeLsum": 0.7449, | |
| "eval_runtime": 36.0479, | |
| "eval_samples_per_second": 20.778, | |
| "eval_steps_per_second": 0.666, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.4766407012939453, | |
| "learning_rate": 0.0, | |
| "loss": 0.1967, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7342413067817688, | |
| "eval_rouge1": 0.7526, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7508, | |
| "eval_rougeLsum": 0.7516, | |
| "eval_runtime": 35.5667, | |
| "eval_samples_per_second": 21.059, | |
| "eval_steps_per_second": 0.675, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.883347058982912e+16, | |
| "train_loss": 0.5770811448717331, | |
| "train_runtime": 4315.4414, | |
| "train_samples_per_second": 16.526, | |
| "train_steps_per_second": 1.033 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.883347058982912e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |