| { | |
| "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": 0.5049543976783752, | |
| "learning_rate": 0.0008, | |
| "loss": 0.8224, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5704124569892883, | |
| "eval_rouge1": 0.6572, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6564, | |
| "eval_rougeLsum": 0.655, | |
| "eval_runtime": 38.0247, | |
| "eval_samples_per_second": 19.724, | |
| "eval_steps_per_second": 0.631, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.47203755378723145, | |
| "learning_rate": 0.0006, | |
| "loss": 0.6196, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5431475043296814, | |
| "eval_rouge1": 0.6602, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6579, | |
| "eval_rougeLsum": 0.6606, | |
| "eval_runtime": 37.5492, | |
| "eval_samples_per_second": 19.974, | |
| "eval_steps_per_second": 0.639, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5097192525863647, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5778, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5372552871704102, | |
| "eval_rouge1": 0.6757, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6756, | |
| "eval_rougeLsum": 0.6745, | |
| "eval_runtime": 38.295, | |
| "eval_samples_per_second": 19.585, | |
| "eval_steps_per_second": 0.627, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.4385349154472351, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5503, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.525574266910553, | |
| "eval_rouge1": 0.659, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6569, | |
| "eval_rougeLsum": 0.6586, | |
| "eval_runtime": 38.268, | |
| "eval_samples_per_second": 19.599, | |
| "eval_steps_per_second": 0.627, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.4007055461406708, | |
| "learning_rate": 0.0, | |
| "loss": 0.5343, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5211541056632996, | |
| "eval_rouge1": 0.6659, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6648, | |
| "eval_rougeLsum": 0.6661, | |
| "eval_runtime": 38.0598, | |
| "eval_samples_per_second": 19.706, | |
| "eval_steps_per_second": 0.631, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.90582346563584e+16, | |
| "train_loss": 0.6208877221351247, | |
| "train_runtime": 3924.9461, | |
| "train_samples_per_second": 18.181, | |
| "train_steps_per_second": 1.136 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.90582346563584e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |