| { | |
| "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": 0.6613966226577759, | |
| "learning_rate": 0.0008, | |
| "loss": 0.7782, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5458054542541504, | |
| "eval_rouge1": 0.7706, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7715, | |
| "eval_rougeLsum": 0.7691, | |
| "eval_runtime": 38.2766, | |
| "eval_samples_per_second": 19.568, | |
| "eval_steps_per_second": 0.627, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6847044825553894, | |
| "learning_rate": 0.0006, | |
| "loss": 0.5952, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5551263689994812, | |
| "eval_rouge1": 0.7627, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7591, | |
| "eval_rougeLsum": 0.7596, | |
| "eval_runtime": 37.6956, | |
| "eval_samples_per_second": 19.87, | |
| "eval_steps_per_second": 0.637, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.6185162663459778, | |
| "learning_rate": 0.0004, | |
| "loss": 0.551, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5163312554359436, | |
| "eval_rouge1": 0.7617, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.761, | |
| "eval_rougeLsum": 0.7609, | |
| "eval_runtime": 38.2354, | |
| "eval_samples_per_second": 19.589, | |
| "eval_steps_per_second": 0.628, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6471401453018188, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5236, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5063948631286621, | |
| "eval_rouge1": 0.7569, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7553, | |
| "eval_rougeLsum": 0.7552, | |
| "eval_runtime": 38.361, | |
| "eval_samples_per_second": 19.525, | |
| "eval_steps_per_second": 0.626, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6893085837364197, | |
| "learning_rate": 0.0, | |
| "loss": 0.5009, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5019394755363464, | |
| "eval_rouge1": 0.762, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7611, | |
| "eval_rougeLsum": 0.7586, | |
| "eval_runtime": 37.6497, | |
| "eval_samples_per_second": 19.894, | |
| "eval_steps_per_second": 0.637, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.922112594935808e+16, | |
| "train_loss": 0.5897949560875315, | |
| "train_runtime": 3934.2037, | |
| "train_samples_per_second": 18.127, | |
| "train_steps_per_second": 1.134 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.922112594935808e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |