| { | |
| "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": 1.8557857275009155, | |
| "learning_rate": 0.0008, | |
| "loss": 2.4132, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.2386975288391113, | |
| "eval_rouge1": 0.5009, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.5, | |
| "eval_rougeLsum": 0.5041, | |
| "eval_runtime": 59.5827, | |
| "eval_samples_per_second": 12.588, | |
| "eval_steps_per_second": 0.403, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.0297691822052, | |
| "learning_rate": 0.0006, | |
| "loss": 1.481, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.9344112277030945, | |
| "eval_rouge1": 0.5062, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.5067, | |
| "eval_rougeLsum": 0.5051, | |
| "eval_runtime": 63.1246, | |
| "eval_samples_per_second": 11.881, | |
| "eval_steps_per_second": 0.38, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.9700727462768555, | |
| "learning_rate": 0.0004, | |
| "loss": 1.194, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7902495265007019, | |
| "eval_rouge1": 0.6449, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6488, | |
| "eval_rougeLsum": 0.6454, | |
| "eval_runtime": 59.7895, | |
| "eval_samples_per_second": 12.544, | |
| "eval_steps_per_second": 0.401, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.7162905931472778, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0257, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.7315613627433777, | |
| "eval_rouge1": 0.4501, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4461, | |
| "eval_rougeLsum": 0.4508, | |
| "eval_runtime": 61.0508, | |
| "eval_samples_per_second": 12.285, | |
| "eval_steps_per_second": 0.393, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.0185036659240723, | |
| "learning_rate": 0.0, | |
| "loss": 0.9184, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6921437382698059, | |
| "eval_rouge1": 0.5367, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.5393, | |
| "eval_rougeLsum": 0.5356, | |
| "eval_runtime": 60.6533, | |
| "eval_samples_per_second": 12.365, | |
| "eval_steps_per_second": 0.396, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.591093554778112e+16, | |
| "train_loss": 1.4064704381831559, | |
| "train_runtime": 6422.8288, | |
| "train_samples_per_second": 11.103, | |
| "train_steps_per_second": 0.694 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.591093554778112e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |