| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 8920, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.2753658294677734, | |
| "learning_rate": 4e-05, | |
| "loss": 1.233, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6087062954902649, | |
| "eval_rouge1": 0.4912, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4859, | |
| "eval_rougeLsum": 0.4881, | |
| "eval_runtime": 112.5011, | |
| "eval_samples_per_second": 6.64, | |
| "eval_steps_per_second": 0.213, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.76861834526062, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7935, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5581778883934021, | |
| "eval_rouge1": 0.4176, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4157, | |
| "eval_rougeLsum": 0.4171, | |
| "eval_runtime": 112.1842, | |
| "eval_samples_per_second": 6.659, | |
| "eval_steps_per_second": 0.214, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.8127230405807495, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7385, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5451030731201172, | |
| "eval_rouge1": 0.4227, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4192, | |
| "eval_rougeLsum": 0.4214, | |
| "eval_runtime": 112.2353, | |
| "eval_samples_per_second": 6.656, | |
| "eval_steps_per_second": 0.214, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.9295331239700317, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7114, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5406314730644226, | |
| "eval_rouge1": 0.4115, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4115, | |
| "eval_rougeLsum": 0.4106, | |
| "eval_runtime": 112.3402, | |
| "eval_samples_per_second": 6.649, | |
| "eval_steps_per_second": 0.214, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.6341142654418945, | |
| "learning_rate": 0.0, | |
| "loss": 0.6996, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5392429828643799, | |
| "eval_rouge1": 0.4255, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4247, | |
| "eval_rougeLsum": 0.4257, | |
| "eval_runtime": 113.1717, | |
| "eval_samples_per_second": 6.601, | |
| "eval_steps_per_second": 0.212, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 8920, | |
| "total_flos": 7.526178534957466e+16, | |
| "train_loss": 0.8352148081689672, | |
| "train_runtime": 5166.7453, | |
| "train_samples_per_second": 13.806, | |
| "train_steps_per_second": 1.726 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 7.526178534957466e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |