| { | |
| "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": 2.7599432468414307, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3987, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.6450759768486023, | |
| "eval_rouge1": 0.4474, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4495, | |
| "eval_rougeLsum": 0.4472, | |
| "eval_runtime": 113.2438, | |
| "eval_samples_per_second": 6.596, | |
| "eval_steps_per_second": 0.212, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.751906633377075, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8591, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5833613276481628, | |
| "eval_rouge1": 0.4222, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4206, | |
| "eval_rougeLsum": 0.4228, | |
| "eval_runtime": 112.1253, | |
| "eval_samples_per_second": 6.662, | |
| "eval_steps_per_second": 0.214, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.9588125944137573, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7949, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5684486031532288, | |
| "eval_rouge1": 0.4462, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4449, | |
| "eval_rougeLsum": 0.4456, | |
| "eval_runtime": 111.7592, | |
| "eval_samples_per_second": 6.684, | |
| "eval_steps_per_second": 0.215, | |
| "step": 5352 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.155653238296509, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7674, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5638545751571655, | |
| "eval_rouge1": 0.4564, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4538, | |
| "eval_rougeLsum": 0.4545, | |
| "eval_runtime": 113.3026, | |
| "eval_samples_per_second": 6.593, | |
| "eval_steps_per_second": 0.212, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.0103201866149902, | |
| "learning_rate": 0.0, | |
| "loss": 0.7542, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 0.5587765574455261, | |
| "eval_rouge1": 0.4494, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.4489, | |
| "eval_rougeLsum": 0.4499, | |
| "eval_runtime": 112.3597, | |
| "eval_samples_per_second": 6.648, | |
| "eval_steps_per_second": 0.214, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 8920, | |
| "total_flos": 7.496541225879552e+16, | |
| "train_loss": 0.914843011009319, | |
| "train_runtime": 5163.073, | |
| "train_samples_per_second": 13.815, | |
| "train_steps_per_second": 1.728 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 7.496541225879552e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |