| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9333333333333336, | |
| "eval_steps": 500, | |
| "global_step": 33, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.662017672688027, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6498, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 21.037803085201503, | |
| "learning_rate": 1.9941379571543597e-05, | |
| "loss": 0.6636, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.8410813209282253, | |
| "learning_rate": 1.796093065705644e-05, | |
| "loss": 0.6676, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_loss": 0.6012949347496033, | |
| "eval_runtime": 3.7225, | |
| "eval_samples_per_second": 4.835, | |
| "eval_steps_per_second": 0.537, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 2.0997223596616403, | |
| "learning_rate": 1.3701381553399147e-05, | |
| "loss": 0.4598, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.5911059539783716, | |
| "learning_rate": 8.382180034472353e-06, | |
| "loss": 0.3797, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "eval_loss": 0.5361791849136353, | |
| "eval_runtime": 3.9906, | |
| "eval_samples_per_second": 4.511, | |
| "eval_steps_per_second": 0.501, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 1.1552775795971262, | |
| "learning_rate": 3.5261371521817247e-06, | |
| "loss": 0.3325, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.998358961831719, | |
| "learning_rate": 5.234682881719766e-07, | |
| "loss": 0.2865, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_loss": 0.525164783000946, | |
| "eval_runtime": 3.8934, | |
| "eval_samples_per_second": 4.623, | |
| "eval_steps_per_second": 0.514, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "step": 33, | |
| "total_flos": 6857183723520.0, | |
| "train_loss": 0.4476780457930131, | |
| "train_runtime": 777.7129, | |
| "train_samples_per_second": 2.727, | |
| "train_steps_per_second": 0.042 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 33, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 6857183723520.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |