| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.73145197997269, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.36198097467422485, | |
| "learning_rate": 6.0716454159077105e-05, | |
| "loss": 0.4455, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_loss": 0.6183120608329773, | |
| "eval_runtime": 195.3588, | |
| "eval_samples_per_second": 2.37, | |
| "eval_steps_per_second": 2.37, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.1898297369480133, | |
| "learning_rate": 0.00012143290831815421, | |
| "loss": 0.0995, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "eval_loss": 0.6644699573516846, | |
| "eval_runtime": 195.5381, | |
| "eval_samples_per_second": 2.368, | |
| "eval_steps_per_second": 2.368, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.25010770559310913, | |
| "learning_rate": 0.00018214936247723133, | |
| "loss": 0.0858, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_loss": 0.6735700368881226, | |
| "eval_runtime": 195.292, | |
| "eval_samples_per_second": 2.371, | |
| "eval_steps_per_second": 2.371, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.021668555215001106, | |
| "learning_rate": 6.0716454159077105e-05, | |
| "loss": 0.0613, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "eval_loss": 0.7165372967720032, | |
| "eval_runtime": 196.7408, | |
| "eval_samples_per_second": 2.353, | |
| "eval_steps_per_second": 2.353, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 0.2751568555831909, | |
| "learning_rate": 0.00012143290831815421, | |
| "loss": 0.0498, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_loss": 0.6982098817825317, | |
| "eval_runtime": 192.741, | |
| "eval_samples_per_second": 2.402, | |
| "eval_steps_per_second": 2.402, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.0906854197382927, | |
| "learning_rate": 0.00018214936247723133, | |
| "loss": 0.0512, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "eval_loss": 0.6751747131347656, | |
| "eval_runtime": 193.0682, | |
| "eval_samples_per_second": 2.398, | |
| "eval_steps_per_second": 2.398, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 54900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 2.5337057909011123e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |