| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "global_step": 3995, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7496871088861077e-05, | |
| "loss": 0.5079, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5189032554626465, | |
| "eval_runtime": 2.8243, | |
| "eval_samples_per_second": 1225.772, | |
| "eval_steps_per_second": 9.914, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 1.4993742177722154e-05, | |
| "loss": 0.4379, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 1.249061326658323e-05, | |
| "loss": 0.3846, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.5475935935974121, | |
| "eval_runtime": 2.8018, | |
| "eval_samples_per_second": 1235.643, | |
| "eval_steps_per_second": 9.994, | |
| "step": 1598 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.987484355444305e-06, | |
| "loss": 0.3219, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5914019346237183, | |
| "eval_runtime": 2.7995, | |
| "eval_samples_per_second": 1236.649, | |
| "eval_steps_per_second": 10.002, | |
| "step": 2397 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 7.484355444305382e-06, | |
| "loss": 0.2964, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 4.981226533166458e-06, | |
| "loss": 0.2524, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.6955985426902771, | |
| "eval_runtime": 2.8279, | |
| "eval_samples_per_second": 1224.237, | |
| "eval_steps_per_second": 9.901, | |
| "step": 3196 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "learning_rate": 2.4780976220275346e-06, | |
| "loss": 0.2253, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.7501208186149597, | |
| "eval_runtime": 2.7917, | |
| "eval_samples_per_second": 1240.104, | |
| "eval_steps_per_second": 10.03, | |
| "step": 3995 | |
| } | |
| ], | |
| "max_steps": 3995, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.804454718241824e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |