| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0991180176396473, |
| "eval_steps": 500, |
| "global_step": 2500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.41999160016799664, |
| "grad_norm": 4.661656589632912e-07, |
| "learning_rate": 4.446666666666667e-06, |
| "logits/chosen": -1.101060390472412, |
| "logits/rejected": -1.163004994392395, |
| "logps/chosen": -1092.45556640625, |
| "logps/rejected": -1472.419189453125, |
| "loss": 0.0307, |
| "rewards/accuracies": 0.9879999756813049, |
| "rewards/chosen": -0.8035958409309387, |
| "rewards/margins": 22.136253356933594, |
| "rewards/rejected": -22.939851760864258, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8399832003359933, |
| "grad_norm": 6.130552776539844e-08, |
| "learning_rate": 3.335555555555556e-06, |
| "logits/chosen": -1.1089242696762085, |
| "logits/rejected": -1.1741572618484497, |
| "logps/chosen": -1095.48779296875, |
| "logps/rejected": -1501.994384765625, |
| "loss": 0.0, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": -0.9832028746604919, |
| "rewards/margins": 24.572891235351562, |
| "rewards/rejected": -25.556095123291016, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.259554808903822, |
| "grad_norm": 1.2126815818191972e-07, |
| "learning_rate": 2.2244444444444447e-06, |
| "logits/chosen": -1.1322691440582275, |
| "logits/rejected": -1.2022215127944946, |
| "logps/chosen": -1206.6405029296875, |
| "logps/rejected": -1637.5062255859375, |
| "loss": 0.0001, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": -12.410662651062012, |
| "rewards/margins": 26.757793426513672, |
| "rewards/rejected": -39.168460845947266, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.6795464090718184, |
| "grad_norm": 3.6735600672521684e-10, |
| "learning_rate": 1.1133333333333334e-06, |
| "logits/chosen": -1.1688064336776733, |
| "logits/rejected": -1.2446866035461426, |
| "logps/chosen": -1242.11181640625, |
| "logps/rejected": -1692.5054931640625, |
| "loss": 0.0, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": -14.96800422668457, |
| "rewards/margins": 29.2105655670166, |
| "rewards/rejected": -44.17856979370117, |
| "step": 2000 |
| }, |
| { |
| "epoch": 2.0991180176396473, |
| "grad_norm": 2.511642094304989e-08, |
| "learning_rate": 2.2222222222222225e-09, |
| "logits/chosen": -1.1509324312210083, |
| "logits/rejected": -1.2300761938095093, |
| "logps/chosen": -1232.4278564453125, |
| "logps/rejected": -1683.3319091796875, |
| "loss": 0.0, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": -15.354888916015625, |
| "rewards/margins": 29.08474349975586, |
| "rewards/rejected": -44.439632415771484, |
| "step": 2500 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 2500, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 1250, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
