| { | |
| "epoch": 3.0, | |
| "eval_log_odds_chosen": 0.23976314067840576, | |
| "eval_log_odds_ratio": -0.6928443908691406, | |
| "eval_logits/chosen": 340.5321350097656, | |
| "eval_logits/rejected": 312.9670104980469, | |
| "eval_logps/chosen": -1.0267834663391113, | |
| "eval_logps/rejected": -1.1786600351333618, | |
| "eval_loss": 1.455647587776184, | |
| "eval_nll_loss": 1.4095592498779297, | |
| "eval_rewards/accuracies": 0.5107913613319397, | |
| "eval_rewards/chosen": -0.051339175552129745, | |
| "eval_rewards/margins": 0.0075938161462545395, | |
| "eval_rewards/rejected": -0.058932989835739136, | |
| "eval_runtime": 90.3296, | |
| "eval_samples": 553, | |
| "eval_samples_per_second": 6.122, | |
| "eval_steps_per_second": 1.539, | |
| "total_flos": 0.0, | |
| "train_loss": 0.968865410152301, | |
| "train_runtime": 16784.8411, | |
| "train_samples": 5034, | |
| "train_samples_per_second": 0.9, | |
| "train_steps_per_second": 0.225 | |
| } |