| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -2.3036882877349854, | |
| "eval_logits/rejected": -2.218935012817383, | |
| "eval_logps/chosen": -341.342041015625, | |
| "eval_logps/rejected": -273.77117919921875, | |
| "eval_loss": 1.0562912225723267, | |
| "eval_rewards/accuracies": 0.72265625, | |
| "eval_rewards/chosen": -6.750503063201904, | |
| "eval_rewards/margins": 3.3230087757110596, | |
| "eval_rewards/rejected": -10.07351303100586, | |
| "eval_runtime": 258.1215, | |
| "eval_samples": 2000, | |
| "eval_samples_per_second": 7.748, | |
| "eval_steps_per_second": 0.062, | |
| "train_loss": 0.07034083745375122, | |
| "train_runtime": 46831.0549, | |
| "train_samples": 62064, | |
| "train_samples_per_second": 3.976, | |
| "train_steps_per_second": 0.031 | |
| } |