all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -4.5593791007995605,
    "eval_logits/rejected": -4.548151016235352,
    "eval_logps/chosen": -340.20562744140625,
    "eval_logps/rejected": -357.3804626464844,
    "eval_loss": 0.6746748089790344,
    "eval_rewards/accuracies": 0.5726666450500488,
    "eval_rewards/chosen": -1.3180568218231201,
    "eval_rewards/margins": 0.11864501982927322,
    "eval_rewards/rejected": -1.436701774597168,
    "eval_runtime": 1558.97,
    "eval_samples": 2994,
    "eval_samples_per_second": 1.92,
    "eval_steps_per_second": 0.241,
    "train_loss": 0.6508825262419409,
    "train_runtime": 40390.1729,
    "train_samples": 19996,
    "train_samples_per_second": 0.495,
    "train_steps_per_second": 0.031
}