{
    "epoch": 1.0,
    "eval_logits/chosen": -2.6617116928100586,
    "eval_logits/rejected": -2.631270170211792,
    "eval_logps/chosen": -283.40789794921875,
    "eval_logps/rejected": -263.1501159667969,
    "eval_loss": 0.8482561111450195,
    "eval_rewards/accuracies": 0.6785714030265808,
    "eval_rewards/chosen": 0.7018745541572571,
    "eval_rewards/margins": 2.07344651222229,
    "eval_rewards/rejected": -1.3715722560882568,
    "eval_runtime": 235.9971,
    "eval_samples": 2000,
    "eval_samples_per_second": 8.475,
    "eval_steps_per_second": 0.267,
    "train_loss": 0.9276265908087324,
    "train_runtime": 18165.5867,
    "train_samples": 61135,
    "train_samples_per_second": 3.365,
    "train_steps_per_second": 0.026
}