all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.8244452476501465,
    "eval_logits/rejected": -2.8161373138427734,
    "eval_logps/chosen": -500.1797790527344,
    "eval_logps/rejected": -503.9600830078125,
    "eval_loss": 0.6877260804176331,
    "eval_preference/chosen": 0.5198387503623962,
    "eval_preference/rejected": 0.48016127943992615,
    "eval_rewards/accuracies": 0.5540000200271606,
    "eval_rewards/chosen": -0.18460142612457275,
    "eval_rewards/margins": 0.08928026258945465,
    "eval_rewards/rejected": -0.2738817036151886,
    "eval_runtime": 21.0768,
    "eval_samples_per_second": 47.446,
    "eval_steps_per_second": 5.931
}