all_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": -2.2986156940460205,
    "eval_logits/rejected": -2.289750814437866,
    "eval_logps/chosen": -257.38525390625,
    "eval_logps/rejected": -221.40939331054688,
    "eval_loss": 0.54385906457901,
    "eval_rewards/accuracies": 0.7291666865348816,
    "eval_rewards/chosen": 1.061666488647461,
    "eval_rewards/margins": 2.0847039222717285,
    "eval_rewards/rejected": -1.023037314414978,
    "eval_runtime": 125.2796,
    "eval_samples_per_second": 15.166,
    "eval_steps_per_second": 0.239,
    "total_flos": 2369906314051584.0,
    "train_loss": 0.32917294554894244,
    "train_runtime": 7373.939,
    "train_samples_per_second": 6.954,
    "train_steps_per_second": 0.027
}