all_results.json

{
    "epoch": 3.97,
    "eval_logits/chosen": 0.8417125940322876,
    "eval_logits/rejected": 0.8990198373794556,
    "eval_logps/chosen": -256.497314453125,
    "eval_logps/rejected": -233.4380340576172,
    "eval_loss": 0.0003914303961209953,
    "eval_rewards/accuracies": 0.49950000643730164,
    "eval_rewards/chosen": 0.001168824383057654,
    "eval_rewards/margins": 0.00016647118900436908,
    "eval_rewards/rejected": 0.001002353266812861,
    "eval_runtime": 491.0166,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.073,
    "eval_steps_per_second": 1.018,
    "train_loss": 0.0003399105843263603,
    "train_runtime": 3335.538,
    "train_samples": 61135,
    "train_samples_per_second": 1.199,
    "train_steps_per_second": 0.074
}