all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": 0.9376290440559387,
    "eval_logits/rejected": 1.0400481224060059,
    "eval_logps/chosen": -280.18701171875,
    "eval_logps/rejected": -252.3540496826172,
    "eval_loss": 0.0390908420085907,
    "eval_rewards/accuracies": 0.6330000162124634,
    "eval_rewards/chosen": -0.013200990855693817,
    "eval_rewards/margins": 0.029489602893590927,
    "eval_rewards/rejected": -0.04269059747457504,
    "eval_runtime": 539.3434,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.708,
    "eval_steps_per_second": 0.927,
    "train_loss": 0.0001892112643969555,
    "train_runtime": 195.8108,
    "train_samples": 61135,
    "train_samples_per_second": 312.215,
    "train_steps_per_second": 19.514
}