all_results.json

{
    "epoch": 1.98,
    "eval_logits/chosen": -0.006220105104148388,
    "eval_logits/rejected": 0.09270644932985306,
    "eval_logps/chosen": -306.42120361328125,
    "eval_logps/rejected": -278.68194580078125,
    "eval_loss": 0.010818002745509148,
    "eval_rewards/accuracies": 0.5,
    "eval_rewards/chosen": -0.0003898772120010108,
    "eval_rewards/margins": -4.452037319424562e-05,
    "eval_rewards/rejected": -0.00034535673330537975,
    "eval_runtime": 710.2034,
    "eval_samples": 2000,
    "eval_samples_per_second": 2.816,
    "eval_steps_per_second": 0.704,
    "train_loss": 0.010306676939850854,
    "train_runtime": 2255.3236,
    "train_samples": 61135,
    "train_samples_per_second": 0.887,
    "train_steps_per_second": 0.055
}