all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": 0.937932550907135,
    "eval_logits/rejected": 1.0401638746261597,
    "eval_logps/chosen": -280.182861328125,
    "eval_logps/rejected": -252.35885620117188,
    "eval_loss": 0.034593481570482254,
    "eval_rewards/accuracies": 0.6290000081062317,
    "eval_rewards/chosen": -0.02636060118675232,
    "eval_rewards/margins": 0.059068720787763596,
    "eval_rewards/rejected": -0.08542931824922562,
    "eval_runtime": 538.4763,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.714,
    "eval_steps_per_second": 0.929,
    "train_loss": 0.038748307645445686,
    "train_runtime": 55741.6245,
    "train_samples": 61135,
    "train_samples_per_second": 1.097,
    "train_steps_per_second": 0.069
}