all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": 0.6529867053031921,
    "eval_logits/rejected": 1.5730761289596558,
    "eval_logps/chosen": -357.2324523925781,
    "eval_logps/rejected": -451.466064453125,
    "eval_loss": 0.5028161406517029,
    "eval_rewards/accuracies": 0.765625,
    "eval_rewards/chosen": -0.9468507170677185,
    "eval_rewards/margins": 0.946345865726471,
    "eval_rewards/rejected": -1.8931965827941895,
    "eval_runtime": 87.2102,
    "eval_samples": 2000,
    "eval_samples_per_second": 22.933,
    "eval_steps_per_second": 0.367,
    "train_loss": 0.07998354963677698,
    "train_runtime": 779.5821,
    "train_samples": 61135,
    "train_samples_per_second": 78.42,
    "train_steps_per_second": 0.613
}