all_results.json

{
    "epoch": 2.810126582278481,
    "eval_logits/chosen": -0.275390625,
    "eval_logits/rejected": -0.154296875,
    "eval_logps/chosen": -156.0,
    "eval_logps/rejected": -203.0,
    "eval_loss": 0.5738985538482666,
    "eval_rewards/accuracies": 1.0,
    "eval_rewards/chosen": -0.011962890625,
    "eval_rewards/margins": 0.259765625,
    "eval_rewards/rejected": -0.271484375,
    "eval_runtime": 6.45,
    "eval_samples_per_second": 37.829,
    "eval_steps_per_second": 1.395
}