eval_results.json

{
    "epoch": 0.998691442030882,
    "eval_logits/chosen": -1.0320225954055786,
    "eval_logits/rejected": -1.019870400428772,
    "eval_logps/chosen": -371.21112060546875,
    "eval_logps/rejected": -372.5772399902344,
    "eval_loss": 0.6026462316513062,
    "eval_rewards/accuracies": 0.656000018119812,
    "eval_rewards/chosen": -0.9395894408226013,
    "eval_rewards/margins": 0.3643726706504822,
    "eval_rewards/rejected": -1.3039621114730835,
    "eval_runtime": 224.7608,
    "eval_samples": 2000,
    "eval_samples_per_second": 8.898,
    "eval_steps_per_second": 0.556
}