all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.244190216064453,
    "eval_logits/rejected": -2.2393178939819336,
    "eval_logps/chosen": -34.020179748535156,
    "eval_logps/rejected": -37.50062561035156,
    "eval_loss": 0.500054657459259,
    "eval_rewards/accuracies": 0.4842192530632019,
    "eval_rewards/chosen": 0.002875381149351597,
    "eval_rewards/margins": -0.0003236339834984392,
    "eval_rewards/rejected": 0.0031990152783691883,
    "eval_runtime": 145.749,
    "eval_samples": 343,
    "eval_samples_per_second": 2.353,
    "eval_steps_per_second": 0.295,
    "train_loss": 0.4997496530607149,
    "train_runtime": 3257.6696,
    "train_samples": 3079,
    "train_samples_per_second": 0.945,
    "train_steps_per_second": 0.118
}