all_results.json

{
    "epoch": 0.99,
    "eval_logits/chosen": -1.22385835647583,
    "eval_logits/rejected": -1.171796202659607,
    "eval_logps/chosen": -407.7206115722656,
    "eval_logps/rejected": -631.7355346679688,
    "eval_loss": 0.6808297634124756,
    "eval_rewards/accuracies": 0.5625,
    "eval_rewards/chosen": 0.00824617501348257,
    "eval_rewards/margins": 0.000521879643201828,
    "eval_rewards/rejected": 0.007724293041974306,
    "eval_runtime": 19.6215,
    "eval_samples": 185,
    "eval_samples_per_second": 9.428,
    "eval_steps_per_second": 0.306,
    "train_loss": 0.6759306937456131,
    "train_runtime": 862.557,
    "train_samples": 4102,
    "train_samples_per_second": 4.756,
    "train_steps_per_second": 0.074
}