all_results.json

{
    "epoch": 0.99,
    "eval_logits/chosen": 0.8018344044685364,
    "eval_logits/rejected": 0.819618821144104,
    "eval_logps/chosen": -91.73626708984375,
    "eval_logps/rejected": -94.22244262695312,
    "eval_loss": 0.010001449845731258,
    "eval_rewards/accuracies": 0.2619999945163727,
    "eval_rewards/chosen": -8.8951361249201e-05,
    "eval_rewards/margins": 0.0001654613733990118,
    "eval_rewards/rejected": -0.00025441276375204325,
    "eval_runtime": 274.1769,
    "eval_samples": 2000,
    "eval_samples_per_second": 7.295,
    "eval_steps_per_second": 0.456,
    "train_loss": 0.009997876410029437,
    "train_runtime": 1153.3315,
    "train_samples": 6113,
    "train_samples_per_second": 5.3,
    "train_steps_per_second": 0.082
}