all_results.json

{
    "epoch": 2.0,
    "eval_logits/chosen": 0.5351993441581726,
    "eval_logits/rejected": 0.5909183621406555,
    "eval_logps/chosen": -269.0498962402344,
    "eval_logps/rejected": -255.128662109375,
    "eval_loss": 2028.09326171875,
    "eval_rewards/accuracies": 0.6899999976158142,
    "eval_rewards/chosen": -0.12432491779327393,
    "eval_rewards/margins": 0.0914725586771965,
    "eval_rewards/rejected": -0.21579748392105103,
    "eval_runtime": 416.5334,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.802,
    "eval_steps_per_second": 1.2,
    "train_loss": 2099.8451463309884,
    "train_runtime": 42790.3459,
    "train_samples": 30567,
    "train_samples_per_second": 1.429,
    "train_steps_per_second": 0.089
}