all_results.json

{
    "epoch": 0.9989071038251366,
    "eval_logits/chosen": -1.3748970031738281,
    "eval_logits/rejected": -1.3346441984176636,
    "eval_logps/chosen": -0.5313604474067688,
    "eval_logps/rejected": -0.7141343355178833,
    "eval_loss": 2.09059739112854,
    "eval_rewards/accuracies": 0.7740963697433472,
    "eval_rewards/chosen": -5.313604354858398,
    "eval_rewards/margins": 1.8277385234832764,
    "eval_rewards/rejected": -7.141342639923096,
    "eval_runtime": 33.4268,
    "eval_samples": 1318,
    "eval_samples_per_second": 39.429,
    "eval_semantic_entropy": 0.997556209564209,
    "eval_steps_per_second": 2.483,
    "total_flos": 0.0,
    "train_loss": 2.4655840506438875,
    "train_runtime": 5955.1851,
    "train_samples": 58558,
    "train_samples_per_second": 9.833,
    "train_steps_per_second": 0.077
}