all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": 0.12150061875581741,
    "eval_logits/rejected": 0.1178746148943901,
    "eval_logps/chosen": -2002.3223876953125,
    "eval_logps/rejected": -1694.0738525390625,
    "eval_loss": 0.06741556525230408,
    "eval_rewards/accuracies": 0.5024999976158142,
    "eval_rewards/chosen": 0.2140834480524063,
    "eval_rewards/margins": 0.03174477815628052,
    "eval_rewards/rejected": 0.1823386698961258,
    "eval_runtime": 510.1036,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.921,
    "eval_steps_per_second": 0.98,
    "train_loss": 0.09710332241058349,
    "train_runtime": 8719.9742,
    "train_samples": 61135,
    "train_samples_per_second": 1.147,
    "train_steps_per_second": 0.072
}