0.001_idpo_noreplacerej_iter_2/train_results.json

{
    "epoch": 1.0,
    "train_loss": 0.6192928140268386,
    "train_runtime": 2654.4881,
    "train_samples": 20378,
    "train_samples_per_second": 7.677,
    "train_steps_per_second": 0.06
}