README.md

    method: chosen_reward_ 

    num_inference_steps: 30

    {metric} : {mean} 

{metric} : {mean} {metric} : {mean} {metric} : {mean} {metric} : {mean}