Update README.md

README.md

@@ -1,4 +1,4 @@
-**Datasets for the Direct Preference for Denoising Diffusion Policy Optimization (D3PO)**
+# Datasets for the Direct Preference for Denoising Diffusion Policy Optimization (D3PO)
 
 **Description**: The dataset for the image distortion experiment of the [`anything-v5`](https://huggingface.co/stablediffusionapi/anything-v5) model in the paper [Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model](https://arxiv.org/abs/2311.13231).