[ICLR'25] HERO: Human-Feedback-Efficient Reinforcement Learning for Online Diffusion Model Finetuning

This repository officially houses the official PyTorch implementation of the paper titled "HERO: Human-Feedback-Efficient Reinforcement Learning for Online Diffusion Model Finetuning", which is presented at ICLR 2025.

• Project Page: https://hero-dm.github.io/
• arXiv: https://arxiv.org/pdf/2410.05116

The code will be released soon. Please stay tuned.

Contacts:

• Ayano Hiranaka: [email protected]
• Shang-Fu Chen: [email protected]
• Chieh-Hsin Lai: [email protected]