Official GitHub repository of DialogGen: Multi-modal Interactive Dialogue System for Multi-turn Text-to-Image Generation ArXiv paper. Project Page
The data is under inspection for safety concerns. We will release the DialogBen and its corresponding evaluation code once it is finished. Please refer to https://github.com/Tencent/HunyuanDiT, which is our newly open-sourced Text-to-Image generation model. All of the following open-source plans adopt HunyuanDiT as our Text-to-Image model. The weights of DialogGen based on llava are also released.
- DialogGen
- Multi-turn T2I Demo (Gradio) https://github.com/Tencent/HunyuanDiT/blob/main/app/multiTurnT2I_app.py
- Data of DialogBen (Coming soon ⏩️)
- Evaluation code on DialogBen (Coming soon ⏩️)