360CVGroup/Qihoo-T2X
An efficient Diffusion Transformer model for text-to-any generation tasks, published at ICLR 2025.
★446 stars Image · Video · Audio
Not currently ranked — collecting fresh signals.
star history
Qihoo-T2X is an efficient Diffusion Transformer architecture designed for text-to-any generation tasks, including text-to-image synthesis. The approach uses proxy-tokenized diffusion to improve computational efficiency while maintaining generation quality. The model was accepted for presentation at ICLR 2025.
Frequently asked
- What is 360CVGroup/Qihoo-T2X?
- An efficient Diffusion Transformer model for text-to-any generation tasks, published at ICLR 2025.
- Is Qihoo-T2X open source?
- Yes — 360CVGroup/Qihoo-T2X is an open-source project tracked on heatdrop.
- How popular is Qihoo-T2X?
- 360CVGroup/Qihoo-T2X has 446 stars on GitHub.
- Where can I find Qihoo-T2X?
- 360CVGroup/Qihoo-T2X is on GitHub at https://github.com/360CVGroup/Qihoo-T2X.