Congrats on the new work UniPic!
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 875
- Forks
- 43
- PR merge metrics
- No merged PRs in 30d
Description
Congrats on the new work UniPic!
We’re the author team of Muddit (Meissonic II). Muddit is the first unified discrete diffusion model trained from visual prior.
Here, we define unified model as that the model conducts different modalities generation with the same parameters and the same paradigm.
- Our Paper: arXiv:2505.23606
- Our Model: Hugging Face - MeissonFlow/Muddit
- Our Demo: Live Demo on Hugging Face Spaces
To our knowledge, Muddit is the best unified discrete diffusion model capable of generating both high-quality images and captions (not in metrics, but in real demos (which we believe matters more than scores alone)).
Feel free to try it out, compare with your models, and let us know what you think!
We welcome discussions on the future of generative paradigms — AutoRegressive vs. Discrete Diffusion.
Best,
Jinbin
MeissonFlow Research
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
This issue is an announcement about Muddit, with links to its paper, Hugging Face model, and demo, rather than a requested change to UniPic. Review those linked resources only if comparing model approaches; there is no implementation target or completion criterion stated.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- machine-learning
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 1/100