Instructions to use Gen-Verse/MMaDA-8B-MixCoT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Gen-Verse/MMaDA-8B-MixCoT with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Gen-Verse/MMaDA-8B-MixCoT", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
|
Download README.md from Gen-Verse/MMaDA-8B-MixCoT: direct link, hf CLI and curl.
- Browser
- Download file 1.57 kB
-
https://huggingface.co/Gen-Verse/MMaDA-8B-MixCoT/resolve/main/README.md
- Command line
-
hf download hf://Gen-Verse/MMaDA-8B-MixCoT/README.md
-
curl -L -o README.md https://huggingface.co/Gen-Verse/MMaDA-8B-MixCoT/resolve/main/README.md
1.57 kB
metadata
license: mit
library_name: transformers
pipeline_tag: any-to-any
MMaDA-8B-MixCoT
We introduce MMaDA, a novel class of multimodal diffusion foundation models designed to achieve superior performance across diverse domains such as textual reasoning, multimodal understanding, and text-to-image generation. MMaDA is distinguished by three key innovations:
- MMaDA adopts a unified diffusion architecture with a shared probabilistic formulation and a modality-agnostic design, eliminating the need for modality-specific components.
- MMaDA introduces a mixed long chain-of-thought (CoT) fine-tuning strategy that curates a unified CoT format across modalities.
- MMaDA adopts a unified policy-gradient-based RL algorithm, which we call UniGRPO, tailored for diffusion foundation models. Utilizing diversified reward modeling, UniGRPO unifies post-training across both reasoning and generation tasks, ensuring consistent performance improvements.
Compared to MMaDA-8B-Base, MMaDA-8B-MixCoT exhibits better instruction-following capabilities and more stable CoT generation performance.
Citation
@article{yang2025mmada,
title={MMaDA: Multimodal Large Diffusion Language Models},
author={Yang, Ling and Tian, Ye and Li, Bowen and Zhang, Xinchen and Shen, Ke and Tong, Yunhai and Wang, Mengdi},
journal={arXiv preprint arXiv:2505.15809},
year={2025}
}