===================================================================================
本模型为：https://huggingface.co/BAAI/Emu3.5-Image 的 NF4 量化版本，可用官方 inference 代码直接加载，需加装 bitsandbytes 依赖。 模型全部加载到显卡的情况下，需占用 24GB，跑图最大需要 32GB 显存。(根据本人测试情况，安装 flash_attn==2.7.4 预编译 whl 也行)
Prompt: "Live shot, close-up, full-body photo, a snow leopard standing on a rock, the body is standing sideways, standing slightly upward on the rock, the tail is slightly cocked, the head is twisted to face the camera, the eyes are looking directly at the camera, the expression is majestic, the background is slightly blurred in the distance, gray rocks and mountains."
Emu3.5-Image 是由国内 BAAI 智源研究院 开源的最新全模态模型，效果比肩 Google-Banana，以下介绍内容引用自官方模型介绍页。本模型为社区研究学习使用，请遵守官方相关版权协议。
===================================================================================
Emu3.5: Native Multimodal Models are World Learners
Emu3.5 Team, BAAI
Project Page | 🤗HF Models | Paper
| 🔹 | Core Concept | Description |
|---|---|---|
| 🧠 | Unified World Modeling | Predicts the next state jointly across vision and language, enabling coherent world modeling and generation. |
| 🧩 | End-to-End Pretraining | Trained with a unified next-token prediction objective over interleaved vision–language sequences. |
| 📚 | Over 10T+ Multimodal Tokens | Pre-trained on over 10 trillion interleaved tokens from video frames and transcripts, capturing spatiotemporal structure. |
| 🔄 | Native Multimodal I/O | Processes and generates interleaved visual–text sequences without modality adapters or task-specific heads. |
| 🎯 | RL Post-Training | Large-scale reinforcement learning enhances reasoning, compositionality, and generation quality. |
| ⚡ | Discrete Diffusion Adaptation (DiDA) | Converts sequential decoding → bidirectional parallel prediction, achieving ≈20× faster inference without performance loss. |
| 🖼️ | Versatile Generation | Excels in long-horizon vision–language generation, any-to-image (X2I) synthesis, and text-rich image creation. |
| 🌐 | Generalizable World Modeling | Enables spatiotemporally consistent world exploration, and open-world embodied manipulation across diverse scenarios. |
| 🏆 | Performance Benchmark | Matches Gemini 2.5 Flash Image (Nano Banana) on image generation/editing, and outperforms on interleaved generation tasks. |
Table of Contents
1. Model & Weights
2. Quick Start
Environment Setup
git clone https://github.com/baaivision/Emu3.5
cd Emu3.5
pip install -r requirements.txt
pip install flash_attn==2.8.3 --no-build-isolation
Configuration
Edit configs/config.py to set:
- Paths:
model_path,vq_path - Task template:
task_type in {t2i, x2i, howto, story, explore, vla},use_imagecontrols<|IMAGE|>usage (set to true when reference images are provided) - Sampling:
sampling_params(classifier_free_guidance, temperature, top_k/top_p, etc.)
Run Inference
python inference.py --cfg configs/config.py
Protobuf outputs are written to outputs/<exp_name>/proto/. For better throughput, we recommend ≥2 GPUs.
Visualize Protobuf Outputs
To visualize generated protobuf files:
python src/utils/vis_proto.py --input <input_proto_file> --output <output_dir>
3. Schedule
- Inference Code
- Advanced Image Decoder
- Discrete Diffusion Adaptation(DiDA)
4. Citation
@misc{cui2025emu35nativemultimodalmodels,
title={Emu3.5: Native Multimodal Models are World Learners},
author={Yufeng Cui and Honghao Chen and Haoge Deng and Xu Huang and Xinghang Li and Jirong Liu and Yang Liu and Zhuoyan Luo and Jinsheng Wang and Wenxuan Wang and Yueze Wang and Chengyuan Wang and Fan Zhang and Yingli Zhao and Ting Pan and Xianduo Li and Zecheng Hao and Wenxuan Ma and Zhuo Chen and Yulong Ao and Tiejun Huang and Zhongyuan Wang and Xinlong Wang},
year={2025},
eprint={2510.26583},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2510.26583},
}