Welcome to VeOmni

VeOmni is a versatile framework for both single- and multi-modal pre-training and post-training. It empowers users to seamlessly scale models of any modality across various accelerators, offering both flexibility and user-friendliness.

---

Get Started

• Installation with Nvidia GPU
• Installation with Ascend NPU (x86)
• Installation with Ascend NPU (ARM)

Usage

• Arguments API Reference
• Basic Modules
• Usage of dyn_bsz
• Support New Models — Guide and Checklist
• Qwen3-VL MoE Integration Example
• Qwen3-Omni-MoE Integration Example
• Support New DiT Models — Guide and Reference
• Checkpoint Conversion
• Trainer
• Agent Workflow Guide

Hardware Support

• Get Started with Ascend NPU

Examples

• Qwen3 training guide
• Qwen3.5 training guide
• Qwen3 MoE training guide
• Qwen3 VL training guide
• Qwen3 Omni MoE training guide
• Wan2.1-I2V training guide
• Qwen3 DPO training guide

Key Features

• Adding a New Model to VeOmni
• Custom Preprocessor Registry
• EP+FSDP2 for Large-scale MoE Model Training
• Long-Sequence Training Using Ulysses

Design

• Kernel Selection in VeOmni

Transformers v5 Updates

• Transformers v5 Compatibility Updates

---

Citation

If you find VeOmni useful for your research and applications, feel free to give us a star ⭐ or cite us using:

@article{ma2025veomni,
  title={VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo},
  author={Ma, Qianli and Zheng, Yaowei and Shi, Zhelun and Zhao, Zhongkai and Jia, Bin and Huang, Ziyue and Lin, Zhiqi and Li, Youjie and Yang, Jiacheng and Peng, Yanghua and others},
  journal={arXiv preprint arXiv:2508.02317},
  year={2025}
}

About ByteDance Seed Team

Founded in 2023, ByteDance Seed Team is dedicated to crafting the industry’s most advanced AI foundation models. The team aspires to become a world-class research team and make significant contributions to the advancement of science and society. You can get to know Bytedance Seed better through the following channels👇