DynMoE-Phi-2-2.7B / README.md
kenshinn's picture
Update README.md
29a2fe1 verified
metadata
license: apache-2.0
pipeline_tag: image-text-to-text

Dynamic Mixture of Experts: An Auto-Tuning Approach for Efficient Transformer Models

If our project helps you, please give us a star ⭐ on GitHub and cite our paper!

πŸ“° News

  • [2024.05.31] πŸ”₯ Our code is released!
  • [2024.05.25] πŸ”₯ Our checkpoints are available now!
  • [2024.05.23] πŸ”₯ Our paper is released!

😎 What's Interesting?

Dynamic Mixture of Experts (DynMoE) incorporates (1) a novel gating method that enables each token to automatically determine the number of experts to activate. (2) An adaptive process automatically adjusts the number of experts during training.

Top-Any Gating

Adaptive Training Process

πŸ’‘ Model Details

  • πŸ€” DynMoE-Phi-2 is a MoE model with dynamic top-k gating, finetuned on LanguageBind/MoE-LLaVA-Phi2-Stage2.
  • πŸš€ Our DynMoE-Phi-2-2.7B has totally 5.3B parameters, but only 3.4B are activated! (average top-k = 1.68)
  • βŒ› With the DynMoE tuning stage, we can complete training on 8 A100 GPUs within 2 days.

πŸ‘ Acknowledgement

We are grateful for the following awesome projects:

πŸ”’ License

This project is released under the Apache-2.0 license as found in the LICENSE file.

✏️ Citation

@misc{guo2024dynamic,
      title={Dynamic Mixture of Experts: An Auto-Tuning Approach for Efficient Transformer Models}, 
      author={Yongxin Guo and Zhenglin Cheng and Xiaoying Tang and Tao Lin},
      year={2024},
      eprint={2405.14297},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}