Unofficial implementation for the paper "Mixture-of-Depths"

Introduction

This is an unofficial implementation for the paper Mixture-of-Depths: Dynamically allocating compute in transformer-based language models

Currently supported models

Model	Supported?
Mistral	✅
Mixtral	✅
LLama	✅
LLama2	✅
LLama3	✅
Gemma	✅
BLOOMZ	✅
BLOOM	✅
DeepSeek	✅
Phi (1.5 & 2)	✅
Qwen2	✅
StarCoder2	✅
Qwen2-MoE	❓
Solar	❓
Baichuan	❌
ChatGLM3	❌
InternLM	❌
Olmo	❌
XVERSE	❌
Yi	❌
Yuan	❌

💾 Installation

pip install mixture-of-depth

Both Linux, Windows and MacOS are supported.

🏁 Quick Start

High-level API (tranformers-compatible)

from transformers import AutoModelForCausalLM
from MoD import apply_mod_to_hf

# Initialize your model from an available hf model
model= AutoModelForCausalLM.from_pretrained("some-repo/some-model")
# Convert the model to include the mixture of depths layers
model = apply_mod_to_hf(model)
# train the model
# ...
# save the model
model.save_pretrained('some_local_directory')

Loading the converted Model

To utilize the converted model, you will need to load the model from the AutoClass. Below is an example demonstrating how to load the model from a local directory:

from MoD import AutoMoDModelForCausalLM

# Replace 'path_to_your_model' with the actual path to your model's directory
model = AutoMoDModelForCausalLM.from_pretrained('path_to_your_model')

Using generate()

Before calling the hf generate() method please explicitly use eval() on the model

🫱🏼‍🫲🏽 Contributing

We welcome contributions from the community, whether it's adding new features, improving documentation, or reporting bugs. Please refer to our contribution guidelines before making a pull request.

📜 License

This repo is open-sourced under the Apache-2.0 license.

Citation

If you use our code in your research, please cite it using the following Bibtex entry:

@article{MoD2024,
  title={Unofficial implementation for the paper "Mixture-of-Depths"},
  author={AstraMind AI},
  journal={https://github.com/astramind-ai/Mixture-of-depths},
  year={2024}
}

Support

For questions, issues, or support, please open an issue on our GitHub repository.