Run large language models at home, BitTorrent-style.
Fine-tuning and inference up to 10x faster than offloading

Generate text with distributed Llama 3.1 (up to 405B), Mixtral (8x22B), Falcon (40B+) or BLOOM (176B) and fine‑tune them for your own tasks — right from your desktop computer or Google Colab:

from transformers import AutoTokenizer
from petals import AutoDistributedModelForCausalLM

# Choose any model available at https://health.petals.dev
model_name = "meta-llama/Meta-Llama-3.1-405B-Instruct"

# Connect to a distributed network hosting model layers
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoDistributedModelForCausalLM.from_pretrained(model_name)

# Run the model as if it were on your computer
inputs = tokenizer("A cat sat", return_tensors="pt")["input_ids"]
outputs = model.generate(inputs, max_new_tokens=5)
print(tokenizer.decode(outputs[0]))  # A cat sat on a mat...

🚀  Try now in Colab

🦙 Want to run Llama? Request access to its weights, then run huggingface-cli login in the terminal before loading the model. Or just try it in our chatbot app.

🔏 Privacy. Your data will be processed with the help of other people in the public swarm. Learn more about privacy here. For sensitive data, you can set up a private swarm among people you trust.

💬 Any questions? Ping us in our Discord!

Connect your GPU and increase Petals capacity

Petals is a community-run system — we rely on people sharing their GPUs. You can help serving one of the available models or host a new model from 🤗 Model Hub!

As an example, here is how to host a part of Llama 3.1 (405B) Instruct on your GPU:

🦙 Want to host Llama? Request access to its weights, then run huggingface-cli login in the terminal before loading the model.

🐧 Linux + Anaconda. Run these commands for NVIDIA GPUs (or follow this for AMD):

conda install pytorch pytorch-cuda=11.7 -c pytorch -c nvidia
pip install git+https://github.com/bigscience-workshop/petals
python -m petals.cli.run_server meta-llama/Meta-Llama-3.1-405B-Instruct

🪟 Windows + WSL. Follow this guide on our Wiki.

🐋 Docker. Run our Docker image for NVIDIA GPUs (or follow this for AMD):

sudo docker run -p 31330:31330 --ipc host --gpus all --volume petals-cache:/cache --rm \
    learningathome/petals:main \
    python -m petals.cli.run_server --port 31330 meta-llama/Meta-Llama-3.1-405B-Instruct

🍏 macOS + Apple M1/M2 GPU. Install Homebrew, then run these commands:

brew install python
python3 -m pip install git+https://github.com/bigscience-workshop/petals
python3 -m petals.cli.run_server meta-llama/Meta-Llama-3.1-405B-Instruct

📚  Learn more (how to use multiple GPUs, start the server on boot, etc.)

🔒 Security. Hosting a server does not allow others to run custom code on your computer. Learn more here.

💬 Any questions? Ping us in our Discord!

🏆 Thank you! Once you load and host 10+ blocks, we can show your name or link on the swarm monitor as a way to say thanks. You can specify them with --public_name YOUR_NAME.

How does it work?

• You load a small part of the model, then join a network of people serving the other parts. Single‑batch inference runs at up to 6 tokens/sec for Llama 2 (70B) and up to 4 tokens/sec for Falcon (180B) — enough for chatbots and interactive apps.
• You can employ any fine-tuning and sampling methods, execute custom paths through the model, or see its hidden states. You get the comforts of an API with the flexibility of PyTorch and 🤗 Transformers.

📜  Read paper            📚  See FAQ

📚 Tutorials, examples, and more

Basic tutorials:

• Getting started: tutorial
• Prompt-tune Llama-65B for text semantic classification: tutorial
• Prompt-tune BLOOM to create a personified chatbot: tutorial

Useful tools:

• Chatbot web app (connects to Petals via an HTTP/WebSocket endpoint): source code
• Monitor for the public swarm: source code

Advanced guides:

• Launch a private swarm: guide
• Run a custom model: guide

Benchmarks

Please see Section 3.3 of our paper.

🛠️ Contributing

Please see our FAQ on contributing.

📜 Citations

Alexander Borzunov, Dmitry Baranchuk, Tim Dettmers, Max Ryabinin, Younes Belkada, Artem Chumachenko, Pavel Samygin, and Colin Raffel. Petals: Collaborative Inference and Fine-tuning of Large Models. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations). 2023.

@inproceedings{borzunov2023petals,
  title = {Petals: Collaborative Inference and Fine-tuning of Large Models},
  author = {Borzunov, Alexander and Baranchuk, Dmitry and Dettmers, Tim and Riabinin, Maksim and Belkada, Younes and Chumachenko, Artem and Samygin, Pavel and Raffel, Colin},
  booktitle = {Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)},
  pages = {558--568},
  year = {2023},
  url = {https://arxiv.org/abs/2209.01188}
}

Alexander Borzunov, Max Ryabinin, Artem Chumachenko, Dmitry Baranchuk, Tim Dettmers, Younes Belkada, Pavel Samygin, and Colin Raffel. Distributed inference and fine-tuning of large language models over the Internet. Advances in Neural Information Processing Systems 36 (2023).

@inproceedings{borzunov2023distributed,
  title = {Distributed inference and fine-tuning of large language models over the {I}nternet},
  author = {Borzunov, Alexander and Ryabinin, Max and Chumachenko, Artem and Baranchuk, Dmitry and Dettmers, Tim and Belkada, Younes and Samygin, Pavel and Raffel, Colin},
  booktitle = {Advances in Neural Information Processing Systems},
  volume = {36},
  pages = {12312--12331},
  year = {2023},
  url = {https://arxiv.org/abs/2312.08361}
}

---

This project is a part of the BigScience research workshop.