Generate text using distributed 176B-parameter [BLOOM](https://huggingface.co/bigscience/bloom) or [BLOOMZ](https://huggingface.co/bigscience/bloomz) and fine-tune them for your own tasks:
🔏 Your data will be processed by other people in the public swarm. Learn more about privacy [here](https://github.com/bigscience-workshop/petals/wiki/Security,-privacy,-and-AI-safety). For sensitive data, you can set up a [private swarm](https://github.com/bigscience-workshop/petals/wiki/Launch-your-own-swarm) among people you trust.
Or use our [Docker](https://www.docker.com) image (works on Linux, macOS, and Windows with [WSL2](https://learn.microsoft.com/en-us/windows/ai/directml/gpu-cuda-in-wsl)):
📚 See [FAQ](https://github.com/bigscience-workshop/petals/wiki/FAQ:-Frequently-asked-questions#running-a-server) to learn how to configure the server to use multiple GPUs, address common issues, etc.
You can also host [BLOOMZ](https://huggingface.co/bigscience/bloomz), a version of BLOOM fine-tuned to follow human instructions in the zero-shot regime — just replace `bloom-petals` with `bloomz-petals`.
🔒 Hosting a server does not allow others to run custom code on your computer. Learn more about security [here](https://github.com/bigscience-workshop/petals/wiki/Security,-privacy,-and-AI-safety).
- Prompt-tune BLOOM to create a personified chatbot: [tutorial](https://colab.research.google.com/github/bigscience-workshop/petals/blob/main/examples/prompt-tuning-personachat.ipynb)
- Prompt-tune BLOOM for text semantic classification: [tutorial](https://colab.research.google.com/github/bigscience-workshop/petals/blob/main/examples/prompt-tuning-sst2.ipynb)
- Petals runs large language models like [BLOOM-176B](https://huggingface.co/bigscience/bloom) **collaboratively** — you load a small part of the model, then team up with people serving the other parts to run inference or fine-tuning.
- Single-batch inference runs at ≈ 1 sec per step (token) — [up to 10x faster](https://github.com/bigscience-workshop/petals#benchmarks) than offloading, enough for [chatbots](http://chat.petals.ml) and other interactive apps. Parallel inference reaches hundreds of tokens/sec.
- Beyond classic language model APIs — 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.
If you don't use Anaconda, you can install PyTorch in [any other way](https://pytorch.org/get-started/locally/). If you want to run models with 8-bit weights, please install PyTorch with CUDA 11.x or newer for compatility with [bitsandbytes](https://github.com/timDettmers/bitsandbytes).
See the instructions for macOS and Windows, the full requirements, and troubleshooting advice in our [FAQ](https://github.com/bigscience-workshop/petals/wiki/FAQ:-Frequently-asked-questions#running-a-client).
<thcolspan="6">Offloading, max. possible speed on 1x A100 <sup>1</sup></th>
</tr>
<tralign="center">
<td>256 Gbit/s</td>
<td></td>
<td>0.18</td>
<td>0.18</td>
<td>2.7</td>
<td>170.3</td>
</tr>
<tralign="center">
<td>128 Gbit/s</td>
<td></td>
<td>0.09</td>
<td>0.09</td>
<td>2.4</td>
<td>152.8</td>
</tr>
<tr>
<thcolspan="6">Petals on 14 heterogeneous servers across Europe and North America <sup>2</sup></th>
</tr>
<tralign="center">
<tdcolspan="2">Real world</td>
<td>0.83</td>
<td>0.79</td>
<td>32.6</td>
<td>179.4</td>
</tr>
<tr>
<thcolspan="6">Petals on 3 servers, with one A100 each <sup>3</sup></th>
</tr>
<tralign="center">
<td>1 Gbit/s</td>
<td>< 5 ms</td>
<td>1.71</td>
<td>1.54</td>
<td>70.0</td>
<td>253.6</td>
</tr>
<tralign="center">
<td>100 Mbit/s</td>
<td>< 5 ms</td>
<td>1.66</td>
<td>1.49</td>
<td>56.4</td>
<td>182.0</td>
</tr>
<tralign="center">
<td>100 Mbit/s</td>
<td>100 ms</td>
<td>1.23</td>
<td>1.11</td>
<td>19.7</td>
<td>112.2</td>
</tr>
</table>
<sup>1</sup>**An upper bound for offloading performance.** We base our offloading numbers on the best possible hardware setup for offloading: CPU RAM offloading via PCIe 4.0 with 16 PCIe lanes per GPU and PCIe switches for pairs of GPUs. We assume zero latency for the upper bound estimation. In 8-bit, the model uses 1 GB of memory per billion parameters. PCIe 4.0 with 16 lanes has a throughput of 256 Gbit/s, so offloading 176B parameters takes 5.5 seconds. The throughput is twice as slow (128 Gbit/s) if we have two GPUs behind the same PCIe switch.
<sup>2</sup>**A real-world distributed setting** with 14 servers holding 2× RTX 3060, 4× 2080Ti, 2× 3090, 2× A4000, and 4× A5000 GPUs. These are personal servers and servers from university labs, spread across Europe and North America and connected to the Internet at speeds of 100–1000 Mbit/s. 4 servers operate from under firewalls.
<sup>3</sup>**An optimistic setup** that requires least communication. The client nodes have 8 CPU cores and no GPU.
We provide more evaluations and discuss these results in more detail in **Section 3.3** of our [paper](https://arxiv.org/pdf/2209.01188.pdf).
## 🛠️ Contributing
Please see our [FAQ](https://github.com/bigscience-workshop/petals/wiki/FAQ:-Frequently-asked-questions#contributing) on contributing.
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.](https://arxiv.org/abs/2209.01188)
_arXiv preprint arXiv:2209.01188,_ 2022.
```bibtex
@article{borzunov2022petals,
title = {Petals: Collaborative Inference and Fine-tuning of Large Models},
author = {Borzunov, Alexander and Baranchuk, Dmitry and Dettmers, Tim and Ryabinin, Max and Belkada, Younes and Chumachenko, Artem and Samygin, Pavel and Raffel, Colin},
