> ## Documentation Index
> Fetch the complete documentation index at: https://www.meilisearch.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Computing Hugging Face embeddings with the GPU

> This guide for experienced users shows you how to compile a Meilisearch binary that generates Hugging Face embeddings with an Nvidia GPU.

This guide is aimed at experienced users working with a self-hosted Meilisearch instance. It shows you how to compile a Meilisearch binary that generates Hugging Face embeddings with an Nvidia GPU.

## Prerequisites

* A [CUDA-compatible Linux distribution](https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#id12)
* An Nvidia GPU with CUDA support
* A modern Rust compiler

## Install CUDA

Follow Nvidia's [CUDA installation instructions](https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html).

## Verify your CUDA install

After you have installed CUDA in your machine, run the following command in your command-line terminal:

```sh theme={null}
nvcc --version | head -1
```

If CUDA is working correctly, you will see the following response:

```
nvcc: NVIDIA (R) Cuda compiler driver
```

## Compile Meilisearch

First, clone Meilisearch:

```sh theme={null}
git clone https://github.com/meilisearch/meilisearch.git
```

Then, compile the Meilisearch binary with `cuda` enabled:

```sh theme={null}
cargo build --release --features cuda
```

This might take a few moments. Once the compiler is done, you should have a CUDA-compatible Meilisearch binary.

## Configure the Hugging Face embedder

Run your freshly compiled binary:

```sh theme={null}
./meilisearch
```

Then add the Hugging Face embedder to your index settings:

```sh theme={null}
curl \
  -X PATCH 'MEILISEARCH_URL/indexes/INDEX_NAME/settings/embedders' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "default": { "source": "huggingFace" } }'
```

Meilisearch will return a summarized task object and place your request on the task queue:

```json theme={null}
{
  "taskUid": 1,
  "indexUid": "INDEX_NAME",
  "status": "enqueued",
  "type": "settingsUpdate",
  "enqueuedAt": "2024-03-04T15:05:43.383955Z"
}
```

Use the task object's `taskUid` to [monitor the task status](/reference/api/tasks/get-task). The Hugging Face embedder will be ready to use once the task is completed.

## Conclusion

You have seen how to compile a Meilisearch binary that uses your Nvidia GPU to compute vector embeddings. Doing this should significantly speed up indexing when using Hugging Face.