concrete/docs/user/basics/installing.md

Installing

Docker image

The easiest way to install the framework is as a docker image. To get the image you need to login to ghcr.io with docker.

FIXME(Arthur): to check this is still valid

docker login ghcr.io

This command will ask for a username and a password. For username, just enter your GitHub username. For password, you should create a personal access token from here selecting read:packages permission. Just paste the generated access token as your password, and you are good to go.

You can then either pull the latest docker image or a specific version:

docker pull ghcr.io/zama-ai/concretefhe:latest
# or
docker pull ghcr.io/zama-ai/concretefhe:v0.1.0

You can then use this image with the following command:

# Without local volume:
docker run --rm -it -p 8888:8888 ghcr.io/zama-ai/concretefhe:v0.1.0

# With local volume to save notebooks on host:
docker run --rm -it -p 8888:8888 -v /host/path:/data ghcr.io/zama-ai/concretefhe:v0.1.0

This will launch a Concrete enabled jupyter server in the docker, that you can access from your browser.

Alternatively you can just open a shell in the docker:

docker run --rm -it ghcr.io/zama-ai/concretefhe:v0.1.0 /bin/bash

python package

To install Concrete from PyPi run the following:

pip install concretefhe

Note that concretefhe has `pygraphviz` as an optional dependency to draw graphs.

`pygraphviz` requires `graphviz` packages being installed on your OS, see <a href="https://pygraphviz.github.io/documentation/stable/install.html">https://pygraphviz.github.io/documentation/stable/install.html</a>

You can install the extra python dependencies for drawing with:

pip install concretefhe[full]
# you may need to force reinstallation
pip install --force-reinstall concretefhe[full]