README.md

metadata

title: Jupyter and Streamlit Docker Template
emoji: 📉
colorFrom: blue
colorTo: green
sdk: docker
python_version: '3.10'
app_port: 7860
app_file: src/app.py
suggested_storage: small
pinned: false
duplicated_from: SpacesExamples/streamlit-docker-example

🧠 Persistent Jupyter and Streamlit Docker Template 🔎

Streamlit Docker Template is a template for creating a Streamlit app with Docker and Hugging Face Spaces.

Code from https://docs.streamlit.io/library/get-started/create-an-app

Local execution

You need Docker installed. On MacOSX we recommand using colima if you do not want to use Docker Desktop for licensing reasons.

• https://docs.docker.com/desktop/install/mac-install/
• https://github.com/abiosoft/colima

$ colima start --cpu 4 --memory 16 --network-address # Adjust ressources as you wish
$ docker build -t persistent-docker-space .
$ docker run -it -p 8501:8501 persistent-docker-space:latest

Setting-up the developpers'tooling

Install poetry

https://python-poetry.org/

Linux and Mac

It should be straightforward with the official documentation

Windows (PowerShell)

(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | py -

The execution will probably be stored at the address: C:\User\<myUserName>\AppData\Roaming\pypoetry\venv\Scripts and this path should be included in the environment path of your machine in order to avoid typing it every time poetry is used. To do so you can execute the following commands:

$Env:Path += ";C:\Users\YourUserName\AppData\Roaming\Python\Scripts"

This will only make the change in the path temporarily. In order to do it permanently you can execute the following command

setx PATH "$Env:Path"

Configuration of the poetry environment

After having installed poetry in your local machine, if there is already a poetry.lock file on your repository, you can execute

poetry install

If it is not the case you can

poetry init
poetry env use "whatever version of python you have in your local machine (compatible with the project)"
poetry shell

pre-commit

https://pre-commit.com/

If there is already a poetry.lock file with pre-commit present in it, you should activate your poetry environment and then install all the pre-commit hooks on your machine

poetry shell
pre-commit install
pre-commit install --install-hooks

If not, you should first add pre-commit to your poetry environment, and follow the steps above

poetry add --group=dev pre-commit

commitizen

https://www.conventionalcommits.org/en/about/

https://commitizen-tools.github.io/commitizen/

Commitizen will be installed as a pre-commit hook. In order for it to be executed before committing you should run the following command (after activating your poetry environment)

pre-commit install --hook-type commit-msg

Finally, every time you will be committing, you should be places in your poetry environment and commitizen hooks should be applied

testing

There are two different kinds of tests that can be run when testing the scripts: unit tests or doctest

These tests can be run by executing the following command:

./scripts/run-tests.sh

pytest

https://docs.pytest.org/en/7.2.x/

These tests should be stored in the directory tests at the root of the project

xdoctest (driven by pytest)

These are the tests that are put in the docstrings of the functions accordingly to the following format:

    def build_greetings(name: Optional[str] = None) -> str:
        """
            Return a greeting message, possibly customize with a name.

            >>> build_greetings()
            'Hello, World!'
            >>> build_greetings('Toto')
            'Nice to meet you, Toto!'
        """
        return name and f"Nice to meet you, {name}!" or "Hello, World!"

The evaluated values would be the ones following the >>>

documentation

https://www.sphinx-doc.org/en/master/

In order to create an automatic documentation of your code you should run the bash script

./scripts/build-clean-docs.sh

And in order to create an interactive session (web-server hosted in your local machine), you can execute the following command

./scripts/interactive-rebuild-docs.sh

Remark: In order to execute a bash script with a Windows OS, it is recommended to use a bash terminal emulator

Hugging Face

See instructions at https://huggingface.co/welcome

Install huggingface_hub into the poetry project.

poetry add --group=dev huggingface_hub

On MacOS, you might probably want to install hugginface-cli from brew :

$ brew install huggingface-cli

In order to deploy the streamlit app you will have to export the poetry config as a requirements.txt :

$ poetry export -o ../requirements.txt --without-hashes --only main