Python tools to submit functions on the Substra platform
Project description
Substra-tools
Substra is an open source federated learning (FL) software. This specific repository, substra-tools, is a Python package defining base classes for Dataset (data opener script) and wrappers to execute functions submitted on the platform.
This repository also contains a Dockerfile to execute the user Python scripts on the Substra platform. This is currently needed to easily have substratools package available inside the Docker image without using a pypi server.
Getting started
To install the substratools Python package, run the following command:
pip install substratools
Pull the substra-tools Docker image
docker pull ghcr.io/substra/substra-tools:0.16.0-nvidiacuda11.8.0-base-ubuntu22.04-python3.10-workflows
Developers
Clone the repository: https://github.com/Substra/substra-tools
Build the Docker image from source
docker build -f Dockerfile .
or for the minimal image (based on alpine):
docker build -f Dockerfile.minimal .
or for the workflows image (contains additional data science dependencies):
docker build -f Dockerfile.workflows .
Setup
To setup the project in development mode, run:
pip install -e ".[test]"
To run all tests, use the following command:
python setup.py test
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file substratools-0.20.0rc1.tar.gz.
File metadata
- Download URL: substratools-0.20.0rc1.tar.gz
- Upload date:
- Size: 21.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.16
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | cf5521785d62011e1a7ed8ab170e7a76ff99c24f7996351c0e16dafde624ed2e |
|
| MD5 | cd332cf51c466aca48174ecfdd413595 |
|
| BLAKE2b-256 | acd1d28f1a3dd060d76d7bebd94b7ebcf3e413a7c7b995cace501b2c8357b54d |
Provenance
File details
Details for the file substratools-0.20.0rc1-py3-none-any.whl.
File metadata
- Download URL: substratools-0.20.0rc1-py3-none-any.whl
- Upload date:
- Size: 28.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.16
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | a045be082a2acefea85e30d520acb956a628674e0b0aabff9539c361640688ce |
|
| MD5 | 520f15c61804e52f965e54ad64ab4029 |
|
| BLAKE2b-256 | 743c3a756fb22cb2b1653ef53c38d9ad29f69f4f3ecf3ccb4f7c86ae0730741e |
