A collection of simple utilities for machine learning workflows.
Project description
icflow
is a Python package with some prototype 'workflow' tools for use in ICHEC.
Goals
The goal is to use it to help convert tools and approaches used by ML 'domain experts' into a more standardized workflow by:
- Analysing incoming Juyter notebooks, models, datasets and runtime environments (conda/container)
- Identifying 'hard-coded' data that can be moved into config files
- Adding utility scripts and methods for fetching data and models as needed
- Describing a study as a workflow, using some scripts here to 'stitch' the workflow together
Ultimately we will end up using some common workflow tools across ICHEC, likely something established and open-source (eg MLFlow) - this package is intended to understand and flesh-out our workflow needs and start transforming how we set up studies to ultimately move to these more standard tools.
Tests
In a Python virtual environment do:
pip install .'[test]'
Unit Tests
pytest
Linting and Static Analysis
black src test
mypy src test
All Tests
Requires tox
:
tox
Project details
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
icflow-0.0.1.tar.gz
(28.0 kB
view hashes)
Built Distribution
icflow-0.0.1-py3-none-any.whl
(33.2 kB
view hashes)