add dependency management for python

[skewballfox]

Pull Request Template

Checklist

• Confirmed that run-checks all script has been executed.
• Made sure the book is up to date with changes in this PR.

Related Issues/PRs

Changes

added a pyproject.toml and a few files generated by rye. a consistent easy to setup virtual environment would at the very least make it easy to setup burn when bouncing between systems, and might help new contributors a bit.

It might be beneficial to make a .vscode/settings.json file to point the python extension to the venv since (at least with rye), it will be in the same spot and vscode won't activate by default otherwise since it's off the default search path. plus we can add recommended extensions and configs but that is out of scope for this PR.

Testing

I set up the environment and created a few ONNX files?

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.07%. Comparing base (5de1517) to head (b5ffd79).
Report is 5 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1887      +/-   ##
==========================================
- Coverage   86.14%   86.07%   -0.07%     
==========================================
  Files         778      781       +3     
  Lines       90814    91616     +802     
==========================================
+ Hits        78232    78861     +629     
- Misses      12582    12755     +173     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[antimora]

It would be helpful to explain how to install the dependencies in README under the onnx-tests project: https://github.com/tracel-ai/burn/blob/main/crates/burn-import/onnx-tests/README.md

[skewballfox]

so slight hiccup. If we use rye like this, the .venv in the directory onnx-test, which is off the search path if opening the full burn project. So setting up for vscode would mean manually selecting the python interpreter the first time.

I could try poetry, but IME for tensor libraries it's hit or miss, mainly because the common scientific computing libraries tend to break spec frequently

[antimora]

LGTM.

Thanks for project set up improvements.

[nathanielsimard]

LGTM