Pixi PyTorch Lightning ML Template¶
A lightweight, reproducible machine learning project template using Pixi, PyTorch Lightning, Aim experiment tracking, YAML configs, pytest smoke tests, and CPU/GPU-ready environments.
This template is designed for research ML projects where you want a clean starting point with reproducible dependencies, structured training and evaluation scripts, local experiment tracking, and easy project reuse.
What this template gives you¶
| Area | Included |
|---|---|
| Environment | Pixi + pixi.lock |
| Deep learning | PyTorch |
| Training framework | PyTorch Lightning |
| Experiment tracking | Local Aim tracking |
| Configuration | YAML config files |
| Checkpointing | Lightning ModelCheckpoint |
| Evaluation | Separate evaluate.py entry point |
| Plot logging | Aim callback for figures |
| Testing | Pytest smoke tests |
| Code quality | Ruff formatting and linting |
| Local outputs | Ignored local/ directory |
Good fit¶
Use this template when you want:
- reproducible Python environments,
- clean training and evaluation entry points,
- config-driven experiments,
- local experiment tracking,
- simple checkpointing,
- smoke tests,
- and a project layout that is easy to copy into a new repository.
Quick example¶
git clone https://github.com/CosmosRedshift7/ml-template.git
cd ml-template
pixi install
pixi run train
pixi run aim-ui
Then open:
http://127.0.0.1:43800
Keywords¶
Pixi PyTorch template, PyTorch Lightning template, machine learning template, ML research template, Aim experiment tracking, reproducible ML environment, Python deep learning template, GPU training template.