julvo/reloading

Change Python code while it's running without losing state

48
/ 100
Emerging

This tool helps deep learning practitioners modify their Python training code on-the-fly without interrupting the model's learning process. You provide your existing training loop or function, and this allows you to tweak things like logging or saving the model while it's actively training, preserving your progress. It's for data scientists or machine learning engineers who need to iterate quickly on model training.

1,127 stars. No commits in the last 6 months. Available on PyPI.

Use this if you are training a deep learning model and want to make live adjustments to your Python code (e.g., adding a print statement, changing how you save the model) without restarting the entire training run.

Not ideal if your primary goal is to change the model architecture, hyperparameters, or dataset during training, as those changes typically require a restart.

deep-learning-training machine-learning-engineering model-debugging data-scientist-workflow
Stale 6m No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 25 / 25
Community 13 / 25

How are scores calculated?

Stars

1,127

Forks

34

Language

Python

License

MIT

Last pushed

Jun 15, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/julvo/reloading"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.