microsoft/tensorwatch

Debugging, monitoring and visualization for Python Machine Learning and Data Science

69
/ 100
Established

This tool helps data scientists and machine learning engineers visualize and debug their model training in real-time within Jupyter Notebooks. You feed it numerical data streams from your live machine learning processes, and it outputs dynamic charts and graphs that update as your model trains. It's designed for anyone actively developing or experimenting with deep learning and reinforcement learning models.

3,466 stars. Used by 1 other package. Actively maintained with 5 commits in the last 30 days. Available on PyPI.

Use this if you need to monitor the performance and internal states of your machine learning models while they are actively training, directly from your Jupyter environment.

Not ideal if you need a production-ready monitoring system, as this tool is specifically for development and debugging and carries security considerations for untrusted environments.

deep-learning-training model-debugging machine-learning-experimentation reinforcement-learning data-science-visualization
Maintenance 13 / 25
Adoption 11 / 25
Maturity 25 / 25
Community 20 / 25

How are scores calculated?

Stars

3,466

Forks

360

Language

Jupyter Notebook

License

MIT

Last pushed

Mar 06, 2026

Commits (30d)

5

Reverse dependents

1

Get this data via API

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

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