msuzen/bristol

Parallel random matrix tools and complexity for deep learning

63
/ 100
Established

This tool helps deep learning researchers and practitioners understand the complexity of their neural networks. It takes a series of weight matrices from a deep learning model or a pre-trained PyTorch model and outputs measures of spectral complexity and ergodicity. The user is typically a deep learning researcher or an ML engineer optimizing model architectures.

Available on PyPI.

Use this if you are a deep learning researcher or engineer interested in the theoretical underpinnings of model complexity and want to determine the optimal number of layers for a deep learning model.

Not ideal if you are looking for a tool that directly improves model performance without needing to understand underlying random matrix theory, or if you need an actively maintained package.

deep-learning neural-networks model-complexity architecture-search random-matrix-theory
No Dependents
Maintenance 13 / 25
Adoption 7 / 25
Maturity 25 / 25
Community 18 / 25

How are scores calculated?

Stars

33

Forks

13

Language

Jupyter Notebook

License

GPL-3.0

Last pushed

Mar 20, 2026

Commits (30d)

0

Get this data via API

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

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