mratsim/Arraymancer

A fast, ergonomic and portable tensor library in Nim with a deep learning focus for CPU, GPU and embedded devices via OpenMP, Cuda and OpenCL backends

50
/ 100
Established

This project provides a robust toolkit for scientists, engineers, and data analysts working with multi-dimensional datasets. It takes in numerical data structured as N-dimensional arrays (tensors) and can perform advanced mathematical operations, machine learning algorithms like classification and regression, and deep learning model training. The output is processed data, insights from models, or trained deep learning networks. It's for users who need high-performance numerical computing.

1,391 stars.

Use this if you need a high-performance framework for complex numerical computations, machine learning tasks, or building deep learning models, especially when working with large, multi-dimensional datasets and prioritizing execution speed.

Not ideal if you prefer an interactive environment like a Python REPL or Jupyter notebooks, as this is a compiled language that requires a different workflow for prototyping.

numerical-computing machine-learning deep-learning data-analysis scientific-modeling
No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 18 / 25

How are scores calculated?

Stars

1,391

Forks

100

Language

Nim

License

Apache-2.0

Last pushed

Jan 02, 2026

Commits (30d)

0

Get this data via API

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

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