Fast & Simple
Tensor Learning
In Python
- Open Source, BSD Licensed
- Pure Python, Tested & Optimized
- Flexible Backends for: NumPy, PyTorch, TensorFlow, JAX, CuPy, and PaddlePaddle
- Thorough Documentation
- Minimal Dependencies
- Tensorized Deep Learning with TensorLy-Torch
Tensor Methods, Made Easy
TensorLy provides all the utilities to easily use tensor methods, whether you are an advanced user or just getting started, from core tensor operations and tensor algebra to tensor decomposition and regression.
Execute Anywhere
TensorLy's backend system lets you write your code once and execute in using any of the supported frameworks, enabling tensor learning on GPU, multi-machines, and deep tensorized learning.
Speedup Your Research
TensorLy is open-source, actively maintained and easily extensible. Its BSD license makes it suitable for use in both industry and academia and it is already used by several university labs and companies all around the world.
If you use TensorLy, please consider citing us:
http://jmlr.org/papers/v20/18-277.html.
@article{tensorly,
author = {Jean Kossaifi and Yannis Panagakis and Anima Anandkumar and Maja Pantic},
title = {TensorLy: Tensor Learning in Python},
journal = {Journal of Machine Learning Research (JMLR)},
volume = {20},
number = {26},
year = {2019},
}