CannyLab / packages / tsnecuda 2.1.0

CUDA Implementation of T-SNE with Python bindings


conda install

  • linux-64  v2.1.0
To install this package with conda run:
conda install -c cannylab tsnecuda




tsnecuda provides an optimized CUDA implementation of the T-SNE algorithm by L Van der Maaten. tsnecuda is able to compute the T-SNE of large numbers of points up to 1200 times faster than other leading libraries, and provides simple python bindings with a SKLearn style interface::

#!/usr/bin/env python

from tsnecuda import TSNE
embeddedX = TSNE(n_components=2).fit_transform(X)

For more information, check out the repository at

© 2021 Anaconda, Inc. All Rights Reserved. (v2.35.4 f94b7cfe) Legal | Privacy Policy