This repository contains a implementation of t-Distributed Stochastic Neighbor Embedding (t-SNE) in PyTorch.
train.py
is a training file for t-SNE. Note that data has to be preprocessed.preprocess.py
preprocesses MNIST data for use in the training file. The preprocessing function can be imported directly for use.tsne_module.py
a t-SNE PyTorch file.
The code is based off of the original paper as well as some other implementations of t-SNE