Skip to content
Code for Learning K-way D-dimensional Discrete Codes For Compact Embedding Representations
Branch: master
Clone or download
Latest commit 509c74d Jun 30, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data init commit Jun 30, 2018
results/pretrains init commit Jun 30, 2018
scripts init commit Jun 30, 2018
README.md init commit Jun 30, 2018
__init__.py init commit Jun 30, 2018
encoder.py
gumbel.py
ptb_word_lm.py init commit Jun 30, 2018
reader.py
util.py init commit Jun 30, 2018

README.md

This repository contains code for paper:

Ting Chen, Martin Renqiang Min, Yizhou Sun, "Learning K-way D-dimensional Discrete Codes For Compact Embedding Representations" Thirty-fifth International Conference on Machine Learning (ICML'18), Stockholm, Sweden, July. 2018.

To run the code, enter scripts folder, execute *.sh, results (log, as well as checkpoints etc.) will be saved into the results folder.

The code is based on tensorflow/models. Tensorflow 1.4 was originally used, but it seems the code also works in higher versions.

You can’t perform that action at this time.