Tensorflow implementation for prodLDA and NVLDA.
Switch branches/tags
Nothing to show
Clone or download
Latest commit 9db5563 Dec 14, 2017
Failed to load latest commit information.
coherence_from_paper_script Update Readme.md Jun 27, 2017
data Create Readme.md Nov 15, 2016
models Update prodlda.py Jun 28, 2017
.gitignore Create .gitignore Nov 15, 2016
LICENSE Create LICENSE Dec 13, 2017
README.md Update README.md Aug 9, 2017
_config.yml Set theme jekyll-theme-cayman May 23, 2017
run.py final before github upload Nov 15, 2016


Autoencoding Variational Inference for Topic Models


  1. As pointed out by @govg, this code depends on a slightly older version of TF. I will try to update it soon, in the meantime you can look up a quick fix here for working with newer version of TF or (3) and (2) below if you'd rather prefer Keras or PyTorch.

  2. @nzw0301 has implemented a Keras version of prodLDA.

  3. @hyqneuron recently implemented a PyTorch version of AVITM. So check out his repo.

  4. Added topic_prop method to both the models. Softmax the output of this method to get the topic proportions.

Code for the ICLR 2017 paper: Autoencoding Variational Inference for Topic Models

> Arxiv

> OpenReview

This is a tensorflow implementation for both of the Autoencoded Topic Models mentioned in the paper.

To run the prodLDA model in the 20Newgroup dataset:

CUDA_VISIBLE_DEVICES=0 python run.py -m prodlda -f 100 -s 100 -t 50 -b 200 -r 0.002 -e 200

Similarly for NVLDA:

CUDA_VISIBLE_DEVICES=0 python run.py -m nvlda -f 100 -s 100 -t 50 -b 200 -r 0.005 -e 300

Check run.py for other options.