Autoencoding Variational Inference for Topic Models
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.
@nzw0301 has implemented a Keras version of prodLDA.
@hyqneuron recently implemented a PyTorch version of AVITM. So check out his repo.
topic_propmethod 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
This is a tensorflow implementation for both of the Autoencoded Topic Models mentioned in the paper.
To run the
prodLDA model in the
CUDA_VISIBLE_DEVICES=0 python run.py -m prodlda -f 100 -s 100 -t 50 -b 200 -r 0.002 -e 200
CUDA_VISIBLE_DEVICES=0 python run.py -m nvlda -f 100 -s 100 -t 50 -b 200 -r 0.005 -e 300
run.py for other options.