Pytorch Recurrent Variational Autoencoder

Model:

This is the implementation of Samuel Bowman's Generating Sentences from a Continuous Space with Kim's Character-Aware Neural Language Models embedding for tokens

Sampling examples:

the new machine could be used to increase the number of ventures block in the company 's <unk> shopping system to finance diversified organizations

u.s. government officials also said they would be willing to consider whether the proposal could be used as urging and programs

men believe they had to go on the <unk> because their <unk> were <unk> expensive important

the companies insisted that the color set could be included in the program

Usage

Before model training it is necessary to train word embeddings:

$ python train_word_embeddings.py

This script train word embeddings defined in Mikolov et al. Distributed Representations of Words and Phrases

Parameters:

--use-cuda

--num-iterations

--batch-size

--num-sample –– number of sampled from noise tokens

To train model use:

$ python train.py

Parameters:

--use-cuda

--num-iterations

--batch-size

--learning-rate

--dropout –– probability of units to be zeroed in decoder input

--use-trained –– use trained before model

To sample data after training use:

$ python sample.py

Parameters:

--use-cuda

--num-sample