TensorFlow implementation of "Generating Sentences from a Continuous Space"

This is NOT an original implementation. There may be some minor differences from the original structure.

Prerequisites

Dataset is not contained. Please prepare your own dataset.

Pickle file of Numpy array of word ids (shape=[batch_size, sentence_length]).

Pickle file of Python dictionary. It should contain "<EOS>", "<PAD>", "<GO>" as meta words.

  dictionary = {word1: id1,
                word2: id2,
                ...}

  python train.py

  python sampling.py

MIT

Ryo Kamoi