VRNN
pytorch (>=0.4 version) code of NIPS 2015 paper 'A Recurrent Latent Variable Model for Sequential Data'
Install / Use
/learn @hjf1997/VRNNREADME
VRNN pytorch version >= 0.4
Introduction
The pytorch version (>=0.4) of variable recurrent neural network (VRNN), which is from the paper 'A Recurrent Latent Variable Model for Sequential Data', Junyoung Chung, Kyle Kastner, Laurent Dinh, Kratarth Goel, Aaron Courville, Yoshua Bengio
The original code is available at Link .
Usage
-
Install pytorch (version >= 0.4). My code is implemented under the version = 1.1.0
-
Run train.py at any IDE (Conda virtual environment and Pycharm is recommend)
-
Change configurations at config.py as you like.
Results
Generating results on MNIST dataset after 160 epoches:
