Plan, Attend, Generate: Planning for Sequence-to-Sequence Models

This is a theano implementation of Plan, Attend, Generate, described in our paper:

Francis Dutil, Caglar Gulcehre, Adam Trischler, Yoshua Bengio, Plan, Attend, Generate: Planning for Sequence-to-Sequence Models (NIPS 2017)

Dependencies

The code for character-level translation is based on the code provided by:

Junyoung Chung, Kyunghyun Cho, Yoshua Bengio, A Character-Level Decoder without Explicit Segmentation for Neural Machine Translation: https://github.com/nyu-dl/dl4mt-cdec

To reproduce the results, theano version 0.9.0 needs to be used.

About

Plan, Attend, Generate: Planning for Sequence-to-Sequence Models

Languages

Language:Python 71.4%Language:Jupyter Notebook 22.4%Language:Perl 3.5%Language:Emacs Lisp 1.3%Language:Shell 0.9%Language:Smalltalk 0.1%Language:Ruby 0.1%Language:NewLisp 0.1%Language:JavaScript 0.1%Language:Slash 0.0%Language:SystemVerilog 0.0%