/seq2seq

Example attention-seq2seq implementations.

Primary LanguagePython

seq2seq

A seq2seq implementation with support for the following:

  • Batch updates
  • Bahdanau and Luong attention-based decoding
  • Teacher forcing via scheduled sampling

Example translation data can be downloaded here.

This codebase started as a modular reimplementation of this seq2seq tutorial.

Get started

virtualenv .venv
source .venv/bin/activate
pip install -r requirements.txt
python run.py

References

  • Bahdanau et al. Neural Machine Translation by Jointly Learning to Align and Translate. 2014. link
  • Luong et al. Effective Approaches to Attention-based Neural Machine Translation. 2015. link
  • Bengio et al. Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks. 2015. link