Neural network machine translation using a sequence to sequence model on a dataset of English and French sentences.
- Small portion of English and French corpus is in the
data/
subdirectory - For more vocabulary train on the WMT10 French-English corpus with a GPU (and once the network is performing well on the limited dataset)
code.ipynb
(Jupyter notebook)
- Implements encoder-decoer RNN using tensorflow
seq2seq
- Testing and output for both datasets is inline in the notebook