This project is a translator assistant that, through terminal, translates English sentences to French sentences.
The goal of this project was to investigate different natural language processing (NLP) models for translating English to French. Models I investigated include: RNN with word index encoding, RNN with one-hot encoding, encoder-decoder, and bidirectional. The model analyses can be found in en_fr.py.
The final model file is translate.py. I was able to achieve an 80% sentences translation accuracy and a 97% word translation accuracy (from English to French).
The data I used was from Susan Li and the code I developed was motivated from Susan Li's code as well. I used Susan's code as starter code and made changes to get my final model and build the terminal interface.
Please see the requirements.txt file for the python requirements. The code to run requires a GPU and can be called with test_job.sh.