In this project we have conducted a literature survey to study different model architectures used for implementing the task of Neural Machine Translation. Using this research, we have implemented this Machine Translation at sentence level using the following deep learning architectures -
- RNN with Bahadanu attention
- Transformer
Following translation tasks have been implemented using the mentioned architectures -
-
English to French
-
English to German
The models have been evaluated using the BLEU score performance metric.
-
Clone the repository and unzip the folder
-
Run the "Neural Machine Translation.ipynb" jupyter notebook