Follow the given instructions to recreate our work :

1. Clone or Install OpenNMT-py from official GitHub repository : "https://github.com/OpenNMT/OpenNMT-py".

2. Download official data from DravidianLangTech2021 website and extra corpus from OPUS website.

3. Install and Use SentencePiece for subword segmentation on training data.

4. Use the config files for appropriate language pairs.
   Config file contains all parameters,layers,model type like RNN,Transformer etc ,batch size , paths to training and valid data

5. Use config file to train a specific model , all models that will be trained are going to be Transformers 


