Skip to content

Latest commit

 

History

History
52 lines (36 loc) · 1.17 KB

File metadata and controls

52 lines (36 loc) · 1.17 KB

Intro to NLP Project


Contributors:

Aryan Chandramania | Lakshmipathi Balaji | Akshit Kumar


Code

The code is in the form of Python. The code is in the code folder.


.
├── code
│   ├── data_analysis.py
│   ├── Indicbart.py
│   ├── LSTM.py
│   └── mT5.py
├── data
│   ├── dev.txt
│   ├── test.txt
│   ├── train_pre_analysis.txt
│   └── train.txt
├── Presentation.pdf
├── README.md
└── Report.pdf

Data

Google Drive

LinCE


Saved Model Checkpoints

Same Google Drive link as above


Running

You can run the Python scripts directly. Any libraries needed are installed by the scripts. The same goes for the pretrained models. All you need is an internet connection and enough compute power.

The input format for training is

<english> \t <hinglish>

The output for testing is the Hinglish sentence