This repository contains the work done for the Kaggle Transformer Sentiment Analysis project. The project aims to perform sentiment analysis using a custom transformer-based encoder and classifier.
Currently, the project has different versions of the notebooks that need to be merged.
- transformer_sentiment_analyzer.ipynb
The main working file that contains all the components of the project.
- data_preprocessing.ipynb
An isolated notebook focusing on the data preprocessing section, aimed at improving this specific component.
- transformer_encoder.ipynb
An isolated notebook dedicated to the transformer encoder section, intended for enhancing this particular part of the project.
- sentiment-analysis.ipynb
An isolated notebook concentrating on the classifier section, with the goal of refining this segment.
The project includes modifications made by the professor and Gebriel. These modifications are denoted by the following naming conventions:
- mod_#: Represents the professor's modifications.
- mod_##: Represents Gebriel's modifications.
The modification versions are labeled with the date they were posted.
Merge the different versions of the notebooks to create a cohesive and unified codebase.
Feel free to contribute to this project by submitting pull requests or opening issues for any suggestions or improvements.
For any questions or inquiries, please reach out to us at [email protected] and [email protected]