This repository contains my work completed during a summer internship, focused on Natural Language Processing (NLP) and transformer-based models for sentiment analysis.
The primary goal of this project is to explore and implement modern NLP techniques using deep learning models such as BERT, RoBERTa, and DeBERTa for sentiment classification tasks, including Hindi text sentiment analysis.
Summer_intern/
│
├── Bert_model.ipynb # BERT implementation
├── Roberta_model.ipynb # RoBERTa implementation
├── Deberta_model_1.ipynb # DeBERTa implementation
├── sentiment_analysis.ipynb # General sentiment analysis pipeline
├── hindi_text_sentiment.csv # Dataset used
├── Summer_Internship_Presentation.pptx # Final presentation
└── freeglut-MinGW-3.0.0-1.mp/ # Supporting files
- BERT (Bidirectional Encoder Representations from Transformers)
- RoBERTa (Robustly Optimized BERT Approach)
- DeBERTa (Decoding-enhanced BERT with disentangled attention)
These models were fine-tuned for sentiment classification tasks.
- Implementation of transformer-based models for NLP tasks
- Sentiment analysis on Hindi text dataset
- Comparative experimentation with multiple architectures
- Data preprocessing and model evaluation
- Python
- Jupyter Notebook
- PyTorch / Transformers (Hugging Face)
- Pandas, NumPy, Scikit-learn
The dataset used:
hindi_text_sentiment.csv- Contains labeled Hindi text for sentiment classification
- Clone the repository:
git clone https://github.com/your-username/your-repo-name.git
- Navigate to the folder:
cd Summer_intern
- Open Jupyter Notebook:
jupyter notebook
- Run any
.ipynbfile step by step
- Successfully implemented multiple transformer models
- Observed performance differences across architectures
- Built a working sentiment classification pipeline