Project Description
A REST API that classifies text as positive, negative, or neutral using a RoBERTa model
fine-tuned on 124 million tweets. Built with FastAPI and served with an interactive web UI,
it understands modern language including slang and informal expressions.
- Returns sentiment label and confidence score (0–1) for any input text
- Understands internet slang, emoji, and informal language via a Twitter-trained model
- Auto-generated interactive API docs at
/docs (Swagger UI)
- Simple web UI — no setup needed for end users
- Deployed on Railway with a public endpoint ready for integration
Technologies
Hugging Face Transformers · RoBERTa
Railway (Cloud Deployment)
