My OpenClassrooms Sentiment Model API

A RESTful API for sentiment analysis built with FastAPI. It predicts the sentiment (positive/negative) of a given text.

Features

• Predict sentiment from text input
• FastAPI-based web server
• Easy to deploy with Uvicorn
• Ready for Docker deployment

Install

pip install -e .

Linting

pip install -e '.[lint]'
pre-commit run --all-files

Test

pip install -e '.[test]'
pytest -sxv tests/

Run

uvicorn oc_sentiment_ml_api.main:app --host 0.0.0.0 --port 8000

Example Usage

curl -X POST "http://localhost:8000/predict" -H "Content-Type: application/json" -d '{"text": "I love this product!"}'

Response:

{
  "sentiment": "positive",
  "confidence": 0.98
}

Contributing

Pull requests are welcome! Please open an issue first to discuss changes.

License

MIT License