Semantic search of UCSD courses built BERT pre-trained models and Pinecone vector database.
to use
pip install -r requirements
streamlit run app.py
to scrape and retrain model
python3 scraper.py
python3 embeddings.py
make sure you have a Pinecone account and store your key in .env
note: scraped data was not thoroughly validated and there could be some name/description mismatches