RAG project planning:
1. Data source: book(full english version of Count of Monte Cristo).
2. Chunking done by size.
3. Retriver bm25 + bi-encoder.
4. Reranker(sentence bert).
5. LiteLLM library (llama3-8b-8192 model).
6. Metadata filtering wasn't used.
7. Citation: user will be provided with the text citation along with Chapter this text belongs to.
8. Vector database wasn't used.
9. Web UI done with gradio