feat: add token usage tracking and display, update sample questions for demo scenarios

feat: expand database from 20K to 34K documents, add processing time tracking and per-phase statistics display

fix: increase max_tokens to 2048 for longer language answers

fix: increase max_tokens from 300 to 1024 to prevent answer truncation

Add FAISS index to repo

Pre-build models in Docker for fast startup

Add pre-built FAISS index via LFS

Add groq dependency

Deploy Bio-RAG