Added current notes to README.md

Added Redis to client query script and attempted to fine tune prompt.

Increased k default to give the model more context.

Added Redis for caching.

Ensured file name is passed through with the file upload.

Added virtual enironment to .gitignore.

Updated logic for dev mode, source name passing, and added proper imports.

Upgraded embedding model.

Added a 'thinking...' loading spinner.

Changed prompt logic and added a dev mode.

Added early stage file upload logic, ui, and vector DB configuration.

Added option to choose intended model. Options of tinyllama, llama3, and mistral.

Updated LLM client to send and recieve real messages.

Fixed index error caused by no logic for responses with no context.

Implemented FAISS-based vectorstore with sentence-transformers.