Update config to include domain-specific embedding models in dropdown

Add F1Score calculation for adherence metric aggregation

Add domain-specific embedding models for financial, law, and customer service

Feature: Background evaluation - run GPT evaluation without keeping UI open

Feature: Auto-save evaluation results to JSON with collection name and timestamp

Fix: Update Qdrant query API for qdrant-client 1.7+

Fix: Load embedding model when loading existing Qdrant collections

Restore full app with all dependencies

Fix: Change SDK from streamlit to docker in README.md

Super minimal test - only streamlit, no deps

Trigger rebuild

Add extensive logging to debug HF Spaces startup

Simple Hello World test app to debug HF Spaces

updated the list of embedding models in config.py to disable some models for a lighter build.

Remove unused deps: openai, langchain, ragas, tenacity, aiohttp

Fix: Remove duplicate torch, remove healthcheck, add logging

Fix HF Spaces timeout: use CPU torch, pre-download model, increase healthcheck timeout

Create chroma_db/temp.txt

updated

Fix: Update Streamlit config to bind to 0.0.0.0 for HuggingFace

Fix: Use CPU-only torch for HuggingFace Spaces free tier

Initial commit for HuggingFace Spaces - RAG Capstone Project with Qdrant Cloud