fix: model singleton cache, dedup guard, Gradio type=messages 9edd318 Bhaskar Ram commited on 4 days ago
fix: sentence-aware chunking, score threshold, DOCX tables, streaming error handling, LLM_MODEL env var 2623b17 Bhaskar Ram commited on 4 days ago
feat: apply all 15 upgrades — BGE embeddings, cosine FAISS, streaming LLM, tenacity retry, dotenv, Dockerfile, tests a465955 Bhaskar Ram commited on 7 days ago