# =============================== # 📦 Embedding + Vector Search # =============================== chromadb sentence-transformers torch # =============================== # 🤖 LLM-Based QA # =============================== transformers accelerate huggingface-hub # =============================== # 📄 PDF Parsing # =============================== pymupdf pdfminer.six # =============================== # 🖼️ OCR + Image Handling # =============================== pytesseract Pillow # =============================== # 🌐 UI Interface # =============================== gradio requests # =============================== # 🛠 Utilities and Fixes # =============================== beautifulsoup4 pydantic # Important: lock to v1.x (chroma not yet compatible with pydantic 2) numpy tqdm # 🔤 NLP & DOCX Parsing nltk spacy python-docx docx2txt