# ===============================
# 📦 Embedding + Vector Search
# ===============================
chromadb
sentence-transformers         # Compatible with huggingface-hub 0.30.1
torch               # Stable with sentence-transformers

# ===============================
# 🤖 LLM-Based QA
# ===============================
transformers                # Works well with huggingface-hub 0.30.1
accelerate
huggingface-hub             # Compatible with transformers 4.37.2

# ===============================
# 📄 PDF Parsing
# ===============================
pymupdf                     # PyMuPDF for full-page text extraction
pdfminer.six              # Optional: structured layout extraction

# ===============================
# 🖼️ OCR + Image Handling
# ===============================
pytesseract                 # Requires separate install of Tesseract binary
Pillow

# ===============================
# 🌐 UI Interface
# ===============================
gradio                       # Gradio 4+ for modern UI
requests

# ===============================
# 🛠 Utilities and Fixes
# ===============================
beautifulsoup4              # Parsing for HTML-in-PDFs (e.g., diagrams/tables)
pydantic                     # Chromadb is not yet compatible with pydantic 2.x
numpy                       # Ensures compatibility with chromadb and transformers
tqdm                       # Progress bar (used in embedding scripts)

# Natural Language Toolkit=========
nltk