| pdfplumber>=0.10.0 | |
| PyPDF2>=3.0.0 | |
| sentence-transformers>=2.2.0 | |
| chromadb>=0.4.0 | |
| mistralai>=0.1.0 | |
| python-dotenv>=1.0.0 | |
| pydantic-settings>=2.0.0 | |
| # Tokenizers / embeddings / transformers | |
| tiktoken>=0.4.0 | |
| sentence-transformers>=2.2.2 | |
| transformers>=4.30.0 | |
| huggingface-hub>=0.13.0 | |
| torch>=2.0.0 # for local model inference; optional if using remote LLMs | |
| tokenizers>=0.13.3 | |
| sentencepiece>=0.1.99 | |
| # Utilities | |
| python-dotenv>=1.0.0 | |
| requests>=2.30.0 | |
| aiohttp>=3.8.0 | |
| pytesseract>=0.3.10 | |
| pdf2image>=1.16.0 | |
| # Web / API (optional - common for demo apps) | |
| fastapi>=0.95.0 | |
| uvicorn>=0.22.0 | |
| # PDF Processing | |
| pymupdf # PyMuPDF for PDF text extraction (replaces pdf2image for better text extraction) | |
| # LLM Integration | |
| mistralai # Mistral API client (inherited from module_a) | |
| # Core utilities | |
| python-dotenv # Environment variable management | |
| requests # HTTP requests | |
| # API Integration | |
| fastapi # FastAPI framework | |
| pydantic # Data validation | |
| # Testing | |
| pytest>=7.0.0 | |
| pytest-asyncio>=0.21.0 | |
| # Supabase & Authentication | |
| supabase>=2.0.0 | |
| PyJWT[crypto]>=2.8.0 | |
| email-validator>=2.0.0 | |
| python-multipart>=0.0.7 | |
| pinecone |