requirements.txt

# Streamlit for the web interface
streamlit==1.24.0  

# Transformers and datasets for working with RAG and language models
transformers==4.30.0  
datasets  

# FAISS for vector search and retrieval
faiss-cpu==1.7.2  

# Torch for model inference
torch==1.11.0  

# PDF handling
pdfplumber==0.7.4  

# Pandas for CSV handling
pandas==2.1.1  

# Hugging Face Hub for downloading and managing models
huggingface-hub>=0.19  

# HF Transfer for model download optimizations
hf-transfer>=0.1.4  

# Protobuf for model file compatibility
protobuf<4  

# Pydantic for data validation
pydantic>=2.0  

# Click for command-line interaction compatibility
click<8.1  

# Numpy for matrix operations and numerical computations
numpy==1.23.0  

# scikit-learn for cosine similarity and other machine learning utilities
scikit-learn==1.2.2  

# Optional: pytesseract and Pillow for OCR if necessary for PDFs without extractable text
pytesseract==0.3.9  
Pillow==9.4.0  

sentence-transformers==2.2.2