requirements.txt

groq

# Core ML and AI
torch==2.0.1
transformers==4.40.1
trl==0.8.6

accelerate==0.24.1
peft==0.10.0
datasets

protobuf<4
click<8.1
pydantic>=2.4.0

datasets
huggingface-hub>=0.19
sentencepiece

# Optional: 8-bit quantization (uncomment if needed)
# bitsandbytes==0.41.1

# Document Parsing & OCR
unstructured
# Optional: add specific extras like [pdf] or [local-inference] depending on your platform
# unstructured[pdf]
pytesseract
pdfminer.six

# Vector DB & Search
faiss-cpu>=1.7.4
llama-index==0.9.48
llama-index-vector-stores-faiss==0.2.0
llama-index-embeddings-huggingface==0.3.0


# Document Formats
PyPDF2>=3.0.1
python-docx>=0.8.11
openpyxl>=3.1.2

# Web Interface
gradio>=4.13.0
fastapi>=0.104.0
uvicorn>=0.24.0

# Data & Utilities
numpy>=1.24.0
pandas>=2.0.0
scipy>=1.11.0
requests>=2.31.0
tqdm>=4.66.0
python-dotenv>=1.0.0
pydantic>=2.4.0

# Async & Performance
asyncio-throttle>=1.0.2
aiofiles>=23.2.1

# Logging
loguru>=0.7.2

# Optional: ONNX & quantization (uncomment if needed)
# onnxruntime>=1.16.0
# optimum>=1.14.0
# auto-gptq>=0.4.2
# autoawq>=0.1.6