SmokeScan / requirements.txt
KinetoLabs's picture
Pin vLLM <0.13.0 to fix V1 engine hang on multi-GPU
a65b765
raw
history blame contribute delete
506 Bytes
# Core ML/AI
torch
transformers>=4.57.0
accelerate
qwen-vl-utils>=0.0.14
torchvision
# vLLM for FP8 quantized model inference (>=0.11.0 required for Qwen3-VL support)
# Pinned <0.13.0: V0 engine deprecated in 0.13.x, VLLM_USE_V1=0 not honored
vllm>=0.11.0,<0.13.0
# UI
gradio>=6.0.0,<7.0.0
# RAG/Vector Store
chromadb
# Data Validation
pydantic
pydantic-settings
# Image Processing
pillow
# PDF Processing
pdfplumber
weasyprint>=60.0
markdown>=3.5
# Utilities
numpy
# Testing
pytest
pytest-asyncio