vllm/requirements.txt

ninja  # For faster builds.
psutil
ray
sentencepiece  # Required for LLaMA tokenizer.
numpy
torch >= 2.0.0
transformers >= 4.28.0  # Required for LLaMA.
xformers >= 0.0.19
fastapi
uvicorn
pydantic  # Required for OpenAI server.
Specify python package dependencies in requirements.txt (#78) 2023-05-08 07:30:43 +08:00			`ninja # For faster builds.`
			`psutil`
			`ray`
			`sentencepiece # Required for LLaMA tokenizer.`
			`numpy`
			`torch >= 2.0.0`
			`transformers >= 4.28.0 # Required for LLaMA.`
			`xformers >= 0.0.19`
			`fastapi`
			`uvicorn`
OpenAI Compatible Frontend (#116) 2023-05-24 12:39:50 +08:00			`pydantic # Required for OpenAI server.`