vllm/requirements.txt

13 lines
314 B
Plaintext
Raw Normal View History

ninja # For faster builds.
psutil
2023-07-20 13:49:31 +08:00
ray >= 2.5.1
sentencepiece # Required for LLaMA tokenizer.
numpy
torch == 2.1.2
transformers >= 4.36.0 # Required for Mixtral.
2023-12-17 18:28:02 +08:00
xformers == 0.0.23.post1 # Required for CUDA 12.1.
fastapi
uvicorn[standard]
2024-01-22 08:05:56 +08:00
pydantic >= 2.0 # Required for OpenAI server.
aioprometheus[starlette]