vllm/requirements.txt

21 lines
510 B
Plaintext
Raw Normal View History

2024-03-19 06:38:33 +08:00
cmake>=3.21
ninja # For faster builds.
psutil
ray >= 2.9
sentencepiece # Required for LLaMA tokenizer.
numpy
torch == 2.2.1
2024-03-29 13:16:12 +08:00
requests
py-cpuinfo
transformers >= 4.39.1 # Required for StarCoder2 & Llava.
xformers == 0.0.25 # Requires PyTorch 2.2.1.
fastapi
uvicorn[standard]
2024-01-22 08:05:56 +08:00
pydantic >= 2.0 # Required for OpenAI server.
prometheus_client >= 0.18.0
pynvml == 11.5.0
triton >= 2.1.0
outlines == 0.0.34
2024-03-28 04:01:46 +08:00
tiktoken == 0.6.0 # Required for DBRX tokenizer
vllm-nccl-cu12>=2.18,<2.19 # for downloading nccl library