mirror of
https://github.com/wassname/vllm.git
synced 2026-08-12 12:30:29 +08:00
2f49f155858faaf82bfd076a821497e41e961658
CacheFlow
Installation
pip install psutil numpy torch transformers
pip install flash-attn # This may take up to 10 mins.
pip install -e .
Run
ray start --head
python server.py [--tensor-parallel-size <N>]
Description
A high-throughput and memory-efficient inference and serving engine for LLMs
40 MiB
Languages
Python
85%
Cuda
10.2%
C++
3.1%
C
0.6%
Shell
0.6%
Other
0.4%