hf_text-generation-inference/server/Makefile-vllm

26 lines
783 B
Plaintext

vllm-cuda:
# Clone vllm
pip install -U ninja packaging --no-cache-dir
git clone https://github.com/Narsil/vllm.git vllm
build-vllm-cuda: vllm-cuda
cd vllm && git fetch && git checkout b5dfc61db88a81069e45b44f7cc99bd9e62a60fa
cd vllm && python setup.py build
install-vllm-cuda: build-vllm-cuda
pip uninstall vllm -y || true
cd vllm && python setup.py install
vllm-rocm:
# Clone vllm
pip install -U ninja packaging --no-cache-dir
git clone https://github.com/fxmarty/rocm-vllm.git vllm
build-vllm-rocm: vllm-rocm
cd vllm && git fetch && git checkout ca6913b3c2ffacdcb7d15e914dc34adbc6c89479
cd vllm && PYTORCH_ROCM_ARCH="gfx90a;gfx942" python setup.py install
install-vllm-rocm: build-vllm-rocm
pip uninstall vllm -y || true
cd vllm && python setup.py install