mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-05-02 07:22:10 +00:00
13 lines
665 B
Plaintext
13 lines
665 B
Plaintext
|
flashinfer_commit := v0.1.5
|
||
|
|
|
||
|
|
build-flashinfer:
|
||
|
|
git clone https://github.com/flashinfer-ai/flashinfer.git flashinfer && \
|
||
|
|
cd flashinfer && git fetch && git checkout $(flashinfer_commit) && \
|
||
|
|
git submodule update --init --recursive && \
|
||
|
|
cd python/ && \
|
||
|
|
CUDA_ARCH_LIST="8.0;9.0a" NVCC_GENCODE="-gencode=arch=compute_80,code=sm_80 -gencode=arch=compute_90a,code=sm_90a" TORCH_CUDA_ARCH_LIST="8.0;9.0a" python setup.py build
|
||
|
|
|
||
|
|
install-flashinfer: build-flashinfer
|
||
|
|
cd flashinfer/python/ && \
|
||
|
|
CUDA_ARCH_LIST="8.0;9.0a" NVCC_GENCODE="-gencode=arch=compute_80,code=sm_80 -gencode=arch=compute_90a,code=sm_90a" TORCH_CUDA_ARCH_LIST="8.0;9.0a" python setup.py install
|