Files
modelscope/docker/scripts/install_tiny_cuda_nn.sh
mulin.lyh 5ba9fd2307 modify auto gptq and vllm env
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/14790283
* upgrade to python3.10

* modify auto gptq and vllm env

* fix lint issue

* Merge remote-tracking branch 'origin/master' into python10_support

* python310 support

* build from repo

* add commit id force install modelscope every build

* add commit id force install modelscope every build

* fix cpu build issue

* fix datahub error message

* Merge branch 'python10_support' of gitlab.alibaba-inc.com:Ali-MaaS/MaaS-lib into python10_support

* add --no-cache-dir install auto_gptq
2023-11-27 20:21:00 +08:00

8 lines
354 B
Bash

export CMAKE_BUILD_PARALLEL_LEVEL=36 && export MAX_JOBS=36 && export TCNN_CUDA_ARCHITECTURES="50;52;60;61;70;75;80;89;90;86" \
&& git clone --recursive https://github.com/nvlabs/tiny-cuda-nn \
&& cd tiny-cuda-nn \
&& cd bindings/torch \
&& python setup.py install \
&& cd ../../.. \
&& rm -rf tiny-cuda-nn