Currently, the workflow that builds CUDA wheels is failing with the following error:
C:\Miniconda3\envs\build\include\crt/host_config.h(153): fatal error C1189: #error: -- unsupported Microsoft Visual Studio version! Only the versions between 2017 and 2022 (inclusive) are supported! The nvcc flag '-allow-unsupported-compiler' can be used to override this version check; however, using an unsupported host compiler may cause compilation failure or incorrect run time execution. Use at your own risk. [C:\Users\runneradmin\AppData\Local\Temp\tmpwbsbwtdg\build\CMakeFiles\CMakeScratch\TryCompile-uh6ciq\cmTC_cbbed.vcxproj]
See:
The relevant workflow is build-wheels-cuda.yaml, which is based on the original workflows by @jllllll in this repository (no longer maintained).
I have tried adding the -allow-unsupported-compiler flag to my own workflows, as suggested in the error message, and while that did allow the wheel to be successfully compiled, it crashed on my Windows tests. A better solution would be to set a speficic past Visual Studio version in the workflow to avoid similar errors in the future.
This ExLlamaV2 workflow does that, but I have been unable to port it to the llama-cpp-python workflow: build-wheels-release.yml.
Currently, the workflow that builds CUDA wheels is failing with the following error:
See:
The relevant workflow is build-wheels-cuda.yaml, which is based on the original workflows by @jllllll in this repository (no longer maintained).
I have tried adding the
-allow-unsupported-compilerflag to my own workflows, as suggested in the error message, and while that did allow the wheel to be successfully compiled, it crashed on my Windows tests. A better solution would be to set a speficic past Visual Studio version in the workflow to avoid similar errors in the future.This ExLlamaV2 workflow does that, but I have been unable to port it to the llama-cpp-python workflow: build-wheels-release.yml.