baalajimaestro/ollama
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Explicitly disable AVX2 on GPU builds

Browse source 
Even though we weren't setting it to on, somewhere in the cmake config
it was getting toggled on.  By explicitly setting it to off, we get `/arch:AVX`
as intended.
This commit is contained in:
Daniel Hiltgen 2024-02-15 14:50:11 -08:00
parent 86808f80a8
commit db2a9ad1fe
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
llm/generate/gen_windows.ps1
View file

@ -185,7 +185,7 @@ if ($null -ne $script:CUDA_LIB_DIR) {
}
init_vars
$script:buildDir="${script:llamacppDir}/build/windows/${script:ARCH}/cuda$script:CUDA_VARIANT"
$script:cmakeDefs += @("-DLLAMA_CUBLAS=ON", "-DLLAMA_AVX=on", "-DCUDAToolkit_INCLUDE_DIR=$script:CUDA_INCLUDE_DIR", "-DCMAKE_CUDA_ARCHITECTURES=${script:CMAKE_CUDA_ARCHITECTURES}")
$script:cmakeDefs += @("-DLLAMA_CUBLAS=ON", "-DLLAMA_AVX=on", "-DLLAMA_AVX2=off", "-DCUDAToolkit_INCLUDE_DIR=$script:CUDA_INCLUDE_DIR", "-DCMAKE_CUDA_ARCHITECTURES=${script:CMAKE_CUDA_ARCHITECTURES}")
build
install
sign