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

disable avx while still allowing gpu support
Some checks failed
release / build-darwin (push) Has been cancelled
Details
release / generate-windows-cpu (push) Has been cancelled
Details
release / generate-windows-rocm (push) Has been cancelled
Details
release / generate-windows-cuda (map[url:https://developer.download.nvidia.com/compute/cuda/11.3.1/local_installers/cuda_11.3.1_465.89_win10.exe version:11]) (push) Has been cancelled
Details
release / generate-windows-cuda (map[url:https://developer.download.nvidia.com/compute/cuda/12.4.0/local_installers/cuda_12.4.0_551.61_windows.exe version:12]) (push) Has been cancelled
Details
release / build-windows (push) Has been cancelled
Details
release / build-linux-amd64 (push) Has been cancelled
Details
release / build-linux-arm64 (push) Has been cancelled
Details
release / build-container-image (linux) (push) Has been cancelled
Details
release / build-container-image (linux-arm64) (push) Has been cancelled
Details
release / merge (push) Has been cancelled
Details
release / build-container-image-rocm (push) Has been cancelled
Details
release / release (push) Has been cancelled
Details

Browse source 
as per discussion for this issue and the most recent comment on how
to fix this issue, at least temporarily, here:

https://github.com/ollama/ollama/issues/2187#issuecomment-2262876198
This commit is contained in:
Gered 2024-09-22 13:27:17 -04:00
parent 504a410f02
commit 47c356a6cd
2 changed files with 4 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
gpu/cpu_common.go
View file

@ -17,7 +17,8 @@ func GetCPUCapability() CPUCapability {
return CPUCapabilityAVX return CPUCapabilityAVX
} }
// else LCD // else LCD
return CPUCapabilityNone //return CPUCapabilityNone
return CPUCapabilityAVX
} }
func IsNUMA() bool { func IsNUMA() bool {

3
llm/generate/gen_linux.sh
View file

@ -52,7 +52,8 @@ if [ -z "${CUDACXX}" ]; then
export CUDACXX=$(command -v nvcc) export CUDACXX=$(command -v nvcc)
fi fi
fi fi
COMMON_CMAKE_DEFS="-DCMAKE_SKIP_RPATH=on -DBUILD_SHARED_LIBS=on -DCMAKE_POSITION_INDEPENDENT_CODE=on -DGGML_NATIVE=off -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_OPENMP=off" # COMMON_CMAKE_DEFS="-DCMAKE_SKIP_RPATH=on -DBUILD_SHARED_LIBS=on -DCMAKE_POSITION_INDEPENDENT_CODE=on -DGGML_NATIVE=off -DGGML_AVX=on -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_OPENMP=off"
COMMON_CMAKE_DEFS="-DCMAKE_SKIP_RPATH=on -DBUILD_SHARED_LIBS=on -DCMAKE_POSITION_INDEPENDENT_CODE=on -DGGML_NATIVE=off -DGGML_AVX=off -DGGML_AVX2=off -DGGML_AVX512=off -DGGML_FMA=off -DGGML_F16C=off -DGGML_OPENMP=off"
source $(dirname $0)/gen_common.sh source $(dirname $0)/gen_common.sh
init_vars init_vars
git_module_setup git_module_setup