mirror of
https://github.com/LostRuins/koboldcpp.git
synced 2026-05-18 23:49:46 +00:00
try remove volta as a dedicated target b (+1 squashed commits)
Squashed commits: [2df689a03] try remove volta as a dedicated target
This commit is contained in:
parent
1a14ae1183
commit
ddba580f00
2 changed files with 2 additions and 4 deletions
|
|
@ -139,10 +139,10 @@ if (LLAMA_CUBLAS)
|
|||
elseif(CUDAToolkit_VERSION VERSION_GREATER 12)
|
||||
add_compile_definitions(GGML_CUDA_USE_GRAPHS) #try enable cuda graphs on cu12 build
|
||||
add_compile_definitions(KCPP_LIMIT_CUDA_MAX_ARCH=800)
|
||||
set(CMAKE_CUDA_ARCHITECTURES "50-virtual;61-virtual;70-virtual;75-virtual;80-virtual") # lowest CUDA 12 standard + lowest for integer intrinsics
|
||||
set(CMAKE_CUDA_ARCHITECTURES "50-virtual;61-virtual;75-virtual;80-virtual") # lowest CUDA 12 standard + lowest for integer intrinsics
|
||||
else()
|
||||
add_compile_definitions(KCPP_LIMIT_CUDA_MAX_ARCH=750) #will cause issues with ggml_cuda_highest_compiled_arch if removed
|
||||
set(CMAKE_CUDA_ARCHITECTURES "35-virtual;50-virtual;61-virtual;70-virtual;75-virtual") # lowest CUDA 12 standard + lowest for integer intrinsics
|
||||
set(CMAKE_CUDA_ARCHITECTURES "35-virtual;50-virtual;61-virtual;75-virtual") # lowest CUDA 12 standard + lowest for integer intrinsics
|
||||
endif()
|
||||
endif()
|
||||
message(STATUS "Using CUDA architectures: ${CMAKE_CUDA_ARCHITECTURES}")
|
||||
|
|
|
|||
2
Makefile
2
Makefile
|
|
@ -226,7 +226,6 @@ NVCCFLAGS += -Wno-deprecated-gpu-targets \
|
|||
-gencode arch=compute_35,code=compute_35 \
|
||||
-gencode arch=compute_50,code=compute_50 \
|
||||
-gencode arch=compute_61,code=compute_61 \
|
||||
-gencode arch=compute_70,code=compute_70 \
|
||||
-gencode arch=compute_75,code=compute_75 \
|
||||
-DKCPP_LIMIT_CUDA_MAX_ARCH=750
|
||||
|
||||
|
|
@ -234,7 +233,6 @@ else ifdef LLAMA_ARCHES_CU12
|
|||
NVCCFLAGS += -Wno-deprecated-gpu-targets \
|
||||
-gencode arch=compute_50,code=compute_50 \
|
||||
-gencode arch=compute_61,code=compute_61 \
|
||||
-gencode arch=compute_70,code=compute_70 \
|
||||
-gencode arch=compute_75,code=compute_75 \
|
||||
-gencode arch=compute_80,code=compute_80 \
|
||||
-DKCPP_LIMIT_CUDA_MAX_ARCH=800
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue