CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-04 09:32:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

limit to GGML_ALLOW_CUDA_GRAPHS defined in llama.cpp cmake

Browse Source
This commit is contained in:
Alan Gray
2024-04-25 00:51:48 -07:00
parent 408759687f
commit 0640427f7b
2 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
CMakeLists.txt
View File

@@ -413,6 +413,7 @@ if (LLAMA_CUDA)
list(APPEND GGML_SOURCES_CUDA "ggml-cuda.cu")
add_compile_definitions(GGML_USE_CUDA)
add_compile_definitions(GGML_ALLOW_CUDA_GRAPHS)
if (LLAMA_CUDA_FORCE_DMMV)
add_compile_definitions(GGML_CUDA_FORCE_DMMV)
endif()

2
ggml-cuda.cu
View File

@@ -2405,7 +2405,7 @@ GGML_CALL static void ggml_backend_cuda_synchronize(ggml_backend_t backend) {
GGML_UNUSED(backend);
}
#if (CUDART_VERSION >= 12000)
#if (CUDART_VERSION >= 12000) && defined(GGML_ALLOW_CUDA_GRAPHS)
#define USE_CUDA_GRAPH
#endif