cmake : fix cuda vars (#0)

diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index fdd4d58670d81ee4621e115fe1f2cadcf7b31b10..25fb1ad170884ecd5335c3736654b816c87c5b45 100644 (file)
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -194,7 +194,7 @@ if (GGML_CUDA)
             # 60 == f16 CUDA intrinsics
             # 61 == integer CUDA intrinsics
             # 70 == compute capability at which unrolling a loop in mul_mat_q kernels is faster
-            if (LLAMA_CUDA_F16 OR LLAMA_CUDA_DMMV_F16)
+            if (GGML_CUDA_F16 OR GGML_CUDA_DMMV_F16)
                 set(CMAKE_CUDA_ARCHITECTURES "60;61;70") # needed for f16 CUDA intrinsics
             else()
                 set(CMAKE_CUDA_ARCHITECTURES "52;61;70") # lowest CUDA 12 standard + lowest for integer intrinsics