sync : llama.cpp (skip)

ggml-ci

diff --git a/ggml-cuda/common.cuh b/ggml-cuda/common.cuh
index 79e1d5424d250917733c8c2d85f7be5d7e374cd5..b98d7cbd0c1c501f9f2e8a1327596af9c0c923de 100644 (file)
--- a/ggml-cuda/common.cuh
+++ b/ggml-cuda/common.cuh
@@ -231,12 +231,6 @@ typedef float dfloat; // dequantize float
 typedef float2 dfloat2;
 #endif //GGML_CUDA_F16
 
-// dmmv = dequantize_mul_mat_vec
-// TODO: remove this?
-#ifndef GGML_CUDA_DMMV_X
-#define GGML_CUDA_DMMV_X 32
-#endif
-
 [[noreturn]]
 static __device__ void no_device_code(
     const char * file_name, const int line, const char * function_name, const int arch, const char * arch_list) {

diff --git a/ggml-cuda/dmmv.cu b/ggml-cuda/dmmv.cu
index ea2950f3d8d9c466750fdf268905348259537616..0b17e3cb961f743d8db4263020d7dd1d49f57ab0 100644 (file)
--- a/ggml-cuda/dmmv.cu
+++ b/ggml-cuda/dmmv.cu
@@ -2,10 +2,6 @@
 #include "dequantize.cuh"
 #include "convert.cuh"
 
-#ifndef GGML_CUDA_MMV_Y
-#define GGML_CUDA_MMV_Y 1
-#endif
-
 #ifndef K_QUANTS_PER_ITERATION
 #define K_QUANTS_PER_ITERATION 2
 #else

diff --git a/ggml-cuda/dmmv.cuh b/ggml-cuda/dmmv.cuh
index 3802678ffceb5cef75de65b08c43132a548533bd..4c5ebd475fdb59ec7f521857c43545fdc7d3b542 100644 (file)
--- a/ggml-cuda/dmmv.cuh
+++ b/ggml-cuda/dmmv.cuh
@@ -1,5 +1,16 @@
 #include "common.cuh"
 
+// dmmv = dequantize_mul_mat_vec
+
+// TODO: remove this?
+#ifndef GGML_CUDA_DMMV_X
+#define GGML_CUDA_DMMV_X 32
+#endif
+
+#ifndef GGML_CUDA_MMV_Y
+#define GGML_CUDA_MMV_Y 1
+#endif
+
 void ggml_cuda_op_dequantize_mul_mat_vec(
     ggml_backend_cuda_context & ctx,
     const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, const char * src0_dd_i, const float * src1_ddf_i,