haraldwolff/llama.cpp
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

Fix compile error on Windows CUDA (#2207)

Browse source
This commit is contained in:
Howard Su 2023-07-13 21:58:09 +08:00 committed by GitHub
parent b782422a3e
commit ff5d58faec
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
ggml-cuda.cu
View file

@ -267,10 +267,9 @@ static __global__ void mul_f32(const float * x, const float * y, float * dst, co
dst[i] = x[i] * y[i%ky];
}
static const float GELU_COEF_A = 0.044715f;
static const float SQRT_2_OVER_PI = 0.79788456080286535587989211986876f;
static __global__ void gelu_f32(const float * x, float * dst, const int k) {
const float GELU_COEF_A = 0.044715f;
const float SQRT_2_OVER_PI = 0.79788456080286535587989211986876f;
const int i = blockDim.x*blockIdx.x + threadIdx.x;
if (i >= k) {
@ -2300,7 +2299,7 @@ inline void ggml_cuda_op_add(
const int64_t ne00 = src0->ne[0];
const int64_t i01_diff = i01_high - i01_low;
const int64_t ne10 = src1->ne[0];
// const int64_t ne10 = src1->ne[0];
// compute
if (src0->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F32) {