1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
|
#include "scale.cuh"
#define MAX_GRIDDIM_X 0x7FFFFFFF
static __global__ void scale_f32(const float * x, float * dst, const float scale, const float bias, const int64_t nelements) {
int64_t tid = (int64_t)blockIdx.x * (int64_t)blockDim.x + (int64_t)threadIdx.x;
int64_t stride = (int64_t)blockDim.x * (int64_t)gridDim.x;
for (int64_t i = tid; i < nelements; i += stride) {
dst[i] = scale * x[i] + bias;
}
}
static void scale_f32_cuda(const float * x, float * dst, const float scale, const float bias, const int64_t nelements, cudaStream_t stream) {
const int64_t num_blocks = (nelements + CUDA_SCALE_BLOCK_SIZE - 1) / CUDA_SCALE_BLOCK_SIZE;
scale_f32<<<MIN(MAX_GRIDDIM_X, num_blocks), CUDA_SCALE_BLOCK_SIZE, 0, stream>>>(x, dst, scale, bias, nelements);
}
void ggml_cuda_op_scale(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
const ggml_tensor * src0 = dst->src[0];
const float * src0_d = (const float *)src0->data;
float * dst_d = (float *)dst->data;
cudaStream_t stream = ctx.stream();
GGML_ASSERT(src0->type == GGML_TYPE_F32);
GGML_ASSERT( dst->type == GGML_TYPE_F32);
float scale;
float bias;
memcpy(&scale, (float *) dst->op_params + 0, sizeof(float));
memcpy(&bias, (float *) dst->op_params + 1, sizeof(float));
scale_f32_cuda(src0_d, dst_d, scale, bias, ggml_nelements(src0), stream);
}
|
