huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-19 03:54:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add the reduce-sum kernel.

Browse Source
This commit is contained in:
laurent
2023-06-26 12:35:26 +01:00
parent 16f0f5b9d2
commit f6104c4b64
3 changed files with 94 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

46
kernels/src/reduce.cu Normal file
View File

@ -0,0 +1,46 @@
#include "cuda_utils.cuh"
#include<stdint.h>
#define SUM_OP(TYPENAME, FN_NAME) \
extern "C" __global__ void FN_NAME( \
const size_t numel, \
const size_t num_dims, \
const size_t num_sum_dims, \
const size_t *info, \
const TYPENAME *inp, \
TYPENAME *out \
) { \
const size_t *dims = info; \
const size_t *strides = info + num_dims; \
const size_t *sum_dims_l = info + 2*num_dims; \
const size_t *sum_dims_s = info + 2*num_dims + num_sum_dims; \
if (is_contiguous(num_dims, dims, strides)) { \
for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
size_t dst_index = i; \
for (unsigned int nd = 0; nd < num_sum_dims; ++nd) { \
size_t stride = sum_dims_s[nd]; \
size_t pre = dst_index / stride; \
size_t post = dst_index / stride; \
dst_index = (pre / sum_dims_l[nd]) * stride + post; \
} \
out[dst_index] += inp[i]; \
} \
} \
else { \
for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
unsigned strided_i = get_strided_index(i, num_dims, dims, strides); \
size_t dst_index = i; \
for (unsigned int nd = 0; nd < num_sum_dims; ++nd) { \
size_t stride = sum_dims_s[nd]; \
size_t pre = dst_index / stride; \
size_t post = dst_index / stride; \
dst_index = (pre / sum_dims_l[nd]) * stride + post; \
} \
out[dst_index] += inp[strided_i]; \
} \
} \
} \
SUM_OP(float, sum_f32)
SUM_OP(double, sum_f64)
SUM_OP(uint32_t, sum_u32)