mirror of
https://github.com/huggingface/candle.git
synced 2025-06-16 10:38:54 +00:00
Different casting ?
This commit is contained in:
@ -13,13 +13,13 @@ extern "C" __global__ void FN_NAME( \
|
|||||||
const size_t *strides = info + num_dims; \
|
const size_t *strides = info + num_dims; \
|
||||||
if (is_contiguous(num_dims, dims, strides)) { \
|
if (is_contiguous(num_dims, dims, strides)) { \
|
||||||
for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
|
for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
|
||||||
out[i] = (DST_TYPENAME) inp[i]; \
|
out[i] = DST_TYPENAME(inp[i]); \
|
||||||
} \
|
} \
|
||||||
} \
|
} \
|
||||||
else { \
|
else { \
|
||||||
for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
|
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); \
|
unsigned strided_i = get_strided_index(i, num_dims, dims, strides); \
|
||||||
out[i] = (DST_TYPENAME) inp[strided_i]; \
|
out[i] = DST_TYPENAME(inp[strided_i]); \
|
||||||
} \
|
} \
|
||||||
} \
|
} \
|
||||||
} \
|
} \
|
||||||
|
Reference in New Issue
Block a user