huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-16 10:38:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

Optimize the index-select cuda kernel. (#976)

Browse Source
This commit is contained in:
Laurent Mazare
2023-09-28 09:05:29 +01:00
committed by GitHub
parent 8a49e01b9d
commit 5e1c595e00
2 changed files with 12 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
candle-core/src/cuda_backend.rs
View File

@ -884,8 +884,6 @@ impl<'a> Map1 for IndexSelect<'a> {
}; };
let ids_shape = ids_l.shape(); let ids_shape = ids_l.shape();
let ids_dims = ids_shape.dims(); let ids_dims = ids_shape.dims();
let ids_el = ids_shape.elem_count();
let cfg = LaunchConfig::for_num_elems(ids_el as u32);
let ds = dev.htod_copy([ids_dims, ids_l.stride()].concat()).w()?; let ds = dev.htod_copy([ids_dims, ids_l.stride()].concat()).w()?;
let src = match src_l.contiguous_offsets() { let src = match src_l.contiguous_offsets() {
Some((o1, o2)) => src.slice(o1..o2), Some((o1, o2)) => src.slice(o1..o2),
@ -894,11 +892,13 @@ impl<'a> Map1 for IndexSelect<'a> {
let left_size: usize = src_l.dims()[..self.2].iter().product(); let left_size: usize = src_l.dims()[..self.2].iter().product();
let right_size: usize = src_l.dims()[self.2 + 1..].iter().product(); let right_size: usize = src_l.dims()[self.2 + 1..].iter().product();
let dim_size = src_l.dims()[self.2]; let dim_size = src_l.dims()[self.2];
let dst_el = ids_shape.elem_count() * left_size * right_size;
let cfg = LaunchConfig::for_num_elems(dst_el as u32);
let func = dev.get_or_load_func(&kernel_name::<T>(name), kernels::INDEXING)?; let func = dev.get_or_load_func(&kernel_name::<T>(name), kernels::INDEXING)?;
// SAFETY: Set later by running the kernel. // SAFETY: Set later by running the kernel.
let out = unsafe { dev.alloc::<T>(ids_el * left_size * right_size) }.w()?; let out = unsafe { dev.alloc::<T>(dst_el) }.w()?;
let params = ( let params = (
ids_el, dst_el,
ids_dims.len(), ids_dims.len(),
&ds, &ds,
ids, ids,

22
candle-kernels/src/indexing.cu
View File

@ -17,20 +17,14 @@ __device__ void index_select(
) { ) {
const size_t *dims = info; const size_t *dims = info;
const size_t *strides = info + num_dims; const size_t *strides = info + num_dims;
if (is_contiguous(num_dims, dims, strides)) { bool b = 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 dst_i = blockIdx.x * blockDim.x + threadIdx.x; dst_i < numel; dst_i += blockDim.x * gridDim.x) {
for (unsigned int j = 0; j < left_size; ++j) { unsigned int left_i = dst_i / (dim_size * right_size);
memcpy(&out[(i + j * numel) * right_size], &inp[(j * dim_size + ids[i]) * right_size], right_size * sizeof(T)); unsigned int id_i = dst_i / right_size % dim_size;
} unsigned int right_i = dst_i % right_size;
} unsigned int src_i = left_i * (dim_size * right_size) + ids[id_i] * right_size + right_i;
} unsigned strided_i = b ? src_i : get_strided_index(src_i, num_dims, dims, strides);
else { out[dst_i] = inp[strided_i];
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);
for (unsigned int j = 0; j < left_size; ++j) {
memcpy(&out[(i + j * numel) * right_size], &inp[(j * dim_size + ids[strided_i]) * right_size], right_size * sizeof(T));
}
}
} }
} }