apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

enable dcu ci (#3402)

Browse Source
This commit is contained in:
lifulll
2025-08-29 10:23:08 +08:00
committed by GitHub
parent 73d60fe64d
commit 72094d4d82
11 changed files with 295 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
custom_ops/gpu_ops/get_padding_offset.cu
View File

@@ -46,7 +46,11 @@ __global__ void GetPaddingOffsetKernel(int *batch_id_per_token,
const int ti = threadIdx.x;
int cum_offset = bi == 0 ? 0 : cum_offsets[bi - 1];
for (int i = ti; i < seq_lens[bi]; i += blockDim.x) {
#ifdef PADDLE_WITH_HIP
batch_id_per_token[bi * max_seq_len - cum_offset + i] = cum_offset;
#else
batch_id_per_token[bi * max_seq_len - cum_offset + i] = bi;
#endif
}
if (ti == 0) {
cum_offsets_out[bi] = cum_offset;