apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-04 08:16:42 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Add custom op declaration for all_reduce (#3473)

Browse Source 
* add custom op declaration

* roll back try except
This commit is contained in:
Ryan
2025-08-20 20:29:58 +08:00
committed by GitHub
parent 33ff0bfe38
commit bcdfc1d6b9
4 changed files with 16 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
custom_ops/gpu_ops/cpp_extensions.cc
View File

@@ -530,7 +530,7 @@ paddle::Tensor FusedHadamardQuantFp8Func(
int64_t init_custom_all_reduce(const std::vector<int64_t>& fake_ipc_ptrs,
paddle::Tensor& rank_data, int64_t rank, bool full_nvlink);
void all_reduce(int64_t _fa, paddle::Tensor& inp, paddle::Tensor& out,
void all_reduce(paddle::Tensor& inp, paddle::Tensor& out, int64_t _fa,
int64_t reg_buffer, int64_t reg_buffer_sz_bytes);
void dispose(int64_t _fa);