apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 08:21:53 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

[XPU] support XPU VL model inference (#4030)

Browse Source 
* [XPU] support XPU VL model inference

* fix image op import and device check

* rebase develop

* fix perf
This commit is contained in:
Lucas
2025-09-25 14:34:15 +08:00
committed by GitHub
parent e36eccfdad
commit 87179cb744
18 changed files with 1300 additions and 146 deletions
Download Patch File Download Diff File Expand all files Collapse all files
custom_ops/xpu_ops/src/plugin/include/xpu/plugin.h
+19
View File
@@ -140,6 +140,25 @@ DLL_EXPORT int quant2d_per_channel(api::Context *ctx, const TX *x,
const TSCALE *scale_in, TY *y,
TSCALE *scale_out, int64_t m, int64_t n);
DLL_EXPORT int text_image_index_out(Context* ctx,
const int* token_type_ids, // x
int* text_index, // y1
int* image_index, // y2
const int64_t token_num);
template <typename T>
DLL_EXPORT int text_image_gather_scatter(api::Context* ctx,
T* input,
T* text_input,
T* image_input,
int* token_type_ids,
int* text_index,
int* image_index,
int64_t token_num,
int64_t text_token_num,
int64_t image_token_num,
int64_t hidden_size,
bool is_scatter);
/*--------------------------------------- MTP being --------------------------------------------*/