apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 17:11:21 +08:00
Code Issues Actions 6 Packages Projects Releases Wiki Activity

[XPU] fix thinking bug where output only contains reasoning_content (#4761)

Browse Source 
Co-authored-by: ddchenhao66 <dhaochen163.com>
This commit is contained in:
ddchenhao66
2025-11-04 14:32:35 +08:00
committed by GitHub
parent 4a4948764d
commit bffa08b74b
8 changed files with 201 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files
custom_ops/xpu_ops/src/plugin/include/xpu/plugin.h
+5 -1
View File
@@ -220,9 +220,12 @@ DLL_EXPORT int limit_thinking_content_length_kernel_v1(
int64_t* next_tokens,
const int* max_think_lens,
const int64_t* step_idx,
const int64_t* eos_token_ids,
int* limit_think_status,
bool* stop_flags,
const int64_t think_end_id,
const int bs);
const int bs,
const int eos_token_id_len);
DLL_EXPORT int limit_thinking_content_length_kernel_v2(
api::Context* ctx,
@@ -230,6 +233,7 @@ DLL_EXPORT int limit_thinking_content_length_kernel_v2(
const int* max_think_lens,
const int64_t* step_idx,
int* limit_think_status,
const bool* stop_flags,
const int64_t think_end_id,
const int64_t line_break_id,
const int bs);