minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[CPU] Add Qwen3.5 model optimization for CPU (#19484)

Browse Source 
Co-authored-by: Zheng, Beilei <beilei.zheng@intel.com>
Co-authored-by: Ma Mingfei <mingfei.ma@intel.com>
Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
This commit is contained in:
jianan-gu
2026-04-26 10:12:36 -07:00
committed by GitHub
co-authored by Zheng, Beilei Ma Mingfei Xinyuan Tong
parent 7d49564431
commit 10fd0faccd
20 changed files with 768 additions and 209 deletions
Download Patch File Download Diff File Expand all files Collapse all files
sgl-kernel/csrc/cpu/torch_extension_cpu.cpp
+15
View File
@@ -374,6 +374,15 @@ std::tuple<at::Tensor, at::Tensor, at::Tensor, at::Tensor> fused_qkvzba_split_re
int64_t head_qk,
int64_t head_v);
// fused_qkvzba_split_reshape_cat_cpu_contiguous
std::tuple<at::Tensor, at::Tensor, at::Tensor, at::Tensor> fused_qkvzba_split_reshape_cat_contiguous_cpu(
const at::Tensor& mixed_qkvz,
const at::Tensor& mixed_ba,
int64_t num_heads_qk,
int64_t num_heads_v,
int64_t head_qk,
int64_t head_v);
// image preprocessor
std::tuple<at::Tensor, at::Tensor> image_preprocess_cpu(
at::TensorList images,
@@ -621,6 +630,12 @@ TORCH_LIBRARY_FRAGMENT(sgl_kernel, m) {
"fused_qkvzba_split_reshape_cat_cpu(Tensor mixed_qkvz, Tensor mixed_ba, int num_heads_qk, int num_heads_v, int "
"head_qk, int head_v) -> (Tensor, Tensor, Tensor, Tensor)");
m.impl("fused_qkvzba_split_reshape_cat_cpu", torch::kCPU, &fused_qkvzba_split_reshape_cat_cpu);
// fused_qkvzba_split_reshape_cat_contiguous_cpu
m.def(
"fused_qkvzba_split_reshape_cat_contiguous_cpu(Tensor mixed_qkvz, Tensor mixed_ba, int num_heads_qk, int "
"num_heads_v, int "
"head_qk, int head_v) -> (Tensor, Tensor, Tensor, Tensor)");
m.impl("fused_qkvzba_split_reshape_cat_contiguous_cpu", torch::kCPU, &fused_qkvzba_split_reshape_cat_contiguous_cpu);
// image preprocessor
m.def(