iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[CPU][Perf]Added tanh AOR for faster gelu activations. (#44639)

Browse Source 
Signed-off-by: Anna Mayne <anna.mayne@arm.com>
Signed-off-by: almayne <anna.mayne@arm.com>
Co-authored-by: Li, Jiang <jiang1.li@intel.com>
This commit is contained in:
almayne
2026-06-30 23:24:40 -07:00
committed by GitHub
co-authored by Li, Jiang <jiang1.li@intel.com>
parent b446792306
commit 89e99202f2
9 changed files with 328 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
csrc/cpu/torch_bindings.cpp
+4
View File
@@ -298,6 +298,10 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
ops.def("gelu_tanh_and_mul(Tensor! out, Tensor input) -> ()");
ops.impl("gelu_tanh_and_mul", torch::kCPU, &gelu_tanh_and_mul);
// GELU tanh implementation.
ops.def("gelu_tanh(Tensor! out, Tensor input) -> ()");
ops.impl("gelu_tanh", torch::kCPU, &gelu_tanh);
// GELU implementation used in GPT-2.
ops.def("gelu_new(Tensor! out, Tensor input) -> ()");
ops.impl("gelu_new", torch::kCPU, &gelu_new);