[Feat][CPU] Enable Gated DeltaNet Attention (Qwen 3.5 / 3.6) (#41025)

Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>
Co-authored-by: Vadim Gimpelson <156319763+vadiklyutiy@users.noreply.github.com>
This commit is contained in:
Fadi Arafeh
2026-05-07 12:57:09 +08:00
committed by GitHub
co-authored by Vadim Gimpelson
parent 20cac26b19
commit 51f22dcfd0
11 changed files with 615 additions and 7 deletions
+3
View File
@@ -213,6 +213,9 @@ class UnquantizedLinearMethod(LinearMethodBase):
def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
if current_platform.is_cpu():
if layer.weight.ndim != 2:
# this is not a linear layer
return
from vllm.model_executor.layers.utils import dispatch_cpu_unquantized_gemm
dispatch_cpu_unquantized_gemm(layer, remove_weight=True)