[MoE Refactor] Introduce MoERunner abstraction and move execution logic from FusedMoE to DefaultMoERunner (#32344)
Signed-off-by: Bill Nell <bnell@redhat.com>
This commit is contained in:
@@ -635,6 +635,7 @@ class GGUFMoEMethod(FusedMoEMethodBase):
|
||||
x: torch.Tensor,
|
||||
topk_weights: torch.Tensor,
|
||||
topk_ids: torch.Tensor,
|
||||
shared_experts_input: torch.Tensor | None,
|
||||
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
|
||||
assert layer.activation == "silu", "Only SiLU activation is supported."
|
||||
if layer.apply_router_weight_on_input:
|
||||
|
||||
Reference in New Issue
Block a user