Enable B12x backend for non-gated MoEs (like Nemotron) (#43328)

Signed-off-by: Andrii Skliar <askliar@nvidia.com>
Co-authored-by: Andrii Skliar <askliar@nvidia.com>
This commit is contained in:
Andrii Skliar
2026-07-06 12:40:07 -07:00
committed by GitHub
co-authored by Andrii Skliar
parent b136cc2c2c
commit b1384f5ec6
3 changed files with 225 additions and 40 deletions
+2 -1
View File
@@ -55,6 +55,7 @@ def make_dummy_moe_config(
intermediate_size: int = 1,
in_dtype: torch.dtype = torch.bfloat16,
max_num_tokens: int = 512,
activation: MoEActivation = MoEActivation.SILU,
) -> FusedMoEConfig:
"""
This is a dummy config for the mk constructor interface
@@ -73,7 +74,7 @@ def make_dummy_moe_config(
else num_experts,
num_logical_experts=num_experts,
moe_parallel_config=FusedMoEParallelConfig.make_no_parallel(),
activation=MoEActivation.SILU,
activation=activation,
in_dtype=in_dtype,
device="cuda",
routing_method=RoutingMethodType.TopK,