[PluggableLayer][3/N] Apply PluggableLayer to moe-related layers. (#33556)

Signed-off-by: whx-sjtu <2952154980@qq.com>
This commit is contained in:
Hexiang Wang
2026-04-14 09:55:00 -04:00
committed by GitHub
parent e1e318af01
commit f02b3269e7
2 changed files with 6 additions and 13 deletions
@@ -24,7 +24,7 @@ import torch.nn as nn
from vllm.config.utils import getattr_iter
from vllm.distributed import get_dp_group, get_ep_group
from vllm.forward_context import ForwardContext, get_forward_context
from vllm.model_executor.custom_op import CustomOp
from vllm.model_executor.custom_op import PluggableLayer
from vllm.model_executor.layers.fused_moe import FusedMoE
from vllm.model_executor.models.interfaces import MixtureOfExperts
from vllm.model_executor.models.utils import maybe_prefix
@@ -38,7 +38,7 @@ if TYPE_CHECKING:
# --8<-- [start:transformers_fused_moe]
@CustomOp.register("transformers_fused_moe")
@PluggableLayer.register("transformers_fused_moe")
class TransformersFusedMoE(FusedMoE):
"""Custom FusedMoE for the Transformers modeling backend."""