fix minor

Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>

Author: tdoublep

vllm/model_executor/layers/mamba/mamba_mixer2.py

@@ -33,7 +33,7 @@
     LoaderFunction, composed_weight_loader, sharded_weight_loader)
 from vllm.model_executor.models.mamba_cache import MambaCacheParams
 from vllm.model_executor.utils import set_weight_attrs
-from vllm.platform import current_platform
+from vllm.platforms import current_platform
 from vllm.utils import direct_register_custom_op
 from vllm.v1.attention.backends.mamba_attn import Mamba2AttentionMetadata
 
@@ -746,7 +746,7 @@ def mamba_mixer2(
     output: torch.Tensor,
     layer_name: str,
     mup_vector: Optional[torch.Tensor] = None,
-):
+) -> None:
     forward_context: ForwardContext = get_forward_context()
     self = forward_context.no_compile_layers[layer_name]
     self.forward_cuda(hidden_states=hidden_states,
@@ -761,7 +761,7 @@ def mamba_mixer2_fake(
     output: torch.Tensor,
     layer_name: str,
     mup_vector: Optional[torch.Tensor] = None,
-):
+) -> None:
     return