[Misc] Clean up type annotation for SupportsMultiModal (#14794)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/model_executor/models/fuyu.py

@@ -18,7 +18,7 @@
 """ PyTorch Fuyu model."""
 import math
 from collections.abc import Iterable, Mapping, Sequence
-from typing import List, Literal, Optional, Set, Tuple, TypedDict, Union
+from typing import List, Literal, Optional, Set, Tuple, TypedDict
 
 import torch
 import torch.nn as nn
@@ -41,7 +41,7 @@ from vllm.multimodal.processing import (BaseMultiModalProcessor,
 from vllm.multimodal.profiling import BaseDummyInputsBuilder, ProcessorInputs
 from vllm.sequence import IntermediateTensors
 
-from .interfaces import SupportsMultiModal, SupportsPP
+from .interfaces import MultiModalEmbeddings, SupportsMultiModal, SupportsPP
 from .utils import (AutoWeightsLoader, flatten_bn, maybe_prefix,
                     merge_multimodal_embeddings)
 
@@ -327,8 +327,7 @@ class FuyuForCausalLM(nn.Module, SupportsMultiModal, SupportsPP):
         return vision_embeddings_flat.split(patches_per_image, dim=0)
 
     def get_multimodal_embeddings(
-        self, **kwargs
-    ) -> Union[list[torch.Tensor], torch.Tensor, tuple[torch.Tensor, ...]]:
+            self, **kwargs: object) -> Optional[MultiModalEmbeddings]:
         image_input = self._parse_and_validate_image_input(**kwargs)
         if image_input is None:
             return None
@@ -338,7 +337,7 @@ class FuyuForCausalLM(nn.Module, SupportsMultiModal, SupportsPP):
     def get_input_embeddings(
         self,
         input_ids: torch.Tensor,
-        multimodal_embeddings: Optional[NestedTensors] = None,
+        multimodal_embeddings: Optional[MultiModalEmbeddings] = None,
     ) -> torch.Tensor:
         inputs_embeds = self.language_model.get_input_embeddings(input_ids)
         if multimodal_embeddings is not None: