[Misc] Minor fix in KVCache type (#3652)

docs/source/models/adding_model.rst

@@ -56,8 +56,8 @@ Next, you need to rewrite the :code:`forward` methods of your model by following
     -    return_dict: Optional[bool] = None,
     -) -> Union[Tuple, CausalLMOutputWithPast]:
     +    positions: torch.Tensor,
-    +    kv_caches: List[KVCache],
-    +    input_metadata: InputMetadata,
+    +    kv_caches: List[torch.Tensor],
+    +    attn_metadata: AttentionMetadata,
     +) -> Optional[SamplerOutput]:
 
 1. Update the code by considering that :code:`input_ids` and :code:`positions` are now flattened tensors.

vllm/model_executor/models/llava.py

-from typing import List, Optional, Tuple
+from typing import List, Optional
 
 import torch
 from torch import nn
@@ -19,8 +19,6 @@ from vllm.model_executor.weight_utils import (default_weight_loader,
                                               hf_model_weights_iterator)
 from vllm.sequence import SamplerOutput
 
-KVCache = Tuple[torch.Tensor, torch.Tensor]
-
 _KEYS_TO_MODIFY_MAPPING = {
     "language_model.lm_head": "lm_head",
     "language_model.model": "language_model",
@@ -102,7 +100,7 @@ class LlavaForConditionalGeneration(nn.Module):
         self,
         input_ids: torch.Tensor,
         positions: torch.Tensor,
-        kv_caches: List[KVCache],
+        kv_caches: List[torch.Tensor],
         attn_metadata: AttentionMetadata,
         image_input: Optional[torch.Tensor] = None
     ) -> SamplerOutput:  # noqa: E501

vllm/worker/neuron_model_runner.py

@@ -14,8 +14,6 @@ from vllm.utils import (async_tensor_h2d, is_pin_memory_available,
 
 logger = init_logger(__name__)
 
-KVCache = Tuple[torch.Tensor, torch.Tensor]
-
 
 class NeuronModelRunner: