remove get_metadata_cls (#4087)

remove get_metadata_cls. It's only used for V0 engine and has been removed from vLLM already.

- vLLM version: v0.11.0
- vLLM main:
83f478bb19

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

vllm_ascend/attention/attention_v1.py

@@ -75,10 +75,6 @@ class AscendAttentionBackend(AttentionBackend):
     def get_impl_cls() -> Type["AscendAttentionBackendImpl"]:
         return AscendAttentionBackendImpl
 
-    @staticmethod
-    def get_metadata_cls() -> Type["AscendMetadata"]:
-        return AscendMetadata
-
     @staticmethod
     def get_builder_cls() -> type["AscendAttentionMetadataBuilder"]:
         return AscendAttentionMetadataBuilder

vllm_ascend/attention/mla_v1.py

@@ -7,9 +7,7 @@ import torch
 import torch.distributed as dist
 import torch_npu
 from torch import nn
-from vllm.attention.backends.abstract import (AttentionBackend,
-                                              AttentionMetadata,
-                                              MLAAttentionImpl)
+from vllm.attention.backends.abstract import AttentionBackend, MLAAttentionImpl
 from vllm.config import VllmConfig, get_current_vllm_config
 from vllm.distributed import (get_dcp_group,
                               get_decode_context_model_parallel_rank,
@@ -69,10 +67,6 @@ class AscendMLABackend(AttentionBackend):
     def get_name() -> str:
         return "ASCEND_MLA"
 
-    @staticmethod
-    def get_metadata_cls() -> type["AttentionMetadata"]:
-        return AscendMLAMetadata
-
     @staticmethod
     def get_builder_cls():
         return AscendMLAMetadataBuilder

vllm_ascend/attention/sfa_v1.py

@@ -4,9 +4,7 @@ from typing import TYPE_CHECKING, ClassVar, Optional, Tuple, Type, TypeVar
 import torch
 import torch_npu
 from torch import nn
-from vllm.attention.backends.abstract import (AttentionBackend,
-                                              AttentionMetadata,
-                                              MLAAttentionImpl)
+from vllm.attention.backends.abstract import AttentionBackend, MLAAttentionImpl
 from vllm.config import VllmConfig
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.linear import (LinearBase,
@@ -35,10 +33,6 @@ class AscendSFABackend(AttentionBackend):
     def get_name() -> str:
         return "ASCEND_SFA"
 
-    @staticmethod
-    def get_metadata_cls() -> type["AttentionMetadata"]:
-        return AscendSFAMetadata
-
     @staticmethod
     def get_builder_cls():
         return AscendSFAMetadataBuilder

vllm_ascend/torchair/torchair_attention.py

@@ -55,10 +55,6 @@ class AscendAttentionTorchairBackend(AscendAttentionBackend):
     def get_impl_cls() -> Type["AscendAttentionTorchairBackendImpl"]:
         return AscendAttentionTorchairBackendImpl
 
-    @staticmethod
-    def get_metadata_cls() -> Type["AscendTorchairMetadata"]:
-        return AscendTorchairMetadata
-
     @staticmethod
     def get_builder_cls() -> type["AscendAttentionTorchairMetadataBuilder"]:
         return AscendAttentionTorchairMetadataBuilder

vllm_ascend/torchair/torchair_mla.py

@@ -6,7 +6,6 @@ import torch
 import torch.nn as nn
 import torch_npu
 from vllm.attention.backends.abstract import (AttentionBackend, AttentionLayer,
-                                              AttentionMetadata,
                                               MLAAttentionImpl)
 from vllm.attention.backends.utils import PAD_SLOT_ID
 from vllm.config import VllmConfig, get_current_vllm_config
@@ -43,10 +42,6 @@ class AscendMLATorchairBackend(AttentionBackend):
     def get_name() -> str:
         return "ASCEND_MLA_TORCHAIR"
 
-    @staticmethod
-    def get_metadata_cls() -> type["AttentionMetadata"]:
-        return AscendMLATorchairMetadata
-
     @staticmethod
     def get_builder_cls():
         return AscendMLATorchairMetadataBuilder

vllm_ascend/torchair/torchair_sfa.py

@@ -6,9 +6,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import torch_npu
-from vllm.attention.backends.abstract import (AttentionBackend,
-                                              AttentionMetadata,
-                                              MLAAttentionImpl)
+from vllm.attention.backends.abstract import AttentionBackend, MLAAttentionImpl
 from vllm.attention.backends.utils import PAD_SLOT_ID
 from vllm.config import VllmConfig, get_current_vllm_config
 from vllm.distributed import get_tensor_model_parallel_world_size, get_tp_group
@@ -43,10 +41,6 @@ class AscendSFATorchairBackend(AttentionBackend):
     def get_name() -> str:
         return "ASCEND_SFA_TORCHAIR"
 
-    @staticmethod
-    def get_metadata_cls() -> type["AttentionMetadata"]:
-        return AscendSFATorchairMetadata
-
     @staticmethod
     def get_builder_cls():
         return AscendSFATorchairMetadataBuilder