[Attention] Remove imports from `vllm/attention/__init__.py` (#29342)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

vllm/model_executor/models/gpt_bigcode.py

@@ -28,7 +28,7 @@ import torch
 from torch import nn
 from transformers import GPTBigCodeConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/gpt_j.py

@@ -26,7 +26,7 @@ import torch
 from torch import nn
 from transformers import GPTJConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/gpt_neox.py

@@ -26,7 +26,7 @@ import torch
 from torch import nn
 from transformers import GPTNeoXConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/gpt_oss.py

@@ -7,7 +7,8 @@ import torch.distributed as dist
 from torch import nn
 from transformers import GptOssConfig
 
-from vllm.attention import Attention, AttentionType
+from vllm.attention.backends.abstract import AttentionType
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/granite.py

@@ -31,7 +31,7 @@ import torch
 from torch import nn
 from transformers import GraniteConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/granitemoe.py

@@ -31,7 +31,7 @@ from typing import Any
 import torch
 from torch import nn
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/grok1.py

@@ -31,7 +31,7 @@ import torch
 import torch.nn.functional as F
 from torch import nn
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/hunyuan_v1.py

@@ -33,7 +33,8 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention import Attention, AttentionType
+from vllm.attention.backends.abstract import AttentionType
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig, get_current_vllm_config
 from vllm.distributed import (

vllm/model_executor/models/internlm2.py

@@ -10,7 +10,7 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/jais.py

@@ -28,7 +28,7 @@ from itertools import islice
 import torch
 from torch import nn
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/lfm2.py

@@ -7,7 +7,7 @@ import torch
 import torch.nn as nn
 from transformers import Lfm2Config
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ModelConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/lfm2_moe.py

@@ -6,7 +6,7 @@ from itertools import islice
 import torch
 import torch.nn as nn
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ModelConfig, VllmConfig, get_current_vllm_config
 from vllm.distributed import (

vllm/model_executor/models/llama.py

@@ -31,7 +31,8 @@ import torch
 from torch import nn
 from transformers import LlamaConfig
 
-from vllm.attention import Attention, AttentionType
+from vllm.attention.backends.abstract import AttentionType
+from vllm.attention.layer import Attention
 from vllm.attention.layers.encoder_only_attention import EncoderOnlyAttention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig

vllm/model_executor/models/llama4.py

@@ -24,7 +24,7 @@ import torch
 from torch import nn
 from transformers import Llama4TextConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.attention.layers.chunked_local_attention import ChunkedLocalAttention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig

vllm/model_executor/models/minicpm.py

@@ -33,7 +33,7 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/minicpm3.py

@@ -29,7 +29,7 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.layernorm import RMSNorm

vllm/model_executor/models/minimax_m2.py

@@ -30,7 +30,7 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ModelConfig, VllmConfig
 from vllm.distributed import (

vllm/model_executor/models/minimax_text_01.py

@@ -14,7 +14,8 @@ import torch
 from torch import nn
 from transformers import MiniMaxConfig
 
-from vllm.attention import Attention, AttentionMetadata
+from vllm.attention.backends.abstract import AttentionMetadata
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ModelConfig, VllmConfig
 from vllm.distributed.parallel_state import (

vllm/model_executor/models/mixtral.py

@@ -32,7 +32,7 @@ import torch
 from torch import nn
 from transformers import MixtralConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig, get_current_vllm_config
 from vllm.distributed import (

vllm/model_executor/models/molmo.py

@@ -17,8 +17,7 @@ from transformers import BatchFeature, PretrainedConfig, ProcessorMixin, TensorT
 from transformers.image_utils import ImageInput
 from transformers.tokenization_utils_base import TextInput
 
-from vllm.attention import Attention
-from vllm.attention.layer import MultiHeadAttention
+from vllm.attention.layer import Attention, MultiHeadAttention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.config.multimodal import BaseDummyOptions