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

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

vllm/model_executor/models/bloom.py

@@ -27,7 +27,7 @@ import torch
 from torch import nn
 from transformers import BloomConfig
 
-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/chameleon.py

@@ -16,7 +16,7 @@ from transformers import (
     ChameleonVQVAEConfig,
 )
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.config import CacheConfig, VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/chatglm.py

@@ -12,7 +12,7 @@ import torch
 from torch import nn
 from torch.nn import LayerNorm
 
-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/clip.py

@@ -14,8 +14,7 @@ from transformers import (
     CLIPVisionConfig,
 )
 
-from vllm.attention import Attention
-from vllm.attention.layer import MultiHeadAttention
+from vllm.attention.layer import Attention, MultiHeadAttention
 from vllm.config import VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
 from vllm.distributed import divide, get_tensor_model_parallel_world_size

vllm/model_executor/models/commandr.py

@@ -30,7 +30,7 @@ import torch
 from torch import nn
 from transformers import Cohere2Config, CohereConfig
 
-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/dbrx.py

@@ -8,7 +8,7 @@ import torch
 import torch.nn as nn
 from transformers import DbrxConfig
 
-from vllm.attention import Attention
+from vllm.attention.layer import Attention
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
     get_pp_group,

vllm/model_executor/models/deepseek_v2.py

@@ -33,8 +33,8 @@ from torch import nn
 from transformers import DeepseekV2Config, DeepseekV3Config
 
 from vllm._aiter_ops import rocm_aiter_ops
-from vllm.attention import Attention
 from vllm.attention.backends.abstract import AttentionBackend
+from vllm.attention.layer import Attention
 from vllm.attention.ops.common import pack_seq_triton, unpack_seq_triton
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ParallelConfig, VllmConfig, get_current_vllm_config

vllm/model_executor/models/dots1.py

@@ -32,7 +32,7 @@ import torch
 from torch import nn
 from transformers import Dots1Config
 
-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/ernie45_moe.py

@@ -32,7 +32,7 @@ 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, get_current_vllm_config
 from vllm.distributed import (

vllm/model_executor/models/ernie45_vl_moe.py

@@ -31,7 +31,7 @@ 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

vllm/model_executor/models/exaone.py

@@ -32,7 +32,7 @@ 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 get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/exaone4.py

@@ -28,7 +28,7 @@ import torch
 from torch import nn
 from transformers import Exaone4Config
 
-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/falcon.py

@@ -30,7 +30,7 @@ from torch import nn
 from torch.nn import LayerNorm
 from transformers import FalconConfig as HF_FalconConfig
 
-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/gemma.py

@@ -26,7 +26,7 @@ import torch
 from torch import nn
 from transformers import GemmaConfig
 
-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/gemma2.py

@@ -23,7 +23,7 @@ import torch
 from torch import nn
 from transformers import Gemma2Config
 
-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/gemma3.py

@@ -23,7 +23,8 @@ import torch.nn.functional as F
 from torch import nn
 from transformers import Gemma3TextConfig
 
-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 get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/gemma3n.py

@@ -21,7 +21,7 @@ import torch
 from torch import nn
 from transformers.models.gemma3n.configuration_gemma3n import Gemma3nTextConfig
 
-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_tensor_model_parallel_world_size

vllm/model_executor/models/glm4.py

@@ -29,7 +29,8 @@ import torch
 from torch import nn
 from transformers import Glm4Config
 
-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 get_pp_group, get_tensor_model_parallel_world_size

vllm/model_executor/models/glm4_moe.py

@@ -31,7 +31,7 @@ import torch
 from torch import nn
 from transformers.models.glm4_moe import Glm4MoeConfig
 
-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/gpt2.py

@@ -27,7 +27,7 @@ import torch
 from torch import nn
 from transformers import GPT2Config
 
-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.parallel_state import (