[5/N][Attention] Finish eliminating `vllm/attention` folder (#32064)

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

vllm/model_executor/models/grok1.py

@@ -32,12 +32,12 @@ import torch
 import torch.nn.functional as F
 from torch import nn
 
-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
 from vllm.logger import init_logger
 from vllm.model_executor.layers.activation import GeluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/hunyuan_v1.py

@@ -33,7 +33,6 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-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 (
@@ -43,6 +42,7 @@ from vllm.distributed import (
     tensor_model_parallel_all_reduce,
 )
 from vllm.model_executor.layers.activation import SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import SharedFusedMoE
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/hunyuan_vision.py

@@ -39,7 +39,7 @@ from vllm.distributed import parallel_state
 from vllm.distributed import utils as dist_utils
 from vllm.logger import init_logger
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/idefics2_vision_model.py

@@ -29,7 +29,7 @@ from transformers.models.idefics2.configuration_idefics2 import (
 
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.conv import Conv2dLayer
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/intern_vit.py

@@ -23,7 +23,7 @@ from vllm.distributed import (
     tensor_model_parallel_all_gather,
 )
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.conv import Conv2dLayer
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/internlm2.py

@@ -10,7 +10,6 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -21,6 +20,7 @@ from vllm.distributed import (
     tensor_model_parallel_all_gather,
 )
 from vllm.model_executor.layers.activation import SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     MergedColumnParallelLinear,

vllm/model_executor/models/interns1_vit.py

@@ -15,7 +15,7 @@ from transformers import PretrainedConfig
 from transformers.utils import torch_int
 
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.conv import Conv2dLayer
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import ColumnParallelLinear, RowParallelLinear

vllm/model_executor/models/iquest_loopcoder.py

@@ -24,10 +24,10 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-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
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/isaac.py

@@ -20,7 +20,7 @@ from vllm.config import VllmConfig
 from vllm.config.model import ModelConfig
 from vllm.distributed import parallel_state
 from vllm.distributed import utils as dist_utils
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/jais.py

@@ -28,7 +28,6 @@ from itertools import islice
 import torch
 from torch import nn
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -36,6 +35,7 @@ from vllm.distributed import (
     get_tensor_model_parallel_rank,
     get_tensor_model_parallel_world_size,
 )
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/jais2.py

@@ -31,7 +31,6 @@ import torch
 from torch import nn
 from transformers import Jais2Config
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -39,6 +38,7 @@ from vllm.distributed import (
     get_tensor_model_parallel_world_size,
 )
 from vllm.model_executor.layers.activation import ReLUSquaredActivation
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/jamba.py

@@ -9,11 +9,11 @@ import torch
 from torch import nn
 from transformers import JambaConfig
 
-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_tensor_model_parallel_world_size
 from vllm.distributed.parallel_state import get_pp_group
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/keye.py

@@ -20,7 +20,7 @@ from vllm.config import VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.logger import init_logger
-from vllm.model_executor.layers.attention.mm_encoder_attention import (
+from vllm.model_executor.layers.attention import (
     MMEncoderAttention,
 )
 from vllm.model_executor.layers.conv import Conv2dLayer

vllm/model_executor/models/lfm2.py

@@ -7,11 +7,11 @@ import torch
 import torch.nn as nn
 from transformers import Lfm2Config
 
-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
 from vllm.model_executor.layers.activation import SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     MergedColumnParallelLinear,

vllm/model_executor/models/lfm2_moe.py

@@ -6,7 +6,6 @@ from itertools import islice
 import torch
 import torch.nn as nn
 
-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 (
@@ -15,6 +14,7 @@ from vllm.distributed import (
     get_tensor_model_parallel_world_size,
 )
 from vllm.model_executor.layers.activation import SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/lfm2_siglip2.py

@@ -13,7 +13,7 @@ from transformers import Siglip2VisionConfig
 from vllm.compilation.decorators import support_torch_compile
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/llama.py

@@ -31,12 +31,12 @@ import torch
 from torch import nn
 from transformers import LlamaConfig
 
-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
 from vllm.model_executor.layers.activation import SiluAndMul
-from vllm.model_executor.layers.attention.encoder_only_attention import (
+from vllm.model_executor.layers.attention import (
+    Attention,
     EncoderOnlyAttention,
 )
 from vllm.model_executor.layers.layernorm import RMSNorm

vllm/model_executor/models/llama4.py

@@ -24,7 +24,6 @@ import torch
 from torch import nn
 from transformers import Llama4TextConfig
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -33,7 +32,8 @@ from vllm.distributed import (
     tensor_model_parallel_all_gather,
 )
 from vllm.logger import init_logger
-from vllm.model_executor.layers.attention.chunked_local_attention import (
+from vllm.model_executor.layers.attention import (
+    Attention,
     ChunkedLocalAttention,
 )
 from vllm.model_executor.layers.fused_moe import SharedFusedMoE

vllm/model_executor/models/mimo_v2_flash.py

@@ -6,7 +6,6 @@ from itertools import islice
 import torch
 from torch import nn
 
-from vllm.attention.layer import Attention
 from vllm.config import (
     CacheConfig,
     VllmConfig,
@@ -22,6 +21,7 @@ from vllm.distributed import (
 )
 from vllm.logger import init_logger
 from vllm.model_executor.layers.activation import SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (

vllm/model_executor/models/minicpm.py

@@ -33,7 +33,6 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -43,6 +42,7 @@ from vllm.distributed import (
     tensor_model_parallel_all_reduce,
 )
 from vllm.model_executor.layers.activation import FatreluAndMul, SiluAndMul
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import fused_experts, fused_topk
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (