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

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

vllm/model_executor/models/minicpm3.py

@@ -29,9 +29,9 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-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.attention import Attention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/minimax_m2.py

@@ -30,7 +30,6 @@ 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, ModelConfig, VllmConfig
 from vllm.distributed import (
@@ -38,6 +37,7 @@ from vllm.distributed import (
     get_tensor_model_parallel_world_size,
     tensor_model_parallel_all_reduce,
 )
+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/minimax_text_01.py

@@ -14,7 +14,6 @@ import torch
 from torch import nn
 from transformers import MiniMaxConfig
 
-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 (
@@ -24,6 +23,7 @@ from vllm.distributed.parallel_state import (
 )
 from vllm.forward_context import get_forward_context
 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/mixtral.py

@@ -32,7 +32,6 @@ import torch
 from torch import nn
 from transformers import MixtralConfig
 
-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 (
@@ -40,6 +39,7 @@ from vllm.distributed import (
     get_pp_group,
     get_tensor_model_parallel_world_size,
 )
+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/mllama4.py

@@ -36,7 +36,7 @@ from vllm.config import VllmConfig, set_current_vllm_config
 from vllm.config.multimodal import BaseDummyOptions
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.forward_context import set_forward_context
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import MMEncoderAttention
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/modernbert.py

@@ -10,7 +10,7 @@ from transformers.activations import ACT2FN
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import ModelConfig, VllmConfig
 from vllm.distributed import get_tensor_model_parallel_world_size
-from vllm.model_executor.layers.attention.encoder_only_attention import (
+from vllm.model_executor.layers.attention import (
     EncoderOnlyAttention,
 )
 from vllm.model_executor.layers.linear import QKVParallelLinear, RowParallelLinear

vllm/model_executor/models/molmo.py

@@ -17,7 +17,6 @@ from transformers import BatchFeature, PretrainedConfig, ProcessorMixin, TensorT
 from transformers.image_utils import ImageInput
 from transformers.tokenization_utils_base import TextInput
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
@@ -29,7 +28,7 @@ from vllm.distributed import (
     tensor_model_parallel_all_gather,
 )
 from vllm.model_executor.layers.activation import MulAndSilu, QuickGELU, SiluAndMul
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import Attention, MMEncoderAttention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/molmo2.py

@@ -23,7 +23,6 @@ from transformers.image_utils import ImageInput
 from transformers.tokenization_utils_base import TextInput
 from transformers.video_utils import VideoInput, VideoMetadata
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.config.multimodal import BaseDummyOptions, VideoDummyOptions
@@ -36,7 +35,7 @@ from vllm.distributed import (
 )
 from vllm.logger import init_logger
 from vllm.model_executor.layers.activation import MulAndSilu, SiluAndMul, get_act_fn
-from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention
+from vllm.model_executor.layers.attention import Attention, MMEncoderAttention
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,

vllm/model_executor/models/moonvit.py

@@ -53,7 +53,7 @@ from transformers.activations import ACT2FN
 from transformers.modeling_utils import PreTrainedModel
 
 from vllm.distributed import divide, get_tensor_model_parallel_world_size
-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/mpt.py

@@ -10,7 +10,6 @@ import torch
 import torch.nn as nn
 from transformers import MptConfig
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import (
@@ -19,6 +18,7 @@ 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 import Attention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/nemotron.py

@@ -30,11 +30,11 @@ 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 get_pp_group, get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.activation import get_act_fn
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/nemotron_h.py

@@ -25,7 +25,6 @@ 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, ModelConfig, VllmConfig
 from vllm.config.parallel import ParallelConfig
@@ -33,6 +32,7 @@ from vllm.distributed import get_ep_group, get_tensor_model_parallel_world_size
 from vllm.distributed.communication_op import tensor_model_parallel_all_gather
 from vllm.distributed.parallel_state import get_pp_group
 from vllm.model_executor.layers.activation import ReLUSquaredActivation
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.fused_moe import FusedMoE, SharedFusedMoE
 from vllm.model_executor.layers.fused_moe.utils import activation_without_mul
 from vllm.model_executor.layers.layernorm import RMSNorm

vllm/model_executor/models/olmo.py

@@ -31,11 +31,11 @@ import torch
 from torch import nn
 from transformers import OlmoConfig
 
-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 import Attention
 from vllm.model_executor.layers.linear import (
     MergedColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/olmo2.py

@@ -32,7 +32,6 @@ import torch
 from torch import nn
 from transformers import Olmo2Config
 
-from vllm.attention.layer import Attention
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
@@ -40,6 +39,7 @@ from vllm.distributed.communication_op import tensor_model_parallel_all_gather
 from vllm.distributed.parallel_state import get_tensor_model_parallel_rank
 from vllm.distributed.utils import split_tensor_along_last_dim
 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/olmoe.py

@@ -21,7 +21,6 @@ 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 VllmConfig
 from vllm.distributed import (
@@ -32,6 +31,7 @@ from vllm.distributed import (
 )
 from vllm.distributed.utils import split_tensor_along_last_dim
 from vllm.logger import init_logger
+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/openpangu.py

@@ -29,7 +29,6 @@ import torch
 from torch import nn
 from transformers import PretrainedConfig
 
-from vllm.attention.layer import Attention, AttentionType
 from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ParallelConfig, VllmConfig
 from vllm.distributed import (
@@ -41,7 +40,8 @@ from vllm.distributed import (
     tensor_model_parallel_all_gather,
 )
 from vllm.model_executor.layers.activation import SiluAndMul
-from vllm.model_executor.layers.attention.static_sink_attention import (
+from vllm.model_executor.layers.attention import (
+    Attention,
     StaticSinkAttention,
 )
 from vllm.model_executor.layers.fused_moe import SharedFusedMoE
@@ -84,6 +84,7 @@ from vllm.model_executor.utils import set_weight_attrs
 from vllm.platforms import current_platform
 from vllm.sequence import IntermediateTensors
 from vllm.transformers_utils.config import set_default_rope_theta
+from vllm.v1.attention.backend import AttentionType
 from vllm.v1.attention.backends.flash_attn_diffkv import FlashAttentionDiffKVBackend
 
 

vllm/model_executor/models/opt.py

@@ -27,11 +27,11 @@ import torch
 from torch import nn
 from transformers import OPTConfig
 
-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 get_act_fn
+from vllm.model_executor.layers.attention import Attention
 from vllm.model_executor.layers.linear import (
     ColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/orion.py

@@ -15,11 +15,11 @@ 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_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.linear import (
     MergedColumnParallelLinear,
     QKVParallelLinear,

vllm/model_executor/models/ouro.py

@@ -33,11 +33,11 @@ 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.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/paddleocr_vl.py

@@ -34,7 +34,7 @@ from vllm.config import VllmConfig
 from vllm.config.multimodal import BaseDummyOptions
 from vllm.distributed import parallel_state
 from vllm.distributed import utils as dist_utils
-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