[Attention] Move Backend enum into registry (#25893)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
@@ -10,8 +10,9 @@ from unittest.mock import patch
|
||||
import pytest
|
||||
import torch
|
||||
|
||||
from vllm.attention.backends.registry import _Backend
|
||||
from vllm.attention.layer import MultiHeadAttention
|
||||
from vllm.attention.selector import _Backend, _cached_get_attn_backend
|
||||
from vllm.attention.selector import _cached_get_attn_backend
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.platforms.cpu import CpuPlatform
|
||||
from vllm.platforms.cuda import CudaPlatform
|
||||
|
||||
@@ -15,10 +15,10 @@ from torch._prims_common import TensorLikeType
|
||||
|
||||
from tests.kernels.quant_utils import native_w8a8_block_matmul
|
||||
from vllm.attention import AttentionBackend, AttentionMetadata, AttentionType
|
||||
from vllm.attention.backends.registry import _Backend
|
||||
from vllm.model_executor.layers.activation import SiluAndMul
|
||||
from vllm.model_executor.layers.fused_moe.utils import (
|
||||
moe_kernel_quantize_input)
|
||||
from vllm.platforms.interface import _Backend
|
||||
from vllm.utils import (STR_BACKEND_ENV_VAR, STR_FLASH_ATTN_VAL,
|
||||
STR_XFORMERS_ATTN_VAL, make_tensor_with_pad)
|
||||
|
||||
|
||||
Reference in New Issue
Block a user