youngkingdom/vllm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[Attention] Move Backend enum into registry (#25893)

Browse Source 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
Matthew Bonanni
2025-10-02 23:32:24 -04:00
committed by GitHub
parent ad2d788016
commit 2aaa423842
31 changed files with 99 additions and 66 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/kernels/utils.py
View File

@ -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)