iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[XPU] allow forcing flash attn for mm_prefix (#48828)

Browse Source 
Signed-off-by: zhenwei-intel <zhenwei.liu@intel.com>
This commit is contained in:
liuzhenwei
2026-07-17 09:44:18 +00:00
committed by GitHub
parent 867ff69733
commit d5b1ec2684
1 changed files with 12 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/platforms/xpu.py
+12 -6
View File
@@ -150,8 +150,18 @@ class XPUPlatform(Platform):
return AttentionBackendEnum.TRITON_ATTN.get_path()
elif attn_selector_config.use_mm_prefix:
# Flash Attention on XPU has no FA4 kernel, so it cannot apply the
# multimodal prefix-LM bidirectional mask. Fall back to Triton
# Attention, which supports mm_prefix.
# multimodal prefix-LM bidirectional mask. Honor an explicit Flash
# Attention request (for text-only workloads); otherwise fall back
# to Triton Attention, which supports mm_prefix.
if selected_backend == AttentionBackendEnum.FLASH_ATTN:
logger.warning_once(
"Using Flash Attention on XPU for a multimodal prefix-LM "
"model because it was explicitly requested. The prefix-LM "
"bidirectional mask cannot be applied, so image/video "
"inputs will produce incorrect results; only use this for "
"text-only workloads."
)
return AttentionBackendEnum.FLASH_ATTN.get_path()
logger.warning_once(
"Flash Attention on XPU does not support multimodal prefix-LM "
"attention. Falling back to Triton Attention backend."
@@ -265,10 +275,6 @@ class XPUPlatform(Platform):
if compilation_config.compile_sizes is None:
compilation_config.compile_sizes = []
attention_config = vllm_config.attention_config
if attention_config.backend is None:
attention_config.backend = AttentionBackendEnum.FLASH_ATTN
# lazy import to avoid circular import
from vllm.utils.torch_utils import supports_xpu_graph