Skip to content

vllm_omni.diffusion.attention.backends.utils.fa ¶

HAS_FLASH_ATTN `module-attribute` ¶

HAS_FLASH_ATTN = (
    flash_attn_func is not None
    or flash_attn_varlen_func is not None
)

logger `module-attribute` ¶

logger = init_logger(__name__)

is_flash_attn_installed `cached` ¶

is_flash_attn_installed() -> bool

Return whether a Flash Attention backend package is importable.

Shared by CUDA/ROCm/MUSA platforms.