[Hardware][Ascend] Integrate MindIE dense and sparse quantized attention for Wan2.2 T2V (#7561)

Signed-off-by: zqxu <xuzhenqiang3@huawei.com>
Co-authored-by: zqxu <xuzhenqiang3@huawei.com>
This commit is contained in:
Zhenqiang Xuandzqxu authored and GitHub committed 2026-09-18 01:28:27 +00:00
1 parent f8197e120f
commit 242c3cddeb
15 files changed
+853 -368

No files matched your search

+2 -1
View File
@@ -941,7 +941,8 @@ class OmniServeCommand(CLISubcommand):
"--diffusion-kv-cache-dtype",
type=str,
default=None,
help="Diffusion attention KV cache dtype (e.g. fp8). Separate from vLLM --kv-cache-dtype.",
help="Diffusion Q/K/V precision: fp8, mxfp8, mxfp4, or float (NPU). "
"Separate from vLLM --kv-cache-dtype. Use --diffusion-attention-config for per-role fallback.",
)
omni_config_group.add_argument(
"--diffusion-kv-cache-skip-steps",