From c6b526bc2e57f5dd351face06f6ca7e578872965 Mon Sep 17 00:00:00 2001
From: Chenguang Zheng <645327136@qq.com>
Date: Mon, 9 Feb 2026 11:46:00 +0800
Subject: [PATCH] fix: add diffusion offload args to OmniConfig group instead
 of serve_parser (#1271)

Signed-off-by: Chenguang ZHENG <645327136@qq.com>
---
 vllm_omni/entrypoints/cli/serve.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/vllm_omni/entrypoints/cli/serve.py b/vllm_omni/entrypoints/cli/serve.py
index ac03de143..1be3b02f1 100644
--- a/vllm_omni/entrypoints/cli/serve.py
+++ b/vllm_omni/entrypoints/cli/serve.py
@@ -194,17 +194,17 @@ class OmniServeCommand(CLISubcommand):
         )
 
         # diffusion model offload parameters
-        serve_parser.add_argument(
+        omni_config_group.add_argument(
             "--enable-cpu-offload",
             action="store_true",
             help="Enable CPU offloading for diffusion models.",
         )
-        serve_parser.add_argument(
+        omni_config_group.add_argument(
             "--enable-layerwise-offload",
             action="store_true",
             help="Enable layerwise (blockwise) offloading on DiT modules.",
         )
-        serve_parser.add_argument(
+        omni_config_group.add_argument(
             "--layerwise-num-gpu-layers",
             type=int,
             default=1,