From 6f4bd3e45cb8a349879959c4a73bffb8eeb7f1db Mon Sep 17 00:00:00 2001 From: NumberWan Date: Sun, 31 May 2026 08:06:18 +0800 Subject: [PATCH] [BugFix]Qwen-Image performance regression by using omni RMSNorm(RMSNorm backend) (#3933) Signed-off-by: NumberWan --- vllm_omni/diffusion/models/qwen_image/qwen_image_transformer.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vllm_omni/diffusion/models/qwen_image/qwen_image_transformer.py b/vllm_omni/diffusion/models/qwen_image/qwen_image_transformer.py index 5e4d526d8..abdca8b1a 100644 --- a/vllm_omni/diffusion/models/qwen_image/qwen_image_transformer.py +++ b/vllm_omni/diffusion/models/qwen_image/qwen_image_transformer.py @@ -17,7 +17,6 @@ from diffusers.models.embeddings import TimestepEmbedding, Timesteps from diffusers.models.modeling_outputs import Transformer2DModelOutput from diffusers.models.normalization import AdaLayerNormContinuous from vllm.logger import init_logger -from vllm.model_executor.layers.layernorm import RMSNorm from vllm.model_executor.layers.linear import ( ColumnParallelLinear, QKVParallelLinear, @@ -44,6 +43,7 @@ from vllm_omni.diffusion.distributed.sp_plan import ( ) from vllm_omni.diffusion.forward_context import get_forward_context from vllm_omni.diffusion.layers.adalayernorm import AdaLayerNorm +from vllm_omni.diffusion.layers.norm import RMSNorm from vllm_omni.diffusion.layers.rope import RotaryEmbedding logger = init_logger(__name__)