[bugfix] fix omni norm_bbox

docs/source/Instruction/命令行参数.md

@@ -694,7 +694,7 @@ App参数继承于[部署参数](#部署参数), [Web-UI参数](#Web-UI参数)
 以下参数的含义可以在对应模型官方repo或者其推理代码中找到相应含义。
 
 ### qwen2_vl, qvq, qwen2_5_vl, mimo_vl, keye_vl, keye_vl_1_5
-参数含义同`qwen_vl_utils<0.0.13`或者`qwen_omni_utils`库，可以查看[这里](https://github.com/QwenLM/Qwen2.5-VL/blob/main/qwen-vl-utils/src/qwen_vl_utils/vision_process.py#L24)。
+参数含义同`qwen_vl_utils<0.0.12`或者`qwen_omni_utils`库，可以查看[这里](https://github.com/QwenLM/Qwen2.5-VL/blob/main/qwen-vl-utils/src/qwen_vl_utils/vision_process.py#L24)。
 
 - IMAGE_FACTOR: 默认为28。
 - MIN_PIXELS: 默认为`4 * 28 * 28`。

docs/source_en/Instruction/Command-line-parameters.md

@@ -712,7 +712,7 @@ Specific model arguments can be set using `--model_kwargs` or environment variab
 The definitions of the parameters listed below can be found in each model’s official repository or in its inference code.
 
 ### qwen2_vl, qvq, qwen2_5_vl, mimo_vl, keye_vl, keye_vl_1_5
-The parameter meanings are the same as in the `qwen_vl_utils<0.0.13` or `qwen_omni_utils` library. You can refer to [here](https://github.com/QwenLM/Qwen2.5-VL/blob/main/qwen-vl-utils/src/qwen_vl_utils/vision_process.py#L24)
+The parameter meanings are the same as in the `qwen_vl_utils<0.0.12` or `qwen_omni_utils` library. You can refer to [here](https://github.com/QwenLM/Qwen2.5-VL/blob/main/qwen-vl-utils/src/qwen_vl_utils/vision_process.py#L24)
 
 - IMAGE_FACTOR: Default is 28
 - MIN_PIXELS: Default is `4 * 28 * 28`

swift/llm/template/template/qwen.py

@@ -790,6 +790,7 @@ def generate(self, model, *args, **kwargs):
 
 class Qwen3OmniTemplate(Qwen2_5OmniTemplate, ThinkingTemplate):
     version = 'omni_v3'
+    norm_bbox = 'norm1000'
     placeholder_tokens = ['<|image_pad|>', '<|audio_pad|>', '<|video_pad|>']
 
     def _post_encode(self, model, inputs: Dict[str, Any]) -> Dict[str, Any]: