试了这个openvino，报了错，不知道怎么解决 https://github.com/zhuo-yoyowz/openvino_notebooks/tree/main/notebooks/405-paddle-ocr-webcam

[nissansz]

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem

• 系统环境/System Environment：
• 版本号/Version：Paddle： PaddleOCR：2.7.5 问题相关组件/Related components：
• 运行指令/Command Code：
• 完整报错/Complete Error Message：

openvino_notebooks/notebooks/405-paddle-ocr-webcam at main · zhuo-yoyowz/openvino_notebooks (github.com)

Exception from src\inference\src\cpp\infer_request.cpp:112: Exception from src\inference\src\cpp\infer_request.cpp:66: Exception from src\plugins\intel_cpu\src\infer_request.cpp:390: Can't set the input tensor with name: x, because the model input (shape=[?,3,?,?]) and the tensor (shape=(3.48.110)) are incompatible

[Sunting78]

推荐用FD 您使用的repo不是我们官方维护的哈

[Liyulingyue]

Can't set the input tensor with name: x, because the model input (shape=[?,3,?,?]) and the tensor (shape=(3.48.110)) are incompatible 应该是你的输入结构和模型推理结构不一致导致的。

[UserWangZz]

Can't set the input tensor with name: x, because the model input (shape=[?,3,?,?]) and the tensor (shape=(3.48.110)) are incompatible 应该是你的输入结构和模型推理结构不一致导致的。 感谢讨论

[nissansz]

onnx推理有什么办法加速吗？速度好像是paddle gpu inference的1/3