Closed lizexu123 closed 7 months ago
Hey, as of right now, this implementation does not support models with dynamic width and heights, only dynamic batch:
If only dynamic batch size is needed, then how to set the yolov8 export ? I use the following code to export, but, height and width are also dynamic.
model = YOLO("../runs/segment/train70/weights/best.pt") model.fuse() model.info(verbose=False) model.export(format="onnx",opset=12,dynamic=True)
I borrowed your code and used it in the paddle framework, but it has not been successful yet. Can you help me take a look?
Hi @Lishumuzixin Please look at v4.1. I have added support for fixed batch size > 1.
Hey, as of right now, this implementation does not support models with dynamic width and heights, only dynamic batch:![image](https://github.com/cyrusbehr/tensorrt-cpp-api/assets/17056751/d307c84c-120f-4262-b092-b60817f8af1f)