[BUG] [ShapeInferenceError] Incompatible dimensions

[stevenwudi]

Describe the bug

When I tried to convert the model of YOLOV5S, there is a bug:

sess = C.InferenceSession(session_options, self._model_bytes, False, self._read_config_from_model)
onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Node (Mul_959) Op (Mul) [ShapeInferenceError] Incompatible dimensions

onnx.version
'1.9.0'
onnxsim.version
'0.3.5'
torch.version
'1.8.1+cu111'

Model

The model is from pytorch model from the official yolov5 model:
https://github.com/ultralytics/yolov5/releases

Looking forward to hear back!

[stevenwudi]

This problem is raised in the issue in yolov5 and solve in PR# 2982
ultralytics/yolov5#2558
PR:
ultralytics/yolov5#2982

[daquexian]

Great! :)