Closed zhang320 closed 7 months ago
Hi @zhang320, you can train YOLOv8 as usual and then pass the new weights file to the launch file using the model
parameter.
If I want to modify the network structure, is this code not feasible
Does it mean that I have to operate on the source code
What do you have to change to make it work? This seems to be more related to Ultralytics. Have you checked issues like this https://github.com/ultralytics/ultralytics/issues/3194?
ok, thank you
Hey @zhang320, how is this going?
Is this version of code unable to modify network structure and train?