Replies: 1 comment 1 reply
-
|
That is the regular dehavior. Some layers are popped from the classifiers in order to use them as feature extractors. |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
Thank you very much for your answer. I understand now. |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Although this may make me appear ignorant, I truly hope that it will be possible to execute the yolov8_tracking project in Windows in the future. When I attempted to run the project in Windows, the result that popped up was:
The following layers are discarded due to unmatched keys or layer size: ['classifier.weight', 'classifier.bias']
Beta Was this translation helpful? Give feedback.
All reactions