You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The weights_path argument allows you to specify initializing the model with pretrained weights. If you don't specify a weights_path argument, it will default to the weights in the file "config/yolov3.weights".
I have created a pull request here that allows you to provide the argument random_initialize to indicate training the model from scratch and not loading any pretrained weights.
Hey :)
Thank you very much for the scripts and the idea and smart solution! :)
For me the same question pop up...
unfortunatly, it seems that the pull request was not included... @cailinw would you mind to share your improvments here?
Thank you very much :)
hey there
thanks a lot for this code
I have a question please
do I need to include the yolov3 weight of a coco dataset during training ??
or what should I include in this line:
parser.add_argument("--weights_path", type=str, default="config/yolov3.weights", help="path to weights file")
The text was updated successfully, but these errors were encountered: