I labeled the frame correctly, but now When I train my dataset, I face the message(TypeError: can’t set attributes of built-in/extension type ‘object’) in constructor.py and I confirmed that the training dataset was generated correctly.
What Should I do to corrct this problem?
Dear Walter, could you please post a more complete error message? Alexander
deeplabcut 2.0.4 / window10 / tensorflow1.12.0
To train the network, I set the configure path and labeled the marker on the selected frames through the GUI. After that, through deeplabcut.check_labels(
config_path') , I confirmed that the markered frames were generated. And throught the command deeplabcut.create_training_dataset(config_path’,num_shuffles=1) I created the training dataset, finally I conducted train_network function(through deeplabcut.train_network(`config_path’,shuffle=1)).
But, the last command’s result message was printed with very long message (see the uploaded image).
And the end of message was as I mentioned (TypeError: can’t set attributes of built-in/extension type ‘object’). I thought the error was occured due to the setattr function of constructor.py.
Why the function setattr connot set the input(maybe I think that the input is marker)?
Am I wrong with the setting of configure path?