Closed HectorAnadon closed 6 years ago
It is fixed with new release v0.4. Thank you for pointing out.
Still have this issue with keras 2.0.6
Hello Taneta, in vgg16.py model line 102, edit include_top=include_top to require_flatten=include_top. The code is below input_shape = _obtain_input_shape(input_shape, default_size=224, min_size=48, data_format=K.image_data_format(), require_flatten=include_top)
this will work for keras 2.2.4
I am using keras 2.0.8. I got that error because apparently include_top is not used anymore. Instead require_flatten is used.
In order to fix it change line 84 to