Hi there,
I don’t know if anyone else already pointed this out.
In Alpaca_model there seems to be small error in the commented out cell notes. The second note after ### START CODE HERE says
“# create the input layer (Same as the imageNetv2 input size)”
That should be MobileNetv2, shouldn’t it?
Thanks for a great course, I’m really learning a lot.
Thanks, but actually I think this is not a bug. As they point out at several points in the text of the notebook, MobilNet is trained on the ImageNet dataset.
But how about this:
In the assertions for the data_augmenter function, they misspell RandomFlip at one point:
assert(augmenter.layers[0].name.startswith('random_flip')), "First layer must be RandomFlip"
assert augmenter.layers[0].mode == 'horizontal', "RadomFlip parameter must be horizontal"
Inconceivable!!!