C4W1 assignment problem

Course Q&A TensorFlow: Advanced Techniques Specialization Generative Deep Learning with TensorFlow
1

Hello, in the cell number 20, i have this line:

stylized_image, display_images = fit_style_transfer(style_image=style_image, content_image=content_image,
style_weight=style_weight, content_weight=content_weight,
optimizer=adam, epochs=10, steps_per_epoch=100)

And i get this error message:

ValueError Traceback (most recent call last)
in
13
14 # start the neural style transfer
—> 15 stylized_image, display_images = fit_style_transfer(style_image=style_image, content_image=content_image,
16 style_weight=style_weight, content_weight=content_weight,
17 optimizer=adam, epochs=10, steps_per_epoch=100)

5 frames
/usr/local/lib/python3.8/dist-packages/tensorflow/python/ops/math_ops.py in add_n(inputs, name)
4051 isinstance(x, (ops.Tensor, indexed_slices.IndexedSlices))
4052 for x in inputs):
→ 4053 raise ValueError("Inputs must be an iterable of at least one "
4054 “Tensor/IndexedSlices with the same dtype and shape.”)
4055

ValueError: Inputs must be an iterable of at least one Tensor/IndexedSlices with the same dtype and shape.

2

You should be looking at the inputs of the of the fit_style_transfer() function ie. style_image, content_image. When you produce them you are probably making some mistake on those functions.

3

Yes, but in the code, those inputs are not supposed to be modified.
They are still the same from the beginning and provided by the assignment, this why it is weird for me.

4

Thats right I saw it carefully, then it means the problem is:

get the style image features

style_targets = get_style_image_features(style_image)

get the content image features

content_targets = get_content_image_features(content_image)

in these functions…or even other functions that you need to write up and are uses in the upstream.

I would suggest at this point to go again from beginning, unless you suspect the problem is somewhere in particular.