Art_Generation_with_Neural_Style_Transfer train_step issues

Course Q&A TensorFlow Developer Professional Convolutional Neural Networks in TensorFlow
,
1

Hi,

I’m working on the train_step of week 4 assignment 2.

I’m in the for loop, directionally i’m:

getting the vgg_model_outputs for the current image

computing the style cost using the style layers and current image

computing the content cost using the content layers and the current image

using those to compute the total cost. it seems straightforward, and when i try to debug my issue, im not getting errors until i try to compute the total cost. it’s giving me the folllowing however:

ValueError                                Traceback (most recent call last)
Input In [84], in <cell line: 7>()
      1 ### you cannot edit this cell
      2 
      3 # You always must run the last cell before this one. You will get an error if not.
      5 generated_image = tf.Variable(generated_image)
----> 7 train_step_test(train_step, generated_image)

File /tf/W4A2/public_tests.py:86, in train_step_test(target, generated_image)
     82 def train_step_test(target, generated_image):
     83     generated_image = tf.Variable(generated_image)
---> 86     J1 = target(generated_image)
     87     print(J1)
     88     assert type(J1) == EagerTensor, f"Wrong type {type(J1)} != {EagerTensor}"

File /usr/local/lib/python3.8/dist-packages/tensorflow/python/util/traceback_utils.py:153, in filter_traceback.<locals>.error_handler(*args, **kwargs)
    151 except Exception as e:
    152   filtered_tb = _process_traceback_frames(e.__traceback__)
--> 153   raise e.with_traceback(filtered_tb) from None
    154 finally:
    155   del filtered_tb

File /tmp/__autograph_generated_filel5ri0qpt.py:17, in outer_factory.<locals>.inner_factory.<locals>.tf__train_step(generated_image)
     15     J_content = ag__.converted_call(ag__.ld(compute_content_cost), (ag__.ld(a_C), ag__.ld(a_G)), None, fscope)
     16     ag__.ld(print)('here2')
---> 17     J = ag__.converted_call(ag__.ld(compute_content_cost), (ag__.ld(J_content), ag__.ld(J_style)), None, fscope)
     18 grad = ag__.converted_call(ag__.ld(tape).gradient, (ag__.ld(J), ag__.ld(generated_image)), None, fscope)
     19 ag__.converted_call(ag__.ld(optimizer).apply_gradients, ([(ag__.ld(grad), ag__.ld(generated_image))],), None, fscope)

File /tmp/__autograph_generated_filequpmd2t3.py:11, in outer_factory.<locals>.inner_factory.<locals>.tf__compute_content_cost(content_output, generated_output)
      9 do_return = False
     10 retval_ = ag__.UndefinedReturnValue()
---> 11 a_C = ag__.ld(content_output)[(- 1)]
     12 a_G = ag__.ld(generated_output)[(- 1)]
     13 (_, n_H, n_W, n_C) = ag__.ld(a_G).shape

ValueError: in user code:

    File "<ipython-input-67-3cdf8296a584>", line 32, in train_step  *
        J = compute_content_cost(J_content, J_style)
    File "<ipython-input-4-07b830dad426>", line 15, in compute_content_cost  *
        a_C = content_output[-1]

    ValueError: Index out of range using input dim 0; input has only 0 dims for '{{node strided_slice}} = StridedSlice[Index=DT_INT32, T=DT_FLOAT, begin_mask=0, ellipsis_mask=0, end_mask=0, new_axis_mask=0, shrink_axis_mask=1](mul_10, strided_slice/stack, strided_slice/stack_1, strided_slice/stack_2)' with input shapes: [], [1], [1], [1] and with computed input tensors: input[3] = <1>.

It seems so straightforward im worried i have an issue in a previous function that the unit tests didnt uncover. Any hints appreciated!

Best,

Matt

2

It looks like you are invoking compute_content_cost incorrectly. What are the arguments to that function? Something other than J_content and J_style, right? :nerd_face:

Although I admit that the exception trace is a bit tricky to read and maybe I’m jumping to the wrong conclusion here. But it seems to be saying that you wrote this line of code:

3

Duh! Thanks Paul, I think I was staring at it too long, was just calling the wrong function!

1 Like
4

Glad that it was helpful. No worries. Anyone who’s been programming for more than a couple of weeks has had the experience of looking right at the bug and not seeing it. A fresh pair of eyes is always a good step if you can invoke that. :laughing: