Error with custom architecture

[fpramunno]

Hi,
I found that with this line of code:

visualtorch/visualtorch/layered.py

Line 324 in 45f5161

z = min(max(int(self_multiply(shape[0:1] + shape[3:]) * scale_z), z), max_z)

I have the following error:

Traceback (most recent call last):

Cell In[3], line 1
img = visualtorch.layered_view(model, input_shape=input_shape, spacing=50)

File ~\anaconda3\lib\site-packages\visualtorch\layered.py:124 in layered_view
layer_y, layer_types, boxes, x_off, img_height, max_right = _create_architecture(

File ~\anaconda3\lib\site-packages\visualtorch\layered.py:324 in _create_architecture
z = min(max(int(self_multiply(shape[0:1] + shape[3:]) * scale_z), z), max_z)

TypeError: 'int' object is not iterable

So switching to:
z = min(max(int(np.prod(self_multiply(shape[0:1] + shape[3:])) * scale_z), z), max_z)

it works.

fpr

[willyfh]

Hi, I’m sorry for my late response. Would it be possible to provide the custom architecture (or a similar one) so we can reproduce the reported bug? Thank you! 🙏