Week3 U-Net HW Q3

Hi, I am having the hard time on how to solve this problem. I passed all the questions before. Thanks for the help.

Hi there,

The error tells you you input to the function needs be a tensor with 4 dimensions and you you have given only three. Perhaps you need a batch size dimension as well for the input.

Hello @Lily007!

The error indicates that input to the ublock6 is not as expected. So, what are the input arguments? They are cblock5[0] and cblock4[1] as expansive_input and contractive_input respectively. Now the question is: how we got cblock5[0] and cblock4[1]?

Check your implementation of cblock5. Why you are using ,_ there? Are you using this,_ in cblock4[1] as well? Remove them.

Best,
Saif.

Hi @saifkhanengr , I changed my code and the problem is solved; however, my output is different from the expected output, can you help me to check where the problem is?


Thanks for you help!

It looks like you specified the wrong number of filters to the last “downsampling” layer. Please check your code versus the instructions.

Note that I figured this out by looking at the file outputs.py to see what the correct model summary should look like.