I successfully reshaped and normalized the training data.
But I’ve got an error at the bottom for the test case.
The error message says I got 0.0039 for the maximum value.
I guess it would be a value after normalizing 1 using 255.
I don’t know what’s happening here.
That’s strange. I just tried it and mine passed the test cases.
Look at your implementation again. You are to reshape by adding an extra dimension to the tensor, then normalize the pixel values by dividing by the maximum. That should give you the answer.
You can refresh and run the cells again, if you are sure of your implementation.
Can you share screenshot of previous grade cell codes by personal DM. Click on my name and then message.
You can also share the codes for reshaping and normalisation.
Don’t post codes here
Hi @vlfdmstmf
The reason for the error is because although your codes were correct but because you used an incorrect grade cell argument for GRADED FUNCTION: reshape_and_normalize for reshaping the image.
You were not suppose to use training_images for recalling the reshape image codes but the recalled args for that grade cell for training_images is images which you were suppose to use for that cell.
Also you do not require to recall double tuple while adding the extra dimension, only use (60000, 28, 28, 1) single tuple.
Let me know if you still continue to have issue.
Regards
DP
1 Like
Oh right! I was not supposed to use training_images for that graded function!
Thanks a lot!
1 Like