So in the MNIST example Notebook the output gets reshaped with this line.
temp = [j.reshape(1, outputSize) for j in output_data]
Why is this necessary because this Netron Graph of the linked model says the Output should look like 1x10 before the download. The shape is only (1,1,10) , so its not like before the last dense layer and not like the last reshape layer ?
So in the MNIST example Notebook the output gets reshaped with this line.
Why is this necessary because this Netron Graph of the linked model says the Output should look like 1x10 before the download. The shape is only (1,1,10) , so its not like before the last dense layer and not like the last reshape layer ?
Greetings