VGG architecture IN MNIST DIGIT

#1

image
image1920×1080 427 KB

The above picture is of vgg so if we want to train MNIST DIGIT dataset with 10 classes then how to change the dense layer dimensions will it be same as 4096, 4096, 1000 or how we will change the dense layers to 10??

please elaborate I am confused how to apply any architecture in other datasets

#2

Refer the Convolutional Neural Nets Topic 26 of the course ASAP.

https://cloudxlab.com/assessment/displayslide/1248/session-11-convolutional-neural-network-april-22-2018?course_id=71&playlist_id=419