$0/5$ points $($graded$)$

Complete the code

in

conv.py to build a convolutional model.

For this, you need to make use of Conv$2$D layers and MaxPool$2$d layers $($and perhaps Dropout$)$ in PyTorch.

Make sure that the last layer of the neural network is a fully connected $($Linear$)$ layer. You may start with the

architecture used in the CNN in the previous tab, but modify to generate $2$ outputs.

Available Functions: You have access to the

module as $,$ to the

as

and to the

layer as

; No need to import anything.current$\_$dir $=$ os$.$path.dirname$($os$.$path.abspath$($file$))$

sys$.$path.append$($current$\_$dir$)$

from train$\_$utils import batchify$\_$data, run$\_$epoch, train$\_$model, Flatten

import utils$\_$multiMNIST as U

path$\_$to$\_$data$\_$dir $=\text{'}../$Datasets$/\text{'}$

use$\_$mini$\_$dataset $=$ True

batch$\_$size $=64$

nb$\_$classes $=10$

nb$\_$epoch $=30$

num$\_$classes $=10$Press ESC then TAB or click outside of the code editor to exit