dataset $=$ datasets.ImageFolder$($root$=$'path$\_$to$\_$your$\_$data', transform$=$transform$)$

I changed to train$\_$data $=$ datasets.ImageFolder$($root$=\text{'}./$train$\text{'},$ transform$=$transform$)$

test$\_$data $=$ datasets.ImageFolder$($root$=\text{'}./$test$\text{'},$ transform$=$transform$).$ Given that I have $2$ folder 'test' : $240$ images and 'train': $960$ images. Each folder will have $4$ subfolder 'Action', 'Comedy', 'Horror', 'Romance', which includes many poster films images for that type. Using ImageFolder load the data into your notebook and create a dataLoader from the data. Using the same CNN architecture defined in the here $($ class CNN$($nn$.$Module$)$:

def $\_\_$init$\_\_($self$)$:

super$($CNN$,$ self$).\_\_$init$\_\_()$

self.conv$1=$ nn$.$Sequential$($

nn$.$Conv$2$d$($

in$\_$channels$=1,$ # $1$ channel, typical for grayscale images

out$\_$channels$=16,$

kernel$\_$size$=5,$ #size of the convolutional filter is $5$x$5$

stride$=1,$ # the filter moves one pixel at a time

padding$=2,$ # adding of $2$ pixels to the input on all sides, ensuring that the output has the same width and height as the input.

$),$

nn$.$ReLU$(),$

nn$.$MaxPool$2$d$($kernel$\_$size$=2),$

$)$

self.conv$2=$ nn$.$Sequential$($

nn$.$Conv$2$d$(16,32,5,1,2),$

nn$.$ReLU$(),$

nn$.$MaxPool$2$d$(2),$

$)$

# fully connected layer, output $10$ classes

self.out $=$ nn$.$Linear$(32*7*7,10)$

def forward$($self$,$ x$)$:

x $=$ self.conv$1($x$)$

x $=$ self.conv$2($x$)$

# flatten the output of conv$2$ to $($batch$\_$size, $32*7*7)$

x $=$ x$.$view$($x$.$size$(0),-1)$

output $=$ self.out$($x$)$

return output$),$ train the model on this

new dataset. Determine the test accuracy of the model.