vertopal.com L 14 PyTorch Datasets.pdf - ECE 329 Lecture 14...

ECE 329 Lecture 14 PyTorch Datasets
Object and Class in Phyton:
• Class: Class is a set or category of things having some property or attribute in
common and differentiated from others by kind, type, or quality.
• Object: object is one of instances of the class. which can perform the functionalities
which are defined in the class.
• self: self represents the instance of the class. By using the self keyword we can
access the attributes and methods of the class in python.
• init :" init " is a reseved method in python classes. It is known as a constructor in
object oriented concepts. This method called when an object is created from the
class and it allow the class to initialize the attributes of a class.
• Object and Class in Phyton
class Robot:
def __init__ ( self , name, color, weight):
self .name = name
self .color = color
self .weight = weight
def introduce_self( self ):
print ( "My name is " + self .name) # this in Java
# r1 = Robot()
# r1.name = "Tom"
# r1.color = "red"
# r1.weight = 30
# r2 = Robot()
# r2.name = "Jerry"
# r2.color = "blue"
# r2.weight = 40
r1 = Robot( "Tom" , "red" , 30 )
r2 = Robot( "Jerry" , "blue" , 40 )
r1.introduce_self()
r2.introduce_self()
Datasets
• If you are working on a real-time project involving Deep Learning, it's common that
most of your time goes into handling data, rather than the neural network that you
would build. This is because data is like fuel for your network: the more appropriate
it is, the faster and the more accurate the results are! One of the main reasons for
your neural network to underperform might be due to bad, or poorly understood

data. Hence it is important to understand, preprocess, and load your data into the
network in a more intuitive way.
• Code for processing data samples can get messy and hard to maintain; we ideally
want our dataset code to be decoupled from our model training code for better
readability and modularity.
• PyTorch provides two data primitives: (1) torch.utils.data.DataLoader and (2)
torch.utils.data.Dataset that allow you to use pre-loaded datasets as well as your
own data.
• Dataset stores the samples and their corresponding labels. You will need a
DataLoader (next lecture) wraps an iterable around the Dataset to enable easy
access to the samples
• PyTorch domain libraries provide a number of pre-loaded datasets, all of which are
pre-loaded in the class torch.datasets .