Assignment-5 Report
CIFAR10- CNN based image classification
Created on June 1|Last edited on June 1
Comment
In this assignment we have used CNN network having 2 hidden layers and 2 max-pool layers of size (2x2) for image classification. The dataset used is CIFAR10 which has 50,000 training and 10,000 test data set containing 10 classes namely:
classes = ('plane', 'car', 'bird', 'cat', 'deer',
'dog', 'frog', 'horse', 'ship', 'truck')
Training an image classifier
We will do the following steps in order:
Load and normalize the CIFAR10 training and test datasets using torchvision
1.Define a Convolutional Neural Network
2.Define a loss function
3.Train the network on the training data
4.Test the network on the test data
5.Evaluate the model using Confusion Matrix
6. Log all the progress in Wandb portal.
Different configurations for which we are evaluating our model:
Configuration-1: SGD optimizer with learning rate = 0.001, momentum = 0.9 and cross-entropy as loss function.
Configuration-2: Adam optimizer with learning rate = 0.01 and cross-entropy as the loss function
Configuration-3: SGD optimizer with learning rate = 0.001, momentum = 0.9 and squared error loss as the loss function.
Configuration-4: Adam optimizer with learning rate = 0.01 and squared error loss as the loss function
Run set
5
Add a comment