CIFAR10 Classification with PyTorch
This project uses the CIFAR10 dataset for training. It consists of 60000
32x32
colour images in 10 classes, with 6000
images per class. There are 50000
training images and 10000
test images.
The dataset is divided into five training batches and one test batch, each with 10000
images. The test batch contains exactly 1000
randomly-selected images from each class. The training batches contain the remaining images in random order, but some training batches may contain more images from one class than another. Between them, the training batches contain exactly 5000
images from each class.
(Source:https://www.cs.toronto.edu/~kriz/cifar.html)
Steps:
- Exploring the dataset
- Defining a neural network architecture
- Hyper-parameter search and training the model
- Evaluating the model's results and making cool graphs!
Results:
Train Accuracy Test Accuracy without data augmentation *81.69% 76.68% with data augmentation 85.15% 79.76% with transfer learning (VGG-16) 92.89% 85.93%
* - running accuracy