Cats vs Dogs classifier

EE4483 Artificial Intelligence & Data Mining Project Option 2

Usage

This code was tested on Python 3.9.6, but should work for most Python version >= 3.7

1. Install required Python packages:

python -m pip install -r requirements.txt

2. Install Pytorch and Torchvision based on your system. Instructions can be found on the Pytorch website here. This code was tested with Pytorch == 1.1.0 with CUDA >= 11.3

3. Extract dataset into same folder as notebook. Folder should look something like this:

./
├─ datasets/
│  ├─ test/
│  ├─ train/
│  │  ├─ cats/
│  │  ├─ dogs/
│  ├─ val/
│  │  ├─ cats/
│  │  ├─ dogs/
├─ main.ipynb
├─ README.md

4. Run main.ipynb. The sections Setup environment and 1 - 3 must be run before any other code.
   • To train and test model for Cats vs Dogs classifier, run sections 4 - 5.
   • To train and test model for CIFAR-10, run sections 6 - 7