hw2

Homework 2 - Skeleton Code

This archive contains everything you need to complete Homework 2. It contains the following files:

download_octmnist.py

Allows you to download the OCTMNIST dataset to a compressed .npz file, which hw2-q2.py can load.

utils.py

Utility functions for hw2-q2.py

hw2-q2.py

Contains skeleton code for Question 2, which covers image classification using CNNs, with implementation in torch.

hw2-q3.ipynb

Contains skeleton code for Question 3, which covers a sequence-to-sequence task using RNNs and attention, with implementation in torch. We suggest using google colab or kaggle to run the notebook with GPU accelerators.

how-to-use-kaggle-tutorial.md

A simple tutorial on how to run the notebook hw2-q3.ipynb in kaggle.