DS4420-final-project

Final Project for DS4420 - Machine Learning and Data Mining 2

Report

https://wandb.ai/dpis-disciples/test-project/reports/DS4420-Machine-Learning-2-Validating-and-Investigating-Self-Organizing-Neural-Networks-SONN---VmlldzoxOTIxNzg0?accessToken=k9iqt6gs6ffehgbv4bx9uast9txud8zaoay13x5gst9vx24o2790zaexikl09nn8

Data Loading

Upon cloning, run python3 mnist.py to download and save a local version of the dataset.

Running Stuff

NOTE: Before you run the first time, make sure you have run pip install wandb and log in using wandb login. Alternatively, you can not log in and run it in offline mode by running wandb offline.

To run our model with MNIST, set the dataset name in config.yaml to "mnist". Run main.py

To run our model with Fashion-MNIST, set the dataset name in config.yaml to "fmnist". Run main.py

To run our model with CIFAR-10, set the dataset name in config.yaml to "cifar10". For CIFAR, look at the cifar config file for the hyper parameter values. Run main.py