Finding the Optimal Learning Rate using Bayesian Optimization on K-MNIST in PyTorch

This repository gives a simple hands-on introduction into Bayesian Optimization for learning rate optimization. I am training a small ResNet implemented in PyTorch on the Kuzushiji-MNIST (or K-MNIST) dataset. It is only meant for educational purposes!

My Blog

I have also a blog explaining everything in detail. Also, checkout the Jupyter notebook, where you'll find the results as well.

Installing Requirements

Please install all required packages.

pip install -r requirements.txt

Run an Example

Simply run the main script with

python main.py

Note that it was test on Python 3.8.