nachiket273/One_Cycle_Policy

Pytorch notebook with One Cycle Policy implementation (https://arxiv.org/abs/1803.09820)

Cyclic Learning Rates and One Cycle Policy.

This notebook contains some of the experiments mentioned in papers Cyclical Learning Rates for Training Neural Networks and A disciplined approach to neural network hyper-parameters: Part 1 - learning rate, batch size, momentum, and weight decay. My blog related to the same can be found here

This notebook works with:

1. Python 3.6+
2. Pytorch 0.3.x