Learning an Adaptive Learning Rate Schedule

This repository contains code implementing adaptive learning rate tuner, as described in "Learning an Adaptive Learning Rate Schedule" (Xu, Dai, Kemp, & Metz 2019). Code written using Python 3 and Tensorflow 1.12.

Running the Code

The code sample provided here is minimally executable. Setup a virtual environment and run the controller using the provided run.sh:

sh adaptive_learning_rate_tuner/run.sh