Skip to content

llSourcell/synthetic_gradients_explained

Repository files navigation

synthetic_gradients_explained

This is the code for "Synthetic Gradients Explained" by Siraj Raval on Youtube

Overview

This is the code for this video on Youtube by Siraj Raval on synthetic gradients. This is an optimization strategy DeepMind released that could overtake backpropagation.

Dependencies

  • numpy

Use pip to install missing dependencies.

Usage

Both versions (backprop and synthetic gradient versions) have their own python file. Run them from command line using

python name_of_file.py

To run the notebook run this in the root directory

jupyter notebook

Install jupyter from here

Credits

Credits for the code go to trask. I've merely created a wrapper to get people started.

About

This is the code for "Synthetic Gradients Explained" by Siraj Raval on Youtube

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published