GithubHelp home page GithubHelp logo

utmist / iclr-2019-initialized-equilibrium-propagation Goto Github PK

View Code? Open in Web Editor NEW
1.0 1.0 0.0 63 KB

ICLR 2019 Reproducibility Challenge: Initialized Equilibrium Propagation for Backprop-Free Training.

Python 100.00%

iclr-2019-initialized-equilibrium-propagation's People

Contributors

imbrem avatar jacobjinkelly avatar mchanchee avatar xwinxu avatar

Stargazers

avatar

Watchers

avatar avatar avatar

iclr-2019-initialized-equilibrium-propagation's Issues

State Gradient Checking Fails

The test_energy_grad_state_check currently fails on all neurons in all layers. Furthermore, the error increases as the value of dh decreases, suggesting this is not simply due to numerical error arising from curvature of the energy.

This error was observed before implementing batching in energy and energy_grad_state, so it unlikely that this is the cause.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.