Git Product home page Git Product logo

Comments (4)

killerducky avatar killerducky commented on July 19, 2024

LZ's policy is to wait until progress stalls, only then lower. Since LZ has regular promotion matches there is an easy way to define stalling. Minigo could keep the always promote policy but add some matches as a way to measure this. Or use CGOS results.

from minigo.

amj avatar amj commented on July 19, 2024

Learning rate of .1/.01/.001 for 200k/400k/800k steps (per AGZ, 2048 batch) equates to 400M/800M/1.6B positions

from minigo.

amj avatar amj commented on July 19, 2024

(for my later reference) On training batch size of 32, that's 12.8M steps / 25.6M steps / 51.2 M steps.

from minigo.

amj avatar amj commented on July 19, 2024

We still don't really know the right answer here. We're using the rates in paper 2 but it seems to be based more off of the l2_cost.

from minigo.

Related Issues (20)

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.