Giter VIP home page Giter VIP logo

Comments (1)

Kaixhin avatar Kaixhin commented on August 21, 2024

The segments number, k, is used for stratified sampling for constructing a minibatch, as you've noted. The partition number, which is a number that I chose myself, and is based on this note from the appendix of the paper:

A small improvement in running time came from avoiding excessive recalculation of partitions for the sampling distribution. We reused the same partition for values of N that are close together and by updating α and β infrequently.

For rank-based PER, N obviously changes as the agent starts and picks up enough memories to fill its replay memory. Once it is full then the partitions don't change, as the buffer replaces old transitions with new ones and N is then static. But with N = 1000000, it'd be expensive to recalculate the indices needed for stratified sampling every time a new transition was added. Thus it should be done a few times until the memory is full. Whether 100 times is the appropriate rate or not, I'm not sure (as it's not given in the paper).

from atari.

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.