https://github.com/cedrickchee/ranger-optimizer

Ranger - a synergistic optimizer using RAdam (Rectified Adam) and Lookahead in one codebase

https://github.com/cedrickchee/ranger-optimizer

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
  • .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: arxiv.org
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (5.3%) to scientific vocabulary

Keywords

deep-neural-network lookahead-methods optimization-algorithms radam ranger-optimizer
Last synced: 5 months ago · JSON representation

Repository

Ranger - a synergistic optimizer using RAdam (Rectified Adam) and Lookahead in one codebase

Basic Info
  • Host: GitHub
  • Owner: cedrickchee
  • License: apache-2.0
  • Language: Python
  • Default Branch: master
  • Homepage:
  • Size: 16.6 KB
Statistics
  • Stars: 2
  • Watchers: 2
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Fork of lessw2020/Ranger-Deep-Learning-Optimizer
Topics
deep-neural-network lookahead-methods optimization-algorithms radam ranger-optimizer
Created over 6 years ago · Last pushed over 6 years ago

https://github.com/cedrickchee/ranger-optimizer/blob/master/

# [Ranger Neural Network Optimizer](https://forums.fast.ai/t/meet-ranger-radam-lookahead-optimizer/52886?u=cedric)

Ranger  a synergistic optimizer combining RAdam (Rectified Adam) and Lookahead in one codebase.

> - RAdam stabilizes training at the start. (reference: ["On the Variance of the Adaptive Learning Rate and Beyond"](https://arxiv.org/abs/1908.03265v1) paper by Liyuan Liu et al.)
>
> - Lookahead stabilizes training and convergence during the rest of training. (reference: ["Lookahead Optimizer: k steps forward, 1 step back"](https://arxiv.org/abs/1907.08610v1) paper by Michael Zhang, Geoffrey Hinton, et al.)

## Examples

_Usage and notebook to test with coming shortly._

## Software Requirements

- Python 3.x
- PyTorch 1.x

## License

[Apache License 2.0](./LICENSE)

## Credits

Based on original work done by [Less Wright (lessw2020)](https://github.com/lessw2020). He built on LonePatient's implementation for the Lookahead part, and the official RAdam code for RAdam:

- Lookahead implementation from LonePatient - https://github.com/lonePatient/lookahead_pytorch/blob/master/optimizer.py
- RAdam code by https://github.com/LiyuanLucasLiu/RAdam/blob/master/radam.py

Owner

  • Name: Cedric Chee
  • Login: cedrickchee
  • Kind: user
  • Location: PID 1
  • Company: InvictusByte

Lead Software Engineer | LLMs | full stack Go/JS dev, backend | product dev @ startups | 🧑‍🎓 CompSci | alumni: fast.ai, Antler.co

GitHub Events

Total
Last Year