https://github.com/stsievert/adadamp

Faster machine learning training via batch size estimation
https://github.com/stsievert/adadamp

hyperparameter-selection machine-learning optimization-algorithms performance-tuning

Last synced: 12 days ago
JSON representation

Faster machine learning training via batch size estimation

• Host: GitHub
• URL: https://github.com/stsievert/adadamp
• Owner: stsievert
• License: other
• Created: 2020-02-07T21:14:05.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2024-08-19T16:08:27.000Z (3 months ago)
• Last Synced: 2024-10-18T12:20:01.372Z (about 1 month ago)
• Topics: hyperparameter-selection, machine-learning, optimization-algorithms, performance-tuning
• Language: Jupyter Notebook
• Homepage: https://docs.stsievert.com/adadamp
• Size: 7.31 MB
• Stars: 2
• Watchers: 2
• Forks: 3
• Open Issues: 5
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.txt

Awesome Lists containing this project

README

        
## adadamp

See the documentation for more detail, https://docs.stsievert.com/adadamp

### Install

https://docs.stsievert.com/adadamp/installation.html

## Benchmarks

Good distributed benchmarks:

* Good convnet benchmarks by PyTorch core dev:

  https://github.com/soumith/convnet-benchmarks/

* Good high level overview on training times:

  https://wrosinski.github.io/deep-learning-speed-vol1/

Less relevant benchmarks

* MLPerf: https://mlperf.org/

* https://github.com/ryujaehun/pytorch-gpu-benchmark