Skip to main content

A learning rate recommending and benchmarking tool.

Project description


GitHub license Version


A learning rate benchmarking and recommending tool, which will help practitioners efficiently select and compose good learning rate policies.

  • Semi-automatic Learning Rate Tuning
  • Evaluation: A set of Useful Metrics, covering Utility, Cost, and Robustness.
  • Verification: Near-optimal Learning Rate

If you find this tool useful, please cite the following paper:

  author = {{Wu}, Yanzhao and {Liu}, Ling and {Bae}, Juhyun and {Chow}, Ka-Ho and
  {Iyengar}, Arun and {Pu}, Calton and {Wei}, Wenqi and {Yu}, Lei and
  {Zhang}, Qi},
  title = "{Demystifying Learning Rate Polices for High Accuracy Training of Deep Neural Networks}",
  journal = {arXiv e-prints},
  keywords = {Computer Science - Machine Learning, Statistics - Machine Learning},
  year = "2019",
  month = "Aug",
  eid = {arXiv:1908.06477},
  pages = {arXiv:1908.06477},
  archivePrefix = {arXiv},
  eprint = {1908.06477},
  primaryClass = {cs.LG},
  adsurl = {},
  adsnote = {Provided by the SAO/NASA Astrophysics Data System}



Supported Platforms

Development / Contributing




See the people page for the full listing of contributors.


Copyright (c) 20XX-20XX Georgia Tech DiSL
Licensed under the Apache License.

Project details

Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for LRBench, version
Filename, size File type Python version Upload date Hashes
Filename, size LRBench- (5.9 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page