# bckenstler/CLR

Repository: https://github.com/bckenstler/CLR
Canonical: https://ross.abutalabs.com/products/clr
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2020-06-05T21:55:36+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3452, "days_push": 2280, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1206, forks 243 (observed 2026-08-28T04:03:59.393837+00:00)

## What it is
A Keras callback implementing cyclical learning rate (CLR) policies from Leslie Smith's paper, with built-in triangular, triangular2, and exp_range modes plus custom scaling functions. It helps neural network training converge faster by cycling the learning rate between bounds.

## Use cases
- implement cyclical learning rate in keras
- speed up neural network convergence
- find optimal learning rate for training
- escape saddle points during training
- experiment with custom lr scheduling policies
- train cifar-10 classifier faster

## When to choose
- You train Keras/TensorFlow models and want faster convergence via cyclical learning rates
- You want to experiment with custom LR scaling policies

## When to avoid
- You use PyTorch or another framework without Keras
- You need actively maintained tooling - the repo hasn't released since 2020

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning
- domain: deep-learning, machine-learning
- platform: python
- tags: keras, learning-rate-scheduling, callback, hyperparameter-tuning

## Member repositories
- bckenstler/CLR (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:59.393837+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:19:16.994728+00:00, confidence not recorded.
  - readme: https://github.com/bckenstler/CLR (fetched 2026-08-28T04:03:59.393837+00:00, sha f27ae8572693)
- Data as of 2026-08-30T08:39:29.467469+00:00.