AI Paper Atlas

Research map / Optimization / Optimizers

L4: Practical loss-based stepsize adaptation for deep learning

Michal Rolinek, Georg Martius

NeurIPS 2018 · Optimization › Optimizers · approach: Adam convergence and adaptive learning rates · 33 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Replace constant tuned stepsizes with default L4 adaptation for Adam or momentum SGDAdam and momentum SGD · MNISTtest accuracy: 97.9 ± 0.3 -> 98.4 ± 0.1; 98.0 ± 0.4 -> 98.4 ± 0.1Mixed
Replace tuned constant stepsizes with default L4 adaptationAdam and momentum SGD · Fashion MNISTtest accuracy: 93.1 ± 0.2 -> 93.35 ± 0.15; 93.0 ± 0.15 -> 93.25 ± 0.2Improved
Increase dropout from p=0.4 to p=0.7 (ablation)L4 Adam and L4 Mom · Fashion MNISTtest accuracy: 93.35 ± 0.15 -> 93.6 ± 0.25; 93.25 ± 0.2 -> 93.45 ± 0.15Improved
Use smaller batch sizes instead of the main-paper batch size (ablation)L4 Adam and L4 Mom · MNISTtraining lossImproved
Use smaller batch sizes (ablation)L4 Adam and L4 Mom · DNC REPEAT-COPYtraining lossImproved
Use batch size 8 (ablation)L4 Mom · DNC REPEAT-COPYtraining lossMixed
Use L4-adapted stepsizes rather than best constant stepsizesL4 Mom and L4 Adam · CIFAR-10 ResNettest accuracyImproved
Use default L4 adaptation instead of the repository's piecewise learning-rate policyL4 Mom and L4 Adam · CIFAR-10 ResNettest accuracyHurt
Use L4 adaptation rather than optimized constant learning ratesL4 Adam and L4 Mom · DNC REPEAT-COPYtraining lossImproved
Use L4 adaptation rather than optimized constant learning ratesL4 Adam and L4 Mom · Badly conditioned regressiontime to convergence: 1.4 ± 0.5 vs 3.0 ± 1.4; 1.0 ± 0.02 vs 3.0 ± 1.4; 2.1± 0.4 vs 123 ± 64; 1.8 ± 0.2 vs 123 ± 64Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy