ParagEkbote·Oct 19What Another Year of Open-Source Contributions Taught Me About Engineering in PracticeIn November 2025, I wrote about what more than 90 open-source pull requests had taught me about software. Nearly a year later, looking back…A response icon1A response icon1
InOptunabyNaoto Mizuno·Oct 15Improving the Default Optimization Algorithm in Optuna v5In Optuna v5, we have updated the default sampler for the first time in several years. This article outlines the discussions and…
Taha Soylu·Sep 24Hyperparameter TuningWe train our models with different parameters. But we do not know which value is more effective. Hyperparameter Tuning helps us identify…
InOptunabyc-bata·Sep 7Rustuna: A Faster Optuna Implementation in RustWe have developed and released Rustuna, a faster Optuna implementation built with Rust. In this post, we introduce Rustuna’s key features…A response icon1A response icon1
InOptunabyYoshihiko Ozaki·Sep 7Releasing Optuna v5.0, an Open-Source Black-Box Optimization LibraryRustuna is also available!
Alejandro Gómez Ruiz·Jul 24Does fancy hyperparameter search beat a humble grid? A bake-off on 8 datasetsPart two of a series putting common ML habits to the test on 8 imbalanced datasets. The habit here: reaching for Optuna “because Bayesian…
Yachika Anand·Jul 6The Mental Model of PyTorch: Build a Neural Network in 5 PhasesI stopped copy-pasting training loops and finally understood what every line does — here’s the whole picture, taught the way I wish someone…A response icon1A response icon1
InData Reply IT | DataTechbyGeard Koci·Jun 10Optuna: Smarter Hyperparameter Optimization Beyond Grid Search.Why define-by-run search spaces, adaptive sampling, and pruning make model tuning faster, cheaper, and more practical than exhaustive search
TechWealthBuzz·Jun 6LightGBM: The Machine Learning Engine That Runs the World’s Production SystemsWhat actually happens inside the fastest gradient boosting framework — and how to make it scream in production
