Pith. sign in

REVIEW 3 cited by

FLAML: A Fast and Lightweight AutoML Library

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1911.04706 v3 pith:3WWPATXX submitted 2019-11-12 cs.LG stat.ML

FLAML: A Fast and Lightweight AutoML Library

classification cs.LG stat.ML
keywords automlflamlcomputationalcosterrorfastlibrarylightweight
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We study the problem of using low computational cost to automate the choices of learners and hyperparameters for an ad-hoc training dataset and error metric, by conducting trials of different configurations on the given training data. We investigate the joint impact of multiple factors on both trial cost and model error, and propose several design guidelines. Following them, we build a fast and lightweight library FLAML which optimizes for low computational resource in finding accurate models. FLAML integrates several simple but effective search strategies into an adaptive system. It significantly outperforms top-ranked AutoML libraries on a large open source AutoML benchmark under equal, or sometimes orders of magnitude smaller budget constraints.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Auto-FP: An Experimental Study of Automated Feature Preprocessing for Tabular Data

    cs.LG 2023-10 unverdicted novelty 7.0

    Experimental comparison of 15 HPO and NAS algorithms for automated feature preprocessing on 45 tabular datasets finds evolution-based methods and random search as top performers.

  2. Closed-loop Auto Research for Molecular Property Prediction: Discovering and Certifying Generalizable Improvements

    cs.AI 2026-06 unverdicted novelty 6.0

    Closed-loop LM-agent auto research finds some transferable gains on molecular property prediction benchmarks via external data but shows non-transfer for model and feature edits selected on validation.

  3. TabPFN-2.5: Advancing the State of the Art in Tabular Foundation Models

    cs.LG 2025-11 unverdicted novelty 6.0

    TabPFN-2.5 scales tabular foundation models to 20x larger datasets, outperforms tuned tree models on TabArena, achieves near-perfect win rates against default XGBoost, and adds a distillation engine for fast productio...