Repository files navigation

README in other languages:中文,deutsch,français,orcontribute your own.

Lale is a Python library for semi-automated data science.Lale makes it easy to automatically select algorithms and tunehyperparameters of pipelines that are compatible withscikit-learn, in a type-safe fashion. Ifyou are a data scientist who wants to experiment with automatedmachine learning, this library is for you!Lale adds value beyond scikit-learn along three dimensions:automation, correctness checks, and interoperability.Forautomation, Lale provides a consistent high-level interface toexisting pipeline search tools including Hyperopt, GridSearchCV, and SMAC.Forcorrectness checks, Lale uses JSON Schema to catch mistakes whenthere is a mismatch between hyperparameters and their type, or betweendata and operators.And forinteroperability, Lale has a growing library of transformersand estimators from popular libraries such as scikit-learn, XGBoost,PyTorch etc.Lale can be installed just like any other Python package and can beedited with off-the-shelf Python tools such as Jupyter notebooks.

• Introductory guide for scikit-learn users
• Installation instructions
• Technical overviewslides,notebook, andvideo
• IBM'sAutoAI SDK uses Lale, see demonotebook
• Guide for wrappingnew operators
• Guide forcontributing to Lale
• FAQ
• Papers
• PythonAPI documentation

The name Lale, pronouncedlaleh, comes from the Persian word fortulip. Similarly to popular machine-learning libraries such asscikit-learn, Lale is also just a Python library, not a new stand-aloneprogramming language. It does not require users to install new toolsnor learn new syntax.

Lale is distributed under the terms of the Apache 2.0 License, seeLICENSE.txt.It is currently in anAlpha release, without warranties of anykind.