skore
Track your Data Science. Skore's open-source Python library accelerates ML model development with automated evaluation reports, smart methodological guidance, a
What is skore?
Track your Data Science. Skore's open-source Python library accelerates ML model development with automated evaluation reports, smart methodological guidance, and comprehensive cross-validation analysis.
Overview of skore
What it is
Skore is an open-source Python library for tracking data science experiments. It accelerates ML model development by providing automated evaluation reports, smart methodological guidance, and comprehensive cross-validation analysis. The library structures experiments, reduces boilerplate code, guides decision-making with built-in warnings, and maintains project clarity. It consists of two components: Skore Lib (the open-source library) and Skore Hub (a collaborative platform for sharing experiments).
Key features
- Automated evaluation reports that generate insights with one line of code - Smart methodological guidance to help avoid common pitfalls - Comprehensive cross-validation analysis for model evaluation - Structured experiment organization for easier understanding and maintenance - Reduced boilerplate code for common evaluations - Decision guidance with built-in methodological warnings - Clear project structure that makes work easier to understand and maintain
Use cases
- Evaluating machine learning models and obtaining structured insights - Tracking and organizing data science experiments - Collaborating with teams on model development via Skore Hub - Improving productivity in analytics and data science workflows
Getting started
Install skore with pip or conda: bash # If you plan to use Skore locally pip install -U skore # If you wish to interact with Skore Hub as well pip install -U skore[hub] # If you wish to log projects to MLflow pip install -U skore[mlflow] bash conda install conda-forge::skore Then use the library to evaluate a model: python from sklearn.datasets import make_classification from sklearn.linear_model import LogisticRegression from skore import evaluate X, y = make_classification(n_classes=2, n_samples=100_000, n_informative=4) clf = LogisticRegression() cv_report = evaluate(clf, X, y, splitter=5) cv_report.help() metrics_summary = cv_report.metrics.summarize().frame() roc_plot = cv_report.metrics.roc() roc_plot.plot()
License and pricing
Skore is licensed under the MIT License and is open source.
Based on the project's public README and repository metadata.