Skip to main content

fseval

Benchmarking framework for Feature Selection and Feature Ranking algorithms 🚀

Get started

Run experiments on distributed systems

Run experiments on HPC clusters like SLURM using Submitit, or use cloud providers like AWS with Ray to run large scale benchmarks.

Easy to use

fseval has an easy to understand API and can easily be extended. Define your own dataset adapters, callbacks and storage providers.

Reproducible experiments

All information to replay an experiment is neatly stored in a config file. Others can easily reproduce your results.

Send experiment results directly to a dashboard

fseval integrates with Weights and Biases, so you can enjoy all the powerful tooling built into the platform to help analyze your data.

See how.

Export your data to any SQL database

Experiment metrics can be sent to one of many SQL databases. Support includes SQLite, PostgreSQL, MySQL, Oracle and more. Support is achieved through integration with SQL ALchemy.

See how.