neptune

An interface to Neptune. A metadata store for MLOps, built for teams that run a lot of experiments. It gives you a single place to log, store, display, organize, compare, and query all your model-building metadata. Neptune is used for: • Experiment tracking: Log, display, organize, and compare ML experiments in a single place. • Model registry: Version, store, manage, and query trained models, and model building metadata. • Monitoring ML runs live: Record and monitor model training, evaluation, or production runs live For more information see .