What is Neptune?¶
Neptune is an experiment tracking tool bringing organization and collaboration to data science projects.
Neptune records your entire experimentation process. Exploratory notebooks, model training runs, code, hyperparameters, metrics, data versions, results exploration visualizations and more. Everything is stored and backed-up in organized knowledge repository, ready to be accessed, analyzed, shared and discussed with your team. No matter what type of problems you are working on, Neptune fits to all, from evaluating credit risk to finding the nuclei in divergent images.
Register here and give it a try!
Is Neptune a tool for me?¶
Neptune is designed with two particular groups in mind:
- Hands-on data scientists, who work in teams.
- Lead data scientists, who need easy-to-use place for her projects’ health-check. She also cares about her team and wants to deliver best possible tooling while allowing teammate to work in most effective way.
What can I use Neptune for?¶
Unconstrained data science team work¶
Neptune allows you to freely choose your Python libraries. It simply mean that you work with your favorite tools while leveraging full neptune’s potential. Check get started tutorial that will guide you - step-by-step - how to adapt your code to fully integrate it with Neptune (in practise: few additional lines of code).
Neptune is light-weight solution that let’s you easily start playing with it in your organization.
Real-time model training tracking¶
Each experiment, once initiated, is immediately available in the project dashboard, which allows you to see its details, track training progress and compare it to all remaining experiments.
Tracking data science work¶
Either you are data scientist, ML researcher or department head, you care about your work artifacts. Neptune facilitates transparency, by introducing feature to compare experiments, see models improvements over time. Power Users can check experiments API, to learn how to build customized comparisons.