How to Track Hyperparameters: Optuna + neptune.ai Integration
In this video, we show how to keep track of your hyperparameter search metadata when using Optuna. Jakub Czakon explains:
- How to use the Neptune-Optuna integration,
- How to track your metadata from hyperparameter optimization sweeps (including visualizations, parameters at each trial, distributions at each trial, best parameters),
- How to analyze the data in the Neptune app,
- And more.
Important: This video was created in June 2021. For the most up-to-date tutorials and examples, please refer to the Neptune-Optuna integration docs.
If you want to try out the integration on your own, check this Neptune-Optuna Colab notebook or check this Neptune-Optuna GitHub repo.
neptune.ai is an ML metadata store primarily used for experiment tracking and model registry.
We are 100% focused on ML metadata management, but we’re making sure that it’s easy to integrate Neptune with other components of the ML stack.
Neptune is already integrated with 25+ tools and libraries, and the list is growing. You can check our roadmap to see what’s currently under development.
Other useful resources
See an Optuna example project in the Neptune app (no registration needed).
Read a case study created with a healthcare startup Theta Tech AI, that uses neptune.ai with Optuna.
Couldn’t find the use case you were looking for?
Just get in touch, and our ML team will create a custom demo for you.