Archive experiment

You can archive an experiment if you don’t want to use it at the moment. The experiment will not be deleted but will not show up in the Evaluation view, and you cannot create deployments using the experiment.

Deep learning is an iterative process. Often you have to try many experiments with different approaches before you find the best performing model. This means that the Evaluation view can become cluttered, and it becomes hard to find the best experiment. That’s why you can archive experiments you’re not interested in.

If you archive an experiment, it will not be shown in the Loss and metrics curve or the Experiment ranking list.

How to archive an experiment

Click on the Experiment options menu (three dots) and select Archive.

Archive an experiment
Figure 1. How to archive an experiment.

Note: You cannot archive a deployed experiment, and you cannot create a deployment using an archived experiment.

Unarchive experiment

Click on the Experiment options menu (three dots) and select Unarchive. ThreeDotsButton_A

Was this page helpful?
Yes No