GuidesRecipesAPI ReferenceChangelogDiscussions
Log In

Discover and test run models


The Modzy Model Library allows your organization to see all the models your organization created or licensed through a single pane of glass in one centralized location. From here, you can create "Featured Models" to highlight for use, filter by different types of models for your use case, as well as view your own models you've created if you are a data scientist.


Model library, complete with search and filtering to support your unique use case

Find a Model

To find the right model for your use case, you will navigate to the "Models" section of the top header. From there, click on the second "All Models" tab. Using the filters, sort by Input Type, Subject, Task, Creator, and Feature to generate a list of models that meet your requirements. As you filter using each criteria, each list will dynamically change to show you what options are available in each category.


All Models view


Filtered list of models using specific criteria

Feature a Model

To feature a model for your organization, you will navigate to the model details page and select the "Make Featured" option under the "Actions" list on the left. This will immediately bump the model to the top of the Model Library under the "Featured Models" list.


Model Details page to feature this model


New featured model appears on the top section of Model Library

Test Run a Model

To determine whether a model is right for your use case, you can run a test job within the interface itself. You will navigate to the model details page of the model you are evaluating, and find the "Test Drive" option on the list on the left of the page. There you can add your input directly into the "Input" box, and hit the "Start Job" button. Depending on the infrastructure setup and model itself, this process could take anywhere from a few seconds to several hours. When your input has completed processing, the result will appear in the output box in JSON where you can view and copy it for use.


Adding input directly in the box


Output in JSON format