Select prediction model

In the second step of the Foundry process, you select the foundational model to use in your model run.

🚧

This page describes features currently in preview. Some improvements may not yet be documented.

When you begin this step, the Model gallery shows available models. By default, the gallery displays models hosted by Labelbox.

The Model Gallery shows the foundational model currently supported by Model Foundry.

If you know the model you want to use, select it from this list.

To locate a specific model, you can:

  • Browse the models displayed in the list.
  • Search for specific models by name (case insensitive)
  • Select individual scenario tags to show appropriate models for a machine-learning task
Select scenario tags to display models that support that scenario.

Scenario tags can filter the model view. Select one to see models associated with that scenario.
Select another tag to change the filter or select the original tag to remove the filter.

If you're not sure which model to select, see Models for an overview of models hosted by Labelbox.

You can also use Model to explore individual models.

Model cards describe the model, display supported scenarios, and provide additional details, such as cost, intended uses, performance ranges, citations, and earlier model runs.

After selecting a model, you need to define and verify your model run.