Skip to main content
Version: v0.3.0

Download a dataset's zero-shot predictions


By default, H2O Label Genie generates zero-shot predictions for specific annotation tasks that can be downloaded. In particular, H2O Label Genie generates zero-shot predictions for the following supported annotation tasks:

  • Image classification
  • Object detection
  • Image instance segmentation
  • Text classification
  • Text summarization

What are zero-shot learning model predictions?​

The labels or suggested labels for a given sample that are provided by a zero-shot learning model are called zero-shot learning model predictions. For example, for an image and text classification annotation task, H2O Label Genie, with a zero-shot learning model activated, offers a percentage probability of an image or text belonging to a certain label (class). For an object detection annotation task, it populates an image with bounding boxes where the bounding boxes capture the desired objects (for example, a car).

To learn more about zero-shot predictions, see Zero-shot learning models.


To download the zero-shot predictions of an annotation task, consider the following instructions:

  • If the Enable zero-shot predictions setting is turned Off, the zero-shot learning model utilized for the annotation is not available during the annotation process while preventing the generation of zero-shot predictions. To turn On the Enable zero-shot predictions setting, see Smart annotation.
  • If the annotation task was recently created (less than a minute ago), you might need to wait a bit before the zero-shot predictions become available for download.
    • H2O Label Genie notifies you to Refresh the instance when zero-shot predictions (suggestions) are available.
  1. In the H2O Label Genie navigation menu, click Annotation tasks.
  2. In the annotation tasks table, double-click the row where the annotation task you want to download its zero-shot predictions is located.
  3. Click the Export tab.
  4. In the Export zero-shot predictions list, select Download ZIP.