Skip to main content
Version: v1.3.0

Release notes

v1.3.0 | April 5, 2023


At the core of the new version, H2O Hydrogen Torch improves and expands on its currently supported problem types. These new improvements and expansions further expand the mission to offer a platform that enables you, with no coding experience, to train state-of-the-art deep neural networks on diverse problem types. The major points of this release are as follows:

  • H2O Hydrogen Torch now supports the following problem types:
    • Speech recognition
    • 3D image classification
    • 3D image regression
    • 3D image semantic segmentation
  • You can now deploy built models to H2O MLOps directly from the H2O Hydrogen Torch UI
  • You can now import data from H2O Drive

To learn more about the new release, observe the below subsections.


  • New: The H2O Hydrogen Torch landing page now has a new design that provides an array of statistics and facts about the instance of the application. For example, on the new landing page, you can now observe statistics about the underlying H2O Hydrogen Torch hardware and infrastructure utilized to run the application instance (for example, you can now monitor the current GPU and CPU load percentage)
  • Improvement: Several new UX improvements are available throughout the application while improving the user experience
  • Improvement: The application's tooltips are now synced with the documentation to provide more details about a particular tooltip



  • New: Now, H2O Hydrogen Torch supports the following problem types:
  • New: Directly from the H2O Hydrogen Torch UI, you can now deploy to H2O MLOps a built model. To learn more, see Deploy a model to H2O MLOps (through the H2O Hydrogen Torch UI)
  • New: For the following experiment types (supported problem types), you can now view random visual samples of the training data after the experiment's augmentations (for NLP models, H2O Hydrogen Torch visualizes in these visual samples the raw input text and how it is tokenized and padded): Text classification, text regression, text metric learning, and text sequence to sequence. To learn more, see Train data insight
  • New: For the following supported problem types, H2O Hydrogen Torch now enables you to utilize/deploy a pre-trained model trained on zero epochs (where H2O Hydrogen Torch does not train the model and the pretrained model (experiment) can be deployed as-is):
    • Speech recognition
    • Text sequence to sequence
    • Text span prediction
  • New: In the Validation prediction insights tab, H2O Hydrogen Torch now displays a sample's image name for computer vision insights (samples)
  • New: For the following supported problem types, H2O Hydrogen Torch can export a model to an open neural network exchange (ONNX) format:
  • New: You can now import datasets for inference only. To do so, when defining the setting for an experiment, set the Train dataframe setting to None while setting the Test dataframe setting to the relevant dataframe (as a result, H2O Hydrogen Torch utilizes the relevant dataset for predictions and not for training)


  • New: The Wheel (.whl) package of a Python scoring pipeline now includes a Dockerfile that enables you to build a dedicated Docker image that can include all requirements to run the scoring pipeline for model inference (production)