Skip to main content
Version: Next 🚧

H2O MLOps Python client

The H2O MLOps Python client lets you use the H2O MLOps API from your Python application. This guide describes how you can install the H2O MLOps Python client, connect to H2O MLOps and carry out tasks using the Python client. After successful installation, you can interact with the H2O MLOps API via the H2O MLOPs gRPC Gateway. For more information, see H2O MLOps gRPC Gateway.

The H2O MLOps Python client documentatin is organized into the following sections:

  • Installation

    The Python client can be easily installed to get started with H2O MLOps programmatically. Follow our installation guide to set up the client in your environment and prepare for integration with the platform.

  • Getting started

    The Python client makes it easy to connect to H2O MLOps, deploy models, and score data from your Python code. Follow our quickstart guide for an end-to-end workflow to create a project, register a model, deploy it, and score against the deployment.

  • Tutorials

    • Connect to H2O MLOps

      Learn recommended methods to establish a secure connection to the H2O MLOps using the Python client.

    • Manage Projects

      Discover how to create, manage, and organize your machine learning projects in H2O MLOps using the Python client with better collaboration and resource management.

    • Manage Experiments

      Track and manage your machine learning experiments using H2O MLOps Python client, including experiment tags. It also describes experiment properties and how to compute Kubernetes options.

    • Handle artifacts

      Learn how to use the Python client to add, retrieve, update, delete, or convert artifacts to strings or dictionaries for an H2O MLOps entity.

    • Manage Models

      Explore comprehensive model management capabilities, including versioning through code.

    • Batch scoring

      Implement efficient batch prediction workflows for processing large volumes of data with your deployed models.

    • Monitoring setup

      Configure comprehensive monitoring for your deployed models to track performance, data drift, and operational metrics.

  • Changelog

    Stay up-to-date with the latest features, improvements, and bug fixes in the Python client through our detailed changelog, which documents version history and updates.

Feedback