Building your ML service with AI Hub

You can reuse the assets on AI Hub to implement your machine learning (ML) service. Here is an overview of some of the ways that you can reuse the assets on AI Hub as components of your ML service:

  • Kubeflow pipelines and components: Kubeflow pipelines are end-to-end machine learning (ML) workflows based on containers. Pipelines are built from components which are self-contained sets of code, packaged as container images. Each component performs a step in the ML workflow, such as preprocessing, data transformation, or training a model.

    You can reuse the pipelines on AI Hub to train a model, or you can reuse pipeline components to build a custom ML workflow.

  • TensorFlow modules: A TensorFlow module is a reusable piece of a TensorFlow graph. With transfer learning, you can use TensorFlow modules to preprocess input feature vectors, or you can incorporate a TensorFlow module into your model as a trainable layer. This can help you train your model faster, using a smaller dataset, while improving generalization. Learn how to use a TensorFlow module from AI Hub.

  • Trained models: Trained models are a representation of what an ML system has learned from its training data. Models on AI Hub can be reused to preprocess input feature vectors, or they can be deployed as ML services. You can download a model from AI Hub to experiment with it locally, or you can deploy a model to AI Platform for predictions. To learn more about using a trained model, read the guide to using a trained model.

  • Services: The collection of services on AI Hub includes APIs, building blocks, ML services, and infrastructure for training and prediction. You can use the ML services and APIs on AI Hub to implement your solution, or you can use services to address key challenges, such as data labeling. For more information, see the guide to finding services and APIs on AI Hub.

  • VM images: You can use the virtual machine (VM) images on AI Hub to quickly implement the infrastructure to train an ML model. To learn more about the VM images on AI Hub, read the guide to finding and using a VM image.

What's next