Skip to content
This repository has been archived by the owner on Apr 2, 2022. It is now read-only.

Latest commit

 

History

History

training

README.md

AI Platform Training

The AI Platform training service allows you to train models using a wide range of different customization options.

You can use the following features:

  • Running a training job using many different machine types
  • Use custom Docker containers with your preferred ML Framework
  • Use GPUs
  • Use TPUs
  • Hyperparameter tuning
  • Distributed training

You can also select different ways to customize your training application. You can submit your input data for AI Platform to train using a built-in algorithm (beta). If the built-in algorithms do not fit your use case, you can submit your own training application to run on AI Platform, or build a custom container (beta) with your training application and its dependencies to run on AI Platform.

ML Frameworks

This folder covers different functionality available in different frameworks:

Samples

This folder covers different functionality available AI Platform Training, the following samples reflect the available features in AI Platform:

AI Platform Training

The AI Platform training service allows you to train models using a wide range of different customization options. You can select many different machine types to power your training jobs, enable distributed training, use hyperparameter tuning, and accelerate with GPUs and TPUs.

  • TensorFlow

    • base Standard code to perform AI Platform Training using TensorFlow Estimators using CPU.
    • census TF Keras A Binary classification model using with TF Keras and AI Platform Trainining.
    • TPU Uses Cloud TPU for Model Training.
    • Hyperparameter tuning Use Hyperparameter tuning.
    • Distributed training Uses Distributed Training using TensorFlow Distribution strategy.

  • scikit-learn

    • base Standard code to perform AI Platform Training using Sci-kit learn using CPU.

  • XGBoost

    • base Standard code to perform AI Platform Training using XGBoost.

AI Platform Training - Custom Containers

Containers on AI Platform is a feature that allows you to run your application within a Docker image. You can build your own custom container to run jobs on AI Platform, using ML frameworks and versions as well as non-ML dependencies, libraries and binaries that are not otherwise supported on AI Platform.

  • PyTorch Train a PyTorch model in AI Platform
  • Horovod How to run Horovod on AI Platform.
  • ResNet How to run custom containers using Hyperparameter tuning.

AI Platform Prediction

Note: These examples use the Chicago Taxi Trips Dataset released by the City of Chicago. Read more about the dataset in Google BigQuery.

Templates

How to contribute?

We welcome external sample contributions! To learn more about contributing new samples, checkout our CONTRIBUTING.md guide. Please feel free to add new samples that are built in notebook form or code form with a README guide.

Want to contribute but don't have an idea? Check out our Sample Request Page and assign the issue to yourself so we know you're working on it!

Documentation

We host AI Platform documentation here