Reference solutions

Note

The managed MLflow integration with Databricks on Google Cloud requires Introduction to Databricks Runtime for Machine Learning 9.1 LTS or above.

Databricks provides a rich collection of tools to help you develop machine learning applications. However, finding these tools and understanding how best to apply them to your machine learning problems can be difficult. The reference solutions provided on this page show detailed examples of how you can use Databricks for some common machine learning applications.

Image processing and computer vision

This article contains a reference solution for distributed image model inference based on a common setup shared by many real-world image applications.

• Reference solution for image applications

Recommender system

This article contains a reference solution for how to build a wide-and-deep model for a recommender system, and how to serve it on Databricks.

• Build a wide-and-deep model in a recommender system

Natural language processing

This article describes performing distributed training and inference for natural language processing applications.

• Natural language processing

Hugging Face Transformers for NLP

This article describes how to use Hugging Face Transformers for NLP. Example notebooks demonstrate text summarization and fine-tuning text classification model for a GPU.

• Hugging Face NLP

Data labeling

This article describes options for labeling data.

• Data labeling