NGC | Catalog
CatalogCollectionsNext Item Prediction Workflow

Next Item Prediction Workflow

For contents of this collection and more information, please view on a desktop device.
Logo for Next Item Prediction Workflow

Features

Description

Utilize session-based recommenders to increase accuracy of predictions when users interests are dynamic, specific to a shorter time frame (i.e., within a session) and when little or no user data is available.

Curator

NVIDIA

Modified

May 23, 2023
Containers
Sorry, your browser does not support inline SVG.
Helm Charts
Sorry, your browser does not support inline SVG.
Models
Sorry, your browser does not support inline SVG.
Resources
Sorry, your browser does not support inline SVG.

Overview

The next item prediction AI workflow shows how to use NVIDIA Merlin, an end-to-end framework for building high-performing recommender systems at scale. A session-based recommendation is the next-generation AI method that predicts the next action of a user – it predicts preferences from contextual user interactions for first-time, early, or anonymous online users.

The next item prediction workflow includes cloud-native Kubernetes services such as an example recommender system deployment based on NVIDIA Merlin, MLflow for model storage, Prometheus monitoring and Grafana dashboards. This reference solution provides a sample dataset as well as Python code and pre-built pipelines for data prep, training, and inference, packaged for deployment via Helm charts, and step-by-step instructions to help organizations quickly get started in building a recommendation system.

image1



Key benefits of the next item prediction AI workflow:

  • Merlin Transformers4Rec library for making state-of-the-art session-based recommendations available for recommender systems.

  • HuggingFace’s Transformers NLP library which makes it easy to use cutting-edge implementations of the latest NLP Transformer architectures in your recommendation systems.

  • Merlin Systems simplifies the deployment of recommender systems to NVIDIA Triton™ Inference Server.

  • Triton Inference Server maximizes inference performance with standardized and optimized model deployment and execution.

Getting Started

To get started, review the documentation linked below for more information on what is included in the workflow, and how to deploy and run the workflow.

Documentation

Additional Resources

Leverage the NVIDIA Merlin SDK to build your own AI-based session based recommendation solutions.

License

By pulling and using the containers or Helm Charts, you accept the terms and conditions of this End User License Agreement.