Machine Learning eXchange (MLX)

Data and AI Assets Catalog and Execution Engine

Allows upload, registration, execution, and deployment of:

AI pipelines and pipeline components
Models
Datasets
Notebooks

Additionally it provides:

Automated sample pipeline code generation to execute registered models, datasets and notebooks
Pipelines engine powered by Kubeflow Pipelines on Tekton, core of Watson AI Pipelines
Components registry for Kubeflow Pipelines
Datasets management by Datashim
Preregistered Datasets from Data Asset Exchange (DAX) and Models from Model Asset Exchange (MAX)
Serving engine by KFServing
Model Metadata schemas

For more details about the project check out this announcement blog post.

1. Deployment

For a simple up-and-running MLX with asset catalog only, we created a Quickstart Guide using Docker Compose.

For a slightly more resource-hungry local deployment that allows pipeline execution, we created the MLX with Kubernetes in Docker (KIND) deployment option.

For a full deployment, we use Kubeflow Kfctl tooling.

2. Access the MLX UI and import Assets to the Catalog

By default, the MLX UI is available at http://<cluster_node_ip>:30380/mlx/

If you deployed on a Kubernetes cluster or using OpenShift, run the following and look for the External-IP column to find the public IP of a node.

kubectl get node -o wide

For information on how to import data and AI assets using MLX's catalog importer, use this guide.

3. Use MLX

For information on how to use MLX and create assets check out this guide.

4. How to Contribute

For information about adding new features, bug fixing, communication or UI and API setup, refer to this document.

5. Troubleshooting

MLX Troubleshooting Instructions

Join the Conversation

Slack: @lfaifoundation/ml-exchange
Mailing lists:
- MLX-Announce for top-level milestone messages and announcements
- MLX-TSC for top-level governance discussions and decissions
- MLX-Technical-Discuss for technical discussions and questions

machine-learning-exchange/mlx