Documentation
-
Momentum
-
MLOps
-
Impulse EDW
-
- 2.1 Create a Warehouse
- 2.2 Edit Warehouse
- 2.3 Datasources In Warehouse
- 2.4 Ingesting Data Into Tables or Datasources
- 2.4.1 Ingesting From Momentum Data Pipeline
- 2.4.2 Uploading File Using Impulse UI
- 2.4.3 Ingesting From External File/Storage System
- 2.5 Add Data to Existing Tables
- 2.5.1 Update Existing Index
- 2.6 Delete Table Records (Rows)
- 2.7 Delete Tables or Datasources
- 2.8 Monitoring Indexing Tasks
- 2.9 View Datasource Stats
-
Inset BI
- Alerts and Reports
- Connecting to a new database
- Registering a new table
- Creating charts in Explore view
- Manage access to Dashboards
-
- Articles coming soon
-
- Articles coming soon
-
- Articles coming soon
-
- Articles coming soon
-
- Articles coming soon
-
APIs
- Articles coming soon
< All Topics
Print
Introduction
MLOps stands for Machine Learning Operations. MLOps streamlines the process of deploying ML models to production, maintaining and monitoring them. MLOps also provides a continuous integration and continuous deployment (CI/CD) for machine learning modeling, testing, and deployment.
MLOps allows deploying models trained using Momentum as well as models built using third party tools, systems, libraries and programming languages. Third party models must comply with one of the following standards:
- PMML or Predictive Model Markup Language
- ONNX or Open Neural Network Exchange
- TensorFlow
ML models can be deployed to MLOps in any of the following ways:
- Momentum ML UI
- MLOps UI
- Restful API
Table of Contents