Model Serving

Share

Model serving refers to the way trained models are made available for others to use. Choosing a model serving strategy can be the first step in model deployment, where factors such as user expectations, production requirements, business rules, and existing technologies are considered. There are four common model serving tactics, namely, batch inference, model as a service, online model as a service, and edge deployment.

Filter terms by

Related resources

deci-infery-updates-blog-featured
Engineering
pytorch-training-sg-new-features-featured-2
Open Source
new-hardware-support-featured
Engineering