ML Create and Verify Inference Endpoint
A 3-step recipe with Amazon Neptune. We can help you get access.
Recipe · 3 steps
ML Create and Verify Inference Endpoint
with Amazon Neptune · Artificial Intelligence Management
Create an inference endpoint, poll until InService, and confirm it is listed.
- Create an inference endpoint backed by the trained model and capture the endpoint id.
- Poll the endpoint status. Retry while it is not yet InService and continue once it reaches InService.
- List the active inference endpoints to confirm the newly created endpoint is registered.
What you'll need to have ready
- mlModelTrainingJobId: Job ID from a completed model training job.
- instanceType: ML instance type for the inference endpoint.
- instanceCount: Minimum number of EC2 instances to deploy.
For your developer
Arazzo workflow amazon-neptune:ml-create-inference-endpoint, 3 operations: createInferenceEndpoint → getInferenceEndpointStatus → listInferenceEndpoints. Specs on APIs.io.