Start a TAO endpoint and query your model.
What it does
Manage TAO inference containers so users can start an endpoint, submit model requests, and stop the service.
When to use it
Use to run inference on a TAO checkpoint through a microservice container, deploy an inference endpoint, or stop a running container.
How to use it
Give it a request to start, query, or stop an inference service; it collects settings, asks for sampling choices before queries, and reports service details or results.
Uses
What you need · 5
Uses Python templates for payload construction, registry management, readiness checks, and stop/request flows; no runtime version is specified.
Inference requires at least one GPU; the requested GPU count defaults to one.
Access · 3
This skill
TAO Inference Microservice
Sends prompts and media for inference
{network_arch}-inference-microservice
Starts and stops inference containers