Describing a Model Endpoint using API
Consider the instructions for describing a Model Endpoint using API.
Use the following command to describe a given model endpoint. The
describeEndpoint API shows you all the detailed configuration and
runtime snapshot of the endpoint. The output below has been elided for brevity.
curl -s -H "Authorization: Bearer ${CDP_TOKEN}" -H "Content-Type: application/json" https://${DOMAIN}/api/v1alpha1/describeEndpoint -d '{"namespace": "serving-default", "name": "[***ENDPOINT_NAME***]"}'|jq
You will receive a response similar to the following:
{
"namespace": "serving-default",
"name": "[***ENDPOINT_NAME***]",
"url": "https://[***DOMAIN***]/namespaces/serving-default/endpoints/[***ENDPOINT_NAME***]/v2/models/[***MODEL_ID***]/infer",
"conditions": [
{
"type": "IngressReady",
"status": "True",
"severity": "",
"last_transition_time": "1726713090",
"reason": "",
"message": ""
},
"observed_generation": 1,
"replica_count": 1,
...
"autoscaling": {
"min_replicas": "1",
"max_replicas": "80",
"autoscalingconfig": {
"metric": "concurrency",
"target": "20",
}
},
...
"api_standard": "oip",
"has_chat_template": false,
"metricFormat": "triton",
"task": "inference"
}