How-To · Resource Management

How-To: Monitor Worker Performance

This guide shows how to read a model's query times, its running worker and its usage in the Lattica Console, and how to check workers with the Python SDK.

Before you start

Load

The average time, in seconds, the model took per query over the last 5 minutes. 0 means no queries in that time.

Average

The average time, in seconds, per query over the last year. It's your baseline.

Compare them

A Load well above Average means queries are running slower than usual right now, for example under a burst of requests.

Read the metrics

Console

console.lattica.ai/models1 / 6

1

Step 1 of 6:

Load

Load on the model's card is the average query time, in seconds, over the last 5 minutes. Here, 3.12 seconds.

Python SDK

Check a model's running worker with studio.workers.active, and list past worker sessions with studio.workers.list_sessions. studio.workers.display prints them as a table.

check_workers.py

import os

from lattica_studio import LatticaStudio

studio = LatticaStudio(os.environ["LATTICA_LICENSE_KEY"])
model = studio.models.get_by_name("DermAI Spotter")

# The model's running worker, with its status and start time.
studio.workers.display(studio.workers.active(model.id))

# The model's worker sessions in October 2026 (ISO 8601 dates).
sessions = studio.workers.list_sessions(
    model_id=model.id,
    from_date="2026-10-01T00:00:00Z",
    to_date="2026-10-31T23:59:59Z",
)
studio.workers.display(sessions)

Load and Average are shown only in the Console. The SDK reports workers and their status, not query times.