querying-mlflow-metrics

What problem does it solve?

This Skill provides a streamlined way to query and aggregate metrics from MLflow tracking servers, enabling efficient monitoring and analysis of machine learning experiments.

Core Features & Use Cases

• Metric Aggregation: Fetch various metrics like token usage, latency, and trace counts with specified aggregations (SUM, AVG, COUNT, etc.).
• Trend Analysis: Analyze metrics over time by bucketing them into hourly or daily intervals.
• Dimensional Grouping: Group results by dimensions such as trace name or status for deeper insights.
• Use Case: Quickly check the average token usage and latency for your model's traces over the last 24 hours to identify performance bottlenecks.

Quick Start

Fetch the sum and average of total tokens from MLflow server http://localhost:5000 for experiment ID 1.