Metrics
Metrics derive actionable insights from observability and evaluation traces.
Collect Actionable Metrics
tmam provides seamless OpenTelemetry auto-instrumentation for a wide range of LLM providers, frameworks, and vector databases. This allows you to effortlessly capture meaningful metrics that reveal the performance and behavior of your AI applications.
This documentation walks you through configuring metrics collection, understanding semantic conventions, and interpreting metric attributes—giving you the tools to improve observability and make data-driven decisions.
Disable Metrics
You have the option to disable the collection of metrics if needed. By default, metrics collection is enabled.
Example:
# Disable metrics collection
tmam.init(disable_metrics=True)