h2oGPTe icon

h2oGPTe

h2oGPTe is an AI-powered search assistant for your internal teams to answer questions gleaned from large volumes of documents, websites and workplace content.

Actions198

Overview

This node operation, "Returns Performance Statistics Grouped by Models," fetches performance statistics aggregated by different AI models. It is useful for monitoring and analyzing how various models perform over a specified time interval. This can help teams understand model efficiency, identify trends, or detect issues in model behavior.

Practical examples include:

  • Tracking the accuracy or response times of multiple language models over the last 24 hours.
  • Comparing performance metrics across different deployed models to decide which to prioritize.
  • Generating reports on model usage and effectiveness for stakeholders.

Properties

Name Meaning
Interval The length of the time interval for which performance stats are obtained. The interval ends at the current time. Example: "24 hours".

Output

The output contains JSON data representing performance statistics grouped by models. This typically includes metrics such as usage counts, success rates, error rates, or other relevant performance indicators for each model within the specified interval.

No binary data output is indicated.

Dependencies

  • Requires an API key credential for authentication with the external service.
  • The node sends a GET request to the endpoint /stats/performance_by_model with the interval query parameter.
  • The base URL for the API is configured via credentials and must be set correctly in the node's credentials configuration.

Troubleshooting

  • Missing or invalid API key: Ensure that the API key credential is properly configured and valid.
  • Invalid interval format: The interval property must be a valid string representing a time duration (e.g., "24 hours"). Incorrect formats may cause errors.
  • API endpoint errors: If the API returns errors, verify network connectivity and that the base URL is correct.
  • Empty or unexpected results: Check if the interval covers a period where data exists; otherwise, the response might be empty.

Links and References

  • No direct links provided in the source code.
  • For more details, consult the API documentation of the service providing the /stats/performance_by_model endpoint.

Discussion