Skip to content

Conversation

jaredoconnell
Copy link
Collaborator

Summary

This PR gets the CSV output to a state comparable to pre-refactor.

Details

Implements the functions to export the required data to the CSV format.

The goal is to include the required information in the CSV without cluttering it, but also without creating too much of a burden to the future maintainers resulting from referencing specific schema elements.

The following columns are new:

  • Profile (an entire JSON dump of the profile)
  • Backend (the entire JSON dump of the internal data structure)
  • Generator Data

You can view these files in the attached output generated by the following test:
guidellm benchmark run --output-path result_7.csv --max-seconds 2 --target=http://localhost:8000 --data "prompt_tokens=256,output_tokens=128" --rate-type constant --rate 1 --output-formats csv

result_7.csv

Test Plan

Run GuideLLM with the following additional args: --output-path result.csv --output-formats csv

The generated file should have all info required.

Related Issues

This is a part of the scheduler refactor.


  • "I certify that all code in this PR is my own, except as noted below."

Use of AI

  • Includes AI-assisted code completion
  • Includes code generated by an AI application
  • Includes AI-generated tests (NOTE: AI written tests should have a docstring that includes ## WRITTEN BY AI ##)

Signed-off-by: jaredoconnell <joconnel@redhat.com>
@jaredoconnell jaredoconnell changed the title Complete CSV output [GuideLLM Refactor] Complete CSV output Sep 29, 2025
sjmonson added a commit that referenced this pull request Sep 30, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant