XMLWordPrintable

    • Icon: Epic Epic
    • Resolution: Done
    • Icon: Normal Normal
    • None
    • None
    • None
    • None
    • LLM load testing enhancements
    • Inference, RHOAI
    • Not Selected
    • False
    • False
    • Hide

      None

      Show
      None
    • 0% To Do, 0% In Progress, 100% Done

      Epic Goal

      Improve our load testing tool llm-load-test and related automation to keep up with best practices / state of the art:

      • Use a dataset that is representative of a wider set of use cases – input/output tokens ranging from 0 to 4096 tokens, with option to configure the bounds for each test
      • Measure time to first token and time per output token (TTFT, TPOT)
      • Develop a load generator that can be used to test models in various runtimes with different interfaces (GRPC, HTTP)
      • Use mlcommons loadgen to drive load
      • Potentially integrate mlcommons loadgen into our load testing tool

      Why is this important?

      Scenarios

      1. ...

      Acceptance Criteria

      • CI - MUST be running successfully with tests automated
      • Release Technical Enablement - Provide necessary release enablement details and documents.
      • ...

      Dependencies (internal and external)

      1. ...

      Previous Work (Optional):

      Open questions::

      Done Checklist

      • CI - CI is running, tests are automated and merged.
      • Release Enablement <link to Feature Enablement Presentation>
      • DEV - Upstream code and tests merged: <link to meaningful PR or GitHub Issue>
      • DEV - Upstream documentation merged: <link to meaningful PR or GitHub Issue>
      • DEV - Downstream build attached to advisory: <link to errata>
      • QE - Test plans in Polarion: <link or reference to Polarion>
      • QE - Automated tests merged: <link or reference to automated tests>
      • DOC - Downstream documentation merged: <link to meaningful PR>

              dagray@redhat.com David Whyte-Gray
              dagray@redhat.com David Whyte-Gray
              David Whyte-Gray, Nikhil Palaskar
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                Resolved: