Skip to content

docs: steady-state metrics reporting design proposal - #447

Open
nv-alicheng wants to merge 1 commit into
mainfrom
doc/alicheng-steady-state-design
Open

docs: steady-state metrics reporting design proposal#447
nv-alicheng wants to merge 1 commit into
mainfrom
doc/alicheng-steady-state-design

Conversation

@nv-alicheng

Copy link
Copy Markdown
Collaborator

Post-processing step (and ad-hoc CLI over events.jsonl) that reports steady-state metrics — super-pass windowing, guarded drain-tail cut, CoV ensemble + trend gate (drift up/down), hairball, edge-case handling.

What does this PR do?

Adds a design doc for a new steady-state window detection and metrics reporting feature

Type of change

  • Bug fix
  • New feature
  • Documentation update
  • Refactor/cleanup

Related issues

cc @arekay-nv to tag any related issues

Testing

  • Tests added/updated
  • All tests pass locally
  • Manual testing completed

Checklist

  • Code follows project style
  • Pre-commit hooks pass
  • Documentation updated (if needed)

Post-processing step (and ad-hoc CLI over events.jsonl) that reports
steady-state metrics — super-pass windowing, guarded drain-tail cut,
CoV ensemble + trend gate (drift up/down), hairball, edge-case handling.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@nv-alicheng
nv-alicheng requested a review from a team August 13, 2026 23:55
@github-actions

Copy link
Copy Markdown

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅

Averaging over these transients understates throughput and overstates tail
latency. The magnitude is workload-dependent and can be large for the tail: in
experiments over recorded runs (single-turn concurrency, offline/max-throughput,
Poisson, and multi-turn agentic, including runs on Hecate), the reported p99 TTFT

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Remove the name of the cluster (and overwrite the commit please)

Comment on lines +64 to +66
a run killed by `SIGKILL`/OOM before the log is flushed yields a truncated log;
the step must degrade to a best-effort result with a status flag rather than
fail (§5.6).

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe I am not reading it in the right way - but I thought it should fail if the original run fails?

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.

2 participants