Skip to content

harness: add inherited presets and auto worker sizing#1539

Open
jioffe502 wants to merge 4 commits intoNVIDIA:mainfrom
jioffe502:retriever-harness-preset-inheritance
Open

harness: add inherited presets and auto worker sizing#1539
jioffe502 wants to merge 4 commits intoNVIDIA:mainfrom
jioffe502:retriever-harness-preset-inheritance

Conversation

@jioffe502
Copy link
Collaborator

Description

  • add preset inheritance to reduce duplicated harness tuning
  • limit auto mode to worker counts and document the user story
  • wire harness tuning into batch requested plan refreshes

Checklist

  • I am familiar with the Contributing Guidelines.
  • New or existing tests cover these changes.
  • The documentation is up to date with these changes.
  • If adjusting docker-compose.yaml environment variables have you ensured those are mimicked in the Helm values.yaml file.

- add preset inheritance to reduce duplicated harness tuning
- limit auto mode to worker counts and document the user story
- wire harness tuning into batch requested plan refreshes

Signed-off-by: Jacob Ioffe <jioffe@nvidia.com>
@jioffe502 jioffe502 requested a review from a team as a code owner March 10, 2026 18:55
@jioffe502 jioffe502 requested a review from drobison00 March 10, 2026 18:55
return {key: value for key, value in env_vars.items() if isinstance(value, str)}


def _normalize_requested_plan_int(value: Any) -> int | None:
Copy link
Collaborator

Choose a reason for hiding this comment

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

What purpose do these added functions in batch.py serve? Can we move them elsewhere?

self._extract_html_kwargs: Dict[str, Any] = {} # noqa: F821
self._use_nemotron_parse_only: bool = False

def _refresh_requested_plan(self, batch_tuning: dict[str, Any]) -> None:
Copy link
Collaborator

Choose a reason for hiding this comment

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

I think most of this can already be accomplished with the existing functions in ray_resource_hueristics.py but if they are needed for some reason lets move them there and name them appropriately not not clutter the core logic

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