RunComparisonSummary
junjo.studio.RunComparisonSummary
Kind: Class
SDK version: 0.69.0
Documentation channel: Stable release
Outcome counts for one side of a comparison.
Members
Section titled “Members”total: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Number of selected cases on this side of the comparison, including errors and queued attempts.
judged
Section titled “judged”judged: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Attempts judged passed or failed; excludes queued attempts and operational errors.
passed
Section titled “passed”passed: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Attempts satisfying the evaluator criteria.
failed
Section titled “failed”failed: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Attempts that did not satisfy the evaluator criteria.
error: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Attempts whose target or evaluator could not complete successfully.
queued
Section titled “queued”queued: int = Field(ge=0, le=MAX_CASES_PER_DATASET)Attempts that have not received a terminal result.
pass_rate
Section titled “pass_rate”pass_rate: float | None = Field(default=None, ge=0.0, le=1.0)Passed divided by judged attempts, as a fraction from 0 to 1; None when nothing was judged.