Skip to content
Judgment Labs
Esc
↑↓navigate↵open⌘Jpreview
On this page

OfflineTestResult

The outcome of an offline test run.

Returned by client.offline_tests.run(). Contains the per-example scoring results plus run-level metadata.

Attributes

PropType
test_run_id?str

The test run ID.

Typestr
status?str

Final run status.

Typestr
ui_results_url?Optional[str]

Link to the results page in the dashboard.

TypeOptional[str]
DefaultNone
results?List[ScoringResult]

One ScoringResult per dataset example, with per-judge ScorerData entries. When a pass_condition_fn was supplied, each ScorerData.success carries the per-row outcome.

TypeList[ScoringResult]
Defaultfield(default_factory=list)
agent_offline_trace_ids?Dict[str, str]

Mapping of example ID to the offline trace produced by the agent entrypoint (agent testing only).

TypeDict[str, str]
Defaultfield(default_factory=dict)
passed?Optional[bool]

Whether every row passed its pass condition.Returns None when no pass condition was evaluated.

TypeOptional[bool]

Was this page helpful?