PromptEvalCalibrationEntry
One sampled verdict of a calibration (1.63.0): the evaluation Call by id, its rubric items and its label. Ids and keys only.
An object with 9 fields.
Fields
| Field | Type | Required | Description |
|---|---|---|---|
arm | PromptEvalArm | yes | |
call_id | CallId | yes | The evaluation Call whose transcript is labelled, read with get_prompt_eval_call and include=transcript. |
index | integer | yes | The entry, by its place in the sample; a label names it. |
items | array of PromptEvalCalibrationItem | yes | |
label_id | PromptEvalLabelId or null | yes | Its label; null until labelled. |
repetition | integer | yes | |
run_id | PromptEvalRunId | yes | |
scenario_id | PromptEvalScenarioId | yes | |
scenario_revision | integer | yes |