evals

Compare Eval

The eval's candidate against its baseline; the verdict is null, with the reason, on fewer than min_trials trials or without a calibrated judge.

get/evals/{eval_id}/compare

Request

  • Base URL
    https://api.os3robotics.com (the document declares no server; this is the origin it was published from)
  • URL
    https://api.os3robotics.com/evals/{eval_id}/compare
  • Auth
    • HTTP bearer
    • or HTTP bearer

Path parameters

eval_idstring uuid required

Response

Successful Response

eval_idstring uuid required
workspace_idstring uuid required
candidate_refstring required
baseline_refstring nullable required
eval_set_refstring required
ninteger required
winsinteger required
lossesinteger nullable required
tiesinteger nullable required
win_ratenumber required
ci_lowernumber required
ci_uppernumber required
strengthnumber nullable required
flip_ratenumber nullable required
p95_latency_msnumber required
min_trialsinteger required
verdict'candidate_better' | 'baseline_better' | 'no_difference' nullable required
reasonstring required
accuracyobject nullable

Changes

No recorded changes to this operation across all 1 revision of this API.