overview
the EvaluationResponse
class represents the results of evaluating a test call. it contains a list of evaluation results and any additional data from the evaluation process.
constructor
EvaluationResponse(
evaluation_results: List[EvaluationResult],
extra_data: Dict[str, Any]
)
parameters
evaluation_results
(List[EvaluationResult]): list of individual evaluation results
extra_data
(Dict[str, Any]): additional data from the evaluation process, such as fixa-observe urls
example usage
response = EvaluationResponse(
evaluation_results=[
EvaluationResult(
name="order_success",
passed=True,
reason="the order was successfully placed and confirmed"
),
EvaluationResult(
name="price_confirmed",
passed=False,
reason="the price was not mentioned during the conversation"
)
],
extra_data={
"fixa_observe_call_url": "https://www.fixa.dev/observe/calls/123"
}
)
notes
- returned by both LocalEvaluator and CloudEvaluator
- evaluation_results contains the pass/fail status and reasoning for each evaluation criterion
- extra_data may contain different information depending on the evaluator used
- with cloudevaluator, extra_data includes the fixa-observe url for detailed analysis
Responses are generated using AI and may contain mistakes.