📄 Step By Step Reasoning Coherency¶

metrics.llm_as_judge.direct.criteria.step_by_step_reasoning_coherency

CriteriaWithOptions(
    name="step_by_step_reasoning_coherency",
    description="On a scale of 1 (sounds like nonsense) to 5 (easy to parse), does the whole generated response make sense? (Ie, does it sound understandable/non-contradictory/sensible, even if it fails to address the context?)",
    prediction_field="generated response",
    context_fields=[
        "question",
        "premise",
        "hypothesis",
        "correct answer",
    ],
    options=[
        CriteriaOption(
            name="1",
            description="",
        ),
        CriteriaOption(
            name="2",
            description="",
        ),
        CriteriaOption(
            name="3",
            description="",
        ),
        CriteriaOption(
            name="4",
            description="",
        ),
        CriteriaOption(
            name="5",
            description="",
        ),
    ],
    option_map={
        "1": 0.0,
        "2": 0.25,
        "3": 0.5,
        "4": 0.75,
        "5": 1.0,
    },
)

[source]

from unitxt.llm_as_judge_constants import CriteriaOption

Explanation about CriteriaWithOptions¶

Criteria used by DirectLLMJudge to run evaluations.

Explanation about CriteriaOption¶

A criteria option.