📄 Fluency¶

metrics.llm_as_judge.direct.criteria.fluency

CriteriaWithOptions(
    name="fluency",
    description="Is the response fluent? The response contains sentences that are well-written and grammatically correct. Consider the quality of the individual sentences and measure the extent to which they are fluent.",
    prediction_field="response",
    context_fields=[],
    options=[
        CriteriaOption(
            name="1",
            description="The response is not fluent at all.",
        ),
        CriteriaOption(
            name="2",
            description="The response is somewhat fluent.",
        ),
        CriteriaOption(
            name="3",
            description="The response is fluent.",
        ),
        CriteriaOption(
            name="4",
            description="The response is very fluent, grammatically correct and well-written.",
        ),
        CriteriaOption(
            name="5",
            description="The response is exceptionally fluent, grammatically correct, and well-written.",
        ),
    ],
    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.