📄 Judge Simplified No Context¶

templates.rag_eval.answer_correctness.judge_simplified_no_context

InputOutputTemplateWithCustomTarget(
    input_format="Question: {question}

Ground-truth answer: {ground_truths}

Prediction: {answer}
",
    output_format="{is_correct}",
    postprocessors=[
        "processors.take_first_word",
        "processors.lower_case",
        "processors.predictions_yes_1_else_0",
        "processors.cast_to_float_return_zero_if_failed",
    ],
    reference="{number_val}",
    target_prefix="Answer: ",
    instruction="You are given a question, the corresponding ground-truth answer and a prediction from a model. Compare the "Ground-truth answer" and the "Prediction" to determine whether the prediction is a correct and complete answer to the question. Answer with only yes/no.
",
)
[source]

References: processors.cast_to_float_return_zero_if_failed, processors.predictions_yes_1_else_0, processors.take_first_word, processors.lower_case