llm_as_judge.py

from typing import Any, Dict, List, Literal, Optional

from .api import evaluate, produce
from .artifact import Artifact, settings
from .inference import InferenceEngine, OpenAiInferenceEngine
from .metrics import BulkInstanceMetric
from .operator import SequentialOperator


class LLMAsJudge(BulkInstanceMetric):
    """LLM as judge based metric class for evaluating correctness.

    Attributes:
        main_score (str): The main score label used for evaluation.
        task (Literal["rating.single_turn"]): The type of task the llm-as-judge runs. This defines the output and input
         format of the jude model.
        template (str): The template used when generating inputs for the judge llm.
        format (str): The format used when generating inputs for judge llm.
        system_prompt (str): The system prompt used when generating inputs for judge llm.
        strip_system_prompt_and_format_from_inputs (bool): Whether to strip the system prompt and formatting from the
         inputs that the models that is being judges received, when they are inserted to the llm-as-judge prompt.
        inference_model (InferenceEngine): the module that creates the inference of the judge llm.
        reduction_map (dict): A dictionary specifying the reduction method for the metric.
        batch_size (int): The size of the bulk.
    """

    main_score: str = "llm_as_judge"
    task: Literal["rating.single_turn", "single_turn_with_reference"]
    template: str
    format: Optional[str] = None
    system_prompt: Optional[str] = None
    strip_system_prompt_and_format_from_inputs: bool = True
    inference_model: InferenceEngine
    reduction_map: Optional[Dict[str, List[str]]] = None
    batch_size: int = 32

    def _get_input_instances(self, task_data: List[Dict]) -> List:
        if self.strip_system_prompt_and_format_from_inputs:
            instances = []
            for task_data_instance in task_data:
                template = task_data_instance["metadata"]["template"]
                instance = SequentialOperator(
                    steps=[template, "formats.empty"]
                ).process_instance(
                    {"inputs": task_data_instance, "outputs": task_data_instance}
                )
                instances.append(instance["source"])
                """
                We also have access to: instance["target"]
                                        instance["references"]
                """
            return instances
        return [t["source"] for t in task_data]

    def _get_instance_for_judge_model(
        self, input_instances: List[str], predictions: List, references: List
    ) -> List[Dict]:
        if self.task == "rating.single_turn":
            instances = [
                {
                    "question": input_instance,
                    "answer": prediction,
                    "rating": 5.0,  # This is a dummy value that is not used in practice
                }
                for input_instance, prediction, reference in zip(
                    input_instances, predictions, references
                )
            ]
        elif self.task == "rating.single_turn_with_reference":
            instances = [
                {
                    "question": input_instance,
                    "answer": prediction,
                    "reference_answer": reference,
                    "rating": 5.0,  # This is a dummy value that is not used in practice
                }
                for input_instance, prediction, reference in zip(
                    input_instances, predictions, references
                )
            ]
        else:
            raise NotImplementedError(
                f"Error in 'LLMAsJudge' metric. {self.task} is not a supported task type."
            )
        return instances

    def prepare(self):
        super().prepare()
        if self.reduction_map is None:
            self.reduction_map = {"mean": [self.main_score]}

        supported_tasks = ["rating.single_turn", "rating.single_turn_with_reference"]
        assert self.task in supported_tasks, (
            f"Error in 'LLMAsJudge' metric. {self.task} is not a supported task type."
            f"The supported tasks types are: {', '.join(supported_tasks)}."
        )

        if isinstance(self.inference_model, OpenAiInferenceEngine):
            if self.format:
                raise ValueError(
                    "Error in 'LLMAsJudge' metric. Inference model 'OpenAiInferenceEngine' does "
                    "not support formatting. Please remove the format definition from the recipe"
                    " (OpenAi Chat API take care of the formatting automatically)."
                )
            if self.system_prompt:
                raise ValueError(
                    "Error in 'LLMAsJudge' metric. Inference model 'OpenAiInferenceEngine' does "
                    "not support system prompt. Please remove the system_prompt definition from the recipe"
                    " (Current implementation of Unitxt does not support this."
                    " Support will be added in future updates)."
                )

    def compute(
        self,
        references: List[List[Any]],
        predictions: List[Any],
        task_data: List[Dict],
    ) -> List[Dict[str, Any]]:
        input_instances = self._get_input_instances(task_data)
        instances = self._get_instance_for_judge_model(
            input_instances, predictions, references
        )

        card = f"cards.dynamic_cards_for_llm_judges.{self.task}"
        recipe_args = {
            "card": card,
            "template": self.template,
            "demos_pool_size": 0,
            "num_demos": 0,
            "__type__": settings.default_recipe,
        }
        if self.system_prompt:
            recipe_args["system_prompt"] = self.system_prompt
        if self.format:
            recipe_args["format"] = self.format
        recipe = Artifact.from_dict(recipe_args)
        dataset = produce(instances, recipe)
        verdicts = self.inference_model.infer(dataset)
        meta_scores = evaluate(predictions=verdicts, data=dataset)
        return [
            {
                self.main_score: instance["processed_prediction"],
                "judge_raw_output": verdict,
            }
            for instance, verdict in zip(meta_scores, verdicts)
        ]