judgeval 0.0.32__py3-none-any.whl → 0.0.33__py3-none-any.whl

judgeval/scorers/judgeval_scorers/local_implementations/summarization/summarization_scorer.py

@@ -1,551 +0,0 @@
-from typing import List, Optional, Union
-import asyncio
-
-from judgeval.constants import APIScorer
-from judgeval.scorers.utils import (
-    get_or_create_event_loop,
-    scorer_progress_meter,
-    create_verbose_logs,
-    parse_response_json,
-    check_example_params
-)
-from judgeval.scorers import JudgevalScorer
-from judgeval.judges import JudgevalJudge
-from judgeval.judges.utils import create_judge
-from judgeval.data import Example, ExampleParams
-from judgeval.scorers.judgeval_scorers.local_implementations.faithfulness.prompts import (
-    FaithfulnessTemplate, 
-    Claims
-)
-from judgeval.scorers.judgeval_scorers.local_implementations.summarization.prompts import *
-
-
-required_params = [
-    ExampleParams.INPUT,
-    ExampleParams.ACTUAL_OUTPUT,
-]
-
-
-class SummarizationScorer(JudgevalScorer):
-    def __init__(
-        self,
-        threshold: float = 0.5,
-        n: int = 5,
-        model: Optional[Union[str, JudgevalJudge]] = None,
-        assessment_questions: Optional[List[str]] = None,
-        include_reason: bool = True,
-        async_mode=True,
-        strict_mode: bool = False,
-        verbose_mode: bool = False,
-    ):
-        super().__init__(
-            score_type=APIScorer.SUMMARIZATION,
-            threshold=1 if strict_mode else threshold,
-            evaluation_model=None,
-            include_reason=include_reason,
-            async_mode=async_mode,
-            strict_mode=strict_mode,
-            verbose_mode=verbose_mode
-        )
-        self.model, self.using_native_model = create_judge(model)
-        self.evaluation_model = self.model.get_model_name()
-
-        if assessment_questions is not None and len(assessment_questions) == 0:
-            self.assessment_questions = None
-        else:
-            self.assessment_questions = assessment_questions
-
-        self.include_reason = include_reason
-        self.n = n
-        self.async_mode = async_mode
-        self.strict_mode = strict_mode
-        self.verbose_mode = verbose_mode
-
-    def score_example(
-        self,
-        example: Example,
-        _show_indicator: bool = True,
-    ) -> float:
-        check_example_params(example, required_params, self)
-        try:
-            with scorer_progress_meter(self, display_meter=_show_indicator):
-                if self.async_mode:
-                    loop = get_or_create_event_loop()
-                    loop.run_until_complete(
-                        self.a_score_example(example, _show_indicator=False)
-                    )
-                else:
-                    self.claims: List[str] = self._generate_claims(
-                        example.actual_output
-                    )
-                    
-                    self.info_coverage_verdicts: List[InfoCoverageVerdict] = (
-                        self._generate_info_coverage_verdicts(example)
-                    )
-                    
-                    self.contradiction_verdicts: List[ContradictionVerdict] = (
-                        self._generate_contradiction_verdicts(example)
-                    )
-                    
-                    contradiction_score = self._calculate_score(ScoreType.CONTRADICTION)
-                    info_coverage_score = self._calculate_score(ScoreType.INFO_COVERAGE)
-                    self.score_breakdown = {
-                        ScoreType.CONTRADICTION.value: contradiction_score,
-                        ScoreType.INFO_COVERAGE.value: info_coverage_score,
-                    }
-                    self.score = min(contradiction_score, info_coverage_score)
-                    self.reason = self._generate_reason()
-                    self.success = self.score >= self.threshold
-                    self.verbose_logs = create_verbose_logs(
-                        self,
-                        steps=[
-                            f"Claims:\n{self.claims}",
-                            f"Assessment Questions:\n{self.assessment_questions}",
-                            f"Info Coverage Verdicts:\n{[v.model_dump() for v in self.info_coverage_verdicts]}",
-                            f"Contradiction Verdicts:\n{[v.model_dump() for v in self.contradiction_verdicts]}",
-                            f"Score: {self.score}\nReason: {self.reason}",
-                        ],
-                    )
-
-                    return self.score
-        except Exception as e:
-            print(f"Error in SummarizationScorer score_example: {e}")
-            raise
-
-    async def a_score_example(
-        self,
-        example: Example,
-        _show_indicator: bool = True,
-    ) -> float:
-        """
-        To score, we take the following steps:
-        1. Generate claims from the actual output
-            - Extract key factual claims from the summary text
-
-        2. Generate info coverage verdicts:
-            a. Generate assessment questions if not provided
-            b. Generate answers to the assessment questions for both summary and original text
-            c. Compare answers to determine if summary adequately covers key information
-            d. Calculate info coverage score based on matching answers
-
-        3. Generate contradiction verdicts:
-            a. Generate claims from the actual output
-            b. Verify each claim against the original text for factual accuracy
-            c. Calculate contradiction score based on verified claims
-
-        4. Calculate final score:
-            - Take minimum of info coverage and contradiction scores
-            - Generate reason explaining the scoring
-            - Check if score meets threshold for success
-        """
-        check_example_params(example, required_params, self)
-        try:
-            with scorer_progress_meter(
-                self,
-                async_mode=True,
-                display_meter=_show_indicator,
-            ):
-                self.claims = await self._a_generate_claims(example.actual_output),
-                
-                self.info_coverage_verdicts, self.contradiction_verdicts = await asyncio.gather(
-                    self._a_generate_info_coverage_verdicts(example),
-                    self._a_generate_contradiction_verdicts(example),
-                )
-                
-                contradiction_score = self._calculate_score(ScoreType.CONTRADICTION)
-                info_coverage_score = self._calculate_score(ScoreType.INFO_COVERAGE)
-                self.score_breakdown = {
-                    ScoreType.CONTRADICTION.value: contradiction_score,
-                    ScoreType.INFO_COVERAGE.value: info_coverage_score,
-                }
-                self.score = min(contradiction_score, info_coverage_score)
-                self.reason = await self._a_generate_reason()
-                self.success = self.score >= self.threshold
-                self.verbose_logs = create_verbose_logs(
-                    self,
-                    steps=[
-                        f"Claims:\n{self.claims}",
-                        f"Assessment Questions:\n{self.assessment_questions}",
-                        f"Info Coverage Verdicts:\n{[v.model_dump() for v in self.info_coverage_verdicts]}",
-                        f"Contradiction Verdicts:\n{[v.model_dump() for v in self.contradiction_verdicts]}",
-                        f"Score: {self.score}\nReason: {self.reason}",
-                    ],
-                )
-
-                return self.score
-        except Exception as e:
-            print(f"Error in SummarizationScorer a_score_example: {e}")
-            raise
-
-    async def _a_generate_reason(self) -> str:
-        if self.include_reason is False:
-            return None
-
-        contradictions = []
-        redundancies = []
-        for verdict in self.contradiction_verdicts:
-            if verdict.verdict.strip().lower() == "no":
-                contradictions.append(verdict.reason)
-            elif verdict.verdict.strip().lower() == "idk":
-                redundancies.append(verdict.reason)
-
-        questions = []
-        if self.info_coverage_verdicts:
-            for verdict in self.info_coverage_verdicts:
-                if (
-                    verdict.original_verdict.strip().lower() == "yes"
-                    and verdict.summary_verdict.strip().lower() == "no"
-                ):
-                    questions.append(verdict.question)
-
-        prompt: dict = SummarizationTemplate.generate_reason(
-            contradictions=contradictions,
-            redundancies=redundancies,
-            questions=questions,
-            score=format(self.score, ".2f"),
-        )
-
-        if len(questions) > 0:
-            prompt += f"""Questions the original text can answer but not the summary:
-{questions}
-
-"""
-        prompt += """JSON:
-"""
-
-        if self.using_native_model:
-            res = await self.model.a_generate(prompt)
-            data = parse_response_json(res, self)
-            return data["reason"]
-        else:
-            try:
-                res: Reason = await self.model.a_generate(prompt, schema=Reason)
-                return res.reason
-            except TypeError:
-                res = await self.model.a_generate(prompt)
-                data = parse_response_json(res, self)
-                return data["reason"]
-
-    def _generate_reason(self) -> str:
-        if self.include_reason is False:
-            return None
-
-        contradictions = []
-        redundancies = []
-        for verdict in self.contradiction_verdicts:
-            if verdict.verdict.strip().lower() == "no":
-                contradictions.append(verdict.reason)
-            elif verdict.verdict.strip().lower() == "idk":
-                redundancies.append(verdict.reason)
-
-        questions = []
-        if self.info_coverage_verdicts:
-            for verdict in self.info_coverage_verdicts:
-                if (
-                    verdict.original_verdict.strip().lower() == "yes"
-                    and verdict.summary_verdict.strip().lower() == "no"
-                ):
-                    questions.append(verdict.question)
-
-        prompt: dict = SummarizationTemplate.generate_reason(
-            contradictions=contradictions,
-            redundancies=redundancies,
-            questions=questions,
-            score=format(self.score, ".2f"),
-        )
-
-        if len(questions) > 0:
-            prompt += f"""Questions the original text can answer but not the summary:
-{questions}
-
-"""
-        prompt += """JSON:
-"""
-
-        if self.using_native_model:
-            res = self.model.generate(prompt)
-            data = parse_response_json(res, self)
-            return data["reason"]
-        else:
-            try:
-                res: Reason = self.model.generate(prompt, schema=Reason)
-                return res.reason
-            except TypeError:
-                res = self.model.generate(prompt)
-                data = parse_response_json(res, self)
-                return data["reason"]
-
-    def _calculate_score(self, score_type: ScoreType) -> float:
-        if score_type == ScoreType.CONTRADICTION:
-            total = len(self.contradiction_verdicts)
-            if total == 0:
-                return 0
-            faithfulness_count = 0
-            for verdict in self.contradiction_verdicts:
-                # Different from the faithfulness score, this
-                # penalizes 'idk' (full of fluff) summaries
-                if verdict.verdict.strip().lower() == "yes":
-                    faithfulness_count += 1
-
-            score = faithfulness_count / total
-
-        else:
-            if self.assessment_questions is None:
-                return 1
-            total = 0
-            coverage_count = 0
-            for verdict in self.info_coverage_verdicts:
-                if verdict.original_verdict.strip().lower() == "yes":
-                    total += 1
-                    if verdict.summary_verdict.strip().lower() == "yes":
-                        coverage_count += 1
-
-            if total == 0:
-                return 0
-
-            score = coverage_count / total
-
-        return 0 if self.strict_mode and score < self.threshold else score
-
-    async def _a_generate_answers(self, text: str) -> List[str]:
-        prompt = SummarizationTemplate.generate_answers(
-            questions=self.assessment_questions, text=text
-        )
-        if self.using_native_model:
-            res = await self.model.a_generate(prompt)
-            data = parse_response_json(res, self)
-            return data["answers"]
-        else:
-            try:
-                res: Answers = await self.model.a_generate(
-                    prompt, schema=Answers
-                )
-                return res.answers
-            except TypeError:
-                res = await self.model.a_generate(prompt)
-                data = parse_response_json(res, self)
-                return data["answers"]
-
-    def _generate_answers(self, text: str) -> List[str]:
-        prompt = SummarizationTemplate.generate_answers(
-            questions=self.assessment_questions, text=text
-        )
-        if self.using_native_model:
-            res = self.model.generate(prompt)
-            data = parse_response_json(res, self)
-            return data["answers"]
-        else:
-            try:
-                res: Answers = self.model.generate(prompt, schema=Answers)
-                return res.answers
-            except TypeError:
-                res = self.model.generate(prompt)
-                data = parse_response_json(res, self)
-                return data["answers"]
-
-    async def _a_generate_assessment_questions(self, text: str):
-        prompt = SummarizationTemplate.generate_questions(text=text, n=self.n)
-        if self.using_native_model:
-            res = await self.model.a_generate(prompt)
-            data = parse_response_json(res, self)
-            return data["questions"]
-        else:
-            try:
-                res: Questions = await self.model.a_generate(
-                    prompt, schema=Questions
-                )
-                return res.questions
-            except TypeError:
-                res = await self.model.a_generate(prompt)
-                data = parse_response_json(res, self)
-                return data["questions"]
-
-    def _generate_assessment_questions(self, text: str):
-        prompt = SummarizationTemplate.generate_questions(text=text, n=self.n)
-        if self.using_native_model:
-            res = self.model.generate(prompt)
-            data = parse_response_json(res, self)
-            return data["questions"]
-        else:
-            try:
-                res: Questions = self.model.generate(prompt, schema=Questions)
-                return res.questions
-            except TypeError:
-                res = self.model.generate(prompt)
-                data = parse_response_json(res, self)
-                return data["questions"]
-
-    async def _a_generate_info_coverage_verdicts(
-        self, example: Example
-    ) -> List[InfoCoverageVerdict]:
-        if self.assessment_questions is None:
-            self.assessment_questions = (
-                await self._a_generate_assessment_questions(example.input)
-            )
-
-        tasks = [
-            self._a_generate_answers(example.input),
-            self._a_generate_answers(example.actual_output),
-        ]
-        results = await asyncio.gather(*tasks)
-        original_answers = results[0]
-        summary_answers = results[1]
-
-        if len(original_answers) != len(summary_answers):
-            raise ValueError("Number of verdicts generated does not equal.")
-
-        coverage_veridcts: List[InfoCoverageVerdict] = []
-        for i in range(len(original_answers)):
-            coverage_veridcts.append(
-                InfoCoverageVerdict(
-                    summary_verdict=summary_answers[i],
-                    original_verdict=original_answers[i],
-                    question=self.assessment_questions[i],
-                )
-            )
-        return coverage_veridcts
-
-    def _generate_info_coverage_verdicts(
-        self, example: Example
-    ) -> List[InfoCoverageVerdict]:
-        if self.assessment_questions is None:
-            self.assessment_questions = self._generate_assessment_questions(
-                example.input
-            )
-
-        original_answers = self._generate_answers(example.input)
-        summary_answers = self._generate_answers(example.actual_output)
-
-        if len(original_answers) != len(summary_answers):
-            raise ValueError("Number of verdicts generated does not equal.")
-
-        coverage_veridcts: List[InfoCoverageVerdict] = []
-        for i in range(len(original_answers)):
-            coverage_veridcts.append(
-                InfoCoverageVerdict(
-                    summary_verdict=summary_answers[i],
-                    original_verdict=original_answers[i],
-                    question=self.assessment_questions[i],
-                )
-            )
-
-        return coverage_veridcts
-
-    async def _a_generate_contradiction_verdicts(
-        self,
-        example: Example,
-    ) -> List[ContradictionVerdict]:
-        if len(self.claims) == 0:
-            return []
-
-        verdicts: List[ContradictionVerdict] = []
-
-        prompt = SummarizationTemplate.generate_contradiction_verdicts(
-            original_text=example.input,
-            summary_claims=self.claims
-        )
-        if self.using_native_model:
-            res = await self.model.a_generate(prompt)
-            data = parse_response_json(res, self)
-            verdicts = [
-                ContradictionVerdict(**item)
-                for item in data["verdicts"]
-            ]
-            return verdicts
-        else:
-            try:
-                res: Verdicts = await self.model.a_generate(
-                    prompt, schema=Verdicts
-                )
-                verdicts = [item for item in res.verdicts]
-                return verdicts
-            except TypeError:
-                res = await self.model.a_generate(prompt)
-                data = parse_response_json(res, self)
-                verdicts = [
-                    ContradictionVerdict(**item)
-                    for item in data["verdicts"]
-                ]
-                return verdicts
-
-    def _generate_contradiction_verdicts(
-        self,
-        example: Example,
-    ) -> List[ContradictionVerdict]:
-        if len(self.claims) == 0:
-            return []
-
-        verdicts: List[ContradictionVerdict] = []
-
-        prompt = SummarizationTemplate.generate_contradiction_verdicts(
-            original_text=example.input,
-            summary_claims=self.claims
-        )
-        if self.using_native_model:
-            res = self.model.generate(prompt)
-            data = parse_response_json(res, self)
-            verdicts = [
-                ContradictionVerdict(**item)
-                for item in data["verdicts"]
-            ]
-            return verdicts
-        else:
-            try:
-                res: Verdicts = self.model.generate(prompt, schema=Verdicts)
-                verdicts = [item for item in res.verdicts]
-                return verdicts
-            except TypeError:
-                res = self.model.generate(prompt)
-                data = parse_response_json(res, self)
-                verdicts = [
-                    ContradictionVerdict(**item)
-                    for item in data["verdicts"]
-                ]
-                return verdicts
-
-    async def _a_generate_claims(self, text: str) -> List[str]:
-        # Borrow faithfulness template since it already works
-        prompt = FaithfulnessTemplate.find_claims(text=text)
-        if self.using_native_model:
-            res = await self.model.a_generate(prompt)
-            data = parse_response_json(res, self)
-            return data["claims"]
-        else:
-            try:
-                res: Claims = await self.model.a_generate(prompt, schema=Claims)
-                return res.claims
-            except TypeError:
-                res = await self.model.a_generate(prompt)
-                data = parse_response_json(res, self)
-                return data["claims"]
-
-    def _generate_claims(self, text: str) -> List[str]:
-        # Borrow faithfulness template
-        prompt = FaithfulnessTemplate.find_claims(text=text)
-        if self.using_native_model:
-            res = self.model.generate(prompt)
-            data = parse_response_json(res, self)
-            return data["claims"]
-        else:
-            try:
-                res: Claims = self.model.generate(prompt, schema=Claims)
-                return res.claims
-            except TypeError:
-                res = self.model.generate(prompt)
-                data = parse_response_json(res, self)
-                return data["claims"]
-
-    def _success_check(self) -> bool:
-        if self.error is not None:
-            self.success = False
-        else:
-            try:
-                self.success = self.score >= self.threshold
-            except:
-                self.success = False
-        return self.success
-
-    @property
-    def __name__(self):
-        return "Summarization"
-    

judgeval-0.0.32.dist-info/RECORD

@@ -1,97 +0,0 @@
-judgeval/__init__.py,sha256=x9HWt4waJwJMAqTuJSg2MezF9Zg-macEjeU-ajbly-8,330
-judgeval/clients.py,sha256=6VQmEqmfCngUdS2MuPBIpHvtDFqOENm8-_BmMvjLyRQ,944
-judgeval/constants.py,sha256=_XmVAkebMyGrDvvanAVlMgVd4p6MLHdEVsTQFI0kz1k,5411
-judgeval/evaluation_run.py,sha256=WGzx-Ug2qhSmunFo8NrmSstBRsOUc5KpKq0Lc51rqsM,6739
-judgeval/judgment_client.py,sha256=k0q2s5A0RkhF9ElD9o-KWN10H36t3Of2PrvNF-silf8,26141
-judgeval/rules.py,sha256=B0ZL0pn72D4Jnlr0zMQ6CPHi7D8AQQRariXCVsiCMiI,20542
-judgeval/run_evaluation.py,sha256=hnEY8QckEviXYNJutf-6tLFq2DWCzqWV1EVyPvrVXyA,28512
-judgeval/version_check.py,sha256=bvJEidB7rAeXozoUbN9Yb97QOR_s2hgvpvj74jJ5HlY,943
-judgeval/common/__init__.py,sha256=7d24BRxtncpMj3AAJCj8RS7TqgjXmW777HVZH6-3sBs,289
-judgeval/common/exceptions.py,sha256=U-TxHLn7oVMezsMuoYouNDb2XuS8RCggfntYf5_6u4E,565
-judgeval/common/logger.py,sha256=KO75wWXCxhUHUMvLaTU31ZzOk6tkZBa7heQ7y0f-zFE,6062
-judgeval/common/tracer.py,sha256=owRRfIZXPUOVCCn0macygnf18mcp8am1eULGnZXD0Kk,68876
-judgeval/common/utils.py,sha256=LUQV5JfDr6wj7xHAJoNq-gofNZ6mjXbeKrGKzBME1KM,33533
-judgeval/data/__init__.py,sha256=xuKx_KCVHGp6CXvQuVmKl3v7pJp-qDaz0NccKxwjtO0,481
-judgeval/data/custom_example.py,sha256=QRBqiRiZS8UgVeTRHY0r1Jzm6yAYsyg6zmHxQGxdiQs,739
-judgeval/data/example.py,sha256=cJrmPGLel_P2sy1UaRvuVSAi35EnA9XMR11Lhp4aDLo,5930
-judgeval/data/result.py,sha256=Gb9tiSDsk1amXgh0cFG6JmlW_BMKxS2kuTwNA0rrHjA,3184
-judgeval/data/scorer_data.py,sha256=JVlaTx1EP2jw2gh3Vgx1CSEsvIFABAN26IquKyxwiJQ,3273
-judgeval/data/sequence.py,sha256=Fkk2HJGnPboH-Fvwgxub_ryG0eUXa3cbsj7ZD0qkeBo,2204
-judgeval/data/sequence_run.py,sha256=RmYjfWKMWg-pcF5PLeiWfrhuDkjDZi5VEmAIEXN3Ib0,2104
-judgeval/data/datasets/__init__.py,sha256=IdNKhQv9yYZ_op0rdBacrFaFVmiiYQ3JTzXzxOTsEVQ,176
-judgeval/data/datasets/dataset.py,sha256=dhLo30hvpmmOK2R6O5wDs_neawUJ4lS8bb4S42SufNQ,13034
-judgeval/data/datasets/eval_dataset_client.py,sha256=xjj66BO9Es9IxXqzQe1RT_e0kpeKlt7OrhRoSuj4KHM,15085
-judgeval/integrations/langgraph.py,sha256=J-cQfFP52TjJewdSTe-fcsUC4HDvjNbXoxmbmF0SgiE,11743
-judgeval/judges/__init__.py,sha256=6X7VSwrwsdxGBNxCyapVRWGghhKOy3MVxFNMQ62kCXM,308
-judgeval/judges/base_judge.py,sha256=ch_S7uBB7lyv44Lf1d7mIGFpveOO58zOkkpImKgd9_4,994
-judgeval/judges/litellm_judge.py,sha256=EIL58Teptv8DzZUO3yP2RDQCDq-aoBB6HPZzPdK6KTg,2424
-judgeval/judges/mixture_of_judges.py,sha256=IJoi4Twk8ze1CJWVEp69k6TSqTCTGrmVYQ0qdffer60,15549
-judgeval/judges/together_judge.py,sha256=l00hhPerAZXg3oYBd8cyMtWsOTNt_0FIqoxhKJKQe3k,2302
-judgeval/judges/utils.py,sha256=9lvUxziGV86ISvVFxYBWc09TWFyAQgUTyPf_a9mD5Rs,2686
-judgeval/scorers/__init__.py,sha256=Z_88Sr45gLFAIbMHzG1BF24TUQGCDiuP9QpmVFvSYJM,1204
-judgeval/scorers/api_scorer.py,sha256=NQ_CrrUPhSUk1k2Q8rKpCG_TU2FT32sFEqvb-Yi54B0,2688
-judgeval/scorers/base_scorer.py,sha256=xdUlY3CnLdCQ1Z5iUeY22Bim5v-OQruZmaVF_4Y1mC0,2183
-judgeval/scorers/exceptions.py,sha256=eGW5CuJgZ5YJBFrE4FHDSF651PO1dKAZ379mJ8gOsfo,178
-judgeval/scorers/judgeval_scorer.py,sha256=79-JJurqHP-qTaWNWInx4SjvQYwXc9lvfPPNgwsh2yA,6773
-judgeval/scorers/prompt_scorer.py,sha256=PaAs2qRolw1P3_I061Xvk9qzvF4O-JR8g_39RqXnHcM,17728
-judgeval/scorers/score.py,sha256=r9QiT4-LIvivcJ6XxByrbswKSO8eQTtAD1UlXT_lcmo,18741
-judgeval/scorers/utils.py,sha256=iHQVTlIANbmCTXz9kTeSdOytgUZ_T74Re61ajqsk_WQ,6827
-judgeval/scorers/judgeval_scorers/__init__.py,sha256=kSmQWKeBvLeZMfLYNQSc2qbJYo1MFIQnf3P-D4ltuSM,6232
-judgeval/scorers/judgeval_scorers/api_scorers/__init__.py,sha256=_sDUBxSG536KGqXNi6dFpaYKghjEAadxBxaaxV9HuuE,1764
-judgeval/scorers/judgeval_scorers/api_scorers/answer_correctness.py,sha256=Fnd9CVIOZ73sWEWymsU5eBrrZqPFjMZ0BKpeW-PDyTg,711
-judgeval/scorers/judgeval_scorers/api_scorers/answer_relevancy.py,sha256=oETeN9K0HSIRdL2SDqn82Vskpwh5SlKnZvs5VDm2OBU,658
-judgeval/scorers/judgeval_scorers/api_scorers/comparison.py,sha256=kuzf9OWvpY38yYSwlBgneLkUZwJNM4FQqvbS66keA90,1249
-judgeval/scorers/judgeval_scorers/api_scorers/contextual_precision.py,sha256=tpSuzFAaW8X9xqA0aLLKwh7qmBK0Pc_bJZMIe_q412U,770
-judgeval/scorers/judgeval_scorers/api_scorers/contextual_recall.py,sha256=pFVhk4pLtQ-FnNlbI-dFF-SIh69Jza7erHqiPkFWoBo,758
-judgeval/scorers/judgeval_scorers/api_scorers/contextual_relevancy.py,sha256=RQ6DZwEhChfecd89Ey-T7ke--7qTaXZlRsNxwH8gaME,823
-judgeval/scorers/judgeval_scorers/api_scorers/derailment_scorer.py,sha256=V9WPuwNMm097V7IknKs8UkmAk0yjnBXTcJha_BHXxTA,475
-judgeval/scorers/judgeval_scorers/api_scorers/execution_order.py,sha256=Pb3CiNF2Ca826B92wJCVAi_68lJjLhqqCKwQKaflSUg,1294
-judgeval/scorers/judgeval_scorers/api_scorers/faithfulness.py,sha256=-BwOapqjryYNKNydtdkUiKIij76dY0O1jBmdc6dKazQ,692
-judgeval/scorers/judgeval_scorers/api_scorers/groundedness.py,sha256=ntEEeTANEOsGlcbiTAF_3r6BeSJEaVDns8po8T0L6Vg,692
-judgeval/scorers/judgeval_scorers/api_scorers/hallucination.py,sha256=k5gDOki-8KXrZXydvdSqDt3NZqQ28hXoOCHQf6jNxr4,686
-judgeval/scorers/judgeval_scorers/api_scorers/instruction_adherence.py,sha256=XnSGEkQfwVqaqnHEGMCsxNiHVzrsrej48uDbLoWc8CQ,678
-judgeval/scorers/judgeval_scorers/api_scorers/json_correctness.py,sha256=mMKEuR87_yanEuZJ5YSGFMHDD_oLVZ6-rQuciFaDOMA,1095
-judgeval/scorers/judgeval_scorers/api_scorers/summarization.py,sha256=QmWB8bVbDYHY5FcF0rYZE_3c2XXgMLRmR6aXJWfdMC4,655
-judgeval/scorers/judgeval_scorers/classifiers/__init__.py,sha256=Qt81W5ZCwMvBAne0LfQDb8xvg5iOG1vEYP7WizgwAZo,67
-judgeval/scorers/judgeval_scorers/classifiers/text2sql/__init__.py,sha256=8iTzMvou1Dr8pybul6lZHKjc9Ye2-0_racRGYkhEdTY,74
-judgeval/scorers/judgeval_scorers/classifiers/text2sql/text2sql_scorer.py,sha256=ly72Z7s_c8NID6-nQnuW8qEGEW2MqdvpJ-5WfXzbAQg,2579
-judgeval/scorers/judgeval_scorers/local_implementations/__init__.py,sha256=k_t-THIAtsk7lNvm9faj0u24dPZjn7qRbZ8YGjQ21xs,1926
-judgeval/scorers/judgeval_scorers/local_implementations/answer_correctness/__init__.py,sha256=cxxUEspgoIdSzJbwIIioamC0-xDqhYVfYAWxaYF-D_Y,177
-judgeval/scorers/judgeval_scorers/local_implementations/answer_correctness/answer_correctness_scorer.py,sha256=3Dpm8BIIe0Th2p0ccO5bb-le93lywjOLSo712HwEIUE,10196
-judgeval/scorers/judgeval_scorers/local_implementations/answer_correctness/prompts.py,sha256=hBUqEd8Hy3g8peOVjpSmRb31fPtpodDzdRUonhKRl30,6686
-judgeval/scorers/judgeval_scorers/local_implementations/answer_relevancy/__init__.py,sha256=r6yae5iaWtlBL_cP8I-1SuhS9dulsy1e7W9Rcz82v6E,169
-judgeval/scorers/judgeval_scorers/local_implementations/answer_relevancy/answer_relevancy_scorer.py,sha256=qoeoFyXXDtqqc7ZSLajqexeSxw5STmrL-uPQIMY3zSw,10529
-judgeval/scorers/judgeval_scorers/local_implementations/answer_relevancy/prompts.py,sha256=-OO3QmkXqGCvdIRKsAuT4wQ1ZqWBQDdb1j3lc3W9q3w,6540
-judgeval/scorers/judgeval_scorers/local_implementations/comparison/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-judgeval/scorers/judgeval_scorers/local_implementations/comparison/comparison_scorer.py,sha256=QnwSTgYx_zyz6K27WTe89MoTcO12WYn_jqE_xj7_H2U,5497
-judgeval/scorers/judgeval_scorers/local_implementations/comparison/prompts.py,sha256=c49aCzyxCogjUTipQUmS13LemFC89X9xEuPNQ_LVHgw,31345
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_precision/__init__.py,sha256=J6tc-T60AVOEaNVuoVU0XIG6dvQri99Q0tnX_Tm-0vc,108
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_precision/contextual_precision_scorer.py,sha256=tRgRyjGpc4Pe3nQ1c-5NeNYFvbulL7YEnoRa9zLp1gc,9649
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_precision/prompts.py,sha256=pN0AURDWSV3iGt11MtJIwzXMuKbM4oC3zdb9yqnjNdU,4875
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_recall/__init__.py,sha256=4kjfqD_95muHZFo75S8_fbTcC1DI1onNIfMmr8gMZaI,99
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_recall/contextual_recall_scorer.py,sha256=hwAv_x3XwGDnSW3a75CTCgIW6eVg8ymdjDdJQvw5p0Y,9260
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_recall/prompts.py,sha256=k5xdCw8gnBvG1_dDSbtBftDDtOZ4qKL2-iQ9AQHsuUI,6541
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_relevancy/__init__.py,sha256=JPCvrekKLbl_xdD49evhtiFIVocuegCpCBkn1auzTSE,184
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_relevancy/contextual_relevancy_scorer.py,sha256=BtVgE7z-9PHfFRcvn96aEG5mXVcWBweVyty934hZdiU,8915
-judgeval/scorers/judgeval_scorers/local_implementations/contextual_relevancy/prompts.py,sha256=uO-8Uo7VrXu4xWpxjIx6_UI3aw5KuJxubSHb71Nzm6Q,4574
-judgeval/scorers/judgeval_scorers/local_implementations/execution_order/__init__.py,sha256=DpOHbjYEhVmP-RiaTEa5PZHpoPvduNXG5p6k9lR0AS0,157
-judgeval/scorers/judgeval_scorers/local_implementations/execution_order/execution_order.py,sha256=y-Ag8YuzEvExUIj4qU7y53INVLH9L_TUTJLIxCIdAQo,5458
-judgeval/scorers/judgeval_scorers/local_implementations/faithfulness/__init__.py,sha256=NbkSqPwxgF4T8KsvuIWhVyRwdOlo7mNHMFuRStTFnvk,154
-judgeval/scorers/judgeval_scorers/local_implementations/faithfulness/faithfulness_scorer.py,sha256=LPVTGHBBJSpE6TrgzZQS2_vw4P9HiUYmykrwo6UMdws,11251
-judgeval/scorers/judgeval_scorers/local_implementations/faithfulness/prompts.py,sha256=vNLjF4NKZJSV4VNenHzoAUB2xVZz6tt_5AzryKmOVrI,11690
-judgeval/scorers/judgeval_scorers/local_implementations/hallucination/__init__.py,sha256=fZk3UQxI9Nljf5qjCRLRkF0D-AERFHElI9cC83_cgV8,158
-judgeval/scorers/judgeval_scorers/local_implementations/hallucination/hallucination_scorer.py,sha256=q9qdmwq96stbTRVA4Egv9eO1KI8zf77jwYkZXaOZePw,9511
-judgeval/scorers/judgeval_scorers/local_implementations/hallucination/prompts.py,sha256=TJWKheQnaJ-pdVzkOTDr3BNZ9bfCrC81S3kvjm4Zjh8,4329
-judgeval/scorers/judgeval_scorers/local_implementations/instruction_adherence/instruction_adherence.py,sha256=keZRmLe5oyIVE98h1nOiW54__xcEv2QInwZcJ34ZKhs,8175
-judgeval/scorers/judgeval_scorers/local_implementations/instruction_adherence/prompt.py,sha256=jv-1Z7K1EhLjy4NGKS35cIcShh7ZDQCXVPqoJnAnDqk,3598
-judgeval/scorers/judgeval_scorers/local_implementations/json_correctness/__init__.py,sha256=xQDw7o9JQ6qajusPnBH0MWBRJ5ct_Ao3pJELXxxVMRo,175
-judgeval/scorers/judgeval_scorers/local_implementations/json_correctness/json_correctness_scorer.py,sha256=RkI-mARc2AYCw5dTb5OSY4UWXIwDcYS3ViiJOVIq0Nw,4339
-judgeval/scorers/judgeval_scorers/local_implementations/summarization/__init__.py,sha256=mv6-XeLSV5yj1H98YYV2iTYVd88zKftZJP42Lgl6R80,89
-judgeval/scorers/judgeval_scorers/local_implementations/summarization/prompts.py,sha256=6GnRz2h-6Fwt4sl__0RgQOyo3n3iDO4MNuHWxdu-rrM,10242
-judgeval/scorers/judgeval_scorers/local_implementations/summarization/summarization_scorer.py,sha256=Qk7lwHgRPYeGoxTOyclAh1VfGItfvHJ6l1t7Nk3SWFM,20927
-judgeval/tracer/__init__.py,sha256=wy3DYpH8U_z0GO_K_gOSkK0tTTD-u5eLDo0T5xIBoAc,147
-judgeval/utils/alerts.py,sha256=O19Xj7DA0YVjl8PWiuH4zfdZeu3yiLVvHfY8ah2wG0g,2759
-judgeval-0.0.32.dist-info/METADATA,sha256=RJzqlHJwfYiOXEcyEEO5WQBM0DC1zQDuoN-Plix6U38,5418
-judgeval-0.0.32.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-judgeval-0.0.32.dist-info/licenses/LICENSE.md,sha256=tKmCg7k5QOmxPK19XMfzim04QiQJPmgIm0pAn55IJwk,11352
-judgeval-0.0.32.dist-info/RECORD,,