azure-ai-evaluation 1.0.0b1__py3-none-any.whl → 1.0.0b3__py3-none-any.whl

azure/ai/evaluation/_evaluators/_chat/_chat.py

@@ -1,350 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-import json
-import logging
-from concurrent.futures import as_completed
-from typing import Dict, List, Union
-
-import numpy as np
-
-from promptflow.tracing import ThreadPoolExecutorWithContext as ThreadPoolExecutor
-
-from .._coherence import CoherenceEvaluator
-from .._fluency import FluencyEvaluator
-from .._groundedness import GroundednessEvaluator
-from ..._model_configurations import AzureOpenAIModelConfiguration, OpenAIModelConfiguration
-from .._relevance import RelevanceEvaluator
-from .retrieval import RetrievalChatEvaluator
-from azure.ai.evaluation._exceptions import EvaluationException, ErrorBlame, ErrorCategory, ErrorTarget
-
-logger = logging.getLogger(__name__)
-
-
-class ChatEvaluator:
-    """
-    Initialize a chat evaluator configured for a specific Azure OpenAI model.
-
-    :param model_config: Configuration for the Azure OpenAI model.
-    :type model_config: Union[~azure.ai.evaluation.AzureOpenAIModelConfiguration,
-        ~azure.ai.evaluation.OpenAIModelConfiguration]
-    :param eval_last_turn: Set to True to evaluate only the most recent exchange in the dialogue,
-        focusing on the latest user inquiry and the assistant's corresponding response. Defaults to False
-    :type eval_last_turn: bool
-    :param parallel: If True, use parallel execution for evaluators. Else, use sequential execution.
-        Default is True.
-    :type parallel: bool
-    :return: A function that evaluates and generates metrics for "chat" scenario.
-    :rtype: Callable
-
-    **Usage**
-
-    .. code-block:: python
-
-        chat_eval = ChatEvaluator(model_config)
-        conversation = [
-            {"role": "user", "content": "What is the value of 2 + 2?"},
-            {"role": "assistant", "content": "2 + 2 = 4", "context": {
-                "citations": [
-                        {"id": "math_doc.md", "content": "Information about additions: 1 + 2 = 3, 2 + 2 = 4"}
-                    ]
-                }
-            }
-        ]
-        result = chat_eval(conversation=conversation)
-
-    **Output format**
-
-    .. code-block:: python
-
-        {
-            "evaluation_per_turn": {
-                "gpt_retrieval": [1.0, 2.0],
-                "gpt_groundedness": [5.0, 2.0],
-                "gpt_relevance": [3.0, 5.0],
-                "gpt_coherence": [1.0, 2.0],
-                "gpt_fluency": [3.0, 5.0]
-            }
-            "gpt_retrieval": 1.5,
-            "gpt_groundedness": 3.5,
-            "gpt_relevance": 4.0,
-            "gpt_coherence": 1.5,
-            "gpt_fluency": 4.0
-        }
-    """
-
-    def __init__(
-        self,
-        model_config: dict,
-        eval_last_turn: bool = False,
-        parallel: bool = True,
-    ):
-        self._eval_last_turn = eval_last_turn
-        self._parallel = parallel
-
-        # TODO: Need a built-in evaluator for retrieval. It needs to be added to `self._rag_evaluators` collection
-        self._rag_evaluators = [
-            GroundednessEvaluator(model_config),
-            RelevanceEvaluator(model_config),
-        ]
-        self._non_rag_evaluators = [
-            CoherenceEvaluator(model_config),
-            FluencyEvaluator(model_config),
-        ]
-        # TODO: Temporary workaround to close the gap of missing retrieval score
-        # https://msdata.visualstudio.com/Vienna/_workitems/edit/3186644
-        # For long term, we need to add a built-in evaluator for retrieval after prompt is generalized for QA and Chat
-        self._retrieval_chat_evaluator = RetrievalChatEvaluator(model_config)
-
-    def __call__(self, *, conversation, **kwargs):
-        """
-        Evaluates chat scenario.
-
-        :keyword conversation: The conversation to be evaluated. Each turn should have "role" and "content" keys.
-            "context" key is optional for assistant's turn and should have "citations" key with list of citations.
-        :paramtype conversation: List[Dict]
-        :return: The scores for Chat scenario.
-        :rtype: dict
-        """
-        self._validate_conversation(conversation)
-
-        # Extract queries, responses and contexts from conversation
-        queries = []
-        responses = []
-        contexts = []
-
-        if self._eval_last_turn:
-            # Process only the last two turns if _eval_last_turn is True
-            conversation_slice = conversation[-2:] if len(conversation) >= 2 else conversation
-        else:
-            conversation_slice = conversation
-
-        for each_turn in conversation_slice:
-            role = each_turn["role"]
-            if role == "user":
-                queries.append(each_turn["content"])
-            elif role == "assistant":
-                responses.append(each_turn["content"])
-                if "context" in each_turn and "citations" in each_turn["context"]:
-                    citations = json.dumps(each_turn["context"]["citations"])
-                    contexts.append(citations)
-
-        # Select evaluators to be used for evaluation
-        compute_rag_based_metrics = True
-        if len(responses) != len(contexts):
-            safe_message = (
-                "Skipping rag based metrics as we need citations or "
-                "retrieved_documents in context key of every assistant's turn"
-            )
-            logger.warning(safe_message)
-            compute_rag_based_metrics = False
-
-        selected_evaluators = []
-        selected_evaluators.extend(self._non_rag_evaluators)
-        if compute_rag_based_metrics:
-            selected_evaluators.extend(self._rag_evaluators)
-
-        # Evaluate each turn
-        per_turn_results = []
-        for turn_num in range(len(queries)):
-            current_turn_result = {}
-
-            if self._parallel:
-                # Parallel execution
-                with ThreadPoolExecutor() as executor:
-                    future_to_evaluator = {
-                        executor.submit(
-                            self._evaluate_turn, turn_num, queries, responses, contexts, evaluator
-                        ): evaluator
-                        for evaluator in selected_evaluators
-                    }
-
-                    for future in as_completed(future_to_evaluator):
-                        result = future.result()
-                        current_turn_result.update(result)
-            else:
-                # Sequential execution
-                for evaluator in selected_evaluators:
-                    async_evaluator = evaluator._to_async()
-                    result = self._evaluate_turn(turn_num, queries, responses, contexts, async_evaluator)
-                    current_turn_result.update(result)
-
-            per_turn_results.append(current_turn_result)
-
-        # Aggregate results
-        # Final aggregated results for a conversation will look like:
-        #     "gpt_groundedness": 2.0, # Mean of all groundedness scores
-        #     "evaluation_per_turn": {
-        #         "gpt_groundedness": {
-        #             "score": [1.0, ...],
-        #             "reason": ["reason1", ...],
-        #         },
-        #     },
-        # }
-        aggregated = self._aggregate_results(per_turn_results)
-
-        # Run RetrievalChatEvaluator and merge the results
-        if compute_rag_based_metrics:
-            retrieval_score = self._retrieval_chat_evaluator(conversation=conversation_slice)
-            aggregated["gpt_retrieval"] = retrieval_score["gpt_retrieval"]
-            aggregated["evaluation_per_turn"]["gpt_retrieval"] = retrieval_score["evaluation_per_turn"]["gpt_retrieval"]
-            aggregated = dict(sorted(aggregated.items()))
-
-        return aggregated
-
-    def _evaluate_turn(self, turn_num, queries, responses, contexts, evaluator):
-        try:
-            query = queries[turn_num] if turn_num < len(queries) else ""
-            response = responses[turn_num] if turn_num < len(responses) else ""
-            context = contexts[turn_num] if turn_num < len(contexts) else ""
-
-            score = evaluator(query=query, response=response, context=context)
-
-            return score
-        except Exception as e:  # pylint: disable=broad-exception-caught
-            logger.warning(
-                f"Evaluator {evaluator.__class__.__name__} failed for turn {turn_num + 1} with exception: {e}"
-            )
-            return {}
-
-    def _aggregate_results(self, per_turn_results: List[Dict]):
-        scores = {}
-        reasons = {}
-
-        for turn in per_turn_results:
-            for metric, value in turn.items():
-                if "reason" in metric:
-                    if metric not in reasons:
-                        reasons[metric] = []
-                    reasons[metric].append(value)
-                else:
-                    if metric not in scores:
-                        scores[metric] = []
-                    scores[metric].append(value)
-
-        aggregated = {}
-        evaluation_per_turn = {}
-
-        for metric, values in scores.items():
-            aggregated[metric] = np.nanmean(values)
-
-            # Prepare per-turn evaluations
-            evaluation_per_turn[metric] = {"score": values}
-            reason_key = f"{metric}_reason"
-            if reason_key in reasons:
-                evaluation_per_turn[metric]["reason"] = reasons[reason_key]
-
-        aggregated["evaluation_per_turn"] = evaluation_per_turn
-
-        return aggregated
-
-    def _validate_conversation(self, conversation: List[Dict]):
-        if conversation is None or not isinstance(conversation, list):
-            msg = "conversation must be a list of dictionaries"
-            raise EvaluationException(
-                message=msg,
-                internal_message=msg,
-                target=ErrorTarget.CHAT_EVALUATOR,
-                category=ErrorCategory.INVALID_VALUE,
-                blame=ErrorBlame.USER_ERROR,
-            )
-
-        expected_role = "user"
-        for turn_num, turn in enumerate(conversation):
-            one_based_turn_num = turn_num + 1
-
-            if not isinstance(turn, dict):
-                msg = f"Each turn in 'conversation' must be a dictionary. Turn number: {one_based_turn_num}"
-                raise EvaluationException(
-                    message=msg,
-                    internal_message=msg,
-                    target=ErrorTarget.CHAT_EVALUATOR,
-                    category=ErrorCategory.INVALID_VALUE,
-                    blame=ErrorBlame.USER_ERROR,
-                )
-
-            if "role" not in turn or "content" not in turn:
-                msg = f"Each turn in 'conversation' must have 'role' and 'content' keys. Turn number: {one_based_turn_num}"
-                raise EvaluationException(
-                    message=msg,
-                    internal_message=msg,
-                    target=ErrorTarget.CHAT_EVALUATOR,
-                    category=ErrorCategory.INVALID_VALUE,
-                    blame=ErrorBlame.USER_ERROR,
-                )
-            
-            if turn["role"] != expected_role:
-                msg = f"Expected role {expected_role} but got {turn['role']}. Turn number: {one_based_turn_num}"
-                raise EvaluationException(
-                    message=msg,
-                    internal_message=msg,
-                    target=ErrorTarget.CHAT_EVALUATOR,
-                    category=ErrorCategory.INVALID_VALUE,
-                    blame=ErrorBlame.USER_ERROR,
-                )
-
-            if not isinstance(turn["content"], str):
-                msg = f"Content in each turn must be a string. Turn number: {one_based_turn_num}"
-                raise EvaluationException(
-                    message=msg,
-                    internal_message=msg,
-                    target=ErrorTarget.CHAT_EVALUATOR,
-                    category=ErrorCategory.INVALID_VALUE,
-                    blame=ErrorBlame.USER_ERROR,
-                )
-
-            if turn["role"] == "assistant" and "context" in turn:
-                if not isinstance(turn["context"], dict):
-                    msg = f"Context in each assistant's turn must be a dictionary. Turn number: {one_based_turn_num}"
-                    raise EvaluationException(
-                        message=msg,
-                        internal_message=msg,
-                        target=ErrorTarget.CHAT_EVALUATOR,
-                        category=ErrorCategory.INVALID_VALUE,
-                        blame=ErrorBlame.USER_ERROR,
-                    )
-
-                if "citations" not in turn["context"]:
-                    msg = f"Context in each assistant's turn must have 'citations' key. Turn number: {one_based_turn_num}"
-                    raise EvaluationException(
-                        message=msg,
-                        internal_message=msg,
-                        target=ErrorTarget.CHAT_EVALUATOR,
-                        category=ErrorCategory.MISSING_FIELD,
-                        blame=ErrorBlame.USER_ERROR,
-                    )
-
-                if not isinstance(turn["context"]["citations"], list):
-                    msg = f"'citations' in context must be a list. Turn number: {one_based_turn_num}"
-                    raise EvaluationException(
-                        message=msg,
-                        internal_message=msg,
-                        target=ErrorTarget.CHAT_EVALUATOR,
-                        category=ErrorCategory.INVALID_VALUE,
-                        blame=ErrorBlame.USER_ERROR,
-                    )
-
-                for citation_num, citation in enumerate(turn["context"]["citations"]):
-                    if not isinstance(citation, dict):
-                        msg = f"Each citation in 'citations' must be a dictionary. Turn number: {one_based_turn_num}, Citation number: {citation_num + 1}"
-                        raise EvaluationException(
-                            message=msg,
-                            internal_message=msg,
-                            target=ErrorTarget.CHAT_EVALUATOR,
-                            category=ErrorCategory.INVALID_VALUE,
-                            blame=ErrorBlame.USER_ERROR,
-                        )
-
-            # Toggle expected role for the next turn
-            expected_role = "user" if expected_role == "assistant" else "assistant"
-
-        # Ensure the conversation ends with an assistant's turn
-        if expected_role != "user":
-            msg = "The conversation must end with an assistant's turn."
-            raise EvaluationException(
-                message=msg,
-                internal_message=msg,
-                target=ErrorTarget.CHAT_EVALUATOR,
-                category=ErrorCategory.INVALID_VALUE,
-                blame=ErrorBlame.USER_ERROR,
-            )

azure/ai/evaluation/_evaluators/_chat/retrieval/__init__.py

@@ -1,9 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-
-from ._retrieval import RetrievalChatEvaluator
-
-__all__ = [
-    "RetrievalChatEvaluator",
-]

azure/ai/evaluation/_evaluators/_content_safety/_content_safety_base.py

@@ -1,66 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-
-from abc import ABC
-
-from azure.ai.evaluation._common.constants import EvaluationMetrics
-from azure.ai.evaluation._common.rai_service import evaluate_with_rai_service
-from azure.ai.evaluation._exceptions import EvaluationException, ErrorBlame, ErrorCategory, ErrorTarget
-from azure.ai.evaluation._model_configurations import AzureAIProject
-
-
-class ContentSafetyEvaluatorBase(ABC):
-    """
-    Initialize a evaluator for a specified Evaluation Metric. Base class that is not
-    meant to be instantiated by users.
-
-
-    :param metric: The metric to be evaluated.
-    :type metric: ~azure.ai.evaluation._evaluators._content_safety.flow.constants.EvaluationMetrics
-    :param azure_ai_project: The scope of the Azure AI project.
-        It contains subscription id, resource group, and project name.
-    :type azure_ai_project: ~azure.ai.evaluation.AzureAIProject
-    :param credential: The credential for connecting to Azure AI project.
-    :type credential: ~azure.core.credentials.TokenCredential
-    """
-
-    def __init__(self, metric: EvaluationMetrics, azure_ai_project: dict, credential=None):
-        self._metric = metric
-        self._azure_ai_project = azure_ai_project
-        self._credential = credential
-
-    async def __call__(self, *, query: str, response: str, **kwargs):
-        """
-        Evaluates content according to this evaluator's metric.
-
-        :keyword query: The query to be evaluated.
-        :paramtype query: str
-        :keyword response: The response to be evaluated.
-        :paramtype response: str
-        :return: The evaluation score computation based on the Content Safety metric (self.metric).
-        :rtype: Any
-        """
-        # Validate inputs
-        # Raises value error if failed, so execution alone signifies success.
-        if not (query and query.strip() and query != "None") or not (
-            response and response.strip() and response != "None"
-        ):
-            msg = "Both 'query' and 'response' must be non-empty strings."
-            raise EvaluationException(
-                message=msg,
-                internal_message=msg,
-                error_category=ErrorCategory.MISSING_FIELD,
-                error_blame=ErrorBlame.USER_ERROR,
-                error_target=ErrorTarget.CONTENT_SAFETY_CHAT_EVALUATOR,
-            )
-
-        # Run score computation based on supplied metric.
-        result = await evaluate_with_rai_service(
-            metric_name=self._metric,
-            query=query,
-            response=response,
-            project_scope=self._azure_ai_project,
-            credential=self._credential,
-        )
-        return result

azure/ai/evaluation/_evaluators/_protected_materials/__init__.py

@@ -1,5 +0,0 @@
-from ._protected_materials import ProtectedMaterialsEvaluator
-
-__all__ = [
-    "ProtectedMaterialsEvaluator",
-]

azure/ai/evaluation/_evaluators/_protected_materials/_protected_materials.py

@@ -1,104 +0,0 @@
-# ---------------------------------------------------------
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# ---------------------------------------------------------
-from promptflow._utils.async_utils import async_run_allowing_running_loop
-from azure.ai.evaluation._common.constants import EvaluationMetrics
-from azure.ai.evaluation._common.rai_service import evaluate_with_rai_service
-from azure.ai.evaluation._exceptions import EvaluationException, ErrorBlame, ErrorCategory, ErrorTarget
-from azure.ai.evaluation._model_configurations import AzureAIProject
-
-
-class _AsyncProtectedMaterialsEvaluator:
-    def __init__(self, azure_ai_project: dict, credential=None):
-        self._azure_ai_project = azure_ai_project
-        self._credential = credential
-
-    async def __call__(self, *, query: str, response: str, **kwargs):
-        """
-        Evaluates content according to this evaluator's metric.
-
-        :keyword query: The query to be evaluated.
-        :paramtype query: str
-        :keyword response: The response to be evaluated.
-        :paramtype response: str
-        :return: The evaluation score computation based on the Content Safety metric (self.metric).
-        :rtype: Any
-        """
-        # Validate inputs
-        # Raises value error if failed, so execution alone signifies success.
-        if not (query and query.strip() and query != "None") or not (
-            response and response.strip() and response != "None"
-        ):
-            msg = "Both 'query' and 'response' must be non-empty strings."
-            raise EvaluationException(
-                message=msg,
-                internal_message=msg,
-                error_category=ErrorCategory.MISSING_FIELD,
-                error_blame=ErrorBlame.USER_ERROR,
-                error_target=ErrorTarget.PROTECTED_MATERIAL_EVALUATOR,
-            )
-
-        # Run score computation based on supplied metric.
-        result = await evaluate_with_rai_service(
-            metric_name=EvaluationMetrics.PROTECTED_MATERIAL,
-            query=query,
-            response=response,
-            project_scope=self._azure_ai_project,
-            credential=self._credential,
-        )
-        return result
-
-
-class ProtectedMaterialsEvaluator:
-    """
-    Initialize a protected materials evaluator to detect whether protected material
-    is present in your AI system's response. Outputs True or False with AI-generated reasoning.
-
-    :param azure_ai_project: The scope of the Azure AI project.
-        It contains subscription id, resource group, and project name.
-    :type azure_ai_project: ~azure.ai.evaluation.AzureAIProject
-    :param credential: The credential for connecting to Azure AI project.
-    :type credential: ~azure.core.credentials.TokenCredential
-    :return: Whether or not protected material was found in the response, with AI-generated reasoning.
-    :rtype: Dict[str, str]
-
-    **Usage**
-
-    .. code-block:: python
-
-        azure_ai_project = {
-            "subscription_id": "<subscription_id>",
-            "resource_group_name": "<resource_group_name>",
-            "project_name": "<project_name>",
-        }
-        eval_fn = ProtectedMaterialsEvaluator(azure_ai_project)
-        result = eval_fn(query="What is the capital of France?", response="Paris.")
-
-    **Output format**
-
-    .. code-block:: python
-
-        {
-            "label": "False",
-            "reasoning": "This query does not contain any protected material."
-        }
-    """
-
-    def __init__(self, azure_ai_project: dict, credential=None):
-        self._async_evaluator = _AsyncProtectedMaterialsEvaluator(azure_ai_project, credential)
-
-    def __call__(self, *, query: str, response: str, **kwargs):
-        """
-        Evaluates protected materials content.
-
-        :keyword query: The query to be evaluated.
-        :paramtype query: str
-        :keyword response: The response to be evaluated.
-        :paramtype response: str
-        :return: A dictionary containing a boolean label and reasoning.
-        :rtype: dict
-        """
-        return async_run_allowing_running_loop(self._async_evaluator, query=query, response=response, **kwargs)
-
-    def _to_async(self):
-        return self._async_evaluator

azure_ai_evaluation-1.0.0b1.dist-info/RECORD

@@ -1,97 +0,0 @@
-azure/ai/evaluation/__init__.py,sha256=aCQXu_6B3tB-WSRtdPrMUODYMytwK1pVoOTwkR5EItg,1984
-azure/ai/evaluation/_constants.py,sha256=MtXK9FV3TgDiq8IYSMkSDbNXVQsZw62D26g7ZXJ62NU,1422
-azure/ai/evaluation/_exceptions.py,sha256=HUMfvguDc7ygcbs3MTK14R3PK7UxGNWQQHH3hYXIV3U,4168
-azure/ai/evaluation/_http_utils.py,sha256=13G64fWveezBdcmjwwwbJrIkdZp2ns3Jvd-addKw1SU,13983
-azure/ai/evaluation/_model_configurations.py,sha256=D02AzOdyO6LQCia0k232Msd7ro35-EcwmlQ0tOD_5H0,652
-azure/ai/evaluation/_user_agent.py,sha256=O2y-QPBAcw7w7qQ6M2aRPC3Vy3TKd789u5lcs2yuFaI,290
-azure/ai/evaluation/_version.py,sha256=oZ93SfxVsREL_UaqV9_yCzfIy_LG5t9AwK_uEE8HjXI,201
-azure/ai/evaluation/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-azure/ai/evaluation/_common/__init__.py,sha256=LHTkf6dMLLxikrGNgbUuREBVQcs4ORHR6Eryo4bm9M8,586
-azure/ai/evaluation/_common/constants.py,sha256=ZKHGVgGA1Fc6Pvo22X-CeOUX6-m0q_UwpOKOWATTSuI,1639
-azure/ai/evaluation/_common/rai_service.py,sha256=Cci_YnX1XeYHiZX4DTVj6_bAj2JADXh1znXS_cDM45M,17540
-azure/ai/evaluation/_common/utils.py,sha256=Ed2pgAXzswrOskWopYyRGdPMOkwbP-cgj32kBOuWMZw,2941
-azure/ai/evaluation/_evaluate/__init__.py,sha256=Yx1Iq2GNKQ5lYxTotvPwkPL4u0cm6YVxUe-iVbu1clI,180
-azure/ai/evaluation/_evaluate/_eval_run.py,sha256=eUUUzF6o3FjkQ6L5aZnESQjWM_UscM5MnCD-5P0dldA,20237
-azure/ai/evaluation/_evaluate/_evaluate.py,sha256=HP7J39YUn9yMKLl2z-k-7_BVpzQIhDQwFOExQk6nb9w,28379
-azure/ai/evaluation/_evaluate/_utils.py,sha256=-8xv0TF3QQPfZlEpLOCGSq2Nvb-esZfpaDySOHLeey4,9680
-azure/ai/evaluation/_evaluate/_batch_run_client/__init__.py,sha256=BkxhojWca3e2QM3hFwO2xrLiiQ0i-3f8wsMfOx1zchs,361
-azure/ai/evaluation/_evaluate/_batch_run_client/batch_run_context.py,sha256=wL_gNM2u8_DYUKopxFnJFc0cUAxWo9HV5gNSfZoluJo,2983
-azure/ai/evaluation/_evaluate/_batch_run_client/code_client.py,sha256=xmck4Wlp0wrzkN6sZsmW-7MMAc1rIw16xwSV52_V8to,7120
-azure/ai/evaluation/_evaluate/_batch_run_client/proxy_client.py,sha256=_2NynIyH_QFZQf5TDZ9x_FtDUXrzvMfGVuTtpZJazpo,1981
-azure/ai/evaluation/_evaluate/_telemetry/__init__.py,sha256=gF0mHhCYjrlOBgR4I6mQ5KIfExpq7W2On31ITmjYN8M,6400
-azure/ai/evaluation/_evaluators/__init__.py,sha256=Yx1Iq2GNKQ5lYxTotvPwkPL4u0cm6YVxUe-iVbu1clI,180
-azure/ai/evaluation/_evaluators/_bleu/__init__.py,sha256=quKKO0kvOSkky5hcoNBvgBuMeeVRFCE9GSv70mAdGP4,260
-azure/ai/evaluation/_evaluators/_bleu/_bleu.py,sha256=q8hTVC8Ubfb0lQmlGnsuEb6J7Qx5c7eYso4ss6RhgOw,2413
-azure/ai/evaluation/_evaluators/_chat/__init__.py,sha256=xOsSHYNGJJiZvBMPbmLd_-ZZs8_15Sblvk-OF7iVoIo,250
-azure/ai/evaluation/_evaluators/_chat/_chat.py,sha256=GsK7EERUNo5uuh025EazxpK6wuN7mzz891_l8WFxqXg,14692
-azure/ai/evaluation/_evaluators/_chat/retrieval/__init__.py,sha256=DmBjBkwDDlCsSGpBeXfpfMM9ekxIJs62dij4rBXND7k,273
-azure/ai/evaluation/_evaluators/_chat/retrieval/_retrieval.py,sha256=X_2UzFd-FP032nT4fjH1EY67bvIcRpmmjv9KQrXpfUE,5739
-azure/ai/evaluation/_evaluators/_chat/retrieval/retrieval.prompty,sha256=NoHNDf_UE8BGAixqZPhRQ4ecxOUi9w9gO8HfHhJuxaY,1761
-azure/ai/evaluation/_evaluators/_coherence/__init__.py,sha256=GRqcSCQse02Spyki0UsRNWMIXiea2lLtPPXNGvkJzQ0,258
-azure/ai/evaluation/_evaluators/_coherence/_coherence.py,sha256=sOP2rYwGe1k_80FhBr5PGG8metmpJwGfx4U7-aVroIs,4121
-azure/ai/evaluation/_evaluators/_coherence/coherence.prompty,sha256=WVEXxKmh_Gbb11_00N2WCIIJSMgPssFxJ5h2--rMG-w,2725
-azure/ai/evaluation/_evaluators/_content_safety/__init__.py,sha256=mR5CbcMyxV9GQoY71Saoi0bQTpEB74HrYmM8gcVhnAg,746
-azure/ai/evaluation/_evaluators/_content_safety/_content_safety.py,sha256=VUazS5AcJ5Z-BOrqfCRt5SdRhmTlxWdxl03FML0zHq4,3998
-azure/ai/evaluation/_evaluators/_content_safety/_content_safety_base.py,sha256=Yb2x-sI5_GivAe8o8BvA2gM64yQCv73v5LS_QyiR9vQ,2802
-azure/ai/evaluation/_evaluators/_content_safety/_content_safety_chat.py,sha256=7cLTnbz0iHderZeXPKYbbBSwvKYZtMiExEMYQSfggd4,11909
-azure/ai/evaluation/_evaluators/_content_safety/_hate_unfairness.py,sha256=Ur1gQPBdr0b_Ov8K0Jflmvua2mgSrk9DhlPOAQR-7w4,2875
-azure/ai/evaluation/_evaluators/_content_safety/_self_harm.py,sha256=MrCOPNmv6PAkVFKDrwNfXeOyhTmyDa2Tik_xEFKxYwg,2676
-azure/ai/evaluation/_evaluators/_content_safety/_sexual.py,sha256=MyRI9QtqElb2gDGPGnyQYfAHsjdw_oK6Ekjz4Voo9rg,2644
-azure/ai/evaluation/_evaluators/_content_safety/_violence.py,sha256=YH0NxossokGX5tFKAIZJvMiprs34KuiNQnMXAbOOMCM,2668
-azure/ai/evaluation/_evaluators/_eci/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-azure/ai/evaluation/_evaluators/_eci/_eci.py,sha256=vm8G-YvZuXipvJSixA6mktEL7xQvqnsHcD1trqh7Cxw,4058
-azure/ai/evaluation/_evaluators/_f1_score/__init__.py,sha256=aEVbO7iMoF20obdpLQKcKm69Yyu3mYnblKELLqu8OGI,260
-azure/ai/evaluation/_evaluators/_f1_score/_f1_score.py,sha256=KNMni1m8dtaC6kLf9JRUVfJ1hZfYBi-g6iZiruJ_Hyc,4553
-azure/ai/evaluation/_evaluators/_fluency/__init__.py,sha256=EEJw39xRa0bOAA1rELTTKXQu2s60n_7CZQRD0Gu2QVw,259
-azure/ai/evaluation/_evaluators/_fluency/_fluency.py,sha256=0KSSYQjWzPpGLBK-NIkIdfp0fPcCOpjIkuKMbYprqI0,4082
-azure/ai/evaluation/_evaluators/_fluency/fluency.prompty,sha256=RparSdDZs-xiGbq7lRifz9z7jaD10ldXDU3E7sO0v2s,2579
-azure/ai/evaluation/_evaluators/_gleu/__init__.py,sha256=Ae2EvQ7gqiYAoNO3LwGIhdAAjJPJDfT85rQGKrRrmbA,260
-azure/ai/evaluation/_evaluators/_gleu/_gleu.py,sha256=qQR5Ic4K4biHrMm8XLe1DtKc3MSwbfCjK_-CLpL9aCc,2326
-azure/ai/evaluation/_evaluators/_groundedness/__init__.py,sha256=UYNJUeRvBwcSVFyZpdsf29un5eyaDzYoo3QvC1gvlLg,274
-azure/ai/evaluation/_evaluators/_groundedness/_groundedness.py,sha256=BZdr2Gqgos1CDTnMPJPUfoa_Pkq9WxgdL-Qem6GbINs,4301
-azure/ai/evaluation/_evaluators/_groundedness/groundedness.prompty,sha256=dDclp_YowBjeiGhwmCxCnS4A3K9r4v2tzsUm-ccLt-I,3199
-azure/ai/evaluation/_evaluators/_meteor/__init__.py,sha256=209na3pPsdmcuYpYHUYtqQybCpc3yZkc93HnRdicSlI,266
-azure/ai/evaluation/_evaluators/_meteor/_meteor.py,sha256=62AD1DVH_T_t9MDK4GEcvujC0Aka-SEggxEAmc68byg,3315
-azure/ai/evaluation/_evaluators/_protected_material/__init__.py,sha256=eRAQIU9diVXfO5bp6aLWxZoYUvOsrDIfy1gnDOeNTiI,109
-azure/ai/evaluation/_evaluators/_protected_material/_protected_material.py,sha256=6eZXjp2iXSaTNCx_8_8A96lhH5cdm3rmeJfvzMJHV9k,4233
-azure/ai/evaluation/_evaluators/_protected_materials/__init__.py,sha256=A12UsRVIebGvy9FtZLBPsOIAWUskBt8iuhRdILyRcSo,112
-azure/ai/evaluation/_evaluators/_protected_materials/_protected_materials.py,sha256=pRB48BwbnCuozVo7hpblh_ZnApWzmrqiTx0JqhH4ioY,4204
-azure/ai/evaluation/_evaluators/_qa/__init__.py,sha256=bcXfT--C0hjym2haqd1B2-u9bDciyM0ThOFtU1Q69sk,244
-azure/ai/evaluation/_evaluators/_qa/_qa.py,sha256=QSpk4wKIBdwVaI0-u4iRVlnezgxhPnSEAWRUxSwOJM8,3792
-azure/ai/evaluation/_evaluators/_relevance/__init__.py,sha256=JlxytW32Nl8pbE-fI3GRpfgVuY9EG6zxIAn5VZGSwyc,265
-azure/ai/evaluation/_evaluators/_relevance/_relevance.py,sha256=YFLxC8VQ9kvEGK6jge7_yQCxXM1yPChy9d-dANVJA60,4507
-azure/ai/evaluation/_evaluators/_relevance/relevance.prompty,sha256=AO70ho2nMhBtKcl_q4nKFW4kA1LjYsmSfymNa-Cbcrw,3735
-azure/ai/evaluation/_evaluators/_rouge/__init__.py,sha256=kusCDaYcXogDugGefRP8MQSn9xv107oDbrMCqZ6K4GA,291
-azure/ai/evaluation/_evaluators/_rouge/_rouge.py,sha256=IbD10QZMYzaXKqr6q4Diy7p06yhJ3kT8pMv2jnyMNtQ,3444
-azure/ai/evaluation/_evaluators/_similarity/__init__.py,sha256=V2Mspog99_WBltxTkRHG5NpN5s9XoiTSN4I8POWEkLA,268
-azure/ai/evaluation/_evaluators/_similarity/_similarity.py,sha256=z_mUu0lL_gZI041PZvIBgSZ9Aj7j8wUgyy5OPFlBJ_w,4495
-azure/ai/evaluation/_evaluators/_similarity/similarity.prompty,sha256=p2Tb4IW6QnP2BaGRQsAicW4V0B23Oezhf5l3Hau0nxU,4770
-azure/ai/evaluation/_evaluators/_xpia/__init__.py,sha256=VMEL8WrpJQeh4sQiOLzP7hRFPnjzsvwfvTzaGCVJPCM,88
-azure/ai/evaluation/_evaluators/_xpia/xpia.py,sha256=-A9EpdZXr5wZzo_-cF_OX8mMAjy8AgKjNXt1D6Up0NU,5819
-azure/ai/evaluation/simulator/__init__.py,sha256=4O7O6esevfOX0Vb7Eo6nl_m9Qx_WDytwfSgZ7-E1HkE,485
-azure/ai/evaluation/simulator/_adversarial_scenario.py,sha256=SxpyMw5wmM5-fiUjl1_oJH0GQEnsa7ASso10MAr2Hjw,1030
-azure/ai/evaluation/simulator/_adversarial_simulator.py,sha256=toUDdMOD4JrsTOh9tsFTLTBz87o6YLgoY6djXnbO7Fk,20577
-azure/ai/evaluation/simulator/_constants.py,sha256=xM-Or2x7RytfoeBM3N7Vt4JQDJX66UdL3CPz0YN5rvE,485
-azure/ai/evaluation/simulator/_direct_attack_simulator.py,sha256=ACd4HZBtY18stKBTGW7iHEq0hn42gNK453V_pioOcxA,11702
-azure/ai/evaluation/simulator/_indirect_attack_simulator.py,sha256=4HX_034m2j8CBNBVSap9CfwKJX00l0fbvuXJyuZYavU,9639
-azure/ai/evaluation/simulator/_tracing.py,sha256=svCnjlI09AQ4aNH8yeXZiE6RKg9Pi4q66oQ8LstKJBk,2968
-azure/ai/evaluation/simulator/_utils.py,sha256=v-_nzpNj8RYnSyBeXcdxqu-jj1Xj_UgYdc21Gty8uwY,4219
-azure/ai/evaluation/simulator/simulator.py,sha256=902a7P5DGrkN1blbCTzHGDEFhA60LlxdjVEJVVa4cSQ,28464
-azure/ai/evaluation/simulator/_conversation/__init__.py,sha256=Qxu2x1cJ2fPTQtYhhNP5C4-d8P33Sru4Zn6xAgroDso,12774
-azure/ai/evaluation/simulator/_conversation/_conversation.py,sha256=9a2plS1QzyuvWbJYj6rY128401e6d_n_yiBNekIDkww,7251
-azure/ai/evaluation/simulator/_conversation/constants.py,sha256=3v7zkjPwJAPbSpJYIK6VOZZy70bJXMo_QTVqSFGlq9A,984
-azure/ai/evaluation/simulator/_helpers/__init__.py,sha256=CD2fGrUgEE3puIm5QYqbkCN1AtntpF-3hRbF98hkhqE,203
-azure/ai/evaluation/simulator/_helpers/_language_suffix_mapping.py,sha256=7BBLH78b7YDelHDLbAIwf-IO9s9cAEtn-RRXmNReHdc,1017
-azure/ai/evaluation/simulator/_helpers/_simulator_data_classes.py,sha256=NjsmHy-bb0yeqKVz9j7f7KIlTEB7MeZe1NzFOLBdBm4,2531
-azure/ai/evaluation/simulator/_model_tools/__init__.py,sha256=aMv5apb7uVjuhMF9ohhA5kQmo652hrGIJlhdl3y2R1I,835
-azure/ai/evaluation/simulator/_model_tools/_identity_manager.py,sha256=qhYmG2r4IAKDePtf9DdqgvNGYlc0xjH4x5sShGxS-lA,5070
-azure/ai/evaluation/simulator/_model_tools/_proxy_completion_model.py,sha256=fHE6OUc1drkABwGQ5L0ctrqARPjv5uwO-u7yOJhdCuQ,8397
-azure/ai/evaluation/simulator/_model_tools/_rai_client.py,sha256=ITPiciPQQkZa5s2E8pa-x3lKrf8rp7sbr_WSPpECOnc,6725
-azure/ai/evaluation/simulator/_model_tools/_template_handler.py,sha256=hpxBYui1eaw9_lDJJMnKj8CSWFDXiTa7XAKXA4lS0J4,5482
-azure/ai/evaluation/simulator/_model_tools/models.py,sha256=5pUESJR5b5rzp9W7B3Aaokz_HLHTSLjR13YKumLAxqs,21813
-azure/ai/evaluation/simulator/_prompty/task_query_response.prompty,sha256=lNSioz2XiGQJb-AXRNYm2JCLMivZKa3JlHfol2Jd7fY,2244
-azure/ai/evaluation/simulator/_prompty/task_simulate.prompty,sha256=00zLVfNgHZdlbC2XvBedSrwDJOaAhl3B1ohE3LKsGg4,928
-azure_ai_evaluation-1.0.0b1.dist-info/METADATA,sha256=1XGkdZ7U-s9NVdHGktn_nk5S8BBUUQA-jrf-_VUpdVU,12268
-azure_ai_evaluation-1.0.0b1.dist-info/WHEEL,sha256=HiCZjzuy6Dw0hdX5R3LCFPDmFS4BWl8H-8W39XfmgX4,91
-azure_ai_evaluation-1.0.0b1.dist-info/top_level.txt,sha256=S7DhWV9m80TBzAhOFjxDUiNbKszzoThbnrSz5MpbHSQ,6
-azure_ai_evaluation-1.0.0b1.dist-info/RECORD,,