eval-ai-library 0.3.0__py3-none-any.whl → 0.3.2__py3-none-any.whl

{eval_ai_library-0.3.0.dist-info → eval_ai_library-0.3.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: eval-ai-library
-Version: 0.3.0
+Version: 0.3.2
 Summary: Comprehensive AI Model Evaluation Framework with support for multiple LLM providers
 Author-email: Aleksandr Meshkov <alekslynx90@gmail.com>
 License: MIT

{eval_ai_library-0.3.0.dist-info → eval_ai_library-0.3.2.dist-info}/RECORD

@@ -1,34 +1,34 @@
-eval_ai_library-0.3.0.dist-info/licenses/LICENSE,sha256=rK9uLDgWNrCHNdp-Zma_XghDE7Fs0u0kDi3WMcmYx6w,1074
-eval_lib/__init__.py,sha256=BA0vmdi5_3Zd1ib8nLyq6pESBVnFX7aKdvRAbR_I9bQ,3029
+eval_ai_library-0.3.2.dist-info/licenses/LICENSE,sha256=rK9uLDgWNrCHNdp-Zma_XghDE7Fs0u0kDi3WMcmYx6w,1074
+eval_lib/__init__.py,sha256=IeDW5pLarPmHCBJu-6vFX71g9VxZTS46UkHjDkrU_Gw,3043
 eval_lib/evaluate.py,sha256=GjlXZb5dnl44LCaJwdkyGCYcC50zoNZn3NrofzNAVJ0,11490
 eval_lib/evaluation_schema.py,sha256=7IDd_uozqewhh7k0p1hKut_20udvRxxkV6thclxKUg0,1904
-eval_lib/llm_client.py,sha256=6uNLIAfSdj4u_n4SAk3UkYmxcfO2Y2y-9Kp0T4bSjPM,13845
+eval_lib/llm_client.py,sha256=3eMcarKLkDLDVh4AOxgWbaIzXlzpqsmEfJXNTBonNic,13633
 eval_lib/metric_pattern.py,sha256=wULgMNDeAqJC_Qjglo7bYzY2eGhA_PmY_hA_qGfg0sI,11730
 eval_lib/price.py,sha256=jbmkkUTxPuXrkSHuaJYPl7jSzfDIzQ9p_swWWs26UJ0,1986
 eval_lib/py.typed,sha256=8PjyZ1aVoQpRVvt71muvuq5qE-jTFZkK-GLHkhdebmc,26
 eval_lib/testcases_schema.py,sha256=qI4o6kX0jH1DR3sHGXUnu3Cyt2oq7rGlsMlOaXSt6F4,696
 eval_lib/utils.py,sha256=-hwagFFn3_QjgyLqF8Qx7JIkpgOEI8-F14eycog3bgc,3141
 eval_lib/agent_metrics/__init__.py,sha256=20Y4BsicD2s7OkOBQPBvB2JKStBDtplv52_q6q35Vgo,525
-eval_lib/agent_metrics/knowledge_retention_metric/knowledge_retention.py,sha256=_9MRWbv8oi6goY-ZXWu7FaNV0vQX1UD1w1Ar7CPVino,8244
-eval_lib/agent_metrics/role_adherence_metric/role_adherence.py,sha256=D9MuKUI8ujh766vF8VQO73m-fPSQxy5u2Tkcr3wHbVk,9180
-eval_lib/agent_metrics/task_success_metric/task_success_rate.py,sha256=mt1PRHadup2k64gF_OyL8d-eQ5zm2EeNwztcE-aGOe0,12452
+eval_lib/agent_metrics/knowledge_retention_metric/knowledge_retention.py,sha256=7j89HOTsu0rMoFnznTjMl-tqQpnZlS6ZIdrHPueEbb8,8289
+eval_lib/agent_metrics/role_adherence_metric/role_adherence.py,sha256=kJsYj9H3W3Mw2iBqj3Br_glP8gU6_diFPiJhRSnHGxg,9225
+eval_lib/agent_metrics/task_success_metric/task_success_rate.py,sha256=v5cO07cymo9GWSZ34ryAx3ya4DDBiRWih9w0bm_j_R8,12497
 eval_lib/agent_metrics/tools_correctness_metric/tool_correctness.py,sha256=qbVMtD6EWKah27FogLEiEh6pBX-k2wwKbwM_kFkvYeQ,4220
 eval_lib/datagenerator/datagenerator.py,sha256=NQZIQuSCmryxIT3lTLS1PpJjENmGqARtR-zTiQ8OvRk,15513
 eval_lib/datagenerator/document_loader.py,sha256=vnQUz_Dxb3SxcVPUmMXZe-rgfPp6OfHb6D2Ie9iqPms,17025
 eval_lib/datagenerator/prompts.py,sha256=iQAYitAbLud3vWJnXGh_OCF4syWYS_S71zZEPI6qYAU,7213
 eval_lib/metrics/__init__.py,sha256=3qClCCjPXt5i0c38g5krfuQnqlAXEl-jhAHy1C_ICMY,1213
 eval_lib/metrics/answer_precision_metric/answer_precision.py,sha256=AxPmwzGFU7tnTrrZuQZ7ow4nNSD-blDHdAGwhMHMxjM,15040
-eval_lib/metrics/answer_relevancy_metric/answer_relevancy.py,sha256=LHf7FI9dYzZcKBdxUH9VpsOceRLh0NMb79qliZRE3Uo,8477
+eval_lib/metrics/answer_relevancy_metric/answer_relevancy.py,sha256=-Xb9I-BVMDf5E55FbJzP6IyvD6IVTUPBI-uCrRnEboc,8522
 eval_lib/metrics/bias_metric/bias.py,sha256=BVH8xlTUTRfVG_F1kauwpGAkVKBkUWhM9rUsrrLhpRU,4020
 eval_lib/metrics/contextual_precision_metric/contextual_precision.py,sha256=CQOb6uR2KeffTkhPSqZae56sX5tXMr0pJVM5W_wU1fU,3993
 eval_lib/metrics/contextual_recall_metric/contextual_recall.py,sha256=iw73_hGLWklHZSBkCRkPDNUt1xD5dknA_7CZ6Efkf5w,3913
-eval_lib/metrics/contextual_relevancy_metric/contextual_relevancy.py,sha256=pjyQPVDzWfTnqTgy5D1OYq1KdsCP1bVeUg3rcV8El4I,6501
-eval_lib/metrics/custom_metric/custom_eval.py,sha256=GrOo39DxHzdMUGnBnCiHfT7Y8VabFFgmYQGcnkyGm1w,11340
+eval_lib/metrics/contextual_relevancy_metric/contextual_relevancy.py,sha256=G1cYlA95YNcpEqQsALVi6ZbyNzWr9ccO2DATcsES5pk,6546
+eval_lib/metrics/custom_metric/custom_eval.py,sha256=Ov3-i6IytaJXlzcMgp46SRSeb8scyhqyuR2BqUtDFoM,11385
 eval_lib/metrics/faithfulness_metric/faithfulness.py,sha256=OqamlhTOps7d-NOStSIK7Tq-UAJXNql8VKjgtuqlDhA,5930
 eval_lib/metrics/geval/geval.py,sha256=mNciHXnqU2drOJsWlYmbwftGiKM89-Ykw2f6XneIGBM,10629
 eval_lib/metrics/restricted_refusal_metric/restricted_refusal.py,sha256=4QqYgGMcp6W9Lw-v4s0AlUhMSOKvBOEgnLvhqVXaT9I,4286
 eval_lib/metrics/toxicity_metric/toxicity.py,sha256=rBE1_fvpbCRdBpBep1y1LTIhofKR8GD4Eh76EOYzxL0,4076
-eval_ai_library-0.3.0.dist-info/METADATA,sha256=s7bRUm49crFB9ICrQaPnJvWHVQ28YUBMlTtlL0L2dWQ,37706
-eval_ai_library-0.3.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-eval_ai_library-0.3.0.dist-info/top_level.txt,sha256=uQHpEd2XI0oZgq1eCww9zMvVgDJgwXMWkCD45fYUzEg,9
-eval_ai_library-0.3.0.dist-info/RECORD,,
+eval_ai_library-0.3.2.dist-info/METADATA,sha256=7kIgiDelzM1wbCdKdy4PXuFudIZ3UCUTPXhlVccFe9k,37706
+eval_ai_library-0.3.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+eval_ai_library-0.3.2.dist-info/top_level.txt,sha256=uQHpEd2XI0oZgq1eCww9zMvVgDJgwXMWkCD45fYUzEg,9
+eval_ai_library-0.3.2.dist-info/RECORD,,

eval_lib/__init__.py

@@ -7,7 +7,7 @@ A powerful library for evaluating AI models with support for multiple LLM provid
 and a wide range of evaluation metrics for RAG systems and AI agents.
 """
 
-__version__ = "0.3.0"
+__version__ = "0.3.2"
 __author__ = "Aleksandr Meshkov"
 
 # Core evaluation functions
@@ -68,12 +68,14 @@ from eval_lib.agent_metrics import (
 
 def __getattr__(name):
     """
-    Ленивый импорт для модулей с тяжёлыми зависимостями.
-    DataGenerator импортируется только когда реально используется.
+    Lazy loading for data generation components.
     """
-    if name == "DataGenerator":
-        from eval_lib.datagenerator.datagenerator import DataGenerator
-        return DataGenerator
+    if name == "DatasetGenerator":
+        from eval_lib.datagenerator.datagenerator import DatasetGenerator
+        return DatasetGenerator
+    if name == "DataGenerator":  # Alias for DatasetGenerator
+        from eval_lib.datagenerator.datagenerator import DatasetGenerator
+        return DatasetGenerator
     if name == "DocumentLoader":
         from eval_lib.datagenerator.document_loader import DocumentLoader
         return DocumentLoader

eval_lib/agent_metrics/knowledge_retention_metric/knowledge_retention.py

@@ -217,6 +217,7 @@ Verdicts:
             "final_score": final_score,
             "comment_final_score": f"Weighted average of verdict scores using softmax aggregation (temperature={self.temperature}).",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the retention score meets the required threshold.",
             "final_reason": summary,

eval_lib/agent_metrics/role_adherence_metric/role_adherence.py

@@ -239,6 +239,7 @@ Return JSON array:
             "final_score": final_score,
             "comment_final_score": f"Weighted average of verdict scores using softmax aggregation (temperature={self.temperature}).",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the role adherence score meets the required threshold.",
             "final_reason": summary,

eval_lib/agent_metrics/task_success_metric/task_success_rate.py

@@ -328,6 +328,7 @@ Criteria: [
             "final_score": final_score,
             "comment_final_score": f"Weighted average of verdict scores using softmax aggregation (temperature={self.temperature}).",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the task success score meets the required threshold.",
             "final_reason": summary,

eval_lib/llm_client.py

@@ -101,7 +101,6 @@ def _get_client(provider: Provider):
     if provider == Provider.AZURE:
         _check_env_var("AZURE_OPENAI_API_KEY", "Azure OpenAI")
         _check_env_var("AZURE_OPENAI_ENDPOINT", "Azure OpenAI")
-        # AZURE_OPENAI_DEPLOYMENT проверяется при вызове, не обязателен здесь
 
         return AsyncAzureOpenAI(
             api_key=os.getenv("AZURE_OPENAI_API_KEY"),
@@ -114,7 +113,6 @@ def _get_client(provider: Provider):
         return genai.Client(api_key=os.getenv("GOOGLE_API_KEY"))
 
     if provider == Provider.OLLAMA:
-        # Ollama может работать без ключа (локальный сервер)
         api_key = _check_env_var(
             "OLLAMA_API_KEY", "Ollama", required=False) or "ollama"
         base_url = _check_env_var(

eval_lib/metrics/answer_relevancy_metric/answer_relevancy.py

@@ -181,6 +181,7 @@ class AnswerRelevancyMetric(MetricPattern):
             "final_score": final_score,
             "comment_final_score": "Score based on the proportion of relevant statements.",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the score exceeds the pass threshold.",
             "final_reason": summary_reason,

eval_lib/metrics/contextual_relevancy_metric/contextual_relevancy.py

@@ -155,6 +155,7 @@ class ContextualRelevancyMetric(MetricPattern):
             "final_score": score,
             "comment_final_score": "Weighted support score from context.",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the score exceeds the threshold.",
             "final_reason": summary,

eval_lib/metrics/custom_metric/custom_eval.py

@@ -312,6 +312,7 @@ JSON:"""
             "final_score": final_score,
             "comment_final_score": f"Weighted average of verdict scores calculated using softmax aggregation (temperature={self.temperature}).",
             "threshold": self.threshold,
+            "temperature": self.temperature,
             "success": success,
             "comment_success": "Whether the final score meets the required threshold.",
             "summary": summary,