test: switch LLM API tests to qwen3.7-max

tests/ops/aggregator/test_entity_attribute_aggregator.py

@@ -40,7 +40,7 @@ def test_default_aggregator(self):
             },
         ]
         op = EntityAttributeAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             attribute='主要经历'
         )
@@ -59,7 +59,7 @@ def test_input_output(self):
             },
         ]
         op = EntityAttributeAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             attribute='身份背景',
             input_key='sub_docs',
@@ -80,7 +80,7 @@ def test_max_token_num(self):
             },
         ]
         op = EntityAttributeAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             attribute='身份背景',
             max_token_num=200
@@ -100,7 +100,7 @@ def test_word_limit_num(self):
             },
         ]
         op = EntityAttributeAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             attribute='身份背景',
             word_limit=20
@@ -128,7 +128,7 @@ def test_example_prompt(self):
             '孙行者、齐天大圣、美猴王\n'
         )
         op = EntityAttributeAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             attribute='另外身份',
             example_prompt=example_prompt,

tests/ops/aggregator/test_meta_tags_aggregator.py

@@ -48,7 +48,7 @@ def test_default_aggregator(self):
             },
         ]
         op = MetaTagsAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             meta_tag_key=MetaKeys.query_sentiment_label,
         )
         self._run_helper(op, samples)
@@ -77,7 +77,7 @@ def test_target_tags(self):
             },
         ]
         op = MetaTagsAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             meta_tag_key=MetaKeys.query_sentiment_label,
             target_tags=['开心', '难过', '其他']
         )
@@ -106,7 +106,7 @@ def test_tag_list(self):
             },
         ]
         op = MetaTagsAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             meta_tag_key=MetaKeys.dialog_sentiment_labels,
             target_tags=['开心', '难过', '其他']
         )

tests/ops/aggregator/test_most_relevant_entities_aggregator.py

@@ -42,7 +42,7 @@ def test_default_aggregator(self):
         ]
 
         op = MostRelevantEntitiesAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             query_entity_type='人物'
         )
@@ -62,7 +62,7 @@ def test_input_output(self):
         ]
 
         op = MostRelevantEntitiesAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             query_entity_type='人物',
             input_key='events',
@@ -83,7 +83,7 @@ def test_max_token_num(self):
             },
         ]
         op = MostRelevantEntitiesAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             entity='李莲花',
             query_entity_type='人物',
             max_token_num=40

tests/ops/aggregator/test_nested_aggregator.py

@@ -41,7 +41,7 @@ def test_default_aggregator(self):
             },
         ]
         op = NestedAggregator(
-            api_model='qwen2.5-72b-instruct'
+            api_model='qwen3.6-plus'
         )
         self._run_helper(op, samples)
 
@@ -58,7 +58,7 @@ def test_input_output(self):
             },
         ]
         op = NestedAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             input_key='sub_docs',
             output_key='text'
         )
@@ -77,7 +77,7 @@ def test_max_token_num_1(self):
             },
         ]
         op = NestedAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             max_token_num=2
         )
         self._run_helper(op, samples)
@@ -95,7 +95,7 @@ def test_max_token_num_2(self):
             },
         ]
         op = NestedAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             max_token_num=90
         )
         self._run_helper(op, samples)
@@ -113,7 +113,7 @@ def test_max_token_num_3(self):
             },
         ]
         op = NestedAggregator(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             max_token_num=200
         )
         self._run_helper(op, samples)

tests/ops/filter/test_llm_analysis_filter.py

@@ -7,7 +7,7 @@
 
 @skip_if_from_fork("Skipping API-based test because running from a fork repo")
 class LLMAnalysisFilterTest(DataJuicerTestCaseBase):
-    api_or_hf_model = 'qwen2.5-72b-instruct'
+    api_or_hf_model = 'qwen3.6-plus'
 
     def _run_test(self, dataset: Dataset, op):
         if Fields.stats not in dataset.features:

tests/ops/filter/test_llm_difficulty_score_filter.py

@@ -10,7 +10,7 @@
 
 @skip_if_from_fork("Skipping API-based test because running from a fork repo")
 class LLMDifficultyScoreFilterTest(DataJuicerTestCaseBase):
-    api_or_hf_model = 'qwen2.5-72b-instruct'
+    api_or_hf_model = 'qwen3.6-plus'
 
     def _run_test(self, dataset: Dataset, op):
         if Fields.stats not in dataset.features:

tests/ops/filter/test_llm_quality_score_filter.py

@@ -10,7 +10,7 @@
 
 @skip_if_from_fork("Skipping API-based test because running from a fork repo")
 class LLMQualityScoreFilterTest(DataJuicerTestCaseBase):
-    api_or_hf_model = 'qwen2.5-72b-instruct'
+    api_or_hf_model = 'qwen3.6-plus'
 
     def _run_test(self, dataset: Dataset, op):
         if Fields.stats not in dataset.features:

tests/ops/filter/test_llm_task_relevance_filter.py

@@ -10,7 +10,7 @@
 
 @skip_if_from_fork("Skipping API-based test because running from a fork repo")
 class LLMTaskRelevanceFilterTest(DataJuicerTestCaseBase):
-    api_or_hf_model = 'qwen2.5-72b-instruct'
+    api_or_hf_model = 'qwen3.6-plus'
 
     def _run_test(self, dataset: Dataset, op, tgt_list):
         if Fields.stats not in dataset.features:

tests/ops/mapper/test_calibrate_qa_mapper.py

@@ -62,12 +62,12 @@ def test(self):
         # before running this test, set below environment variables:
         # export OPENAI_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1/
         # export OPENAI_API_KEY=your_dashscope_key
-        op = CalibrateQAMapper(api_model='qwen2.5-72b-instruct')
+        op = CalibrateQAMapper(api_model='qwen3.6-plus')
         self._run_op(op)
 
     def test_args(self):
         op = CalibrateQAMapper(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             api_endpoint=
             'https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions',
             response_path='choices.0.message.content')

tests/ops/mapper/test_calibrate_query_mapper.py

@@ -66,7 +66,7 @@ def test(self):
         # before running this test, set below environment variables:
         # export OPENAI_API_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
         # export OPENAI_API_KEY=your_key
-        self._run_op('qwen2.5-72b-instruct')
+        self._run_op('qwen3.6-plus')
 
 
 if __name__ == '__main__':

tests/ops/mapper/test_calibrate_response_mapper.py

@@ -68,7 +68,7 @@ def test(self):
         # before running this test, set below environment variables:
         # export OPENAI_API_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
         # export OPENAI_API_KEY=your_key
-        self._run_op('qwen2.5-72b-instruct')
+        self._run_op('qwen3.6-plus')
 
 
 if __name__ == '__main__':

tests/ops/mapper/test_dialog_intent_detection_mapper.py

@@ -54,7 +54,7 @@ def test_default(self):
             ]
         }]
 
-        op = DialogIntentDetectionMapper(api_model='qwen2.5-72b-instruct')
+        op = DialogIntentDetectionMapper(api_model='qwen3.6-plus')
         self._run_op(op, samples)
 
     def test_max_round(self):
@@ -80,7 +80,7 @@ def test_max_round(self):
             ]
         }]
 
-        op = DialogIntentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogIntentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -107,7 +107,7 @@ def test_max_round_zero(self):
             ]
         }]
 
-        op = DialogIntentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogIntentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=0)
         self._run_op(op, samples)
 
@@ -132,7 +132,7 @@ def test_query(self):
             'response': '「委屈」我也没说什么呀，就是觉得你有点冤枉我了'
         }]
 
-        op = DialogIntentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogIntentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -160,7 +160,7 @@ def test_intent_candidates(self):
         }]
 
         op = DialogIntentDetectionMapper(
-            api_model='qwen2.5-72b-instruct',
+            api_model='qwen3.6-plus',
             intent_candidates=['评价', '讽刺', '表达困惑']
             )
         self._run_op(op, samples)
@@ -190,7 +190,7 @@ def test_rename_keys(self):
 
         labels_key = 'my_label'
         analysis_key = 'my_analysis'
-        op = DialogIntentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogIntentDetectionMapper(api_model='qwen3.6-plus',
                                         labels_key=labels_key,
                                         analysis_key=analysis_key)
         self._run_op(op, samples, labels_key=labels_key, analysis_key=analysis_key)

tests/ops/mapper/test_dialog_sentiment_detection_mapper.py

@@ -54,7 +54,7 @@ def test_default(self):
             ]
         }]
 
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct')
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus')
         self._run_op(op, samples)
 
     def test_max_round(self):
@@ -80,7 +80,7 @@ def test_max_round(self):
             ]
         }]
 
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -107,7 +107,7 @@ def test_max_round_zero(self):
             ]
         }]
 
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=0)
         self._run_op(op, samples)
 
@@ -132,7 +132,7 @@ def test_query(self):
             'response': '「委屈」我也没说什么呀，就是觉得你有点冤枉我了'
         }]
 
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -159,7 +159,7 @@ def test_sentiment_candidates(self):
             ]
         }]
 
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus',
                                 sentiment_candidates=['认可', '不满', '困惑'])
         self._run_op(op, samples)
 
@@ -188,7 +188,7 @@ def test_rename_keys(self):
 
         labels_key = 'my_label'
         analysis_key = 'my_analysis'
-        op = DialogSentimentDetectionMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentDetectionMapper(api_model='qwen3.6-plus',
                                             labels_key=labels_key,
                                             analysis_key=analysis_key)
         self._run_op(op, samples, labels_key=labels_key, analysis_key=analysis_key)

tests/ops/mapper/test_dialog_sentiment_intensity_mapper.py

@@ -53,7 +53,7 @@ def test_default(self):
             ]
         }]
 
-        op = DialogSentimentIntensityMapper(api_model='qwen2.5-72b-instruct')
+        op = DialogSentimentIntensityMapper(api_model='qwen3.6-plus')
         self._run_op(op, samples)
 
     def test_max_round(self):
@@ -79,7 +79,7 @@ def test_max_round(self):
             ]
         }]
 
-        op = DialogSentimentIntensityMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentIntensityMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -106,7 +106,7 @@ def test_max_round_zero(self):
             ]
         }]
 
-        op = DialogSentimentIntensityMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentIntensityMapper(api_model='qwen3.6-plus',
                                             max_round=0)
         self._run_op(op, samples)
 
@@ -131,7 +131,7 @@ def test_query(self):
             'response': '「委屈」我也没说什么呀，就是觉得你有点冤枉我了'
         }]
 
-        op = DialogSentimentIntensityMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentIntensityMapper(api_model='qwen3.6-plus',
                                             max_round=1)
         self._run_op(op, samples)
 
@@ -160,7 +160,7 @@ def test_rename_keys(self):
 
         intensities_key = 'my_intensity'
         analysis_key = 'my_analysis'
-        op = DialogSentimentIntensityMapper(api_model='qwen2.5-72b-instruct',
+        op = DialogSentimentIntensityMapper(api_model='qwen3.6-plus',
                                             intensities_key=intensities_key,
                                             analysis_key=analysis_key)
         self._run_op(op, samples, intensities_key=intensities_key, analysis_key=analysis_key)