feat: backend model load balancing support (#4927)

api/tests/integration_tests/utils/test_module_import_helper.py

@@ -1,6 +1,6 @@
 import os
 
-from core.utils.module_import_helper import import_module_from_source, load_single_subclass_from_source
+from core.helper.module_import_helper import import_module_from_source, load_single_subclass_from_source
 from tests.integration_tests.utils.parent_class import ParentClass
 
 

api/tests/integration_tests/workflow/nodes/test_llm.py

@@ -92,7 +92,8 @@ def test_execute_llm(setup_openai_mock):
                 provider=CustomProviderConfiguration(
                     credentials=credentials
                 )
-            )
+            ),
+            model_settings=[]
         ),
         provider_instance=provider_instance,
         model_type_instance=model_type_instance
@@ -206,10 +207,11 @@ def test_execute_llm_with_jinja2(setup_code_executor_mock, setup_openai_mock):
                 provider=CustomProviderConfiguration(
                     credentials=credentials
                 )
-            )
+            ),
+            model_settings=[]
         ),
         provider_instance=provider_instance,
-        model_type_instance=model_type_instance
+        model_type_instance=model_type_instance,
     )
 
     model_instance = ModelInstance(provider_model_bundle=provider_model_bundle, model='gpt-3.5-turbo')

api/tests/integration_tests/workflow/nodes/test_parameter_extractor.py

@@ -42,7 +42,8 @@ def get_mocked_fetch_model_config(
                 provider=CustomProviderConfiguration(
                     credentials=credentials
                 )
-            )
+            ),
+            model_settings=[]
         ),
         provider_instance=provider_instance,
         model_type_instance=model_type_instance