update Runner call to check for supports_base_model attribute instead of try except

guardrails/llm_providers.py

@@ -32,6 +32,8 @@ class PromptCallableBase:
     failed, and how to fix it.
     """
 
+    supports_base_model = False
+
     def __init__(self, *args, **kwargs):
         self.init_args = args
         self.init_kwargs = kwargs
@@ -119,6 +121,8 @@ def _invoke_llm(
 
 
 class OpenAIChatCallable(PromptCallableBase):
+    supports_base_model = True
+
     def _invoke_llm(
         self,
         text: Optional[str] = None,
@@ -588,6 +592,8 @@ async def invoke_llm(
 
 
 class AsyncOpenAIChatCallable(AsyncPromptCallableBase):
+    supports_base_model = True
+
     async def invoke_llm(
         self,
         text: Optional[str] = None,

guardrails/run.py

@@ -1,5 +1,6 @@
 import copy
 import json
+from functools import partial
 from typing import Any, Dict, List, Optional, Sequence, Tuple, Type, Union
 
 from eliot import add_destinations, start_action
@@ -549,36 +550,24 @@ def call(
         3. Log the output
         """
 
+        # If the API supports a base model, pass it in.
+        api_fn = api
+        if api is not None:
+            supports_base_model = getattr(api, "supports_base_model", False)
+            if supports_base_model:
+                api_fn = partial(api, base_model=self.base_model)
+
         with start_action(action_type="call", index=index, prompt=prompt) as action:
             if output is not None:
-                llm_response = LLMResponse(
-                    output=output,
-                )
-            elif api is None:
+                llm_response = LLMResponse(output=output)
+            elif api_fn is None:
                 raise ValueError("API or output must be provided.")
             elif msg_history:
-                try:
-                    llm_response = api(
-                        msg_history=msg_history_source(msg_history),
-                        base_model=self.base_model,
-                    )
-                except Exception:
-                    # If the API call fails, try calling again without the base model.
-                    llm_response = api(msg_history=msg_history_source(msg_history))
+                llm_response = api_fn(msg_history=msg_history_source(msg_history))
             elif prompt and instructions:
-                try:
-                    llm_response = api(
-                        prompt.source,
-                        instructions=instructions.source,
-                        base_model=self.base_model,
-                    )
-                except Exception:
-                    llm_response = api(prompt.source, instructions=instructions.source)
+                llm_response = api_fn(prompt.source, instructions=instructions.source)
             elif prompt:
-                try:
-                    llm_response = api(prompt.source, base_model=self.base_model)
-                except Exception:
-                    llm_response = api(prompt.source)
+                llm_response = api_fn(prompt.source)
             else:
                 raise ValueError("'prompt' or 'msg_history' must be provided.")
 

tests/unit_tests/validators/test_competitor_check.py

@@ -4,15 +4,7 @@
 from guardrails.validators import CompetitorCheck, FailResult
 
 
-class TestCompetitorCheck:  # (unittest.TestCase):
-    def setUp(self):
-        # You can initialize any required resources or configurations here.
-        pass
-
-    def tearDown(self):
-        # You can clean up any resources used during the tests here.
-        pass
-
+class TestCompetitorCheck:
     def test_perform_ner(self, mocker):
         # Create a mock NLP object
         mock_util_is_package = mocker.patch("spacy.util.is_package")