together, standard-tests: specify tool_choice in standard tests (#25548)

Here we allow standard tests to specify a value for `tool_choice` via a
`tool_choice_value` property, which defaults to None.

Chat models [available in
Together](https://docs.together.ai/docs/chat-models) have issues passing
standard tool calling tests:
- llama 3.1 models currently [appear to rely on user-side
parsing](https://docs.together.ai/docs/llama-3-function-calling) in
Together;
- Mixtral-8x7B and Mistral-7B (currently tested) consistently do not
call tools in some tests.

Specifying tool_choice also lets us remove an existing `xfail` and use a
smaller model in Groq tests.

libs/standard-tests/langchain_standard_tests/integration_tests/chat_models.py

@@ -170,7 +170,11 @@ class ChatModelIntegrationTests(ChatModelTests):
     def test_tool_calling(self, model: BaseChatModel) -> None:
         if not self.has_tool_calling:
             pytest.skip("Test requires tool calling.")
-        model_with_tools = model.bind_tools([magic_function])
+        if self.tool_choice_value == "tool_name":
+            tool_choice: Optional[str] = "magic_function"
+        else:
+            tool_choice = self.tool_choice_value
+        model_with_tools = model.bind_tools([magic_function], tool_choice=tool_choice)
 
         # Test invoke
         query = "What is the value of magic_function(3)? Use the tool."
@@ -188,7 +192,13 @@ class ChatModelIntegrationTests(ChatModelTests):
         if not self.has_tool_calling:
             pytest.skip("Test requires tool calling.")
 
-        model_with_tools = model.bind_tools([magic_function_no_args])
+        if self.tool_choice_value == "tool_name":
+            tool_choice: Optional[str] = "magic_function_no_args"
+        else:
+            tool_choice = self.tool_choice_value
+        model_with_tools = model.bind_tools(
+            [magic_function_no_args], tool_choice=tool_choice
+        )
         query = "What is the value of magic_function()? Use the tool."
         result = model_with_tools.invoke(query)
         _validate_tool_call_message_no_args(result)
@@ -212,7 +222,11 @@ class ChatModelIntegrationTests(ChatModelTests):
             name="greeting_generator",
             description="Generate a greeting in a particular style of speaking.",
         )
-        model_with_tools = model.bind_tools([tool_])
+        if self.tool_choice_value == "tool_name":
+            tool_choice: Optional[str] = "greeting_generator"
+        else:
+            tool_choice = self.tool_choice_value
+        model_with_tools = model.bind_tools([tool_], tool_choice=tool_choice)
         query = "Using the tool, generate a Pirate greeting."
         result = model_with_tools.invoke(query)
         assert isinstance(result, AIMessage)

libs/standard-tests/langchain_standard_tests/unit_tests/chat_models.py

@@ -96,6 +96,11 @@ class ChatModelTests(BaseStandardTests):
     def has_tool_calling(self) -> bool:
         return self.chat_model_class.bind_tools is not BaseChatModel.bind_tools
 
+    @property
+    def tool_choice_value(self) -> Optional[str]:
+        """Value to use for tool choice when used in tests."""
+        return None
+
     @property
     def has_structured_output(self) -> bool:
         return (