core: bind_tools interface on basechatmodel (#20360)

langchain-ai · Apr 12, 2024 · 2928237 · 2928237
1 parent e6806a0
commit 2928237
Show file tree

Hide file tree

Showing 3 changed files with 129 additions and 114 deletions.
diff --git a/libs/core/langchain_core/language_models/chat_models.py b/libs/core/langchain_core/language_models/chat_models.py
@@ -9,11 +9,13 @@
     TYPE_CHECKING,
     Any,
     AsyncIterator,
+    Callable,
     Dict,
     Iterator,
     List,
     Optional,
     Sequence,
+    Type,
     Union,
     cast,
 )
@@ -53,7 +55,9 @@
 from langchain_core.tracers.log_stream import LogStreamCallbackHandler
 
 if TYPE_CHECKING:
-    from langchain_core.runnables import RunnableConfig
+    from langchain_core.pydantic_v1 import BaseModel
+    from langchain_core.runnables import Runnable, RunnableConfig
+    from langchain_core.tools import BaseTool
 
 
 def generate_from_stream(stream: Iterator[ChatGenerationChunk]) -> ChatResult:
@@ -599,16 +603,18 @@ def _generate_with_cache(
         # astream_events() or astream_log(). Bail out if _stream not implemented
         if type(self)._stream != BaseChatModel._stream and kwargs.pop(
             "stream",
-            next(
-                (
-                    True
-                    for h in run_manager.handlers
-                    if isinstance(h, LogStreamCallbackHandler)
-                ),
-                False,
-            )
-            if run_manager
-            else False,
+            (
+                next(
+                    (
+                        True
+                        for h in run_manager.handlers
+                        if isinstance(h, LogStreamCallbackHandler)
+                    ),
+                    False,
+                )
+                if run_manager
+                else False
+            ),
         ):
             chunks: List[ChatGenerationChunk] = []
             for chunk in self._stream(messages, stop=stop, **kwargs):
@@ -680,16 +686,18 @@ async def _agenerate_with_cache(
             or type(self)._stream != BaseChatModel._stream
         ) and kwargs.pop(
             "stream",
-            next(
-                (
-                    True
-                    for h in run_manager.handlers
-                    if isinstance(h, LogStreamCallbackHandler)
-                ),
-                False,
-            )
-            if run_manager
-            else False,
+            (
+                next(
+                    (
+                        True
+                        for h in run_manager.handlers
+                        if isinstance(h, LogStreamCallbackHandler)
+                    ),
+                    False,
+                )
+                if run_manager
+                else False
+            ),
         ):
             chunks: List[ChatGenerationChunk] = []
             async for chunk in self._astream(messages, stop=stop, **kwargs):
@@ -896,6 +904,13 @@ def dict(self, **kwargs: Any) -> Dict:
         starter_dict["_type"] = self._llm_type
         return starter_dict
 
+    def bind_tools(
+        self,
+        tools: Sequence[Union[Dict[str, Any], Type[BaseModel], Callable, BaseTool]],
+        **kwargs: Any,
+    ) -> Runnable[LanguageModelInput, BaseMessage]:
+        raise NotImplementedError()
+
 
 class SimpleChatModel(BaseChatModel):
     """A simplified implementation for a chat model to inherit from."""

diff --git a/libs/standard-tests/langchain_standard_tests/unit_tests/chat_models.py b/libs/standard-tests/langchain_standard_tests/unit_tests/chat_models.py
@@ -32,7 +32,7 @@ def chat_model_params(self) -> dict:
     def chat_model_has_tool_calling(
         self, chat_model_class: Type[BaseChatModel]
     ) -> bool:
-        return hasattr(chat_model_class, "bind_tools")
+        return chat_model_class.bind_tools is not BaseChatModel.bind_tools
 
     @pytest.fixture
     def chat_model_has_structured_output(