core: In RunnableSequence pass kwargs to the first step (#22393)

- This is a pattern that shows up occasionally in langgraph questions,
people chain a graph to something else after, and want to pass the graph
some kwargs (eg. stream_mode)

libs/core/langchain_core/runnables/base.py

@@ -2379,7 +2379,9 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
                 name=self.name,
             )
 
-    def invoke(self, input: Input, config: Optional[RunnableConfig] = None) -> Output:
+    def invoke(
+        self, input: Input, config: Optional[RunnableConfig] = None, **kwargs: Any
+    ) -> Output:
         from langchain_core.beta.runnables.context import config_with_context
 
         # setup callbacks and context
@@ -2396,13 +2398,14 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         # invoke all steps in sequence
         try:
             for i, step in enumerate(self.steps):
-                input = step.invoke(
-                    input,
-                    # mark each step as a child run
-                    patch_config(
-                        config, callbacks=run_manager.get_child(f"seq:step:{i+1}")
-                    ),
+                # mark each step as a child run
+                config = patch_config(
+                    config, callbacks=run_manager.get_child(f"seq:step:{i+1}")
                 )
+                if i == 0:
+                    input = step.invoke(input, config, **kwargs)
+                else:
+                    input = step.invoke(input, config)
         # finish the root run
         except BaseException as e:
             run_manager.on_chain_error(e)
@@ -2433,13 +2436,14 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         # invoke all steps in sequence
         try:
             for i, step in enumerate(self.steps):
-                input = await step.ainvoke(
-                    input,
-                    # mark each step as a child run
-                    patch_config(
-                        config, callbacks=run_manager.get_child(f"seq:step:{i+1}")
-                    ),
+                # mark each step as a child run
+                config = patch_config(
+                    config, callbacks=run_manager.get_child(f"seq:step:{i+1}")
                 )
+                if i == 0:
+                    input = await step.ainvoke(input, config, **kwargs)
+                else:
+                    input = await step.ainvoke(input, config)
         # finish the root run
         except BaseException as e:
             await run_manager.on_chain_error(e)
@@ -2519,7 +2523,7 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
                             if i not in failed_inputs_map
                         ],
                         return_exceptions=return_exceptions,
-                        **kwargs,
+                        **(kwargs if stepidx == 0 else {}),
                     )
                     # If an input failed, add it to the map
                     for i, inp in zip(remaining_idxs, inputs):
@@ -2549,6 +2553,8 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
                             )
                             for rm, config in zip(run_managers, configs)
                         ],
+                        return_exceptions=return_exceptions,
+                        **(kwargs if i == 0 else {}),
                     )
 
         # finish the root runs
@@ -2646,7 +2652,7 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
                             if i not in failed_inputs_map
                         ],
                         return_exceptions=return_exceptions,
-                        **kwargs,
+                        **(kwargs if stepidx == 0 else {}),
                     )
                     # If an input failed, add it to the map
                     for i, inp in zip(remaining_idxs, inputs):
@@ -2676,6 +2682,8 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
                             )
                             for rm, config in zip(run_managers, configs)
                         ],
+                        return_exceptions=return_exceptions,
+                        **(kwargs if i == 0 else {}),
                     )
         # finish the root runs
         except BaseException as e:
@@ -2704,6 +2712,7 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         input: Iterator[Input],
         run_manager: CallbackManagerForChainRun,
         config: RunnableConfig,
+        **kwargs: Any,
     ) -> Iterator[Output]:
         from langchain_core.beta.runnables.context import config_with_context
 
@@ -2714,14 +2723,14 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         # steps that don't natively support transforming an input stream will
         # buffer input in memory until all available, and then start emitting output
         final_pipeline = cast(Iterator[Output], input)
-        for step in steps:
-            final_pipeline = step.transform(
-                final_pipeline,
-                patch_config(
-                    config,
-                    callbacks=run_manager.get_child(f"seq:step:{steps.index(step)+1}"),
-                ),
+        for idx, step in enumerate(steps):
+            config = patch_config(
+                config, callbacks=run_manager.get_child(f"seq:step:{idx+1}")
             )
+            if idx == 0:
+                final_pipeline = step.transform(final_pipeline, config, **kwargs)
+            else:
+                final_pipeline = step.transform(final_pipeline, config)
 
         for output in final_pipeline:
             yield output
@@ -2731,6 +2740,7 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         input: AsyncIterator[Input],
         run_manager: AsyncCallbackManagerForChainRun,
         config: RunnableConfig,
+        **kwargs: Any,
     ) -> AsyncIterator[Output]:
         from langchain_core.beta.runnables.context import aconfig_with_context
 
@@ -2742,14 +2752,15 @@ class RunnableSequence(RunnableSerializable[Input, Output]):
         # steps that don't natively support transforming an input stream will
         # buffer input in memory until all available, and then start emitting output
         final_pipeline = cast(AsyncIterator[Output], input)
-        for step in steps:
-            final_pipeline = step.atransform(
-                final_pipeline,
-                patch_config(
-                    config,
-                    callbacks=run_manager.get_child(f"seq:step:{steps.index(step)+1}"),
-                ),
+        for idx, step in enumerate(steps):
+            config = patch_config(
+                config,
+                callbacks=run_manager.get_child(f"seq:step:{idx+1}"),
             )
+            if idx == 0:
+                final_pipeline = step.atransform(final_pipeline, config, **kwargs)
+            else:
+                final_pipeline = step.atransform(final_pipeline, config)
         async for output in final_pipeline:
             yield output
 

libs/core/tests/unit_tests/runnables/test_runnable.py

@@ -182,6 +182,7 @@ class FakeRunnable(Runnable[str, int]):
         self,
         input: str,
         config: Optional[RunnableConfig] = None,
+        **kwargs: Any,
     ) -> int:
         return len(input)
 
@@ -1409,26 +1410,136 @@ async def test_passthrough_tap_async(mocker: MockerFixture) -> None:
     fake = FakeRunnable()
     mock = mocker.Mock()
 
-    seq: Runnable = fake | RunnablePassthrough(mock)
+    seq: Runnable = RunnablePassthrough(mock) | fake | RunnablePassthrough(mock)
 
-    assert await seq.ainvoke("hello") == 5
-    assert mock.call_args_list == [mocker.call(5)]
+    assert await seq.ainvoke("hello", my_kwarg="value") == 5
+    assert mock.call_args_list == [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call(5),
+    ]
+    mock.reset_mock()
+
+    assert await seq.abatch(["hello", "byebye"], my_kwarg="value") == [5, 6]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
+    mock.reset_mock()
+
+    assert await seq.abatch(
+        ["hello", "byebye"], my_kwarg="value", return_exceptions=True
+    ) == [
+        5,
+        6,
+    ]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
+    mock.reset_mock()
+
+    assert sorted(
+        [
+            a
+            async for a in seq.abatch_as_completed(
+                ["hello", "byebye"], my_kwarg="value", return_exceptions=True
+            )
+        ]
+    ) == [
+        (0, 5),
+        (1, 6),
+    ]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
     mock.reset_mock()
 
     assert [
-        part async for part in seq.astream("hello", dict(metadata={"key": "value"}))
+        part
+        async for part in seq.astream(
+            "hello", dict(metadata={"key": "value"}), my_kwarg="value"
+        )
     ] == [5]
-    assert mock.call_args_list == [mocker.call(5)]
-    mock.reset_mock()
-
-    assert seq.invoke("hello") == 5
-    assert mock.call_args_list == [mocker.call(5)]
-    mock.reset_mock()
-
-    assert [part for part in seq.stream("hello", dict(metadata={"key": "value"}))] == [
-        5
+    assert mock.call_args_list == [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call(5),
+    ]
+    mock.reset_mock()
+
+    assert seq.invoke("hello", my_kwarg="value") == 5  # type: ignore[call-arg]
+    assert mock.call_args_list == [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call(5),
+    ]
+    mock.reset_mock()
+
+    assert seq.batch(["hello", "byebye"], my_kwarg="value") == [5, 6]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
+    mock.reset_mock()
+
+    assert seq.batch(["hello", "byebye"], my_kwarg="value", return_exceptions=True) == [
+        5,
+        6,
+    ]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
+    mock.reset_mock()
+
+    assert sorted(
+        a
+        for a in seq.batch_as_completed(
+            ["hello", "byebye"], my_kwarg="value", return_exceptions=True
+        )
+    ) == [
+        (0, 5),
+        (1, 6),
+    ]
+    assert len(mock.call_args_list) == 4
+    for call in [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call("byebye", my_kwarg="value"),
+        mocker.call(5),
+        mocker.call(6),
+    ]:
+        assert call in mock.call_args_list
+    mock.reset_mock()
+
+    assert [
+        part
+        for part in seq.stream(
+            "hello", dict(metadata={"key": "value"}), my_kwarg="value"
+        )
+    ] == [5]
+    assert mock.call_args_list == [
+        mocker.call("hello", my_kwarg="value"),
+        mocker.call(5),
     ]
-    assert mock.call_args_list == [mocker.call(5)]
     mock.reset_mock()