上一篇详细介绍了人在回路中,下面应该是Breakpoints 断点、Time travel 时间旅行、Tools 工具这几个章节,但是我觉得这些都是调试使用的,使用技巧也是中断的概念,所以大家去官网自己学习,我这边直接进入子图学习,为后续的多智能体学习做准备。
子图是一个图 ,它被用作另一个图中的节点 -这是应用于 LangGraph 的封装概念。子图允许您构建具有多个组件的复杂系统,这些组件本身就是图。
使用子图的一些原因是:
添加子图时,需要定义父图和子图的通信方式:
共享状态模式 -父图和子图在其状态模式中具有共享状态键
不同的状态模式 - 父模式和子模式中没有共享状态键
下面就详细介绍这两种状态模式
一种常见的情况是父图和子图通过模式中的共享状态键(通道)进行通信。例如,在多代理系统中,代理通常通过共享消息密钥进行通信。
如果您的子图与父图共享状态键,您可以按照以下步骤将其添加到您的图中:
的 subgraph_builder
)并对其进行编译.add_node
方法例子如下:
from typing_extensions import TypedDict
from langgraph.graph import StateGraph,START,END
# 定义子图状态
class SubGraphState(TypedDict):
foo:str
bar:str
# 定义子图节点
def subgraph_node1(state:SubGraphState):
return{"bar":"bar"}
def subgraph_node2(state:SubGraphState):
return{"foo":state["foo"]+state["bar"]} # 这里需要注意,使用了子图独立的bar键,父图是没有这个键的
# 构建子图
subgraph_builder = StateGraph(SubGraphState)
subgraph_builder.add_node("subgraph_node1", subgraph_node1)
subgraph_builder.add_node("subgraph_node2", subgraph_node2)
subgraph_builder.add_edge(START, "subgraph_node1")
subgraph_builder.add_edge("subgraph_node1", "subgraph_node2")
subgraph_builder.add_edge("subgraph_node2", END)
subgraph = subgraph_builder.compile()
# 定义父图
class ParentState(TypedDict):
foo:str # 父图只定义了foo,没有bar
def node_1(state:ParentState):
return{"foo":" hello! "+ state["foo"]}
# 构建父图
parent_graph = StateGraph(ParentState)
parent_graph.add_node("node_1", node_1)
parent_graph.add_node("node_2", subgraph)
parent_graph.add_edge(START, "node_1")
parent_graph.add_edge("node_1", "node_2")
parent_graph.add_edge("node_2", END)
graph = parent_graph.compile()
for chunk in graph.stream({"foo": "foo"},config):
print(chunk)
{'node_1': {'foo': ' hello! foo'}}
{'node_2': {'foo': ' hello! foobar'}}
对于更复杂的系统,您可能希望定义与父图具有完全不同模式的子图(没有共享键)。例如,您可能希望为多代理系统中的每个代理保留一个私有消息历史记录。
如果您的应用程序是这种情况,则需要定义一个调用子图的节点函数 。该函数需要在调用子图之前将输入(父)状态转换为子图状态,并在从节点返回状态更新之前将结果转换回父状态。
一层子图
from typing_extensions import TypedDict
from langgraph.graph.state import StateGraph, START
# 定义子图
class SubgraphState(TypedDict):
# 这里定义的子图状态和父图的完全不同了
bar: str
baz: str
def subgraph_node_1(state: SubgraphState):
return {"baz": "baz"}
def subgraph_node_2(state: SubgraphState):
return {"bar": state["bar"] + state["baz"]}
subgraph_builder = StateGraph(SubgraphState)
subgraph_builder.add_node(subgraph_node_1)
subgraph_builder.add_node(subgraph_node_2)
subgraph_builder.add_edge(START, "subgraph_node_1")
subgraph_builder.add_edge("subgraph_node_1", "subgraph_node_2")
subgraph = subgraph_builder.compile()
# 定义父图
class ParentState(TypedDict):
foo: str
def node_1(state: ParentState):
return {"foo": "hi! " + state["foo"]}
def node_2(state: ParentState):
response = subgraph.invoke({"bar": state["foo"]}) # 在父图的节点函数内部调用子图,数据进行了转换才能调用
return {"foo": response["bar"]}
builder = StateGraph(ParentState)
builder.add_node("node_1", node_1)
builder.add_node("node_2", node_2)
builder.add_edge(START, "node_1")
builder.add_edge("node_1", "node_2")
graph = builder.compile()
for chunk in graph.stream({"foo": "foo"}, subgraphs=True):
print(chunk)
((), {'node_1': {'foo': 'hi! foo'}})
(('node_2:f916dc5a-e340-d03f-fde5-24825fb2c914',), {'subgraph_node_1': {'baz': 'baz'}})
(('node_2:f916dc5a-e340-d03f-fde5-24825fb2c914',), {'subgraph_node_2': {'bar': 'hi! foobaz'}})
((), {'node_2': {'foo': 'hi! foobaz'}})
二级子图
from typing_extensions import TypedDict
from langgraph.graph.state import StateGraph, START, END
# 孙子图
class GrandChildState(TypedDict):
my_grandchild_key: str
def grandchild_1(state: GrandChildState) -> GrandChildState:
# NOTE: 这里是无法访问子图和父图的状态的
return {"my_grandchild_key": state["my_grandchild_key"] + ", how are you"}
grandchild = StateGraph(GrandChildState)
grandchild.add_node("grandchild_1", grandchild_1)
grandchild.add_edge(START, "grandchild_1")
grandchild.add_edge("grandchild_1", END)
grandchild_graph = grandchild.compile()
# 子图
class ChildState(TypedDict):
my_child_key: str
def call_grandchild_graph(state: ChildState) -> ChildState:
# NOTE: 子图和孙子图的数据在这里进行交互
grandchild_graph_input = {"my_grandchild_key": state["my_child_key"]}
grandchild_graph_output = grandchild_graph.invoke(grandchild_graph_input)
return {"my_child_key": grandchild_graph_output["my_grandchild_key"] + " today?"}
child = StateGraph(ChildState)
child.add_node("child_1", call_grandchild_graph)
child.add_edge(START, "child_1")
child.add_edge("child_1", END)
child_graph = child.compile()
#父图
class ParentState(TypedDict):
my_key: str
def parent_1(state: ParentState) -> ParentState:
# NOTE: 父图不包含子图和父图
return {"my_key": "hi " + state["my_key"]}
def parent_2(state: ParentState) -> ParentState:
return {"my_key": state["my_key"] + " bye!"}
def call_child_graph(state: ParentState) -> ParentState:
child_graph_input = {"my_child_key": state["my_key"]}
child_graph_output = child_graph.invoke(child_graph_input)
return {"my_key": child_graph_output["my_child_key"]}
parent = StateGraph(ParentState)
parent.add_node("parent_1", parent_1)
parent.add_node("child", call_child_graph)
parent.add_node("parent_2", parent_2)
parent.add_edge(START, "parent_1")
parent.add_edge("parent_1", "child")
parent.add_edge("child", "parent_2")
parent.add_edge("parent_2", END)
parent_graph = parent.compile()
for chunk in parent_graph.stream({"my_key": "Bob"}, subgraphs=True):
print(chunk)
((), {'parent_1': {'my_key': 'hi Bob'}})
(('child:991b3f8f-449a-e406-ef14-49951b36d4e6', 'child_1:62c9f423-eb44-d618-0693-fbc4489aef5f'), {'grandchild_1': {'my_grandchild_key': 'hi Bob, how are you'}})
(('child:991b3f8f-449a-e406-ef14-49951b36d4e6',), {'child_1': {'my_child_key': 'hi Bob, how are you today?'}})
((), {'child': {'my_key': 'hi Bob, how are you today?'}})
((), {'parent_2': {'my_key': 'hi Bob, how are you today? bye!'}})
你只需要在编译父图时提供checkpointer 。LangGraph 将自动将检查指针传播到子图。
from langgraph.graph import START, StateGraph
from langgraph.checkpoint.memory import InMemorySaver
from typing_extensions import TypedDict
class State(TypedDict):
foo: str
# Subgraph
def subgraph_node_1(state: State):
return {"foo": state["foo"] + "bar"}
subgraph_builder = StateGraph(State)
subgraph_builder.add_node(subgraph_node_1)
subgraph_builder.add_edge(START, "subgraph_node_1")
subgraph = subgraph_builder.compile()
# Parent graph
builder = StateGraph(State)
builder.add_node("node_1", subgraph)
builder.add_edge(START, "node_1")
checkpointer = InMemorySaver()
graph = builder.compile(checkpointer=checkpointer)
如果你想让子图拥有自己的内存 ,你可以用 checkpointer=True
编译它。这在多代理系统中非常有用,如果您希望代理跟踪其内部消息历史记录:
subgraph_builder = StateGraph(...)
subgraph = subgraph_builder.compile(checkpointer=True)
当您启用持久性时,您可以通过 graph.get_state(config)
检查图形状态 (检查点)。要查看子图状态,可以使用 graph.get_state(config, subgraphs=True)
。
子图状态只能在子图中断时查看。恢复图形后,将无法访问子图形状态。
from langgraph.graph import START, StateGraph
from langgraph.checkpoint.memory import InMemorySaver
from langgraph.types import interrupt, Command
from typing_extensions import TypedDict
class State(TypedDict):
foo: str
# Subgraph
def subgraph_node_1(state: State):
value = interrupt("Provide value:")
return {"foo": state["foo"] + value}
subgraph_builder = StateGraph(State)
subgraph_builder.add_node(subgraph_node_1)
subgraph_builder.add_edge(START, "subgraph_node_1")
subgraph = subgraph_builder.compile()
# Parent graph
builder = StateGraph(State)
builder.add_node("node_1", subgraph)
builder.add_edge(START, "node_1")
checkpointer = InMemorySaver()
graph = builder.compile(checkpointer=checkpointer)
config = {"configurable": {"thread_id": "1"}}
graph.invoke({"foo": ""}, config)
parent_state = graph.get_state(config)
subgraph_state = graph.get_state(config, subgraphs=True).tasks[0].state
# resume the subgraph
graph.invoke(Command(resume="bar"), config)
要在流输出中包含子图的输出,可以在父图的 .stream()
方法中设置 subgraphs=True
。这将从父图和任何子图流输出。
for chunk in graph.stream(
{"foo": "foo"},
subgraphs=True,
stream_mode="updates",
):
print(chunk)