欧日韩无套内射变态,日本在线视频www色,日本肉体xxxx裸交

? ? ? ?這就是在LangGraph中創(chuàng)建代理執(zhí)行器的方式，和LangChain的執(zhí)行器功能類似。我們將進(jìn)一步探討狀態(tài)圖的接口以及返回結(jié)果的不同流式傳輸方法。

四、探索聊天代理執(zhí)行器

? ? ? ?我們將在LangGraph中探索聊天代理執(zhí)行器，這是一個(gè)設(shè)計(jì)用于處理基于聊天的模型的工具。此執(zhí)行器是唯一的，因?yàn)樗耆鶕?jù)輸入消息的列表進(jìn)行操作，通過向該列表中添加新消息來隨著時(shí)間的推移更新代理的狀態(tài)。

讓我們深入了解設(shè)置過程：

4.1 安裝軟件包：

? ? ? ?同樣需要LangChain軟件包，LangChain OpenAI用于模型，Tavily軟件包用于搜索工具，并為這些服務(wù)設(shè)置API密鑰。

!pip install --quiet -U langchain langchain_openai tavily-python

import osimport getpass
os.environ["OPENAI_API_KEY"] = getpass.getpass("OpenAI API Key:")os.environ["TAVILY_API_KEY"] = getpass.getpass("Tavily API Key:")os.environ["LANGCHAIN_TRACING_V2"] = "true"os.environ["LANGCHAIN_API_KEY"] = getpass.getpass("LangSmith API Key:")

4.2 設(shè)置工具和模型：

? ? ? ?我們將使用Tavily Search作為我們的工具，并設(shè)置一個(gè)工具執(zhí)行器來調(diào)用這些工具。對(duì)于模型，我們將使用LangChain集成中的Chat OpenAI模型，確保其在啟用流式進(jìn)行初始化。這使我們能夠流式返回tokens，并附加我們希望模型調(diào)用的函數(shù)。

from langchain_community.tools.tavily_search import TavilySearchResultsfrom langchain_openai import ChatOpenAIfrom langgraph.prebuilt import ToolExecutorfrom langchain.tools.render import format_tool_to_openai_functiontools = [TavilySearchResults(max_results=1)]tool_executor = ToolExecutor(tools)# We will set streaming=True so that we can stream tokens# See the streaming section for more information on this.model = ChatOpenAI(temperature=0, streaming=True)functions = [format_tool_to_openai_function(t) for t in tools]model = model.bind_functions(functions)

4.3 定義代理狀態(tài)：

代理狀態(tài)是一個(gè)簡單的字典，其中包含消息列表的鍵。我們將使用“add to”標(biāo)記，這樣隨著時(shí)間的推移，節(jié)點(diǎn)對(duì)此消息列表的任何更新都會(huì)累積。

from typing import TypedDict, Annotated, Sequenceimport operatorfrom langchain_core.messages import BaseMessage

class AgentState(TypedDict): messages: Annotated[Sequence[BaseMessage], operator.add]

4.4 創(chuàng)建節(jié)點(diǎn)和邊：

節(jié)點(diǎn)表示具體的工作任務(wù)，邊連接節(jié)點(diǎn)。我們需要一個(gè)代理節(jié)點(diǎn)來調(diào)用語言模型并獲得響應(yīng)，一個(gè)操作節(jié)點(diǎn)來查看是否有任何工具需要調(diào)用，以及一個(gè)函數(shù)來確定我們是否應(yīng)該繼續(xù)調(diào)用工具或完成。

from langgraph.prebuilt import ToolInvocationimport jsonfrom langchain_core.messages import FunctionMessage
# Define the function that determines whether to continue or notdef should_continue(state): messages = state['messages'] last_message = messages[-1] # If there is no function call, then we finish if "function_call" not in last_message.additional_kwargs: return "end" # Otherwise if there is, we continue else: return "continue"
# Define the function that calls the modeldef call_model(state): messages = state['messages'] response = model.invoke(messages) # We return a list, because this will get added to the existing list return {"messages": [response]}
# Define the function to execute toolsdef call_tool(state): messages = state['messages'] # Based on the continue condition # we know the last message involves a function call last_message = messages[-1] # We construct an ToolInvocation from the function_call action = ToolInvocation( tool=last_message.additional_kwargs["function_call"]["name"], tool_input=json.loads(last_message.additional_kwargs["function_call"]["arguments"]), ) # We call the tool_executor and get back a response response = tool_executor.invoke(action) # We use the response to create a FunctionMessage function_message = FunctionMessage(content=str(response), name=action.tool) # We return a list, because this will get added to the existing list return {"messages": [function_message]}

4.5 構(gòu)建圖：

我們創(chuàng)建一個(gè)具有代理狀態(tài)的圖，為代理和動(dòng)作添加節(jié)點(diǎn)，并將入口點(diǎn)設(shè)置為代理節(jié)點(diǎn)。條件邊是根據(jù)代理應(yīng)該繼續(xù)還是結(jié)束來添加的，并且正常邊總是在動(dòng)作后返回到代理。

from langgraph.graph import StateGraph, END# Define a new graphworkflow = StateGraph(AgentState)
# Define the two nodes we will cycle betweenworkflow.add_node("agent", call_model)workflow.add_node("action", call_tool)
# Set the entrypoint as agent# This means that this node is the first one calledworkflow.set_entry_point("agent")
# We now add a conditional edgeworkflow.add_conditional_edges( # First, we define the start node. We use agent. # This means these are the edges taken after the agent node is called. "agent", # Next, we pass in the function that will determine which node is called next. should_continue, # Finally we pass in a mapping. # The keys are strings, and the values are other nodes. # END is a special node marking that the graph should finish. # What will happen is we will call should_continue, and then the output of that # will be matched against the keys in this mapping. # Based on which one it matches, that node will then be called. { # If tools, then we call the tool node. "continue": "action", # Otherwise we finish. "end": END })
# We now add a normal edge from tools to agent.# This means that after tools is called, agent node is called next.workflow.add_edge('action', 'agent')
# Finally, we compile it!# This compiles it into a LangChain Runnable,# meaning you can use it as you would any other runnableapp = workflow.compile()

4.6 編譯和使用圖形：

編譯圖形后，我們創(chuàng)建一個(gè)帶有消息鍵的輸入字典。運(yùn)行圖形將處理這些消息，將AI響應(yīng)、功能結(jié)果和最終輸出添加到消息列表中。

from langchain_core.messages import HumanMessage
inputs = {"messages": [HumanMessage(content="what is the weather in sf")]}app.invoke(inputs)

4.7 觀察執(zhí)行過程：

? ? ? ?使用LangSmith，我們可以看到我們的代理所采取的詳細(xì)步驟，包括對(duì)OpenAI的調(diào)用和由此產(chǎn)生的輸出。

流式功能：LangGraph還提供流式功能。

五、如何在循環(huán)中修改humans操作

? ? ? ?讓我們修改LangGraph中的聊天代理執(zhí)行器，使其包含一個(gè)“human in the loop”組件，這樣在執(zhí)行工具操作之前可以進(jìn)行人工驗(yàn)證。

設(shè)置：初始設(shè)置保持不變。不需要額外安裝。我們將創(chuàng)建我們的工具，設(shè)置工具執(zhí)行器，準(zhǔn)備我們的模型，將工具綁定到模型，并定義代理狀態(tài)——所有這些都與我們?cè)谇耙粋€(gè)會(huì)話中所做的一樣。

關(guān)鍵修改——調(diào)用工具功能：主要的變化來自調(diào)用工具功能。我們添加了一個(gè)步驟，系統(tǒng)在交互式IDE中提示用戶（即您?。儐柺欠窭^續(xù)執(zhí)行特定操作。如果用戶響應(yīng)“否”，則會(huì)引發(fā)錯(cuò)誤，進(jìn)程將停止。這是我們的人工驗(yàn)證步驟。

# Define the function to execute toolsdef call_tool(state): messages = state['messages'] # Based on the continue condition # we know the last message involves a function call last_message = messages[-1] # We construct an ToolInvocation from the function_call action = ToolInvocation( tool=last_message.additional_kwargs["function_call"]["name"], tool_input=json.loads(last_message.additional_kwargs["function_call"]["arguments"]), ) response = input(prompt=f"[y/n] continue with: {action}?") if response == "n": raise ValueError # We call the tool_executor and get back a response response = tool_executor.invoke(action) # We use the response to create a FunctionMessage function_message = FunctionMessage(content=str(response), name=action.tool) # We return a list, because this will get added to the existing list return {"messages": [function_message]}

使用修改的執(zhí)行器：當(dāng)我們運(yùn)行這個(gè)修改的執(zhí)行程序時(shí)，它會(huì)在執(zhí)行任何工具操作之前請(qǐng)求批準(zhǔn)。如果我們同意說“是”，它將正常進(jìn)行。然而，如果我們說“不”，則會(huì)引發(fā)錯(cuò)誤并停止該過程。

utput from node 'agent':---{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\n "query": "weather in San Francisco"\n}', 'name': 'tavily_search_results_json'}})]}
---
---------------------------------------------------------------------------ValueError Traceback (most recent call last)Cell In[10], line 4 1 from langchain_core.messages import HumanMessage 3 inputs = {"messages": [HumanMessage(content="what is the weather in sf")]}----> 4 for output in app.stream(inputs): 5 # stream() yields dictionaries with output keyed by node name 6 for key, value in output.items(): 7 print(f"Output from node '{key}':")

? ? ? ?這是一個(gè)基本的實(shí)現(xiàn)。在現(xiàn)實(shí)世界中，您可能希望用更復(fù)雜的響應(yīng)來代替錯(cuò)誤，并使用更用戶友好的界面，而不是Jupyter筆記本。但這讓您清楚地了解了如何將一個(gè)簡單而有效的人工循環(huán)組件添加到LangGraph代理中。

六、修改管理代理步驟

? ? ? ?讓我們來看看在LangGraph中修改聊天代理執(zhí)行器，以在處理消息時(shí)操縱代理的內(nèi)部狀態(tài)。

本教程建立在基本的聊天代理執(zhí)行程序設(shè)置的基礎(chǔ)上，因此，如果您還沒有在基本筆記本中完成初始設(shè)置，請(qǐng)先完成。我們?cè)谶@里只關(guān)注新的修改。

關(guān)鍵修改——過濾消息：我們引入的主要更改是過濾傳遞給模型的消息的方法。現(xiàn)在，您可以自定義代理考慮的消息。例如：

def call_model(state): messages = state['messages'][-5:] response = model.invoke(messages) # We return a list, because this will get added to the existing list return {"messages": [response]}

僅選擇最近的五條消息。
包括系統(tǒng)消息加上五條最新消息。
總結(jié)比最近五條消息舊的消息。

此修改是一個(gè)小而強(qiáng)大的添加，允許您控制代理如何與其消息歷史進(jìn)行交互，并改進(jìn)其決策過程。

使用修改的執(zhí)行器：實(shí)現(xiàn)非常簡單。僅有一條輸入消息不同，但重要的是，您希望應(yīng)用于代理步驟的任何邏輯都可以插入到這個(gè)新的修改部分。

此方法非常適合修改聊天代理執(zhí)行器，但如果使用標(biāo)準(zhǔn)代理執(zhí)行器時(shí)，同樣的原理也適用。

七、強(qiáng)制調(diào)用工具

? ? ? ?我們將對(duì)LangGraph中的聊天代理執(zhí)行器進(jìn)行簡單但有效的修改，確保始終首先調(diào)用一個(gè)工具。這是建立在基本的聊天代理執(zhí)行器筆記本上的，所以請(qǐng)確保您已經(jīng)檢查了背景信息。

關(guān)鍵修改——強(qiáng)制工具調(diào)用優(yōu)先：我們這里的重點(diǎn)是設(shè)置聊天代理調(diào)用特定工具作為其第一個(gè)操作。為此，我們將添加一個(gè)新節(jié)點(diǎn)，并將其命名為“first model node”。該節(jié)點(diǎn)將被編程為返回一條消息，指示代理調(diào)用特定工具，如“Tavil search results Json”工具，并將最新的消息內(nèi)容作為查詢。

# This is the new first - the first call of the model we want to explicitly hard-code some actionfrom langchain_core.messages import AIMessageimport json
def first_model(state): human_input = state['messages'][-1].content return { "messages": [ AIMessage( content="", additional_kwargs={ "function_call": { "name": "tavily_search_results_json", "arguments": json.dumps({"query": human_input}) } } ) ] }

更新圖：我們將修改現(xiàn)有的圖，將這個(gè)新的“first agent”節(jié)點(diǎn)作為入口點(diǎn)。這樣可以確保始終首先調(diào)用第一個(gè)代理節(jié)點(diǎn)，然后調(diào)用動(dòng)作節(jié)點(diǎn)。我們?cè)O(shè)置了一個(gè)從代理到動(dòng)作或結(jié)束的條件節(jié)點(diǎn)，以及一個(gè)從動(dòng)作回到代理的直接節(jié)點(diǎn)。關(guān)鍵的添加是從第一個(gè)代理到操作的一個(gè)新節(jié)點(diǎn)，確保工具調(diào)用一開始就發(fā)生。

from langgraph.graph import StateGraph, END# Define a new graphworkflow = StateGraph(AgentState)
# Define the new entrypointworkflow.add_node("first_agent", first_model)
# Define the two nodes we will cycle betweenworkflow.add_node("agent", call_model)workflow.add_node("action", call_tool)
# Set the entrypoint as agent# This means that this node is the first one calledworkflow.set_entry_point("first_agent")
# We now add a conditional edgeworkflow.add_conditional_edges( # First, we define the start node. We use agent. # This means these are the edges taken after the agent node is called. "agent", # Next, we pass in the function that will determine which node is called next. should_continue, # Finally we pass in a mapping. # The keys are strings, and the values are other nodes. # END is a special node marking that the graph should finish. # What will happen is we will call should_continue, and then the output of that # will be matched against the keys in this mapping. # Based on which one it matches, that node will then be called. { # If tools, then we call the tool node. "continue": "action", # Otherwise we finish. "end": END })
# We now add a normal edge from tools to agent.# This means that after tools is called, agent node is called next.workflow.add_edge('action', 'agent')
# After we call the first agent, we know we want to go to actionworkflow.add_edge('first_agent', 'action')
# Finally, we compile it!# This compiles it into a LangChain Runnable,# meaning you can use it as you would any other runnableapp = workflow.compile()

使用修改的執(zhí)行器：當(dāng)我們運(yùn)行這個(gè)更新的執(zhí)行器時(shí)，第一個(gè)結(jié)果會(huì)很快返回，因?yàn)槲覀兝@過了初始的語言模型調(diào)用，直接調(diào)用該工具。通過觀察LangSmith中的過程可以證實(shí)這一點(diǎn)，在LangSmith中，我們可以看到工具是第一個(gè)被調(diào)用的東西，然后是最后的語言模型調(diào)用。