WEBDIENSTE/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-08-22 15:42:25 +02:00
Code Issues Packages Projects Releases Wiki Activity

cr

Browse Source
This commit is contained in:
Harrison Chase
2024-01-15 15:35:11 -08:00
parent 1ffdd7b93d
commit 7bfeb1758a
7 changed files with 0 additions and 3176 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/chat_executor_with_function_calling/base.ipynb
-606
View File
@@ -1,606 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Chat Executor\n",
"\n",
"In this example we will build a chat executor that uses function calling from scratch."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"tools = [TavilySearchResults(max_results=1)]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "3b541bb9-900c-40d0-964d-7b5dfee30667",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage\n",
"\n",
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we continue\n",
" else:\n",
" return \"continue\"\n",
"\n",
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages']\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}\n",
"\n",
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" action = ToolInvocation(\n",
" tool=last_message.additional_kwargs[\"function_call\"][\"name\"],\n",
" tool_input=json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"]),\n",
" )\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "8edb04b9-40b6-46f1-a7a8-4b2d8aba7752",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'messages': [HumanMessage(content='what is the weather in sf'),\n",
" AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}),\n",
" FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'),\n",
" AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"app.invoke(inputs)"
]
},
{
"cell_type": "markdown",
"id": "5a9e8155-70c5-4973-912c-dc55104b2acf",
"metadata": {},
"source": [
"This may take a little bit - it's making a few calls behind the scenes.\n",
"In order to start seeing some intermediate results as they happen, we can use streaming - see below for more information on that.\n",
"\n",
"## Streaming\n",
"\n",
"LangGraph has support for several different types of streaming.\n",
"\n",
"### Streaming Node Output\n",
"\n",
"One of the benefits of using LangGraph is that it is easy to stream output as it's produced by each node.\n"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "f544977e-31f7-41f0-88c4-ec9c27b8cecb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "markdown",
"id": "2a1b56c5-bd61-4192-8bdb-458a1e9f0159",
"metadata": {},
"source": [
"### Streaming LLM Tokens\n",
"\n",
"You can also access the LLM tokens as they are produced by each node. \n",
"In this case only the \"agent\" node produces LLM tokens.\n",
"In order for this to work properly, you must be using an LLM that supports streaming as well as have set it when constructing the LLM (e.g. `ChatOpenAI(model=\"gpt-3.5-turbo-1106\", streaming=True)`)\n"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "cfd140f0-a5a6-4697-8115-322242f197b5",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"content='' additional_kwargs={'function_call': {'arguments': '', 'name': 'tavily_search_results_json'}}\n",
"content='' additional_kwargs={'function_call': {'arguments': '{\\n', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' ', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' \"', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': 'query', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': '\":', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' \"', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': 'weather', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' in', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' San', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': ' Francisco', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': '\"\\n', 'name': ''}}\n",
"content='' additional_kwargs={'function_call': {'arguments': '}', 'name': ''}}\n",
"content=''\n",
"content=''\n",
"content='I'\n",
"content=\"'m\"\n",
"content=' sorry'\n",
"content=','\n",
"content=' but'\n",
"content=' I'\n",
"content=' couldn'\n",
"content=\"'t\"\n",
"content=' find'\n",
"content=' the'\n",
"content=' current'\n",
"content=' weather'\n",
"content=' in'\n",
"content=' San'\n",
"content=' Francisco'\n",
"content='.'\n",
"content=' However'\n",
"content=','\n",
"content=' you'\n",
"content=' can'\n",
"content=' check'\n",
"content=' the'\n",
"content=' historical'\n",
"content=' weather'\n",
"content=' data'\n",
"content=' for'\n",
"content=' January'\n",
"content=' '\n",
"content='202'\n",
"content='4'\n",
"content=' in'\n",
"content=' San'\n",
"content=' Francisco'\n",
"content=' ['\n",
"content='here'\n",
"content=']('\n",
"content='https'\n",
"content='://'\n",
"content='we'\n",
"content='athers'\n",
"content='park'\n",
"content='.com'\n",
"content='/h'\n",
"content='/m'\n",
"content='/'\n",
"content='557'\n",
"content='/'\n",
"content='202'\n",
"content='4'\n",
"content='/'\n",
"content='1'\n",
"content='/H'\n",
"content='istorical'\n",
"content='-'\n",
"content='Weather'\n",
"content='-in'\n",
"content='-Jan'\n",
"content='uary'\n",
"content='-'\n",
"content='202'\n",
"content='4'\n",
"content='-in'\n",
"content='-S'\n",
"content='an'\n",
"content='-F'\n",
"content='r'\n",
"content='anc'\n",
"content='isco'\n",
"content='-Cal'\n",
"content='ifornia'\n",
"content='-'\n",
"content='United'\n",
"content='-'\n",
"content='States'\n",
"content=').'\n",
"content=''\n"
]
}
],
"source": [
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"async for output in app.astream_log(inputs, include_types=[\"llm\"]):\n",
" # astream_log() yields the requested logs (here LLMs) in JSONPatch format\n",
" for op in output.ops:\n",
" if op[\"path\"] == \"/streamed_output/-\":\n",
" # this is the output from .stream()\n",
" ...\n",
" elif op[\"path\"].startswith(\"/logs/\") and op[\"path\"].endswith(\n",
" \"/streamed_output/-\"\n",
" ):\n",
" # because we chose to only include LLMs, these are LLM tokens\n",
" print(op[\"value\"])"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "08ae8246-11d5-40e1-8567-361e5bef8917",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/dynamically-returning-directly.ipynb
-565
View File
@@ -1,565 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Dynamically Returning Directly\n",
"\n",
"In this example we will build a chat executor where the LLM can optionally decide to return the result of a tool call as the final answer. This is useful in cases where you have tools that can sometimes generate responses that are acceptable as final answers, and you want to use the LLM to determine when that is the case\n",
"\n",
"This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example [here](./base.ipynb).\n",
"\n",
"Any modifications of that example are called below with **MODIFICATION**, so if you are looking for the differences you can just search for that."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n",
"\n",
"**MODIFICATION**\n",
"\n",
"We overwrite the default schema of the input tool to have an additional parameter for returning directly."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "4a1b9990-3b11-4a51-bd51-76117afd38b9",
"metadata": {},
"outputs": [],
"source": [
"from langchain_core.pydantic_v1 import BaseModel, Field\n",
"\n",
"class SearchTool(BaseModel):\n",
" \"\"\"Look up things online, optionally returning directly\"\"\"\n",
" query: str = Field(description=\"query to look up online\")\n",
" return_direct: bool = Field(\n",
" description=\"Whether or the result of this should be returned directly to the user without you seeing what it is\", \n",
" default = False\n",
" )"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"search_tool = TavilySearchResults(max_results=1, args_schema=SearchTool)\n",
"tools = [search_tool]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "03308b6b-de72-4cdc-b6c6-47e654df340e",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage"
]
},
{
"cell_type": "markdown",
"id": "50bf356c-2dbd-4f66-8fa3-133e9c2e371e",
"metadata": {},
"source": [
"**MODIFICATION**\n",
"\n",
"We change the `should_continue` function to check whether return_direct was set to True"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "55e088b1-f3c8-4798-9ca8-5b0be961b49a",
"metadata": {},
"outputs": [],
"source": [
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we check if it's suppose to return direct\n",
" else:\n",
" arguments = json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"])\n",
" if arguments.get(\"return_direct\", False):\n",
" return \"final\"\n",
" else:\n",
" return \"continue\""
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "2b45da72-1afa-4cd7-9b7f-49a7c99cdb8a",
"metadata": {},
"outputs": [],
"source": [
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages']\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}"
]
},
{
"cell_type": "markdown",
"id": "8535a36c-3ced-401e-98b5-ec1d1b434bbc",
"metadata": {},
"source": [
"**MODIFICATION**\n",
"\n",
"We change the tool calling to get rid of the `return_direct` parameter (not used in the actual tool call)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "dd876f5d-88d6-4f93-b1d0-f2f0b6f4d991",
"metadata": {},
"outputs": [],
"source": [
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" tool_name = last_message.additional_kwargs[\"function_call\"][\"name\"]\n",
" arguments = json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"])\n",
" if tool_name == \"tavily_search_results_json\":\n",
" if \"return_direct\" in arguments:\n",
" del arguments[\"return_direct\"]\n",
" action = ToolInvocation(\n",
" tool=tool_name,\n",
" tool_input=arguments,\n",
" )\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!\n",
"\n",
"**MODIFICATION**\n",
"\n",
"We add a separate node for any tool call where `return_direct=True`. The reason this is needed is that after this node we want to end, while after other tool calls we want to go back to the LLM. "
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"workflow.add_node(\"final\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Final call\n",
" \"final\": \"final\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"workflow.add_edge('final', END)\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "f544977e-31f7-41f0-88c4-ec9c27b8cecb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "08ae8246-11d5-40e1-8567-361e5bef8917",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\",\\n \"return_direct\": true\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'final':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf? return this result directly by setting return_direct = True'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\",\\n \"return_direct\": true\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf? return this result directly by setting return_direct = True\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "49ccc134-4abe-4982-8ecd-d70fc56a4d2d",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/force-calling-a-tool-first.ipynb
-485
View File
@@ -1,485 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Force Calling a Tool First\n",
"\n",
"In this example we will build a chat executor that always calls a certain tool first. In this example, we will create an agent with a search tool. However, at the start we will force the agent to call the search tool (and then let it do whatever it wants after). This is useful when you want to force agents to call particular tools, but still want flexibility of what happens after that.\n",
"\n",
"This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example [here](./base.ipynb).\n",
"\n",
"Any modifications of that example are called below with **MODIFICATION**, so if you are looking for the differences you can just search for that."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"tools = [TavilySearchResults(max_results=1)]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "3b541bb9-900c-40d0-964d-7b5dfee30667",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage\n",
"\n",
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we continue\n",
" else:\n",
" return \"continue\"\n",
"\n",
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages']\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}\n",
"\n",
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" action = ToolInvocation(\n",
" tool=last_message.additional_kwargs[\"function_call\"][\"name\"],\n",
" tool_input=json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"]),\n",
" )\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "7c3e0ac2-0c89-4751-bc2c-f644654841d1",
"metadata": {},
"source": [
"**MODIFICATION**\n",
"\n",
"Here we create a node that returns an AIMessage with a tool call - we will use this at the start to force it call a tool"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "1bfd2b22-292a-4f4d-91a0-46bb704f5e38",
"metadata": {},
"outputs": [],
"source": [
"# This is the new first - the first call of the model we want to explicitly hard-code some action\n",
"from langchain_core.messages import AIMessage\n",
"import json\n",
"\n",
"def first_model(state):\n",
" human_input = state['messages'][-1].content\n",
" return {\n",
" \"messages\": [\n",
" AIMessage(\n",
" content=\"\", \n",
" additional_kwargs={\n",
" \"function_call\": {\n",
" \"name\": \"tavily_search_results_json\", \n",
" \"arguments\": json.dumps({\"query\": human_input})\n",
" }\n",
" }\n",
" )\n",
" ]\n",
" }"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!\n",
"\n",
"**MODIFICATION**\n",
"\n",
"We will define a `first_agent` node which we will set as the entrypoint."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the new entrypoint\n",
"workflow.add_node(\"first_agent\", first_model)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"first_agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"\n",
"# After we call the first agent, we know we want to go to action\n",
"workflow.add_edge('first_agent', 'action')\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "8edb04b9-40b6-46f1-a7a8-4b2d8aba7752",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'first_agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'name': 'tavily_search_results_json', 'arguments': '{\"query\": \"what is the weather in sf\"}'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoJanuary 2024 Weather History in San Francisco California, United States. The data for this report comes from the San Francisco International Airport. ... frigid 15°F freezing 32°F very cold 45°F cold 55°F cool 65°F comfortable 75°F warm 85°F hot 95°F sweltering. The hourly reported temperature, color coded into bands. ...'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco on this [website](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'name': 'tavily_search_results_json', 'arguments': '{\"query\": \"what is the weather in sf\"}'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoJanuary 2024 Weather History in San Francisco California, United States. The data for this report comes from the San Francisco International Airport. ... frigid 15°F freezing 32°F very cold 45°F cold 55°F cool 65°F comfortable 75°F warm 85°F hot 95°F sweltering. The hourly reported temperature, color coded into bands. ...'}]\", name='tavily_search_results_json'), AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco on this [website](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "08ae8246-11d5-40e1-8567-361e5bef8917",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/high-level.ipynb
-136
View File
@@ -1,136 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "8bcd1a3d-7c50-4f58-be4e-1ed654aa33be",
"metadata": {},
"source": [
"# Chat Executor: with function calling\n",
"\n",
"This notebook walks through an example creating a chat executor that uses function calling.\n",
"This is useful for getting started quickly.\n",
"However, it is highly likely you will want to customize the logic - for information on that, check out the other examples in this folder."
]
},
{
"cell_type": "markdown",
"id": "e130cf70-a30e-47d7-8fd5-464f1a92e374",
"metadata": {},
"source": [
"## Set up the chat model and tools\n",
"\n",
"Here we will define the chat model and tools that we want to use.\n",
"Importantly, this model MUST support OpenAI function calling."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "efb7e3c0-c63f-40f6-93ce-19681d650fc2",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"from langgraph.prebuilt import chat_agent_executor\n",
"from langchain_core.messages import HumanMessage"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "a7025f33-3160-41cf-868b-17ebc916fb1d",
"metadata": {},
"outputs": [],
"source": [
"tools = [TavilySearchResults(max_results=1)]\n",
"model = ChatOpenAI()"
]
},
{
"cell_type": "markdown",
"id": "43064805-2ac9-4b5a-850c-a68dd7282350",
"metadata": {},
"source": [
"## Create executor\n",
"\n",
"We can now use the high level interface to create the executor"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "32b4ae66-f667-4a8b-a602-503fd0effcd9",
"metadata": {},
"outputs": [],
"source": [
"app = chat_agent_executor.create_function_calling_executor(model, tools)"
]
},
{
"cell_type": "markdown",
"id": "d63dbfc7-a5c1-4a03-991c-f0789ba52c52",
"metadata": {},
"source": [
"We can now invoke this executor. The input to this must be a dictionary with a single `messsages` key that contains a list of messages."
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "0abc5655-d772-450c-832f-1fee1111a5f6",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"----\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"----\n",
"{'messages': [AIMessage(content=\"I couldn't find the current weather in San Francisco. However, you can check historical weather data for January 2024 [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"----\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content=\"I couldn't find the current weather in San Francisco. However, you can check historical weather data for January 2024 [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"----\n"
]
}
],
"source": [
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for s in app.stream(inputs):\n",
" print(list(s.values())[0])\n",
" print(\"----\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "87f147e3-f96f-4b96-a3cc-ec7affd7a57f",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/human-in-the-loop.ipynb
-471
View File
@@ -1,471 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Chat Executor\n",
"\n",
"In this example we will build a chat executor that has a human in the loop. We will use the human to approve specific actions.\n",
"\n",
"This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example [here](./base.ipynb).\n",
"\n",
"Any modifications of that example are called below with **MODIFICATION**, so if you are looking for the differences you can just search for that."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"tools = [TavilySearchResults(max_results=1)]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "b547109f-f9e8-4e77-a7e7-ed2bae7a72ab",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage\n",
"\n",
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we continue\n",
" else:\n",
" return \"continue\"\n",
"\n",
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages']\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}"
]
},
{
"cell_type": "markdown",
"id": "ac402f66-4442-4a1f-9f9b-4a5d97532ceb",
"metadata": {},
"source": [
"**MODIFICATION**\n",
"\n",
"We modify the function that is calling the tool to first ask for user approval to continue. Note that this is a simple example and we could modify it to change the tool input, use some other channel besides input, etc."
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "73fd6432-42e8-472a-89ca-bb5ddbbcc35a",
"metadata": {},
"outputs": [],
"source": [
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" action = ToolInvocation(\n",
" tool=last_message.additional_kwargs[\"function_call\"][\"name\"],\n",
" tool_input=json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"]),\n",
" )\n",
" response = input(prompt=f\"[y/n] continue with: {action}?\")\n",
" if response == \"n\":\n",
" raise ValueError\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "f544977e-31f7-41f0-88c4-ec9c27b8cecb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n"
]
},
{
"name": "stdin",
"output_type": "stream",
"text": [
"[y/n] continue with: tool='tavily_search_results_json' tool_input={'query': 'weather in San Francisco'}? y\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "08ae8246-11d5-40e1-8567-361e5bef8917",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/managing-agent-steps.ipynb
-463
View File
@@ -1,463 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Chat Executor\n",
"\n",
"In this example we will build a chat executor that better manages intermediate steps. The base chat executor will just put all messages into the model, but if the intermediate steps an agent is taking start to get long, you may want to modify that. In this example we will only include the ten most recent messages.\n",
"\n",
"This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example [here](./base.ipynb).\n",
"\n",
"Any modifications of that example are called below with **MODIFICATION**, so if you are looking for the differences you can just search for that."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"tools = [TavilySearchResults(max_results=1)]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e718a9c5-6596-457f-ac25-a25d8cb8c259",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage\n",
"\n",
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we continue\n",
" else:\n",
" return \"continue\""
]
},
{
"cell_type": "markdown",
"id": "a763aa63-701c-40fa-a9d3-9d992ebe7e4d",
"metadata": {},
"source": [
"**MODIFICATION**\n",
"\n",
"Here we don't pass all messages to the model but rather only pass the five most recent. Note that this is a pretty simplistic way to handle messages, and there may be other methods you may want to look into depending on your use case"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "714e4135-7cb5-4f17-b2ae-46f7e98bde61",
"metadata": {},
"outputs": [],
"source": [
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages'][-5:]\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "b3ca9564-63cc-4309-b158-5e8d3e907164",
"metadata": {},
"outputs": [],
"source": [
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" action = ToolInvocation(\n",
" tool=last_message.additional_kwargs[\"function_call\"][\"name\"],\n",
" tool_input=json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"]),\n",
" )\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "f544977e-31f7-41f0-88c4-ec9c27b8cecb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content=\"I'm sorry, but I couldn't find the current weather in San Francisco. However, you can check the historical weather data for January 2024 in San Francisco [here](https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States).\")]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "08ae8246-11d5-40e1-8567-361e5bef8917",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
examples/chat_executor_with_function_calling/respond-in-format.ipynb
-450
View File
@@ -1,450 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "51466c8d-8ce4-4b3d-be4e-18fdbeda5f53",
"metadata": {},
"source": [
"# Chat Executor\n",
"\n",
"In this example we will build a chat executor that responds in a specific format. We will do this by using OpenAI function calling. This is useful when you want to enforce that an agent's response is in a specific format. In this example, we will ask it respond as if a weatherman, so to return the temperature and then any other additional info.\n",
"\n",
"This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example [here](./base.ipynb).\n",
"\n",
"Any modifications of that example are called below with **MODIFICATION**, so if you are looking for the differences you can just search for that."
]
},
{
"cell_type": "markdown",
"id": "7cbd446a-808f-4394-be92-d45ab818953c",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"First we need to install the packages required"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "af4ce0ba-7596-4e5f-8bf8-0b0bd6e62833",
"metadata": {},
"outputs": [],
"source": [
"!pip install --quiet -U langchain langchain_openai tavily-python"
]
},
{
"cell_type": "markdown",
"id": "0abe11f4-62ed-4dc4-8875-3db21e260d1d",
"metadata": {},
"source": [
"Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c903a1cf-2977-4e2d-ad7d-8b3946821d89",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import getpass\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"OpenAI API Key:\")\n",
"os.environ[\"TAVILY_API_KEY\"] = getpass.getpass(\"Tavily API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "f0ed46a8-effe-4596-b0e1-a6a29ee16f5c",
"metadata": {},
"source": [
"Optionally, we can set API key for [LangSmith tracing](https://smith.langchain.com/), which will give us best-in-class observability."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "95e25aec-7c9f-4a63-b143-225d0e9a79c3",
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
"os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass(\"LangSmith API Key:\")"
]
},
{
"cell_type": "markdown",
"id": "21ac643b-cb06-4724-a80c-2862ba4773f1",
"metadata": {},
"source": [
"## Set up the tools\n",
"\n",
"We will first define the tools we want to use.\n",
"For this simple example, we will use a built-in search tool via Tavily.\n",
"However, it is really easy to create your own tools - see documentation [here](https://python.langchain.com/docs/modules/agents/tools/custom_tools) on how to do that.\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d7ef57dd-5d6e-4ad3-9377-a92201c1310e",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.tools.tavily_search import TavilySearchResults\n",
"\n",
"tools = [TavilySearchResults(max_results=1)]"
]
},
{
"cell_type": "markdown",
"id": "01885785-b71a-44d1-b1d6-7b5b14d53b58",
"metadata": {},
"source": [
"We can now wrap these tools in a simple ToolExecutor.\n",
"This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output.\n",
"A ToolInvocation is any class with `tool` and `tool_input` attribute.\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5cf3331e-ccb3-41c8-aeb9-a840a94d41e7",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolExecutor\n",
"\n",
"tool_executor = ToolExecutor(tools)"
]
},
{
"cell_type": "markdown",
"id": "5497ed70-fce3-47f1-9cad-46f912bad6a5",
"metadata": {},
"source": [
"## Set up the model\n",
"\n",
"Now we need to load the chat model we want to use.\n",
"Importantly, this should satisfy two criteria:\n",
"\n",
"1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.\n",
"2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.\n",
"\n",
"Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "892b54b9-75f0-4804-9ed0-88b5e5532989",
"metadata": {},
"outputs": [],
"source": [
"from langchain_openai import ChatOpenAI\n",
"\n",
"# We will set streaming=True so that we can stream tokens\n",
"# See the streaming section for more information on this.\n",
"model = ChatOpenAI(temperature=0, streaming=True)"
]
},
{
"cell_type": "markdown",
"id": "a77995c0-bae2-4cee-a036-8688a90f05b9",
"metadata": {},
"source": [
"\n",
"After we've done this, we should make sure the model knows that it has these tools available to call.\n",
"We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.\n",
"\n",
"\n",
"**MODIFICATION**\n",
"\n",
"We also want to define a response schema for the language model and bind it to the model as a function as well"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "cd3cbae5-d92c-4559-a4aa-44721b80d107",
"metadata": {},
"outputs": [],
"source": [
"from langchain.tools.render import format_tool_to_openai_function\n",
"from langchain_core.pydantic_v1 import BaseModel, Field\n",
"from langchain_core.utils.function_calling import convert_pydantic_to_openai_function\n",
"\n",
"class Response(BaseModel):\n",
" \"\"\"Final response to the user\"\"\"\n",
" temperature: float = Field(description=\"the temperature\")\n",
" other_notes: str = Field(description=\"any other notes about the weather\")\n",
"\n",
"\n",
"functions = [format_tool_to_openai_function(t) for t in tools]\n",
"functions.append(convert_pydantic_to_openai_function(Response))\n",
"model = model.bind_functions(functions)"
]
},
{
"cell_type": "markdown",
"id": "8e8b9211-93d0-4ad5-aa7a-9c09099c53ff",
"metadata": {},
"source": [
"## Define the agent state\n",
"\n",
"The main type of graph in `langgraph` is the `StatefulGraph`.\n",
"This graph is parameterized by a state object that it passes around to each node.\n",
"Each node then returns operations to update that state.\n",
"These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute.\n",
"Whether to set or add is denoted by annotating the state object you construct the graph with.\n",
"\n",
"For this example, the state we will track will just be a list of messages.\n",
"We want each node to just add messages to that list.\n",
"Therefore, we will use a `TypedDict` with one key (`messages`) and annotate it so that the `messages` attribute is always added to.\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ea793afa-2eab-4901-910d-6eed90cd6564",
"metadata": {},
"outputs": [],
"source": [
"from typing import TypedDict, Annotated, Sequence\n",
"import operator\n",
"from langchain_core.messages import BaseMessage\n",
"\n",
"\n",
"class AgentState(TypedDict):\n",
" messages: Annotated[Sequence[BaseMessage], operator.add]"
]
},
{
"cell_type": "markdown",
"id": "e03c5094-9297-4d19-a04e-3eedc75cefb4",
"metadata": {},
"source": [
"## Define the nodes\n",
"\n",
"We now need to define a few different nodes in our graph.\n",
"In `langgraph`, a node can be either a function or a [runnable](https://python.langchain.com/docs/expression_language/).\n",
"There are two main nodes we need for this:\n",
"\n",
"1. The agent: responsible for deciding what (if any) actions to take.\n",
"2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.\n",
"\n",
"We will also need to define some edges.\n",
"Some of these edges may be conditional.\n",
"The reason they are conditional is that based on the output of a node, one of several paths may be taken.\n",
"The path that is taken is not known until that node is run (the LLM decides).\n",
"\n",
"1. Conditional Edge: after the agent is called, we should either:\n",
" a. If the agent said to take an action, then the function to invoke tools should be called\n",
" b. If the agent said that it was finished, then it should finish\n",
"2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next\n",
"\n",
"Let's define the nodes, as well as a function to decide how what conditional edge to take.\n",
"\n",
"**MODIFICATION**\n",
"\n",
"We will change the `should_continue` function to check what function was called. If the function `Response` was called - that is the function that is NOT a tool, but rather the formatted response, so we should NOT continue in that case."
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "3b541bb9-900c-40d0-964d-7b5dfee30667",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.prebuilt import ToolInvocation\n",
"import json\n",
"from langchain_core.messages import FunctionMessage\n",
"\n",
"# Define the function that determines whether to continue or not\n",
"def should_continue(state):\n",
" messages = state['messages']\n",
" last_message = messages[-1]\n",
" # If there is no function call, then we finish\n",
" if \"function_call\" not in last_message.additional_kwargs:\n",
" return \"end\"\n",
" # Otherwise if there is, we need to check what type of function call it is\n",
" elif last_message.additional_kwargs[\"function_call\"][\"name\"] == \"Response\":\n",
" return \"end\"\n",
" # Otherwise we continue\n",
" else:\n",
" return \"continue\"\n",
"\n",
"# Define the function that calls the model\n",
"def call_model(state):\n",
" messages = state['messages']\n",
" response = model.invoke(messages)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [response]}\n",
"\n",
"# Define the function to execute tools\n",
"def call_tool(state):\n",
" messages = state['messages']\n",
" # Based on the continue condition\n",
" # we know the last message involves a function call\n",
" last_message = messages[-1]\n",
" # We construct an ToolInvocation from the function_call\n",
" action = ToolInvocation(\n",
" tool=last_message.additional_kwargs[\"function_call\"][\"name\"],\n",
" tool_input=json.loads(last_message.additional_kwargs[\"function_call\"][\"arguments\"]),\n",
" )\n",
" # We call the tool_executor and get back a response\n",
" response = tool_executor.invoke(action)\n",
" # We use the response to create a FunctionMessage\n",
" function_message = FunctionMessage(content=str(response), name=action.tool)\n",
" # We return a list, because this will get added to the existing list\n",
" return {\"messages\": [function_message]}"
]
},
{
"cell_type": "markdown",
"id": "ffd6e892-946c-4899-8cc0-7c9291c1f73b",
"metadata": {},
"source": [
"## Define the graph\n",
"\n",
"We can now put it all together and define the graph!"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "813ae66c-3b58-4283-a02a-36da72a2ab90",
"metadata": {},
"outputs": [],
"source": [
"from langgraph.graph import StateGraph, END\n",
"# Define a new graph\n",
"workflow = StateGraph(AgentState)\n",
"\n",
"# Define the two nodes we will cycle between\n",
"workflow.add_node(\"agent\", call_model)\n",
"workflow.add_node(\"action\", call_tool)\n",
"\n",
"# Set the entrypoint as `agent`\n",
"# This means that this node is the first one called\n",
"workflow.set_entry_point(\"agent\")\n",
"\n",
"# We now add a conditional edge\n",
"workflow.add_conditional_edges(\n",
" # First, we define the start node. We use `agent`.\n",
" # This means these are the edges taken after the `agent` node is called.\n",
" \"agent\",\n",
" # Next, we pass in the function that will determine which node is called next.\n",
" should_continue,\n",
" # Finally we pass in a mapping.\n",
" # The keys are strings, and the values are other nodes.\n",
" # END is a special node marking that the graph should finish.\n",
" # What will happen is we will call `should_continue`, and then the output of that\n",
" # will be matched against the keys in this mapping.\n",
" # Based on which one it matches, that node will then be called.\n",
" {\n",
" # If `tools`, then we call the tool node.\n",
" \"continue\": \"action\",\n",
" # Otherwise we finish.\n",
" \"end\": END\n",
" }\n",
")\n",
"\n",
"# We now add a normal edge from `tools` to `agent`.\n",
"# This means that after `tools` is called, `agent` node is called next.\n",
"workflow.add_edge('action', 'agent')\n",
"\n",
"# Finally, we compile it!\n",
"# This compiles it into a LangChain Runnable,\n",
"# meaning you can use it as you would any other runnable\n",
"app = workflow.compile()"
]
},
{
"cell_type": "markdown",
"id": "547c3931-3dae-4281-ad4e-4b51305594d4",
"metadata": {},
"source": [
"## Use it!\n",
"\n",
"We can now use it!\n",
"This now exposes the [same interface](https://python.langchain.com/docs/expression_language/) as all other LangChain runnables."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "f544977e-31f7-41f0-88c4-ec9c27b8cecb",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node 'action':\n",
"---\n",
"{'messages': [FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json')]}\n",
"\n",
"---\n",
"\n",
"Output from node 'agent':\n",
"---\n",
"{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"temperature\": 60,\\n \"other_notes\": \"Partly cloudy\"\\n}', 'name': 'Response'}})]}\n",
"\n",
"---\n",
"\n",
"Output from node '__end__':\n",
"---\n",
"{'messages': [HumanMessage(content='what is the weather in sf'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"query\": \"weather in San Francisco\"\\n}', 'name': 'tavily_search_results_json'}}), FunctionMessage(content=\"[{'url': 'https://weatherspark.com/h/m/557/2024/1/Historical-Weather-in-January-2024-in-San-Francisco-California-United-States', 'content': 'January 2024 Weather History in San Francisco California, United States Daily Precipitation in January 2024 in San Francisco Observed Weather in January 2024 in San Francisco San Francisco Temperature History January 2024 Hourly Temperature in January 2024 in San Francisco Hours of Daylight and Twilight in January 2024 in San FranciscoThis report shows the past weather for San Francisco, providing a weather history for January 2024. It features all historical weather data series we have available, including the San Francisco temperature history for January 2024. You can drill down from year to month and even day level reports by clicking on the graphs.'}]\", name='tavily_search_results_json'), AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\\n \"temperature\": 60,\\n \"other_notes\": \"Partly cloudy\"\\n}', 'name': 'Response'}})]}\n",
"\n",
"---\n",
"\n"
]
}
],
"source": [
"from langchain_core.messages import HumanMessage\n",
"\n",
"inputs = {\"messages\": [HumanMessage(content=\"what is the weather in sf\")]}\n",
"for output in app.stream(inputs):\n",
" # stream() yields dictionaries with output keyed by node name\n",
" for key, value in output.items():\n",
" print(f\"Output from node '{key}':\")\n",
" print(\"---\")\n",
" print(value)\n",
" print(\"\\n---\\n\")"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}