langchain/docs/docs/integrations/chat/zhipuai.ipynb

{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "---\n",
    "sidebar_label: ZHIPU AI\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ZHIPU AI\n",
    "\n",
    "This notebook shows how to use [ZHIPU AI API](https://open.bigmodel.cn/dev/api) in LangChain with the langchain.chat_models.ChatZhipuAI.\n",
    "\n",
    ">[*GLM-4*](https://open.bigmodel.cn/) is a multi-lingual large language model aligned with human intent, featuring capabilities in Q&A, multi-turn dialogue, and code generation. The overall performance of the new generation base model GLM-4 has been significantly improved compared to the previous generation, supporting longer contexts; Stronger multimodality; Support faster inference speed, more concurrency, greatly reducing inference costs; Meanwhile, GLM-4 enhances the capabilities of intelligent agents.\n",
    "\n",
    "## Getting started\n",
    "### Installation\n",
    "First, ensure the zhipuai package is installed in your Python environment. Run the following command:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install --upgrade httpx httpx-sse PyJWT"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing the Required Modules\n",
    "After installation, import the necessary modules to your Python script:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.chat_models import ChatZhipuAI\n",
    "from langchain_core.messages import AIMessage, HumanMessage, SystemMessage"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setting Up Your API Key\n",
    "Sign in to [ZHIPU AI](https://open.bigmodel.cn/login?redirect=%2Fusercenter%2Fapikeys) for the an API Key to access our models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "os.environ[\"ZHIPUAI_API_KEY\"] = \"zhipuai_api_key\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Initialize the ZHIPU AI Chat Model\n",
    "Here's how to initialize the chat model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "chat = ChatZhipuAI(\n",
    "    model=\"glm-4\",\n",
    "    temperature=0.5,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Basic Usage\n",
    "Invoke the model with system and human messages like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "messages = [\n",
    "    AIMessage(content=\"Hi.\"),\n",
    "    SystemMessage(content=\"Your role is a poet.\"),\n",
    "    HumanMessage(content=\"Write a short poem about AI in four lines.\"),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "response = chat(messages)\n",
    "print(response.content)  # Displays the AI-generated poem"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Advanced Features\n",
    "### Streaming Support\n",
    "For continuous interaction, use the streaming feature:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.callbacks.manager import CallbackManager\n",
    "from langchain_core.callbacks.streaming_stdout import StreamingStdOutCallbackHandler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "streaming_chat = ChatZhipuAI(\n",
    "    model=\"glm-4\",\n",
    "    temperature=0.5,\n",
    "    streaming=True,\n",
    "    callback_manager=CallbackManager([StreamingStdOutCallbackHandler()]),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "streaming_chat(messages)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Asynchronous Calls\n",
    "For non-blocking calls, use the asynchronous approach:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "async_chat = ChatZhipuAI(\n",
    "    model=\"glm-4\",\n",
    "    temperature=0.5,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "response = await async_chat.agenerate([messages])\n",
    "print(response)"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Using With Functions Call\n",
    "\n",
    "GLM-4 Model can be used with the function call as well，use the following code to run a simple LangChain json_chat_agent."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "os.environ[\"TAVILY_API_KEY\"] = \"tavily_api_key\""
   ],
   "metadata": {
    "collapsed": false
   },
   "execution_count": null
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "from langchain import hub\n",
    "from langchain.agents import AgentExecutor, create_json_chat_agent\n",
    "from langchain_community.tools.tavily_search import TavilySearchResults\n",
    "\n",
    "tools = [TavilySearchResults(max_results=1)]\n",
    "prompt = hub.pull(\"hwchase17/react-chat-json\")\n",
    "llm = ChatZhipuAI(temperature=0.01, model=\"glm-4\")\n",
    "\n",
    "agent = create_json_chat_agent(llm, tools, prompt)\n",
    "agent_executor = AgentExecutor(\n",
    "    agent=agent, tools=tools, verbose=True, handle_parsing_errors=True\n",
    ")"
   ],
   "metadata": {
    "collapsed": false
   },
   "execution_count": null
  },
  {
   "cell_type": "code",
   "outputs": [],
   "source": [
    "agent_executor.invoke({\"input\": \"what is LangChain?\"})"
   ],
   "metadata": {
    "collapsed": false
   },
   "execution_count": null
  }
 ],
 "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}