05-LangChainAgents.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LangChain Agents!\n",
    "\n",
    "NOTE: You will need a Google Search API Key.\n",
    "You can get a key for free from https://serpapi.com/\n",
    "\n",
    "Make sure that you have an environment variable named: SERPAPI_API_KEY\n",
    "with the value being the value of your key"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pip -q install langchain openai google-search-results tiktoken wikipedia numexpr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.agents import load_tools\n",
    "from langchain.agents import initialize_agent\n",
    "from langchain.llms import OpenAI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "llm = OpenAI(temperature=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tools = load_tools([\"serpapi\", \"llm-math\"], llm=llm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# What is a tool like this\n",
    "tools[0].name, tools[0].description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent = initialize_agent(tools, \n",
    "                         llm, \n",
    "                         agent=\"zero-shot-react-description\", \n",
    "                         verbose=True,\n",
    "                         handle_parsing_errors=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# take a look at the template!\n",
    "# NOTE!!!! : The agent already has built up a prompt template for us based on the description of the tool\n",
    "agent.agent.llm_chain.prompt.template"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chain Of Thought\n",
    "\n",
    "Running these question on the agents will cause it to get through a series of observations and thoughts.\n",
    "See below!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent.run(\"Hi, How are you today?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "agent.run(\"Who is the current president of the United States?\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent.run(\"What is the average age in the United States?\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Smarter Agent\n",
    "\n",
    "Let's make an even smarter agent that can do more things!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tools = load_tools([\"serpapi\", \"llm-math\",\"wikipedia\",\"terminal\"], llm=llm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent = initialize_agent(tools, \n",
    "                         llm, \n",
    "                         agent=\"zero-shot-react-description\", \n",
    "                         verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent.agent.llm_chain.prompt.template"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "agent.run(\"Where is Apple's office headquarters?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "agent.run(\"What files are in my current directory?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent.run(\"Does my current directory have python notebook files and if so how many?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}