{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "ZRGlUb8O4fPV"
   },
   "outputs": [],
   "source": [
    "%pip install -U langgraph langchain_google_genai langchain_community langgraph-checkpoint-postgres  openai langchain_groq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "cMfPUmHIxqTi"
   },
   "outputs": [],
   "source": [
    "%%capture --no-stderr\n",
    "%pip install --upgrade --quiet  playwright > /dev/null\n",
    "%pip install --upgrade --quiet  lxml browser-use langchain_openai"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kkZ7jVUOUV7Q"
   },
   "outputs": [],
   "source": [
    "!playwright install"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "-_T1MhnGUl2q"
   },
   "outputs": [],
   "source": [
    "!pip install \"anyio<4\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "yARYirp1UhDR"
   },
   "outputs": [],
   "source": [
    "# This import is required only for jupyter notebooks, since they have their own eventloop\n",
    "import nest_asyncio\n",
    "\n",
    "nest_asyncio.apply()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "jyVP10O_5Qck"
   },
   "outputs": [],
   "source": [
    "from google.colab import userdata\n",
    "from langchain_openai import ChatOpenAI\n",
    "\n",
    "llm = ChatOpenAI(model='gpt-4o-mini', temperature=0, api_key=userdata.get('Open_api_key'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e9duizdv5cOH",
    "outputId": "a07b1702-d485-4641-c307-601e6ab34b9b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AIMessage(content='Hello! How can I assist you today?', additional_kwargs={'refusal': None}, response_metadata={'token_usage': {'completion_tokens': 10, 'prompt_tokens': 8, 'total_tokens': 18, 'completion_tokens_details': {'accepted_prediction_tokens': 0, 'audio_tokens': 0, 'reasoning_tokens': 0, 'rejected_prediction_tokens': 0}, 'prompt_tokens_details': {'audio_tokens': 0, 'cached_tokens': 0}}, 'model_name': 'gpt-4o-mini-2024-07-18', 'system_fingerprint': 'fp_bd83329f63', 'finish_reason': 'stop', 'logprobs': None}, id='run-28a9088f-7539-412a-aa80-1663be40e74f-0', usage_metadata={'input_tokens': 8, 'output_tokens': 10, 'total_tokens': 18, 'input_token_details': {'audio': 0, 'cache_read': 0}, 'output_token_details': {'audio': 0, 'reasoning': 0}})"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "llm.invoke('hi')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "wS8ouhiVQ2dL",
    "outputId": "653879a8-b3ac-4178-edee-5cd834e3404a"
   },
   "outputs": [],
   "source": [
    "import asyncio\n",
    "\n",
    "from langchain_openai import ChatOpenAI\n",
    "\n",
    "from browser_use import Agent\n",
    "from browser_use.browser import BrowserProfile, BrowserSession\n",
    "\n",
    "# Basic configuration for the browser\n",
    "browser_profile = BrowserProfile(\n",
    "\theadless=True,  # Run in headless mode\n",
    "\t# disable_security=True  # Uncomment if you want to disable security\n",
    ")\n",
    "\n",
    "# Initialize the browser session with the specified profile\n",
    "browser_session = BrowserSession(browser_profile=browser_profile)\n",
    "\n",
    "\n",
    "async def main():\n",
    "\t# Initialize the agent with the task and language model\n",
    "\tagent = Agent(\n",
    "\t\ttask='What is Langgraph',\n",
    "\t\tllm=llm,  # Replace with your LLM configuration\n",
    "\t\tbrowser_session=browser_session,\n",
    "\t\tgenerate_gif=False,  # Disable GIF generation\n",
    "\t)\n",
    "\n",
    "\t# Run the agent and get results asynchronously\n",
    "\tresult = await agent.run()\n",
    "\n",
    "\t# Process results token-wise\n",
    "\tfor action in result.action_results():\n",
    "\t\tprint(action.extracted_content, end='\\r', flush=True)\n",
    "\t\tprint('\\n\\n')\n",
    "\t\t# if action.is_done:\n",
    "\t\t#     print(action.extracted_content)\n",
    "\n",
    "\t# Close the browser after completion\n",
    "\tawait browser_session.close()\n",
    "\n",
    "\n",
    "# Run the asynchronous main function\n",
    "asyncio.run(main())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "TFK-fNoLDFcF",
    "outputId": "d78fbeae-c8f0-4c26-e0e3-7a0a683d3fc1"
   },
   "outputs": [],
   "source": [
    "# from browser_use import Agent\n",
    "import asyncio\n",
    "\n",
    "from langchain_openai import ChatOpenAI\n",
    "\n",
    "from browser_use.browser import BrowserProfile, BrowserSession\n",
    "\n",
    "# Basic configuration\n",
    "browser_profile = BrowserProfile(\n",
    "\theadless=True,\n",
    "\t# disable_security=True\n",
    ")\n",
    "# Reuse existing browser\n",
    "browser_session = BrowserSession(browser_profile=browser_profile)\n",
    "# async def main():\n",
    "agent = Agent(\n",
    "\ttask='what is langchain',\n",
    "\tllm=llm,\n",
    "\tbrowser_session=browser_session,\n",
    "\tgenerate_gif=False,  # Browser instance will be reused\n",
    ")\n",
    "\n",
    "result = await agent.run()\n",
    "print(result)\n",
    "# Manually close the browser\n",
    "# asyncio.run(main())\n",
    "await browser_session.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "nKGC936xODry",
    "outputId": "de70d715-c30a-4d5b-9d25-40bd79d410de"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LangChain is a composable framework designed for building applications with large language models (LLMs). It simplifies the integration of language models with external data sources and is open-source, supported by an active community. LangChain provides tools for developers to streamline the application lifecycle of LLMs.\n"
     ]
    }
   ],
   "source": [
    "# display(result.action_results())\n",
    "for action in result.action_results():\n",
    "\tif action.is_done:\n",
    "\t\tprint(action.extracted_content)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "name": "python3"
  },
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}