langchain/docs/docs/integrations/tools/dalle_image_generator.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dall-E Image Generator\n",
    "\n",
    ">[OpenAI Dall-E](https://openai.com/dall-e-3) are text-to-image models developed by `OpenAI` using deep learning methodologies to generate digital images from natural language descriptions, called \"prompts\".\n",
    "\n",
    "This notebook shows how you can generate images from a prompt synthesized using an OpenAI LLM. The images are generated using `Dall-E`, which uses the same OpenAI API key as the LLM."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Needed if you would like to display images in the notebook\n",
    "%pip install --upgrade --quiet  opencv-python scikit-image langchain-community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "q-k8wmp0zquh"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "from langchain_openai import OpenAI\n",
    "\n",
    "os.environ[\"OPENAI_API_KEY\"] = \"insertapikey\""
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run as a chain"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.chains import LLMChain\n",
    "from langchain_community.utilities.dalle_image_generator import DallEAPIWrapper\n",
    "from langchain_core.prompts import PromptTemplate\n",
    "from langchain_openai import OpenAI\n",
    "\n",
    "llm = OpenAI(temperature=0.9)\n",
    "prompt = PromptTemplate(\n",
    "    input_variables=[\"image_desc\"],\n",
    "    template=\"Generate a detailed prompt to generate an image based on the following description: {image_desc}\",\n",
    ")\n",
    "chain = LLMChain(llm=llm, prompt=prompt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_url = DallEAPIWrapper().run(chain.run(\"halloween night at a haunted museum\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_url"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# You can click on the link above to display the image\n",
    "# Or you can try the options below to display the image inline in this notebook\n",
    "\n",
    "try:\n",
    "    import google.colab\n",
    "\n",
    "    IN_COLAB = True\n",
    "except ImportError:\n",
    "    IN_COLAB = False\n",
    "\n",
    "if IN_COLAB:\n",
    "    from google.colab.patches import cv2_imshow  # for image display\n",
    "    from skimage import io\n",
    "\n",
    "    image = io.imread(image_url)\n",
    "    cv2_imshow(image)\n",
    "else:\n",
    "    import cv2\n",
    "    from skimage import io\n",
    "\n",
    "    image = io.imread(image_url)\n",
    "    cv2.imshow(\"image\", image)\n",
    "    cv2.waitKey(0)  # wait for a keyboard input\n",
    "    cv2.destroyAllWindows()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run as a tool with an agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.tools.openai_dalle_image_generation import (\n",
    "    OpenAIDALLEImageGenerationTool,\n",
    ")\n",
    "from langchain_community.utilities.dalle_image_generator import DallEAPIWrapper\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langgraph.prebuilt import create_react_agent\n",
    "\n",
    "llm = ChatOpenAI(model=\"gpt-3.5-turbo\", temperature=0)\n",
    "api_wrapper = DallEAPIWrapper()\n",
    "dalle_tool = OpenAIDALLEImageGenerationTool(api_wrapper=api_wrapper)\n",
    "\n",
    "tools = [dalle_tool]\n",
    "\n",
    "agent = create_react_agent(llm, tools, debug=True)\n",
    "\n",
    "# User prompt\n",
    "prompt = \"Create an image of a halloween night at a haunted museum\"\n",
    "\n",
    "messages = [\n",
    "    # \"role\": \"user\" Indicates message is coming from user\n",
    "    # \"content\": prompt is where the user's input is placed\n",
    "    {\"role\": \"user\", \"content\": prompt}\n",
    "]\n",
    "\n",
    "# Sending the message to be processed and adjusted by ChatGPT, after which is sent through DALL-E\n",
    "response = agent.invoke({\"messages\": messages})\n",
    "\n",
    "print(response)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python (langchain-env)",
   "language": "python",
   "name": "langchain-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}