langchain/docs/modules/utils/examples/dalle_image_generator.ipynb

{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Dall-E Image Generator\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 opencv-python scikit-image"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "q-k8wmp0zquh"
      },
      "outputs": [],
      "source": [
        "from langchain.llms import OpenAI\n",
        "import os\n",
        "os.environ[\"OPENAI_API_KEY\"] = \"<your-key-here>\""
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Run as a chain"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [],
      "source": [
        "from langchain.utilities.dalle_image_generator import DallEAPIWrapper\n",
        "from langchain.prompts import PromptTemplate\n",
        "from langchain.chains import LLMChain\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": 7,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "https://oaidalleapiprodscus.blob.core.windows.net/private/org-rocrupyvzgcl4yf25rqq6d1v/user-WsxrbKyP2c8rfhCKWDyMfe8N/img-mg1OWiziXxQN1aR2XRsLNndg.png?st=2023-01-31T07%3A34%3A15Z&se=2023-01-31T09%3A34%3A15Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-01-30T22%3A19%3A44Z&ske=2023-01-31T22%3A19%3A44Z&sks=b&skv=2021-08-06&sig=XDPee5aEng%2BcbXq2mqhh39uHGZTBmJgGAerSd0g%2BMEs%3D\n"
          ]
        }
      ],
      "source": [
        "image_url = DallEAPIWrapper().run(chain.run(\"halloween night at a haunted museum\"))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "# You can click on the link above to display the image for\n",
        "# Or you can try the options below to display the image inline in this notebook\n",
        "\n",
        "try:\n",
        "    import google.colab\n",
        "    IN_COLAB = True\n",
        "except:\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()\n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Run as a tool with an agent"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
            "\u001b[32;1m\u001b[1;3m What is the best way to turn this description into an image?\n",
            "Action: Dall-E Image Generator\n",
            "Action Input: A spooky Halloween night at a haunted museum\u001b[0mhttps://oaidalleapiprodscus.blob.core.windows.net/private/org-rocrupyvzgcl4yf25rqq6d1v/user-WsxrbKyP2c8rfhCKWDyMfe8N/img-ogKfqxxOS5KWVSj4gYySR6FY.png?st=2023-01-31T07%3A38%3A25Z&se=2023-01-31T09%3A38%3A25Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-01-30T22%3A19%3A36Z&ske=2023-01-31T22%3A19%3A36Z&sks=b&skv=2021-08-06&sig=XsomxxBfu2CP78SzR9lrWUlbask4wBNnaMsHamy4VvU%3D\n",
            "\n",
            "Observation: \u001b[36;1m\u001b[1;3mhttps://oaidalleapiprodscus.blob.core.windows.net/private/org-rocrupyvzgcl4yf25rqq6d1v/user-WsxrbKyP2c8rfhCKWDyMfe8N/img-ogKfqxxOS5KWVSj4gYySR6FY.png?st=2023-01-31T07%3A38%3A25Z&se=2023-01-31T09%3A38%3A25Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-01-30T22%3A19%3A36Z&ske=2023-01-31T22%3A19%3A36Z&sks=b&skv=2021-08-06&sig=XsomxxBfu2CP78SzR9lrWUlbask4wBNnaMsHamy4VvU%3D\u001b[0m\n",
            "Thought:\u001b[32;1m\u001b[1;3m With the image generated, I can now make my final answer.\n",
            "Final Answer: An image of a Halloween night at a haunted museum can be seen here: https://oaidalleapiprodscus.blob.core.windows.net/private/org-rocrupyvzgcl4yf25rqq6d1v/user-WsxrbKyP2c8rfhCKWDyMfe8N/img-ogKfqxxOS5KWVSj4gYySR6FY.png?st=2023-01-31T07%3A38%3A25Z&se=2023-01-31T09%3A38%3A25Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-01-30T22\u001b[0m\n",
            "\n",
            "\u001b[1m> Finished chain.\u001b[0m\n"
          ]
        }
      ],
      "source": [
        "from langchain.agents import load_tools\n",
        "from langchain.agents import initialize_agent\n",
        "\n",
        "tools = load_tools(['dalle-image-generator'])\n",
        "agent = initialize_agent(tools, llm, agent=\"zero-shot-react-description\", verbose=True)\n",
        "output = agent.run(\"Create an image of a halloween night at a haunted museum\")"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "langchain",
      "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": "3.9.16"
    },
    "vscode": {
      "interpreter": {
        "hash": "3570c8892273ffbeee7ead61dc7c022b73551d9f55fb2584ac0e8e8920b18a89"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}