DALL_E_CLIP_VAE_Explained.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "DALL-E_CLIP-VAE_Explained",
      "provenance": [],
      "collapsed_sections": [
        "P301l12fEY6A",
        "KAcixx9Z3XYH",
        "JeMCHwDdHIcu",
        "UlIUu0jK3S19",
        "GHB7BcR1zNLZ",
        "1yryNdZUUXnl",
        "XaocGDQXz3Zx",
        "WztSrRF23Rqg"
      ],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "fc3b748138564b8cba66722ad838042a": {
          "model_module": "@jupyter-widgets/output",
          "model_name": "OutputModel",
          "model_module_version": "1.0.0",
          "state": {
            "_view_name": "OutputView",
            "msg_id": "",
            "_dom_classes": [],
            "_model_name": "OutputModel",
            "outputs": [
              {
                "output_type": "display_data",
                "metadata": {
                  "tags": [],
                  "needs_background": "light"
                },
                "image/png": "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\n",
                "text/plain": "<Figure size 864x864 with 1 Axes>"
              }
            ],
            "_view_module": "@jupyter-widgets/output",
            "_model_module_version": "1.0.0",
            "_view_count": null,
            "_view_module_version": "1.0.0",
            "layout": "IPY_MODEL_c30e317e87f745fc92aad293738f5a1f",
            "_model_module": "@jupyter-widgets/output"
          }
        },
        "c30e317e87f745fc92aad293738f5a1f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "967fd7256f2849228b49a57ef64e97a0": {
          "model_module": "@jupyter-widgets/output",
          "model_name": "OutputModel",
          "model_module_version": "1.0.0",
          "state": {
            "_view_name": "OutputView",
            "msg_id": "",
            "_dom_classes": [],
            "_model_name": "OutputModel",
            "outputs": [
              {
                "output_type": "display_data",
                "metadata": {
                  "tags": []
                },
                "text/html": "\n    <video width=400 controls>\n          <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\">\n    </video>\n    ",
                "text/plain": "<IPython.core.display.HTML object>"
              },
              {
                "output_type": "stream",
                "metadata": {
                  "tags": []
                },
                "text": "\u001b[31mInterrupted at 150 iterations\u001b[0m\n",
                "stream": "stdout"
              }
            ],
            "_view_module": "@jupyter-widgets/output",
            "_model_module_version": "1.0.0",
            "_view_count": null,
            "_view_module_version": "1.0.0",
            "layout": "IPY_MODEL_768719392ba04ec58e602b17ace429ce",
            "_model_module": "@jupyter-widgets/output"
          }
        },
        "768719392ba04ec58e602b17ace429ce": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/simonsanvil/DALL-E-Explained/blob/main/DALL_E_CLIP_VAE_Explained.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O78RfTZfh7ji"
      },
      "source": [
        "\n",
        "<h1>DALL-E and Zero-Shot Text-to-Image Generation Explained</h1>\n",
        "\n",
        "<p>Description and applications of OpenAI's paper about the DALL-E model and other text-to-image generation schemes (CLIP-dVAE)</p>\n",
        "\n",
        "By Simon S. Viloria (Github: [simonsanvil](https://github.com/simonsanvil))\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "----\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2aDcCMUnjQLh"
      },
      "source": [
        "# What is this notebook about?\n",
        "\n",
        "The primary reason of this notebook is to give a brief explanation about OpenAI's *Zero Shot Text-to-Image Generation* <a href=\"#scrollTo=iKP0tnHaiTyl\">(1)</sub></a> paper where they introduce *DALL-E*, a deep-leaning model to generate images directly from  a text-prompt. I will also showcase some of the outputs that can be accomplished with the model described in their paper and walkthrough how you can generate your own images from text-captions with this notebook (althought using a different methology than the one described in the paper).\n",
        "\n",
        "**Note:** Most of the code implementation here was taken or adapted from other sources listed at the bottom of this notebook such as OpenAI's [DALL-E pytorch package](https://github.com/openai/DALL-E)  and [Ryan Murdoch's (Advadnoun)](https://twitter.com/advadnoun) LatentVisions notebooks.\n",
        "\n",
        "[Click here](#scrollTo=iKP0tnHaiTyl) to go to the references cell."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FQULmg_lpMNq"
      },
      "source": [
        "\n",
        "### Skip directly to the fun part?\n",
        "\n",
        "Go [to this part](#scrollTo=_UYDSkamzJtx) of the notebook to see the implementation of a text-to-image generation scheme that uses CLIP <a href=\"#scrollTo=iKP0tnHaiTyl\">(2)</sub></a> and a pre-trained generator network <a href=\"#scrollTo=iKP0tnHaiTyl\">(3)</sub></a> to generate images from your own text inputs."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vjmNf_oNRVJv"
      },
      "source": [
        "---------------\n",
        "\n",
        "# What is DALL-E and Zero-Shot Text-to-Image Generation? \n",
        "\n",
        "On January 5th of 2021, [OpenAI](https://openai.com/about/) released a [blog post](https://openai.com/blog/dall-e/) introducing their new deep learning model DALL-E<sup>[1]</sup>, a transformer language model trained to generate images from text captions with precise coherence. A few months after, they published the paper ***Zero-Shot Text-to-Image Generation*** describing their approach with creating this model along with code to replicate the discrete Variational Auto Encoder (dVAE) used in their research.\n",
        "\n",
        "\n",
        "Zero-Shot Text-to-Image generation refers to the concept of generating an image from a text input in a way that makes the image **consistent** with the text. If the prompt \"*A giraffe wearing a red scarf*\" is given then one would expect the output to be an image that assimilates a giraffe with a red piece of cloth aroung its neck. The Zero-Shot part comes from the fact that the model wasn't explicitly trained with a fixed set of text-prompts meaning that it can, in principle, generalize to any text input (with mixed degrees of performance).\n",
        "\n",
        "\n",
        "## How does DALLE work?\n",
        "\n",
        "DALL-E is a language model that is at its core an autoregressive network with 12 *billion* parameter trained on 250 million image-text pairs. In the paper they explained the methodology used to make this model by dividing it into two parts to describe the two stages of learning they had to model: \n",
        "\n",
        "- The first part was about **learning the vocabulary of the image-text pairs**. What they did is to train a discrete Variational Auto-Encoder (VAE) to compress the 256x256x3 training images into 32x32 grids  of discrete *image tokens* of vocabulary size 8192. That is, they learnt to map and reconstruct an image to and from a embedding (or latent) space of 32*32=1024 integers (image tokens). \n",
        "\n",
        "| ![VQ-VAE example](https://i.imgur.com/R9VMWD6.png) | ![DALL-E dvae reconstruction example](https://www.dropbox.com/s/nb9nvznxverq16h/dalle-vae-img-reconstruction.jpg?raw=1) |\n",
        "|:--:|  :--: |\n",
        "| Example of a VQ-VAE taken from Van den Oord et al. 2017 <sup>[2]</sup> | Reconstruction of an original image by DALL-E's dVAE |\n",
        "\n",
        "- The second part was about **learning the prior** distribution over the text and image tokens. What they did here is concatenate 256 tokens obtained from encoding the input text prompts with the encoded 1024 tokens from their correponding image and training a transformer to model this autoregressively as a single stream of data of $1024+256 = 1080$ tokens.  The result is that from an initial set of at least 256 tokens, the model will \"autocomplete\" the remaining ones such that an image is generated that is consistent to the initial tokens <sup>[3]</sup>.\n",
        "\n",
        "In summary, with the dVAE from the first stage and the autoregressive transformer from the second one, a single step of DALL-E would have to (1) use the transformer to predict the following 1024 image tokens from the first 256 tokens obtained from the input text-prompt and (2) take the full stream of 1024 image tokens that are generated by the transformer and generate an image using the dVAE to map from the embedding space onto the image space. \n",
        "\n",
        "-----\n",
        "[1] The name DALL-E comes from a wordplay combining <strong>WALL-E</strong>, the Disney's Pixar character, and <strong>Dalí</strong> from <i>Salvador Dalí</i>, the famous spanish painter. \n",
        "\n",
        "[2] Oord, Aaron van den, Oriol Vinyals, and Koray Kavukcuoglu. \"Neural discrete representation learning.\" (2017) [[Link]](https://arxiv.org/pdf/1711.00937.pdf)\n",
        "\n",
        "[3] This is similar to what GTP-3 (another language model by OpenAI) does to generate text from an initial text-input. Although GTP-3 is more than 10 times larger than DALL-E with 175 billion parameters ([Source](https://arxiv.org/abs/2005.14165))."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pxYkH0DRuyim"
      },
      "source": [
        "## Results\n",
        "\n",
        "The results published in their blog and paper show an extremely good capability of generating completely new images that are coherent to the input text prompt. The model is also capable of reconstructing images that have their bottom part missing or understanding the relationship between a given top image and generating a new image from it at the bottom.  \n",
        "\n",
        "\n",
        "\n",
        "| ![armchair in the shape of an avocado](https://www.dropbox.com/s/fff2odffn5ujvk2/armchair-in-the-shape-avocado.jpg?raw=1) | ![text-to-image examples](https://www.dropbox.com/s/vs22itzf47ygdhx/text-to-img-dalle-examples.jpg?raw=1) |\n",
        "|:--:|  :--: |\n",
        "| ![exact same cat at the top as a sketch in the bottom](https://www.dropbox.com/s/5co90nh6qnlhtjo/cat-top-bottom.jpg?raw=1) | ![Bust of Homer](https://www.dropbox.com/s/xu7etvz3anx6mcu/bust-of-homer.jpg?raw=1) |\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kbji1EoJJwbn"
      },
      "source": [
        "# Implementation of a Text-to-Image Generation Scheme\n",
        "\n",
        "Even though a lot of people would love to play with DALL-E and/or see more of it in action, OpenAI hasn't (sadly) fully released it to the public yet and they haven't expressed any plans to do so in the nearby future. They've only released the dVAE described in the first stage of their paper. But, even thought it can be used to map and reconstruct images to and from the latent space perfectly, is missing the important part that is actually able to represent text as images (the transformer). \n",
        "\n",
        "It is not needed to mention also that for most people and companies it is prohibitely expensive to attempt to train a model as large as DALL-E for themselves (would cost more than a hundred thousands of dollars to train such model!). \n",
        "\n",
        "Because of that and until they release the full model (if ever), we are bound to look or come up with other schemes that are able to do text-to-image generation in a different way. Ryan Murdoch <a href=\"#scrollTo=iKP0tnHaiTyl\">(4)</sub></a> is one that has come up with a simple scheme to accomplish this. He implemented a method that combines [CLIP](https://arxiv.org/abs/2103.00020) and a generative model (such as the dVAE DALL-E uses), to iteratively generate images that assimilate a text input. \n",
        "\n",
        "## Text-to-Image generation with CLIP\n",
        "\n",
        "**What is CLIP?:** CLIP was introduced by OpenAI in [another blog post](https://openai.com/blog/clip/) the same day that they introduced DALL-E. CLIP is a neural network that is extremely good at telling whether an image and a text label fit together, that is, given an image and a any set of text labels, CLIP will output how likely each label is to be representative of the image. So if you show CLIP an image of a cat and the labels `[\"a dog\",\"a giraffe\",\"a house\", \"a cat\"]` it will assign more probability to the labels related to the cat picture (`a cat` in this case). \n",
        "\n",
        "| <img src=\"https://www.dropbox.com/s/4ucl5y878kaddor/clip_example.png?raw=1\" width = \"500px\"></img> | <img src=\"https://www.dropbox.com/s/vemxyq9fks119yl/is_fully_differentiable.png?raw=1\" width = \"500px\"></img>|\n",
        "|:--:|  :--: |\n",
        "|CLIP is really good at telling you whether an image fits a text label* | It is fully differentiable\\*|\n",
        "\n",
        "\n",
        "The beauty about CLIP is that the network is fully differentiable and therefore if we have a generator that feeds every image that it creates to CLIP and define our loss function as obtaining a high value from it, the \"error\" between the given label(s) and image can be backpropagated through the generator model to incrementally get closer and closer to an image that CLIP recognizes as one that assimilates the text label. So if we start with any image obtained from the generator (it can be random, or just noise) we just need to traverse through the embedding space in the direction that minimizes CLIP's error until we get to an image that is good enough at emulating the text (by CLIP standards).\n",
        "\n",
        "| <img src=\"https://www.dropbox.com/s/99w6ckad7ud9qbg/full_clip_example.png?raw=1\" width = \"8500px\">|\n",
        "|:--: |\n",
        "|Backpropagating through CLIP and the generator network*|\n",
        "\n",
        "----\n",
        "\n",
        "* ***Image Sources:** Youtube: Yannic Kilcher's - [What Happens when OpenAI's CLIP meets BigGAN](https://www.youtube.com/watch?v=rR5_emVeyBk&t=364s)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_UYDSkamzJtx"
      },
      "source": [
        "**Below is the implementation of the methodology described above to do Zero-Shot Text-to-Image generation**. Most of the code in the remaining of this notebook was adapted from other notebooks published by Ryan Murdoch <a href=\"#scrollTo=iKP0tnHaiTyl\">(4)</sub></a>. I've only expanded on the ways that the outputs are visualized and integrated the implementations of two different Generators into one notebook, such that is possible to choose between the dVAE that is used by DALL-E <a href=\"#scrollTo=iKP0tnHaiTyl\">(5)</sub></a> and a VQGAN created by CompVis that uses Taming Transformers <a href=\"#scrollTo=iKP0tnHaiTyl\">(3)</sub></a>.\n",
        "\n",
        "*I recommend to use either when generating images from scratch but the VQGAN tends to work better when an input image is introduced.*\n",
        "\n",
        "**The following are some examples of media I've been able to generate with this:**\n",
        "\n",
        "| Input: \"A city landscape in the style of Van Gogh\" (DALL-E dVAE) | Input: Selfie of me + \"A cat\" (progression video, VQGAN) |\n",
        "|:--:|  :--: |\n",
        "|![cityscape in the style of Van Gogh](https://www.dropbox.com/s/53j6v0cm4gj9tx2/cityscape-van-gogh-dalledvae.jpeg?raw=1) |  ![me+cat = gif](https://media.giphy.com/media/FK08sYn8tLA6O1jOOd/giphy.gif) |\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r8gNRZUUMrWM"
      },
      "source": [
        "**NOTE: This part of the notebook is meant to run with Google Colab on a GPU runtime**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EUOAL4UcytFb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "outputId": "d5559958-0563-4e27-e27f-a66d93ebfba4"
      },
      "source": [
        "#@markdown ## 0. GPU Information\n",
        "#@markdown ###Were you lucky today?\n",
        "#@markdown | V100 | P100 | T4 | K80 |\n",
        "#@markdown |:----:|:---:|:--:|:----:|\n",
        "#@markdown | 🤩 | 😊 | 😬 | 💩\n",
        "\n",
        "from subprocess import PIPE, run\n",
        "result = run(['nvidia-smi','-L'],stdout=PIPE,stderr=PIPE,universal_newlines=True)\n",
        "meanings = [('V100','🤩'),('P100','😊'),('T4','😬'),('K80','💩')]\n",
        "out = result.stdout.split(\"(UUID\")[0] + ''.join([emoji for graphic,emoji in meanings if graphic in result.stdout][0])\n",
        "print(out.replace('\\n',' '))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "GPU 0: Tesla T4 😬\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P301l12fEY6A"
      },
      "source": [
        "## 1. Load imports and define functions and variables\n",
        "\n",
        "**Execute all the following continuously until you get to the section to define the input parameters. You only need to run the cells inside this section once per session.**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KAcixx9Z3XYH"
      },
      "source": [
        "### 1.1 Top (imports)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "piJOg9MY7khd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5d8fe3c2-c015-456d-9585-6991e1162aa1"
      },
      "source": [
        "# don't use half of these lol\n",
        "\n",
        "import torch\n",
        "import numpy as np\n",
        "import torchvision\n",
        "import torchvision.transforms.functional as TF\n",
        "!pip install kornia\n",
        "import kornia\n",
        "\n",
        "import PIL\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import os\n",
        "import random\n",
        "import imageio\n",
        "from IPython import display\n",
        "from IPython.core.interactiveshell import InteractiveShell\n",
        "InteractiveShell.ast_node_interactivity = \"all\"\n",
        "\n",
        "import glob\n",
        "\n",
        "from google.colab import output\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: kornia in /usr/local/lib/python3.7/dist-packages (0.5.6)\n",
            "Requirement already satisfied: torch>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from kornia) (1.9.0+cu102)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from torch>=1.6.0->kornia) (3.7.4.3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JeMCHwDdHIcu"
      },
      "source": [
        "### 1.2 Perceptor (CLIP)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tm3_VmxpAiB1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7c7baa60-3787-4cf2-f8ab-475aa209dc3a"
      },
      "source": [
        "!pip install --no-deps ftfy regex tqdm\n",
        "!git clone https://github.com/openai/CLIP.git\n",
        "\n",
        "#Import CLIP and load the model\n",
        "from CLIP import clip\n",
        "perceptor, preprocess = clip.load('ViT-B/32', jit=False)\n",
        "perceptor.eval()\n",
        "\n",
        "# clip.available_models()\n",
        "\n",
        "perceptor.visual.input_resolution\n",
        "\n",
        "scaler = 1"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: ftfy in /usr/local/lib/python3.7/dist-packages (6.0.3)\n",
            "Requirement already satisfied: regex in /usr/local/lib/python3.7/dist-packages (2019.12.20)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (4.41.1)\n",
            "fatal: destination path 'CLIP' already exists and is not an empty directory.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "CLIP(\n",
              "  (visual): VisualTransformer(\n",
              "    (conv1): Conv2d(3, 768, kernel_size=(32, 32), stride=(32, 32), bias=False)\n",
              "    (ln_pre): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "    (transformer): Transformer(\n",
              "      (resblocks): Sequential(\n",
              "        (0): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (1): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (2): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (3): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (4): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (5): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (6): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (7): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (8): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (9): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (10): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (11): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (ln_post): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "  )\n",
              "  (transformer): Transformer(\n",
              "    (resblocks): Sequential(\n",
              "      (0): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (1): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (2): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (3): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (4): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (5): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (6): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (7): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (8): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (9): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (10): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (11): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=512, out_features=2048, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=2048, out_features=512, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "    )\n",
              "  )\n",
              "  (token_embedding): Embedding(49408, 512)\n",
              "  (ln_final): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "224"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UlIUu0jK3S19"
      },
      "source": [
        "### 1.3 Define Custom Visualization functions"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gCj1wr8pOqOe"
      },
      "source": [
        "import glob\n",
        "from moviepy.editor import VideoFileClip\n",
        "import cv2\n",
        "class ImageSaver():\n",
        "  def __init__(self,imgdir=None):\n",
        "    self.generated_images = []\n",
        "    self.imgdir = imgdir\n",
        "\n",
        "  def save_images_in_dir(self,dirname):\n",
        "    for i,img in enumerate(self.generated_images):\n",
        "      imageio.imwrite(os.path.join(dirname,f\"{i}.png\"),img)\n",
        "\n",
        "  def make_video(self,video_name,fps=8, in_self_dir = True):\n",
        "    if in_self_dir:\n",
        "      video_name = os.path.join(self.imgdir,video_name)\n",
        "    height, width, layers = self.generated_images[0].shape\n",
        "    size = (width,height)\n",
        "    out = cv2.VideoWriter(video_name,cv2.VideoWriter_fourcc(*'mp4v'), fps, size)\n",
        "    for i in range(len(self.generated_images)):\n",
        "      img_np = (self.generated_images[i]*255).astype(np.uint8)\n",
        "      frame = cv2.cvtColor(img_np, cv2.COLOR_RGB2BGR)\n",
        "      out.write(frame)\n",
        "    out.release()\n",
        "  \n",
        "  def show_video(self,video_path_or_name):\n",
        "    if os.path.isdir(os.path.basename(video_path_or_name)):\n",
        "      fpath = video_path_or_name\n",
        "    elif os.path.isfile(os.path.join(self.imgdir,video_path_or_name)):\n",
        "      fpath  = os.path.join(self.imgdir,video_path_or_name)\n",
        "    else:\n",
        "      raise Exception(\"video not found in the given directory\")\n",
        "    \n",
        "    my_vid=VideoFileClip(fpath)\n",
        "    return my_vid\n",
        "    \n",
        "def displ(img,imSaver, pre_scaled=True):\n",
        "  img = np.array(img)[:,:,:]\n",
        "  img = np.transpose(img, (1, 2, 0))\n",
        "  if not pre_scaled:\n",
        "    img = scale(img, 48*4, 32*4)\n",
        "  imageio.imwrite(str(3) + '.png', (np.array(img)*255).astype(np.uint8) )\n",
        "  imSaver.generated_images += [np.array(img)]\n",
        "  # return display.Image(str(3)+'.png')\n",
        "\n",
        "def gallery(array, ncols=2):\n",
        "    nindex, height, width, intensity = array.shape\n",
        "    nrows = nindex//ncols\n",
        "    assert nindex == nrows*ncols\n",
        "    # want result.shape = (height*nrows, width*ncols, intensity)\n",
        "    result = (array.reshape(nrows, ncols, height, width, intensity)\n",
        "              .swapaxes(1,2)\n",
        "              .reshape(height*nrows, width*ncols, intensity))\n",
        "    return result\n",
        "\n",
        "def card_padded(im, to_pad=3):\n",
        "  return np.pad(np.pad(np.pad(im, [[1,1], [1,1], [0,0]],constant_values=0), [[2,2], [2,2], [0,0]],constant_values=1),\n",
        "            [[to_pad,to_pad], [to_pad,to_pad], [0,0]],constant_values=0)\n",
        "\n",
        "def get_all(img):\n",
        "  img = np.transpose(img, (0,2,3,1))\n",
        "  cards = np.zeros((img.shape[0], sideX+12, sideY+12, 3))\n",
        "  for i in range(len(img)):\n",
        "    cards[i] = card_padded(img[i])\n",
        "  print(img.shape)\n",
        "  cards = gallery(cards)\n",
        "  imageio.imwrite(str(3) + '.png', np.array(cards))\n",
        "  return display.Image(str(3)+'.png')\n",
        "  "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MXIVvrehzlS0"
      },
      "source": [
        "# from google.colab import drive\n",
        "# drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GHB7BcR1zNLZ"
      },
      "source": [
        "### 1.4 Generator Utils"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WtlDVVMvzMUd"
      },
      "source": [
        "%%capture\n",
        "\n",
        "!pip uninstall torchtext --yes\n",
        "\n",
        "# %cd /content/\n",
        "# !git clone https://github.com/CompVis/taming-transformers  \n",
        "# %cd /content/taming-transformers\n",
        "\n",
        "\n",
        "\n",
        "# download a VQGAN with a larger codebook (16384 entries)\n",
        "# !mkdir -p logs/vqgan_imagenet_f16_16384/checkpoints\n",
        "# !mkdir -p logs/vqgan_imagenet_f16_16384/configs\n",
        "\n",
        "# if len(os.listdir('logs/vqgan_imagenet_f16_16384/checkpoints/')) == 0:\n",
        "#   !wget 'https://heibox.uni-heidelberg.de/f/867b05fc8c4841768640/?dl=1' -O 'logs/vqgan_imagenet_f16_16384/checkpoints/last.ckpt' \n",
        "#   !wget 'https://heibox.uni-heidelberg.de/f/274fb24ed38341bfa753/?dl=1' -O 'logs/vqgan_imagenet_f16_16384/configs/model.yaml' \n",
        "\n",
        "%pip install omegaconf==2.0.0 pytorch-lightning==1.0.8\n",
        "import sys\n",
        "sys.path.append(\".\")\n",
        "\n",
        "DEVICE = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "import yaml\n",
        "import torch\n",
        "from omegaconf import OmegaConf\n",
        "# from taming.models.vqgan import VQModel\n",
        "\n",
        "def load_config(config_path, display=False):\n",
        "  config = OmegaConf.load(config_path)\n",
        "  if display:\n",
        "    print(yaml.dump(OmegaConf.to_container(config)))\n",
        "  return config\n",
        "\n",
        "def load_vqgan(config, ckpt_path=None):\n",
        "  model = VQModel(**config.model.params)\n",
        "  if ckpt_path is not None:\n",
        "    sd = torch.load(ckpt_path, map_location=\"cpu\")[\"state_dict\"]\n",
        "    missing, unexpected = model.load_state_dict(sd, strict=False)\n",
        "  return model.eval()\n",
        "\n",
        "def preprocess_vqgan(x):\n",
        "  x = 2.*x - 1.\n",
        "  return x\n",
        "\n",
        "def custom_to_pil(x,is_dalle=False):\n",
        "  if is_dalle:\n",
        "    return T.ToPILImage(mode='RGB')(x)\n",
        "  x = x.detach().cpu()\n",
        "  x = torch.clamp(x, -1., 1.)\n",
        "  x = (x + 1.)/2.\n",
        "  x = x.permute(1,2,0).numpy()\n",
        "  x = (255*x).astype(np.uint8)\n",
        "  x = PIL.Image.fromarray(x)\n",
        "  if not x.mode == \"RGB\":\n",
        "    x = x.convert(\"RGB\")\n",
        "  return x\n",
        "\n",
        "def reconstruct_with_vqgan(x, model):\n",
        "  # could also use model(x) for reconstruction but use explicit encoding and decoding here\n",
        "  z, _, [_, _, indices] = model.encode(preprocess_vqgan(x))\n",
        "  print(f\"VQGAN: latent shape: {z.shape[2:]}\")\n",
        "  xrec = model.decode(z)\n",
        "  return xrec\n",
        "\n",
        "import requests, io\n",
        "import torchvision.transforms as T\n",
        "import torchvision.transforms.functional as TF\n",
        "import torch.nn.functional as F\n",
        "\n",
        "def download_image(url):\n",
        "    resp = requests.get(url)\n",
        "    resp.raise_for_status()\n",
        "    return PIL.Image.open(io.BytesIO(resp.content))\n",
        "\n",
        "def preprocess_img(img, target_image_size=256):\n",
        "    s = min(img.size)\n",
        "    \n",
        "    if s < target_image_size:\n",
        "        raise ValueError(f'min dim for image {s} < {target_image_size}')\n",
        "        \n",
        "    r = target_image_size / s\n",
        "    s = (round(r * img.size[1]), round(r * img.size[0]))\n",
        "    img = TF.resize(img, s, interpolation=PIL.Image.LANCZOS)\n",
        "    img = TF.center_crop(img, output_size=2 * [target_image_size])\n",
        "    img = torch.unsqueeze(T.ToTensor()(img), 0)\n",
        "    return map_pixels(img)\n",
        "\n",
        "def stack_reconstructions(x1, x2, titles=[]):\n",
        "  assert x1.size == x2.size\n",
        "  w, h = x1.size[0], x1.size[1]\n",
        "  img = PIL.Image.new(\"RGB\", (2*w, h))\n",
        "  img.paste(x1, (0*w,0))\n",
        "  img.paste(x2, (1*w,0))\n",
        "  font = PIL.ImageFont.truetype(\"/usr/share/fonts/truetype/liberation/LiberationSans-BoldItalic.ttf\", 22)\n",
        "  for i, title in enumerate(titles):\n",
        "    PIL.ImageDraw.Draw(img).text((i*w, 0), f'{title}', (255, 255, 255), font=font) # coordinates, text, color, font\n",
        "  return img\n",
        "\n",
        "def reconstruct_with_dalle(x, encoder, decoder, do_preprocess=False):\n",
        "  # takes in tensor (or optionally, a PIL image) and returns a PIL image\n",
        "  if do_preprocess:\n",
        "    x = preprocess(x)\n",
        "  z_logits = encoder(x)\n",
        "  z = torch.argmax(z_logits, axis=1)\n",
        "  \n",
        "  print(f\"DALL-E: latent shape: {z.shape}\")\n",
        "  z = F.one_hot(z, num_classes=encoder.vocab_size).permute(0, 3, 1, 2).float()\n",
        "\n",
        "  x_stats = decoder(z).float()\n",
        "  x_rec = unmap_pixels(torch.sigmoid(x_stats[:, :3]))\n",
        "\n",
        "  return x_rec\n",
        "\n",
        "%cd /content/"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ITiS_drbKqdu"
      },
      "source": [
        "# torch.cuda.empty_cache()\n",
        "# url = 'https://media-exp1.licdn.com/dms/image/C4D03AQF4a7byPVzwug/profile-displayphoto-shrink_800_800/0/1596310205648?e=1627516800&v=beta&t=2mryZb343ucG3TOadMFWgfzdbqpQF_5mGZah6D1dWeM'\n",
        "# x = preprocess_img(download_image(url))\n",
        "# x = x.to('cuda')\n",
        "# x1 = reconstruct_with_vqgan(x, model16384)\n",
        "# x2 = reconstruct_with_dalle(x, encoder_dalle, model)\n",
        "# stack_reconstructions(custom_to_pil(x1[0]),custom_to_pil(x2[0],True),[\"VQGAN\",\"DALL-E\"])\n",
        "\n",
        "# del x1\n",
        "# del x2\n",
        "# import gc\n",
        "# gc.collect()\n",
        "# torch.cuda.empty_cache()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mw0KBLebMywW"
      },
      "source": [
        "## 2. Define Input Parameters\n",
        "\n",
        "Input parameters to fill in the input text-prompt and/or image paths that will be used to guide the text-to-image generation. **Run this as many times as you want per session with the inputs you prefer.**\n",
        "\n",
        "Images default to being uploaded to `/content/<image_name.png>`. \n",
        "\n",
        "<br>\n",
        "\n",
        "**Detailed Instructions:** You'll need to first, type an image description such as \"*A wet capybara in the rain*\" into `text_input` and/or upload an image and enter its path in `input_img_path`. Adding the path to an image allows you to optimize based on that image and a text prompt. \n",
        "\n",
        "You may also tinker with the text and input-image weights and the learning rate to control the quality of the output image.\n",
        "\n",
        "Your output will start appearing at the bottom of this page near the [Generate Output](#scrollTo=Y2-ro0FFhUh9&line=1&uniqifier=1) heading as it processes after a short while. Scroll down below it to see newly generated images appear.\n",
        "\n",
        "I've also made it such that a **progress video** is created from all the images generated during training. It will appear in the same area where the training loop starts and updates every 500 training iterations.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YIFGIINZ6pFi",
        "cellView": "form"
      },
      "source": [
        "# @markdown <h3>Optional: Take a selfie with your camera!</h3>\n",
        "\n",
        "#@markdown Uncheck `disable` and run this cell to take a picture with your camera and save it to colab as `file_name`. Or leave it as disable if you prefer to upload (or not) your own pictures.\n",
        "\n",
        "\n",
        "file_name = \"myselfie.jpg\" #@param {type:\"string\"}\n",
        "\n",
        "disable = True #@param {type:\"boolean\"}\n",
        "\n",
        "import os\n",
        "from IPython.display import Javascript\n",
        "from google.colab.output import eval_js\n",
        "from base64 import b64decode\n",
        "\n",
        "def take_photo(filename='photo.jpg', quality=0.8):\n",
        "  js = Javascript('''\n",
        "    async function takePhoto(quality) {\n",
        "      const div = document.createElement('div');\n",
        "      const capture = document.createElement('button');\n",
        "      capture.textContent = 'Capture';\n",
        "      div.appendChild(capture);\n",
        "\n",
        "      const video = document.createElement('video');\n",
        "      video.style.display = 'block';\n",
        "      const stream = await navigator.mediaDevices.getUserMedia({video: true});\n",
        "\n",
        "      document.body.appendChild(div);\n",
        "      div.appendChild(video);\n",
        "      video.srcObject = stream;\n",
        "      await video.play();\n",
        "\n",
        "      // Resize the output to fit the video element.\n",
        "      google.colab.output.setIframeHeight(document.documentElement.scrollHeight, true);\n",
        "\n",
        "      // Wait for Capture to be clicked.\n",
        "      await new Promise((resolve) => capture.onclick = resolve);\n",
        "\n",
        "      const canvas = document.createElement('canvas');\n",
        "      canvas.width = video.videoWidth;\n",
        "      canvas.height = video.videoHeight;\n",
        "      canvas.getContext('2d').drawImage(video, 0, 0);\n",
        "      stream.getVideoTracks()[0].stop();\n",
        "      div.remove();\n",
        "      return canvas.toDataURL('image/jpeg', quality);\n",
        "    }\n",
        "    ''')\n",
        "  display.display(js)\n",
        "  data = eval_js('takePhoto({})'.format(quality))\n",
        "  binary = b64decode(data.split(',')[1])\n",
        "  with open(filename, 'wb') as f:\n",
        "    f.write(binary)\n",
        "  return filename\n",
        "\n",
        "from IPython.display import Image\n",
        "\n",
        "if not disable:\n",
        "  try:\n",
        "    filename = take_photo(filename=file_name)\n",
        "    print('Saved to {}'.format(filename))\n",
        "    \n",
        "    # Show the image which was just taken.\n",
        "    display.display(Image(filename))\n",
        "  except Exception as err:\n",
        "    # Errors will be thrown if the user does not have a webcam or if they do not\n",
        "    # grant the page permission to access it.\n",
        "    print(str(err))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nq0wA-wc-P-s"
      },
      "source": [
        "#@title Input Parameters { display-mode: \"form\" }\n",
        "#@markdown #### You should execute this cells and the ones below everytime you adjust a new parameter.\n",
        "\n",
        "text_input = \"\" #@param {type:\"string\"}\n",
        "weight_text_input = 2.7 #@param {type:\"slider\", min:-5, max:5, step:0.1}\n",
        "w0 = weight_text_input\n",
        "\n",
        "#@markdown ---------\n",
        "text_to_add = \"\" #@param {type:\"string\"}\n",
        "w1 = 0.3 #@param {type:\"slider\", min:-5, max:5, step:0.1}\n",
        "text_to_remove = \"incoherent, confusing, cropped, watermarks, moustache, glasses\" #@param {type:\"string\"}\n",
        "#@markdown --------\n",
        "# @markdown You can input a picture by adding the path as the parameter `input_img_path`, and combine your picture with a text input to generate a distortionate image based on the text input (such as the generated image of me as a cat). **If you do this you should also set `weight_text_input` to a low value like 0.9 and `weight_input_img` to a larger one such as 3.5 or so to get better results.**\n",
        "\n",
        "input_img_path = \"\" #@param {type:\"string\"}\n",
        "img_enc_path = input_img_path\n",
        "weight_input_img = 0 #@param {type:\"slider\", min:-5, max:5, step:0.1}\n",
        "w2 = weight_input_img\n",
        "\n",
        "#@markdown --------\n",
        "# @markdown Initial Image to start optimizing with. Only works with DALL-E dVAE\n",
        "\n",
        "init_img_path = \"\" #@param {type:\"string\"}\n",
        "disable_init_img =   True #@param {type:\"boolean\"}\n",
        "ne_img_enc_path = \"\"\n",
        "w3 = 0 #param {type:\"slider\", min:-5, max:5, step:0.1}\n",
        "\n",
        "#@markdown --------\n",
        "generator = \"DALL-E dVAE\" #@param [\"DALL-E dVAE\", \"CompVis' VQGAN\"]\n",
        "\n",
        "learning_rate = 0.015 #@param {type:\"slider\", min:0, max:5, step:0.005}\n",
        "batch_size = 1 #@param {type:\"slider\", min:1, max:5,step:1}\n",
        "max_epochs = 10000 #@param {type:\"number\"}\n",
        "\n",
        "#@markdown --------\n",
        "# How to weight the 2 texts (w0 and w1) and the images (w3 & w3)\n",
        "im_shape = [512, 512, 3] #@param {type:\"raw\"}\n",
        "sideX, sideY, channels = im_shape\n",
        "\n",
        "\n",
        "progression_video_save_path = \"progression.mp4\" #@param {type:\"string\"}\n",
        "progression_video_fps = 8 #@param {type:\"integer\"}\n",
        "\n",
        "restart_images = True #param {type:\"boolean\"}\n",
        "\n",
        "import ipywidgets as widgets"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4jl-jCwmO53r"
      },
      "source": [
        "**After setting new input parameters you have to run all the cells below  in order to initiate the training loop and generate your images.**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1yryNdZUUXnl"
      },
      "source": [
        "## 3. Download/Reload Selected Generator"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ncV3NPFMP234",
        "outputId": "6ade7624-8f72-4093-bf6f-6967af744830"
      },
      "source": [
        "#download generator model\n",
        "from termcolor import colored\n",
        "is_dall_e = False\n",
        "if generator==\"DALL-E dVAE\":\n",
        "  print(colored(\"Installing OpenAI's DALL-E dVAE...\",\"blue\"))\n",
        "  !pip install git+https://github.com/openai/DALL-E.git\n",
        "  from dall_e import map_pixels, unmap_pixels, load_model\n",
        "    # For faster load times, download these files locally and use the local paths instead.\n",
        "  encoder_dalle = load_model(\"https://cdn.openai.com/dall-e/encoder.pkl\", DEVICE)\n",
        "  model = load_model(\"https://cdn.openai.com/dall-e/decoder.pkl\", DEVICE)\n",
        "  # model = decoder_dalle #load_model(\"https://cdn.openai.com/dall-e/decoder.pkl\", 'cuda')\n",
        "  is_dall_e = True\n",
        "else:\n",
        "  print(colored(\"Installing Compvis' VQGAN...\",\"blue\"))\n",
        "  %cd /content/\n",
        "  !pip install einops\n",
        "  !git clone https://github.com/CompVis/taming-transformers  \n",
        "  %cd /content/taming-transformers\n",
        "\n",
        "\n",
        "  #download a VQGAN with a larger codebook (16384 entries)\n",
        "  if not os.path.isdir('taming-transformers'):\n",
        "    !mkdir -p logs/vqgan_imagenet_f16_16384/checkpoints\n",
        "    !mkdir -p logs/vqgan_imagenet_f16_16384/configs\n",
        "\n",
        "  from taming.models.vqgan import VQModel\n",
        "\n",
        "  if len(os.listdir('logs/vqgan_imagenet_f16_16384/checkpoints/')) == 0:\n",
        "    !wget 'https://heibox.uni-heidelberg.de/f/867b05fc8c4841768640/?dl=1' -O 'logs/vqgan_imagenet_f16_16384/checkpoints/last.ckpt' \n",
        "    !wget 'https://heibox.uni-heidelberg.de/f/274fb24ed38341bfa753/?dl=1' -O 'logs/vqgan_imagenet_f16_16384/configs/model.yaml' \n",
        "\n",
        "  config16384 = load_config(\"logs/vqgan_imagenet_f16_16384/configs/model.yaml\", display=False)\n",
        "  model16384 = load_vqgan(config16384, ckpt_path=\"logs/vqgan_imagenet_f16_16384/checkpoints/last.ckpt\").to(DEVICE)\n",
        "  def model(x):\n",
        "    o_i2 = x\n",
        "    o_i3 = model16384.post_quant_conv(o_i2)\n",
        "    i = model16384.decoder(o_i3)\n",
        "    return i\n",
        "  %cd /content/\n",
        "\n",
        "# encoder = load_model(\"https://cdn.openai.com/dall-e/encoder.pkl\", 'cuda')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[34mInstalling OpenAI's DALL-E dVAE...\u001b[0m\n",
            "Collecting git+https://github.com/openai/DALL-E.git\n",
            "  Cloning https://github.com/openai/DALL-E.git to /tmp/pip-req-build-8yv9qsfh\n",
            "  Running command git clone -q https://github.com/openai/DALL-E.git /tmp/pip-req-build-8yv9qsfh\n",
            "Requirement already satisfied: Pillow in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (7.1.2)\n",
            "Collecting blobfile\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/01/54/9e01c570475b7ea16a4e489bba85b7736d6eac5f6de1fb9081564eb1dfac/blobfile-1.2.3-py3-none-any.whl (61kB)\n",
            "\u001b[K     |████████████████████████████████| 71kB 10.0MB/s \n",
            "\u001b[?25hCollecting mypy\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/37/38/1f0771d818f8a1fbb34e15cf694cc619a6d537b88d36162021fd7bc4d964/mypy-0.910-cp37-cp37m-manylinux2010_x86_64.whl (21.5MB)\n",
            "\u001b[K     |████████████████████████████████| 21.5MB 1.6MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (1.19.5)\n",
            "Requirement already satisfied: pytest in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (3.6.4)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (2.23.0)\n",
            "Requirement already satisfied: torch in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (1.9.0+cu102)\n",
            "Requirement already satisfied: torchvision in /usr/local/lib/python3.7/dist-packages (from DALL-E==0.1) (0.10.0+cu102)\n",
            "Requirement already satisfied: filelock~=3.0 in /usr/local/lib/python3.7/dist-packages (from blobfile->DALL-E==0.1) (3.0.12)\n",
            "Collecting xmltodict~=0.12.0\n",
            "  Downloading https://files.pythonhosted.org/packages/28/fd/30d5c1d3ac29ce229f6bdc40bbc20b28f716e8b363140c26eff19122d8a5/xmltodict-0.12.0-py2.py3-none-any.whl\n",
            "Collecting pycryptodomex~=3.8\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/e5/9d/99a949925b5fc9604cb65219951fd270ef30d0fd4f064d1b363eb8bb5e9b/pycryptodomex-3.10.1-cp35-abi3-manylinux2010_x86_64.whl (1.9MB)\n",
            "\u001b[K     |████████████████████████████████| 1.9MB 46.8MB/s \n",
            "\u001b[?25hCollecting urllib3~=1.25\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/5f/64/43575537846896abac0b15c3e5ac678d787a4021e906703f1766bfb8ea11/urllib3-1.26.6-py2.py3-none-any.whl (138kB)\n",
            "\u001b[K     |████████████████████████████████| 143kB 47.4MB/s \n",
            "\u001b[?25hCollecting typed-ast<1.5.0,>=1.4.0; python_version < \"3.8\"\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/65/b3/573d2f1fecbbe8f82a8d08172e938c247f99abe1be3bef3da2efaa3810bf/typed_ast-1.4.3-cp37-cp37m-manylinux1_x86_64.whl (743kB)\n",
            "\u001b[K     |████████████████████████████████| 747kB 43.5MB/s \n",
            "\u001b[?25hRequirement already satisfied: toml in /usr/local/lib/python3.7/dist-packages (from mypy->DALL-E==0.1) (0.10.2)\n",
            "Collecting mypy-extensions<0.5.0,>=0.4.3\n",
            "  Downloading https://files.pythonhosted.org/packages/5c/eb/975c7c080f3223a5cdaff09612f3a5221e4ba534f7039db34c35d95fa6a5/mypy_extensions-0.4.3-py2.py3-none-any.whl\n",
            "Requirement already satisfied: typing-extensions>=3.7.4 in /usr/local/lib/python3.7/dist-packages (from mypy->DALL-E==0.1) (3.7.4.3)\n",
            "Requirement already satisfied: attrs>=17.4.0 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (21.2.0)\n",
            "Requirement already satisfied: py>=1.5.0 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (1.10.0)\n",
            "Requirement already satisfied: more-itertools>=4.0.0 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (8.8.0)\n",
            "Requirement already satisfied: atomicwrites>=1.0 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (1.4.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (57.2.0)\n",
            "Requirement already satisfied: pluggy<0.8,>=0.5 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (0.7.1)\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.7/dist-packages (from pytest->DALL-E==0.1) (1.15.0)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests->DALL-E==0.1) (2.10)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests->DALL-E==0.1) (2021.5.30)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests->DALL-E==0.1) (3.0.4)\n",
            "Building wheels for collected packages: DALL-E\n",
            "  Building wheel for DALL-E (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for DALL-E: filename=DALL_E-0.1-cp37-none-any.whl size=6011 sha256=d62e0165d6dffb548311f76fc00220ed10c8bc0fe5fa90b71d073d48297a1726\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-_gf4obph/wheels/e9/f5/e7/efa7ddb4c5899f6e6ffbbd112b8c7a030872274a5cba9ccf04\n",
            "Successfully built DALL-E\n",
            "\u001b[31mERROR: requests 2.23.0 has requirement urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1, but you'll have urllib3 1.26.6 which is incompatible.\u001b[0m\n",
            "\u001b[31mERROR: datascience 0.10.6 has requirement folium==0.2.1, but you'll have folium 0.8.3 which is incompatible.\u001b[0m\n",
            "Installing collected packages: xmltodict, pycryptodomex, urllib3, blobfile, typed-ast, mypy-extensions, mypy, DALL-E\n",
            "  Found existing installation: urllib3 1.24.3\n",
            "    Uninstalling urllib3-1.24.3:\n",
            "      Successfully uninstalled urllib3-1.24.3\n",
            "Successfully installed DALL-E-0.1 blobfile-1.2.3 mypy-0.910 mypy-extensions-0.4.3 pycryptodomex-3.10.1 typed-ast-1.4.3 urllib3-1.26.6 xmltodict-0.12.0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "urllib3"
                ]
              }
            }
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XaocGDQXz3Zx"
      },
      "source": [
        "## 4. Latent coordinate & Text"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GdCh2D8Dt8Xd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "98876bdb-2207-4aee-a99d-cf4382b132f8"
      },
      "source": [
        "if restart_images:\n",
        "  output_widget = widgets.Output()\n",
        "  video_out_widget = widgets.Output()\n",
        "  imSaver = ImageSaver(\"generated\")\n",
        "\n",
        "torch.cuda.empty_cache()\n",
        "\n",
        "#. A detailed, high-quality photo without distortions\n",
        "t = 0\n",
        "if text_input != '':\n",
        "  tx = clip.tokenize(text_input)\n",
        "  t = perceptor.encode_text(tx.cuda()).detach().clone()\n",
        "\n",
        "text_add = 0\n",
        "if text_to_add != '':\n",
        "  text_add = clip.tokenize(text_to_add)\n",
        "  text_add = perceptor.encode_text(text_add.cuda()).detach().clone()\n",
        "\n",
        "nom = torchvision.transforms.Normalize((0.48145466, 0.4578275, 0.40821073), (0.26862954, 0.26130258, 0.27577711))\n",
        "img_enc = 0\n",
        "if img_enc_path != '':\n",
        "  if 'http' in img_enc_path:\n",
        "    img_enc = np.asarray(download_image(img_enc_path))\n",
        "  else:\n",
        "    img_enc = imageio.imread(img_enc_path)\n",
        "  img_enc = (torch.nn.functional.interpolate(torch.tensor(img_enc).unsqueeze(0).permute(0, 3, 1, 2), (224, 224)) / 255).cuda()[:,:3]\n",
        "  img_enc = nom(img_enc)\n",
        "  img_enc = perceptor.encode_image(img_enc.cuda()).detach().clone()\n",
        "\n",
        "init_image = False\n",
        "if init_img_path != '' and not disable_init_img:\n",
        "  init_image = True\n",
        "  if 'http' in init_img_path:\n",
        "    init_image = download_image(init_img_path)\n",
        "  else:\n",
        "    init_image = PIL.Image.open(init_img_path)\n",
        "  init_x = preprocess_img(init_img,min([sideX,sideY]))\n",
        "  if is_dall_e:\n",
        "    z_logits = encoder_dalle(init_x.cuda())\n",
        "    z = torch.nn.functional.softmax(z_logits * 100000, dim=1)\n",
        "    z = z.detach().requires_grad_(True)\n",
        "  else:\n",
        "    z = 0\n",
        "\n",
        "ne_img_enc = 0\n",
        "# if ne_img_enc_path != '':\n",
        "#   if 'http' in ne_img_enc_path:\n",
        "#     ne_img_enc = np.asarray(download_image(ne_img_enc_path))\n",
        "#   else:\n",
        "#     ne_img_enc = imageio.imread(ne_img_enc_path)\n",
        "#   ne_img_enc = (torch.nn.functional.interpolate(torch.tensor(ne_img_enc).unsqueeze(0).permute(0, 3, 1, 2), (224, 224)) / 255).cuda()[:,:3]\n",
        "#   ne_img_enc = nom(ne_img_enc)\n",
        "#   ne_img_enc = perceptor.encode_image(ne_img_enc.cuda()).detach().clone()\n",
        "\n",
        "class Pars(torch.nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Pars, self).__init__()\n",
        "        #DALL-E\n",
        "        if is_dall_e:\n",
        "          if init_image:\n",
        "            self.normu = z\n",
        "          else:\n",
        "            hots = torch.nn.functional.one_hot((torch.arange(0, 8192).to(torch.int64)), num_classes=8192)\n",
        "            rng = torch.zeros(batch_size, 64*64, 8192).uniform_()**torch.zeros(batch_size, 64*64, 8192).uniform_(.1,1)\n",
        "            for b in range(batch_size):\n",
        "              for i in range(64**2):\n",
        "                rng[b,i] = hots[[np.random.randint(8191)]]\n",
        "            rng = rng.permute(0, 2, 1)\n",
        "            self.normu = torch.nn.Parameter(rng.cuda().view(batch_size, 8192, 64, 64))\n",
        "        else:\n",
        "          normu = .5*torch.randn(batch_size, 256, sideX//16, sideY//16).cuda()       \n",
        "          self.normu = torch.nn.Parameter(torch.sinh(1.9*torch.arcsinh(normu)))\n",
        "\n",
        "    def forward(self):\n",
        "      if is_dall_e:\n",
        "        if init_image:\n",
        "            return torch.nn.functional.softmax(self.normu * 10, dim=1)\n",
        "        return torch.nn.functional.gumbel_softmax(hadies*self.normu.reshape(batch_size, 8192//2, -1), dim=1, tau=1.873).view(batch_size, 8192, 64, 64) \n",
        "      else:\n",
        "        return self.normu.clip(-6, 6)\n",
        "\n",
        "dec = .1\n",
        "lats = Pars().cuda()\n",
        "mapper = [lats.normu]\n",
        "\n",
        "augs = torch.nn.Sequential(\n",
        "    torchvision.transforms.RandomHorizontalFlip(),\n",
        "    torchvision.transforms.RandomAffine(24, (.1, .1), fill=0)\n",
        ").cuda()\n",
        "up_noise = .11\n",
        "\n",
        "optimizer = torch.optim.AdamW([{'params': mapper, 'lr': learning_rate}], weight_decay=dec)\n",
        "eps = 0\n",
        "hadies = 1.\n",
        "if restart_images:\n",
        "  itt = 0\n",
        "  with torch.no_grad():\n",
        "    if is_dall_e:\n",
        "      al = unmap_pixels(torch.sigmoid(model(lats()).cpu().float())).numpy()\n",
        "    else:\n",
        "      al = (model(lats()).cpu().clip(-1, 1) + 1) / 2\n",
        "    for allls in al:\n",
        "      displ(allls[:3],imSaver)\n",
        "      print('\\n')  "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WztSrRF23Rqg"
      },
      "source": [
        "## 5. Define train functions and media outputs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oq_Xg1XhXN7o"
      },
      "source": [
        "from fastprogress import progress_bar\n",
        "from IPython.display import HTML\n",
        "from base64 import b64encode\n",
        "import matplotlib.image as mpimg\n",
        "\n",
        "plt.rcParams['axes.facecolor'] = 'white'\n",
        "\n",
        "t_not = clip.tokenize(text_to_remove)\n",
        "t_not = perceptor.encode_text(t_not.cuda()).detach().clone()\n",
        "\n",
        "if not os.path.isdir(\"generated\"):\n",
        "  os.mkdir(\"generated\")\n",
        "\n",
        "def augment(into, cutn=32):\n",
        "  into = torch.nn.functional.pad(into, (sideX//2, sideX//2, sideX//2, sideX//2), mode='constant', value=0)\n",
        "  into = augs(into)\n",
        "  p_s = []\n",
        "  for ch in range(cutn):\n",
        "    # size = torch.randint(int(.5*sideX), int(1.9*sideX), ())\n",
        "    size = int(torch.normal(1.2, .3, ()).clip(.43, 1.9) * sideX) \n",
        "    if ch > cutn - 4:\n",
        "      size = int(sideX*1.4)\n",
        "    offsetx = torch.randint(0, int(sideX*2 - size), ())\n",
        "    offsety = torch.randint(0, int(sideX*2 - size), ())\n",
        "    apper = into[:, :, offsetx:offsetx + size, offsety:offsety + size]\n",
        "    apper = torch.nn.functional.interpolate(apper, (int(224*scaler), int(224*scaler)), mode='bilinear', align_corners=True)\n",
        "    p_s.append(apper)\n",
        "  into = torch.cat(p_s, 0)\n",
        "  into = into + up_noise*torch.rand((into.shape[0], 1, 1, 1)).cuda()*torch.randn_like(into, requires_grad=False)\n",
        "  return into\n",
        "\n",
        "def checkin(loss, display_image=True):\n",
        "  global up_noise  \n",
        "  global hadies\n",
        "  with torch.no_grad():    \n",
        "    if is_dall_e:\n",
        "      alnot = torch.tensor(unmap_pixels(torch.sigmoid(model(lats())[:, :3]).cpu().float())).cuda()\n",
        "    else:\n",
        "      alnot = model(lats()).float()\n",
        "    alnot = augment((((alnot).clip(-1, 1) + 1) / 2), cutn=1)\n",
        "    if display_image:\n",
        "      with output_widget:\n",
        "        output_widget.clear_output(wait=True)\n",
        "        for allls in alnot.cpu():\n",
        "            displ(allls,imSaver)\n",
        "\n",
        "        if is_dall_e:\n",
        "          alnot = torch.tensor(unmap_pixels(torch.sigmoid(model(lats())[:, :3]).cpu().float())).cuda() \n",
        "        else:\n",
        "          alnot = (model(lats()).cpu().clip(-1, 1) + 1) / 2\n",
        "\n",
        "        for allls in alnot.cpu():          \n",
        "          displ(allls,imSaver) \n",
        "          plt.figure(figsize=(12,12))\n",
        "          img=mpimg.imread(f\"3.png\")\n",
        "          plt.imshow(img)\n",
        "          ax = plt.gca()\n",
        "          ax.axes.xaxis.set_ticks([])\n",
        "          ax.axes.yaxis.set_visible(False)\n",
        "          xlab = f'\"{text_input}\"'\n",
        "          plt.xlabel(xlab,fontsize=12)\n",
        "          plt.show()\n",
        "          # print('\\n')\n",
        "    else:\n",
        "      for allls in alnot.cpu():\n",
        "          displ(allls,imSaver)\n",
        "\n",
        "  # \"ding\"\n",
        "  #output.eval_js('new Audio(\"https://freesound.org/data/previews/80/80921_1022651-lq.ogg\").play()')\n",
        "\n",
        "def ascend_txt():\n",
        "  global up_noise\n",
        "  if is_dall_e:\n",
        "    out = unmap_pixels(torch.sigmoid(model(lats())[:, :3].float()))\n",
        "  else:\n",
        "    out = model(lats())\n",
        "\n",
        "  if init_image:\n",
        "    into = augment(out)\n",
        "  else:\n",
        "    into = augment((out.clip(-1, 1) + 1) / 2) #augment(out) #\n",
        "  into = nom(into)\n",
        "  iii = perceptor.encode_image(into)\n",
        "\n",
        "  q = w0*t + w1*text_add + w2*img_enc + w3*ne_img_enc\n",
        "  q = q / q.norm(dim=-1, keepdim=True)\n",
        "  all_s = torch.cosine_similarity(q, iii, -1)\n",
        "\n",
        "  return [0, -10*all_s + 5 * torch.cosine_similarity(t_not, iii, -1)]\n",
        "  \n",
        "def train(i):\n",
        "  global dec\n",
        "  global up_noise\n",
        "\n",
        "  loss1 = ascend_txt()\n",
        "  loss = loss1[0] + loss1[1]\n",
        "  loss = loss.mean()\n",
        "  optimizer.zero_grad()\n",
        "  loss.backward()\n",
        "  optimizer.step()\n",
        "\n",
        "  if torch.abs(lats()).max() > 5:\n",
        "    for g in optimizer.param_groups:\n",
        "      g['weight_decay'] = dec\n",
        "  else:\n",
        "    for g in optimizer.param_groups:\n",
        "      g['weight_decay'] = 0\n",
        "  \n",
        "  \n",
        "  if itt % 4 == 0:\n",
        "    checkin(loss1)\n",
        "  elif itt%10 == 0:\n",
        "    checkin(loss1,False)\n",
        "  \n",
        "  if itt%50==0:\n",
        "    imSaver.make_video(progression_video_save_path, progression_video_fps)\n",
        "    video_out_widget.clear_output()\n",
        "    with video_out_widget:\n",
        "      display_video(\"generated/\"+progression_video_save_path)\n",
        "\n",
        "import IPython\n",
        "from google.colab import output\n",
        "\n",
        "def display_video(video_path):\n",
        "  # Compressed video path\n",
        "  compressed_path = \"result_compressed.mp4\"\n",
        "\n",
        "  if os.path.isfile(compressed_path):\n",
        "    os.remove(compressed_path)\n",
        "\n",
        "  os.system(f\"ffmpeg -i {video_path} -vcodec libx264 {compressed_path}\")\n",
        "\n",
        "  #show video\n",
        "  mp4 = open(compressed_path,'rb').read()\n",
        "  data_url = \"data:video/mp4;base64,\" + b64encode(mp4).decode()\n",
        "  display.display(HTML(\"\"\"\n",
        "    <video width=400 controls>\n",
        "          <source src=\"%s\" type=\"video/mp4\">\n",
        "    </video>\n",
        "    \"\"\" % data_url))\n",
        "\n",
        "def loop():\n",
        "  global itt\n",
        "  try:\n",
        "    for asatreat in progress_bar(range(itt,max_epochs)):\n",
        "      train(itt)\n",
        "      itt+=1\n",
        "  except KeyboardInterrupt:\n",
        "    with video_out_widget:\n",
        "      print(colored(f\"Interrupted at {itt} iterations\",\"red\"))\n",
        "    checkin(-1)\n",
        "    return"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1YrfTRWaiXe4"
      },
      "source": [
        "## 6. Training Loop + Output!\n",
        "\n",
        "### Run the two following cells to initiate the Training Loop\n",
        "\n",
        "Your generated images should start to appear shortly after"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 700,
          "referenced_widgets": [
            "fc3b748138564b8cba66722ad838042a",
            "c30e317e87f745fc92aad293738f5a1f"
          ]
        },
        "id": "Y2-ro0FFhUh9",
        "outputId": "fe1d7aaf-3d44-4747-c129-918fc8b920e7"
      },
      "source": [
        "#@title ### Your images will appear below here when the training loop is running\n",
        "output.register_callback('notebook.display_video', display_video)\n",
        "display.display(output_widget)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "fc3b748138564b8cba66722ad838042a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "Output()"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 456,
          "referenced_widgets": [
            "967fd7256f2849228b49a57ef64e97a0",
            "768719392ba04ec58e602b17ace429ce"
          ]
        },
        "id": "anOhsHa5bRIu",
        "outputId": "01a4460f-1739-4e44-e4e5-c1c98bf81785"
      },
      "source": [
        "#@title ## This cell will initiate the training loop!\n",
        "#@markdown #### Your generated images should appear above. It renders one every 25 epochs/iterations.\n",
        "#@markdown #### The progression video will appear below and it updates every 100 epochs. \n",
        "from termcolor import colored #colored prints\n",
        "display.display(video_out_widget)\n",
        "try:\n",
        "  loop()\n",
        "except KeyboardInterrupt:\n",
        "  print(colored(f\"Interrupted at {itt} epochs\",\"red\"))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "967fd7256f2849228b49a57ef64e97a0",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "Output()"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "    <div>\n",
              "        <style>\n",
              "            /* Turns off some styling */\n",
              "            progress {\n",
              "                /* gets rid of default border in Firefox and Opera. */\n",
              "                border: none;\n",
              "                /* Needs to be in here for Safari polyfill so background images work as expected. */\n",
              "                background-size: auto;\n",
              "            }\n",
              "            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n",
              "                background: #F44336;\n",
              "            }\n",
              "        </style>\n",
              "      <progress value='150' class='' max='10000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
              "      1.50% [150/10000 02:09<2:22:04]\n",
              "    </div>\n",
              "    "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "D-l9iElCDX-K"
      },
      "source": [
        "## Final Video:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wZ8f3kf8YXHf"
      },
      "source": [
        "imSaver.make_video(progression_video_save_path,fps=8)\n",
        "video = imSaver.show_video(progression_video_save_path)\n",
        "video.ipython_display(width=360)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iKP0tnHaiTyl"
      },
      "source": [
        "# Credits and References:\n",
        "\n",
        "1. **Zero-Shot Text-to-Image Generation:** https://paperswithcode.com/paper/zero-shot-text-to-image-generation (Aditya Ramesh, Mikhail Pavlov, Gabriel Goh, Scott Gray, Chelsea Voss, Alec Radford, Mark Chen, Ilya Sutskever)\n",
        "\n",
        "2. **OpenAI CLIP:** https://github.com/openai/CLIP (Alec Radford, Jong Wook Kim,Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal,\n",
        "Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever)\n",
        "\n",
        "3. **CompVis Taming Transformers:** https://github.com/CompVis/taming-transformers (Patrick Esser, Robin Rombach, Bjorn Ommer) \n",
        "\n",
        "4. **Ryan Murdoch's work ([@advadnoun](https://twitter.com/advadnoun) on Twitter).** Most of the code implementations here are taken and/or adapted from some of his notebooks. \n",
        "\n",
        "5. **OpenAI DALL-E's dVAE:** https://github.com/openai/DALL-E/ (Aditya Ramesh, Mikhail Pavlov, Gabriel Goh, Scott Gray, Chelsea Voss, Alec Radford, Mark Chen, Ilya Sutskever)\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MSYPs_9lqgng"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}