CSE_676_Text2Image_final.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4CKLxVcE71-E"
      },
      "outputs": [],
      "source": [
        "#!cd '/content/drive/MyDrive/Sem 2/CSE 676 - DL/Project/Final'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "g1dNbxuvfLtA",
        "outputId": "416cdabe-0ec4-4154-ebc4-09a72a04bac9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kTl3E5BsB0Xo"
      },
      "source": [
        "# **Text to Image 🧙🏻‍♂️ ✨**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oa3boQMtT_YO"
      },
      "outputs": [],
      "source": [
        "#Abhishek\n",
        "path='/content/drive/MyDrive/abhi_DL_GAN/Project/'\n",
        "#Saj\n",
        "#path='/content/drive/MyDrive/Sem 2/CSE 676 - DL/Project/'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9i7SMZcn8XBq",
        "outputId": "74ea7548-cc97-4224-9a6b-1b302b31083d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Unzipping tokenizers/punkt.zip.\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "\n",
        "import pickle\n",
        "import gensim\n",
        "\n",
        "import matplotlib.pyplot as pyplot\n",
        "from keras.preprocessing.image import array_to_img\n",
        "from numpy import expand_dims, linspace\n",
        "\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "from numpy import zeros\n",
        "from numpy import ones\n",
        "from numpy import vstack\n",
        "from numpy import asarray\n",
        "\n",
        "import os\n",
        "from keras.callbacks import ModelCheckpoint\n",
        "from keras.initializers import RandomNormal\n",
        "from numpy.random import random\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras import Model\n",
        "from numpy.random import randn\n",
        "from random import randint, choice\n",
        "\n",
        "import time\n",
        "from keras.layers.advanced_activations import PReLU\n",
        "from tensorflow.keras.utils import plot_model\n",
        "from IPython.display import clear_output\n",
        "\n",
        "import matplotlib.pyplot as pyplot\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "from nltk.tokenize import word_tokenize"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VvNtFzMviPaL"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HVVa5TaS9Uam"
      },
      "source": [
        "## **Importing Pickle and Word2Vec model 🗃️**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tvUzENKX8fOm"
      },
      "outputs": [],
      "source": [
        "#for Birds\n",
        "# data = pickle.load(open(path + \"Final/Bird/Pickles/image_vectors.p\", \"rb\"))\n",
        "# word_vector = pickle.load(open(path + \"Final/Bird/Pickles/word_vector_min_bird.p\", \"rb\"))\n",
        "# model = gensim.models.KeyedVectors.load_word2vec_format(path + 'Final/word2vec_pretrained_model/GoogleNews-vectors-negative300.bin', binary=True)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#For flickr8K\n",
        "data = pickle.load(open(path + \"Final/Pickles/image_vectors.p\", \"rb\"))\n",
        "word_vector = pickle.load(open(path + \"Final/Pickles/flickr8k_embeddings.p\", \"rb\"))\n",
        "model = gensim.models.KeyedVectors.load_word2vec_format(path + 'Final/word2vec_pretrained_model/GoogleNews-vectors-negative300.bin', binary=True)"
      ],
      "metadata": {
        "id": "Qa_eymo5Js_Z"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tr4EhbbD91TQ"
      },
      "source": [
        "## **Setting lables, Caption for corresponding images | Adding jitter (zoom) 🔊**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "-Ax5c8Mz89tJ",
        "outputId": "4c6bebab-140f-41ed-bda2-fa02d21d8c0e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "8127\n",
            "8127\n",
            "391723162_3bdeb7ea33.jpg\n",
            "391723162_3bdeb7ea33.jpg\n",
            "375171241_0302ad8481.jpg\n",
            "380034515_4fbdfa6b26.jpg\n",
            "3717531382_e1e05e22c5.jpg\n",
            "381976882_0063d16d88.jpg\n",
            "380537190_11d6c0a412.jpg\n",
            "379006645_b9a2886b51.jpg\n",
            "397451339_76a84bd310.jpg\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 9 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "def zoom(image):\n",
        "  image = expand_dims(image, 0) #add additional dimension necessary for zooming\n",
        "  image = image_augmentation_generator.flow(image, batch_size=1)\n",
        "  result = image[0].reshape(image[0].shape[1:]) #remove additional dimension (1, 64, 64, 3) to (64, 64, 3)\n",
        "  return result\n",
        "\n",
        "image_augmentation_generator = tf.keras.preprocessing.image.ImageDataGenerator(zoom_range=[0.8, 1.0]) # random zoom proves to be helpful in capturing more details https://machinelearningmastery.com/how-to-configure-image-data-augmentation-when-training-deep-learning-neural-networks/\n",
        "\n",
        "n = 227\n",
        "image_embeddings = []\n",
        "captions = []\n",
        "labels = []\n",
        "for i, k in enumerate(data.keys()):\n",
        "  if k not in  word_vector:\n",
        "    continue\n",
        "  image_embeddings.append(data[k])\n",
        "  captions.append(word_vector[k])\n",
        "  labels.append(k)\n",
        "  if i % n == 0:\n",
        "   image_embeddings.append(zoom(data[k]))\n",
        "   captions.append(word_vector[k])\n",
        "   labels.append(k)\n",
        "\n",
        "print(len(captions))\n",
        "print(len(image_embeddings))\n",
        "\n",
        "pyplot.axis('off')\n",
        "slcie = image_embeddings[3:33]\n",
        "for i in range(9):\n",
        "  pyplot.subplot(3, 3, i+1)\n",
        "  pyplot.imshow(array_to_img(slcie[i]))\n",
        "  print(labels[i])\n",
        "  pyplot.axis('off')\n",
        "\n",
        "pyplot.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iROpkYfBcU3J"
      },
      "source": [
        "RUN THE BELOW CODES FOR FLICKR ONLY"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Qk56pbMY-lYY"
      },
      "outputs": [],
      "source": [
        "#for flickr8k only\n",
        "# Replacing NaN and appending 1 to match dimension=\n",
        "a=np.argwhere(np.isnan(captions))\n",
        "for i in a:\n",
        "  captions[i[0]][i[1]]=0\n",
        "\n",
        "image_embeddings.append(image_embeddings[8126])\n",
        "captions.append(captions[8126])\n",
        "labels.append(labels[8126])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6MCsC5OQuYwN",
        "outputId": "37caf1ee-41bb-4e78-d307-272d32c9ebb9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "8128 8128 8128\n"
          ]
        }
      ],
      "source": [
        "print(len(image_embeddings),\n",
        "len(captions),\n",
        "len(labels))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HjmmqGA3AkzG"
      },
      "source": [
        "## **Vectorizing**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "exX8nQFgAAhr"
      },
      "outputs": [],
      "source": [
        "def get_random_word_vectors_from_dataset(n_samples):\n",
        "  ix = np.random.randint(0, len(captions), n_samples)\n",
        "  return np.asarray(captions)[ix]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "h4OlXt8iAVqX"
      },
      "outputs": [],
      "source": [
        "def generate_random_vectors(n_samples):  \n",
        "  vectorized_random_captions = []\n",
        "\n",
        "  for n in range(n_samples):\n",
        "    rnd = randint(8, 25)\n",
        "    result_array = np.empty((0, 300))\n",
        "    for i in range(rnd):\n",
        "      result_array = np.append(result_array, [model[choice(model.index2entity)]], axis=0)\n",
        "    vectorized_random_captions.append(np.mean(result_array, axis=0).astype('float32'))\n",
        "\n",
        "  return np.array(vectorized_random_captions)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hVoxf3d-BON1"
      },
      "source": [
        "## **Discriminator Model ⚙️**\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4JqucuWPBL0h"
      },
      "outputs": [],
      "source": [
        "# Discriminator model\n",
        "def define_discriminator():\n",
        "    kernel_init = tf.random_normal_initializer(stddev=0.02)\n",
        "    #batch_init = tf.random_normal_initializer(1., 0.02)\n",
        "    word_vector_dim = 300\n",
        "    dropout_prob = 0.25\n",
        "\n",
        "    in_label = layers.Input(shape=(300,))\n",
        "  \n",
        "    n_nodes = 3 * 64 * 64\n",
        "    li = layers.Dense(n_nodes)(in_label)\n",
        "    li = layers.Reshape((64, 64, 3))(li)\n",
        "    dis_input = layers.Input(shape=(64, 64, 3))\n",
        "\n",
        "    merge = layers.Concatenate()([dis_input, li])\n",
        "    #check output\n",
        "    discriminator = layers.Conv2D(filters=64, kernel_size=(3, 3), padding=\"same\")(merge)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "    discriminator = layers.GaussianNoise(0.2)(discriminator)\n",
        "    \n",
        "    #regularization layer to avoid overfitting\n",
        "    discriminator = layers.Conv2D(filters=96, kernel_size=(3, 3), strides=(2, 2), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU()(discriminator)\n",
        "\n",
        "    discriminator = layers.Conv2D(filters=192, kernel_size=(3, 3), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Conv2D(filters=384, kernel_size=(3, 3), strides=(2, 2), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Conv2D(filters=192, kernel_size=(3, 3), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Conv2D(filters=384, kernel_size=(3, 3), strides=(2, 2), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Conv2D(filters=768, kernel_size=(3, 3), padding=\"same\",kernel_initializer=kernel_init)(discriminator)\n",
        "    discriminator = layers.BatchNormalization(momentum=0.8)(discriminator)\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Flatten()(discriminator)\n",
        "\n",
        "    discriminator = layers.Dense(1024)(discriminator)\n",
        "\n",
        "    discriminator = layers.LeakyReLU(0.2)(discriminator)\n",
        "\n",
        "    discriminator = layers.Dense(1)(discriminator)\n",
        "\n",
        "    discriminator_model = Model(inputs=[dis_input, in_label], outputs=discriminator)\n",
        "\n",
        "    discriminator_model.summary()\n",
        "\n",
        "    return discriminator_model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p3GtIZSIBgn3"
      },
      "source": [
        "## **Generator Model 🚀**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qY8Nwu7kBU-4"
      },
      "outputs": [],
      "source": [
        "# Using ResNet to improve image feature laerning\n",
        "def resnet_block(model, kernel_size, filters, strides):\n",
        "    gen = model #identity\n",
        "    model = layers.Conv2D(filters=filters, kernel_size=kernel_size, strides=strides, padding=\"same\")(model)\n",
        "    model = layers.BatchNormalization(momentum=0.8)(model)\n",
        "    model = tf.keras.layers.PReLU(alpha_initializer='zeros', alpha_regularizer=None, alpha_constraint=None, shared_axes=[1, 2])(model)\n",
        "    model = layers.Conv2D(filters=filters, kernel_size=kernel_size, strides=strides, padding=\"same\")(model)\n",
        "    model = layers.BatchNormalization(momentum=0.8)(model)\n",
        "    model = layers.Add()([gen, model])\n",
        "    return model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6qFJKPpdBB7U"
      },
      "outputs": [],
      "source": [
        "# Generator model\n",
        "def define_generator():\n",
        "    kernel_init = tf.random_normal_initializer(stddev=0.02)\n",
        "    #batch_init = tf.random_normal_initializer(1., 0.02)\n",
        "\n",
        "    random_input = layers.Input(shape=(100,))\n",
        "    text_input1 = layers.Input(shape=(300,))\n",
        "    text_layer1 = layers.Dense(8192)(text_input1)\n",
        "    text_layer1 = layers.Reshape((8, 8, 128))(text_layer1)\n",
        "\n",
        "    n_nodes = 128 * 8 * 8\n",
        "    gen_input_dense = layers.Dense(n_nodes)(random_input)\n",
        "    generator = layers.Reshape((8, 8, 128))(gen_input_dense)\n",
        "\n",
        "    merge = layers.Concatenate()([generator, text_layer1])\n",
        "\n",
        "    model = layers.Conv2D(filters=64, kernel_size=9, strides=1, padding=\"same\")(merge)\n",
        "    model = tf.keras.layers.PReLU(alpha_initializer='zeros', alpha_regularizer=None, alpha_constraint=None, shared_axes=[1, 2])(model)\n",
        "\n",
        "    gen_model = model\n",
        "\n",
        "    for _ in range(4):\n",
        "      model = resnet_block(model, 3, 64, 1)\n",
        "\n",
        "    model = layers.Conv2D(filters=64, kernel_size=3, strides=1, padding=\"same\")(model)\n",
        "    model = layers.BatchNormalization(momentum=0.8)(model)\n",
        "    model = layers.Add()([gen_model, model])\n",
        "    #check summary here\n",
        "    \n",
        "    # model = layers.Conv2DTranspose(filters=256, kernel_size=(3, 3), strides=(2, 2), padding=\"same\", kernel_initializer=kernel_init)(model)\n",
        "    # model = layers.LeakyReLU(0.2)(model)\n",
        "    model = layers.Conv2DTranspose(filters=768, kernel_size=(3, 3), strides=(2, 2), padding=\"same\", kernel_initializer=kernel_init)(model)\n",
        "    model = layers.LeakyReLU(0.2)(model)\n",
        "\n",
        "    model = layers.Conv2DTranspose(filters=384, kernel_size=(3, 3), strides=(2, 2), padding=\"same\", kernel_initializer=kernel_init)(model)\n",
        "    model = layers.LeakyReLU(0.2)(model)\n",
        "\n",
        "    model = layers.Conv2DTranspose(filters=192, kernel_size=(3, 3), strides=(2, 2), padding=\"same\", kernel_initializer=kernel_init)(model)\n",
        "    model = layers.LeakyReLU(0.2)(model)\n",
        "\n",
        "    model = layers.Conv2DTranspose(filters=64, kernel_size=(3, 3), strides=(1, 1), padding=\"same\", kernel_initializer=kernel_init)(model)\n",
        "    model = layers.LeakyReLU(0.2)(model)\n",
        "\n",
        "    model = layers.Conv2D(3, (3, 3), padding='same', activation='tanh')(model)\n",
        "\n",
        "    generator_model = Model(inputs=[random_input, text_input1], outputs=model)\n",
        "\n",
        "    generator_model.summary()\n",
        "\n",
        "    return generator_model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "epOATN86Gw0O"
      },
      "source": [
        "\n",
        "Generating Latent Noise"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Ivs-3umuFeqp"
      },
      "outputs": [],
      "source": [
        "def generate_latent_points(latent_dim, n_samples):\n",
        "    x_input  = tf.random.normal([n_samples, latent_dim])\n",
        "    text_captions = get_random_word_vectors_from_dataset(n_samples)\n",
        "    return [x_input, text_captions]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qAbvL7YAGKCm"
      },
      "outputs": [],
      "source": [
        "def noisy_labels(y, p_flip):\n",
        "    n_select = int(p_flip * int(y.shape[0]))\n",
        "    flip_ix = np.random.choice([i for i in range(int(y.shape[0]))], size=n_select)\n",
        "\n",
        "    op_list = []\n",
        "    for i in range(int(y.shape[0])):\n",
        "        if i in flip_ix:\n",
        "            op_list.append(tf.subtract(1.0, y[i]))\n",
        "        else:\n",
        "            op_list.append(y[i])\n",
        "\n",
        "    outputs = tf.stack(op_list)\n",
        "    return outputs"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vuUAY1GwGNIo"
      },
      "outputs": [],
      "source": [
        "def load_data():\n",
        "    return asarray(image_embeddings), asarray(captions).astype('float32')\n",
        "\n",
        "#label smoothing\n",
        "def smooth_positive_labels(y):\n",
        "\treturn y - 0.3 + (np.random.random(y.shape) * 0.5)\n",
        " \n",
        "def smooth_negative_labels(y):\n",
        "\treturn y + np.random.random(y.shape) * 0.3"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RdG04PhSHBin"
      },
      "source": [
        "## **Generating and Saving images 🎉**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hWeeVmrDGUo9"
      },
      "outputs": [],
      "source": [
        "def generate_and_save_images(model, epoch, test_input):\n",
        "  predictions = model(test_input, training=False)\n",
        "\n",
        "  print(predictions.shape)\n",
        "  pyplot.figure(figsize=[7, 7])\n",
        "\n",
        "  for i in range(predictions.shape[0]):\n",
        "      pyplot.subplot(5, 5, i+1)\n",
        "      pyplot.imshow(array_to_img(predictions.numpy()[i]))\n",
        "      pyplot.axis('off')\n",
        "\n",
        "  #Birds\n",
        "  #pyplot.savefig('/content/drive/MyDrive/abhi_DL_GAN/Project/Final/Bird/bird_gen_images/bird_image_at_epoch_{:04d}.png'.format(epoch))\n",
        "  \n",
        "  #Flick8k\n",
        "  pyplot.savefig('/content/drive/MyDrive/abhi_DL_GAN/Project/Final/flickr_gen_images/Flickr8k_images_at_epoch_{:04d}.png'.format(epoch))\n",
        "  pyplot.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CJvOGdEJHOZL"
      },
      "source": [
        "## **Discriminator Loss 📈**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QfoMMM9tGpwZ"
      },
      "outputs": [],
      "source": [
        "def discriminator_loss(r_real_output_real_text, f_fake_output_real_text_1, f_real_output_fake_text):\n",
        "   alpha = 0.5\n",
        "   real_output_noise = smooth_positive_labels(noisy_labels(tf.ones_like(r_real_output_real_text), 0.10))\n",
        "   fake_output_real_text_noise_1 = smooth_negative_labels(tf.zeros_like(f_fake_output_real_text_1))\n",
        "   real_output_fake_text_noise = smooth_negative_labels(tf.zeros_like(f_real_output_fake_text))\n",
        "\n",
        "   real_loss = tf.reduce_mean(binary_cross_entropy(real_output_noise, r_real_output_real_text))\n",
        "   fake_loss_ms_1 = tf.reduce_mean(binary_cross_entropy(fake_output_real_text_noise_1, f_fake_output_real_text_1))\n",
        "   fake_loss_2 = tf.reduce_mean(binary_cross_entropy(real_output_fake_text_noise, f_real_output_fake_text))\n",
        "\n",
        "   total_loss = real_loss + alpha * fake_loss_2 + (1-alpha) * fake_loss_ms_1 \n",
        "   return total_loss"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6OoduDviHeH8"
      },
      "source": [
        "## **Generator Loss📉**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "CIZTre_rHZTT"
      },
      "outputs": [],
      "source": [
        "def generator_loss(f_fake_output_real_text):\n",
        "   return tf.reduce_mean(binary_cross_entropy(tf.ones_like(f_fake_output_real_text), f_fake_output_real_text))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ne0pUbf2Dr21"
      },
      "source": [
        "## Plot Loss"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PDUJRLo4Dvtr"
      },
      "outputs": [],
      "source": [
        "def plot_history(d1_hist, g_hist):\n",
        "\tpyplot.subplot(2, 1, 1)\n",
        "\tpyplot.plot(d1_hist, label='disc')\n",
        "\tpyplot.plot(g_hist, label='gen')\n",
        "\tpyplot.legend()\n",
        "\tpyplot.show()\n",
        "\tpyplot.savefig('/content/drive/MyDrive/abhi_DL_GAN/Project/Final/flickr_gen_images') #flickr\n",
        "\t#pyplot.savefig('/content/drive/MyDrive/abhi_DL_GAN/Project/Final/Bird/bird_gen_images') #birds\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "myc4AtqJIQj-"
      },
      "source": [
        "## **Training train 👷🏼**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        },
        "id": "ujwnLtjlHi0L",
        "outputId": "1e78d5de-2d78-4057-cec2-902f8cc06c49"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n",
            "Time for epoch 540 is 27.34685182571411 sec\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "@tf.function\n",
        "def train_step(images, epoch):\n",
        "\n",
        "  #define half_batch\n",
        "  latent_dim = 100\n",
        "  n_batch = 64\n",
        "\n",
        "  noise_1 = tf.random.normal([32, latent_dim])\n",
        "  noise_2 = tf.random.normal([32, latent_dim])\n",
        "  real_captions = images[1]\n",
        "  print(real_captions)\n",
        "  real_images = images[0]\n",
        "  print(real_images)\n",
        "\n",
        "  random_captions = generate_random_vectors(n_batch)\n",
        "  random_captions_1, random_captions_2  = tf.split(random_captions, 2, 0)\n",
        "  real_captions_1, real_captions_2  = tf.split(real_captions, 2 ,0)\n",
        "  real_images_1, real_images_2 = tf.split(real_images, 2, 0)\n",
        "  temp_d,temp_g=[],[]\n",
        "  with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:\n",
        "    \n",
        "    noise = tf.concat([noise_1, noise_2], 0)\n",
        "\n",
        "    generated_images = generator([noise, real_captions], training=True)\n",
        "\n",
        "    fake_1, fake_2 = tf.split(generated_images, 2, 0)\n",
        "\n",
        "    f_fake_output_real_text_1 = discriminator([fake_1, real_captions_1], training=True)\n",
        "    f_fake_output_real_text_2 = discriminator([fake_2, real_captions_2], training=True)\n",
        "\n",
        "    r_real_output_real_text_1 = discriminator([real_images_1, real_captions_1], training=True)\n",
        "    r_real_output_real_text_2 = discriminator([real_images_2, real_captions_2], training=True)\n",
        "\n",
        "    f_real_output_fake_text_1 = discriminator([real_images_1, random_captions_1], training=True)\n",
        "    f_real_output_fake_text_2 = discriminator([real_images_2, random_captions_2], training=True)\n",
        "\n",
        "    #### Calculating losses ####\n",
        "\n",
        "    gen_loss = generator_loss(f_fake_output_real_text_1) + generator_loss(f_fake_output_real_text_2) \n",
        "    # mode seeking loss\n",
        "    \n",
        "    lz = tf.math.reduce_mean(tf.math.abs(fake_2-fake_1)) / tf.math.reduce_mean(tf.math.abs(noise_2-noise_1))\n",
        "\n",
        "    eps = 1 * 1e-5\n",
        "    loss_lz = 1 / (eps+lz) * ms_loss_weight\n",
        "    total_gen_loss = gen_loss + loss_lz\n",
        "\n",
        "    tf.print('G_loss', [total_gen_loss])\n",
        "    temp_g.append(total_gen_loss)\n",
        "    disc_loss_1 = discriminator_loss(r_real_output_real_text_1, f_fake_output_real_text_1, f_real_output_fake_text_1)\n",
        "    disc_loss_2 = discriminator_loss(r_real_output_real_text_2, f_fake_output_real_text_2, f_real_output_fake_text_2)\n",
        "    \n",
        "    total_disc_loss = disc_loss_1 + disc_loss_2\n",
        "    temp_d.append(total_disc_loss)\n",
        "    tf.print('D_loss', [total_disc_loss])\n",
        "\n",
        "    #### Done calculating losses ####\n",
        "\n",
        "  gradients_of_discriminator = disc_tape.gradient(total_disc_loss, discriminator.trainable_variables)  \n",
        "\n",
        "  gradients_of_generator = gen_tape.gradient(total_gen_loss, generator.trainable_variables)    \n",
        "\n",
        "  generator_optimizer.apply_gradients(zip(gradients_of_generator, generator.trainable_variables))\n",
        "\n",
        "  discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, discriminator.trainable_variables))\n",
        "  return temp_d,temp_g\n",
        "\n",
        "def train(dataset, epochs = 540):\n",
        "\n",
        "  #checkpoint_dir = path + 'Final/Bird/checkpoints' #Birds\n",
        "  checkpoint_dir = path + 'Final/checkpoints'. #Flickr\n",
        "  checkpoint_prefix = os.path.join(checkpoint_dir, \"ckpt\")\n",
        "  checkpoint = tf.train.Checkpoint(generator_optimizer=generator_optimizer,\n",
        "                                discriminator_optimizer=discriminator_optimizer,\n",
        "                                  generator=generator,\n",
        "                                  discriminator=discriminator)\n",
        "  \n",
        "  ckpt_manager = tf.train.CheckpointManager(checkpoint, checkpoint_dir, max_to_keep=3)\n",
        "  if ckpt_manager.latest_checkpoint:\n",
        "    checkpoint.restore(ckpt_manager.latest_checkpoint)  #ckpt_manager.checkpoints[3]\n",
        "    print ('Latest checkpoint restored!!')\n",
        "  d1_hist, g_hist= list(), list()\n",
        "  \n",
        "  for epoch in range(epochs):\n",
        "    start = time.time()\n",
        "\n",
        "    for image_batch in dataset:\n",
        "      total_disc_loss,total_gen_loss=train_step(image_batch, epoch) \n",
        "    d1_hist.extend(total_disc_loss)\n",
        "    g_hist.extend(total_gen_loss)\n",
        "  \n",
        "    if (epoch +1) % 10 == 0:\n",
        "      [z_input, labels_input] = generate_latent_points(100, 25)\n",
        "      generate_and_save_images(generator,\n",
        "                            epoch + 1,\n",
        "                             [z_input, labels_input])\n",
        "\n",
        "    if (epoch + 1) % 40 == 0:\n",
        "    #if epoch> 0:\n",
        "      ckpt_save_path = ckpt_manager.save()\n",
        "      print ('Saving checkpoint for epoch {} at {}'.format(epoch+1,ckpt_save_path))\n",
        "\n",
        "    if (epoch +1) % 60 == 0:\n",
        "    #if epoch> 0:\n",
        "      clear_output(wait=True)\n",
        "      #generator.save(path + 'Final/Bird/Weights/bird_model_weights_%03d.h5' % (epoch + 1)) #Birds\n",
        "      generator.save(path + 'Final/Weights/new_flickr8k_model_weights_%03d.h5' % (epoch + 1))     #Flickr\n",
        "\n",
        "    print ('Time for epoch {} is {} sec'.format(epoch + 1, time.time()-start))\n",
        "  return d1_hist, g_hist\n",
        "\n",
        "ms_loss_weight = 1.0\n",
        "\n",
        "binary_cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)\n",
        "\n",
        "generator_optimizer = tf.keras.optimizers.Adam(learning_rate=0.00004, beta_1 = 0.5)\n",
        "discriminator_optimizer = tf.keras.optimizers.Adam(learning_rate=0.00004, beta_1 = 0.5)\n",
        "\n",
        "discriminator = define_discriminator()\n",
        "generator = define_generator()\n",
        "\n",
        "images, lbs = load_data()\n",
        "BUFFER_SIZE = images.shape[0]\n",
        "BATCH_SIZE = 64\n",
        "\n",
        "train_dataset = tf.data.Dataset.from_tensor_slices((images,lbs)).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)\n",
        "\n",
        "\n",
        "d1_hist, g_hist=train(train_dataset)\n",
        "plot_history(d1_hist, g_hist)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Sac7zEJ6OAik"
      },
      "source": [
        "## **Visualization 💯**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E2a43G_5DuVI",
        "outputId": "54171d5a-73f9-43af-b333-8fe461e905ae"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n"
          ]
        }
      ],
      "source": [
        "#\n",
        "from keras.preprocessing.image import array_to_img\n",
        "\n",
        "from numpy import expand_dims\n",
        "from numpy import zeros\n",
        "from numpy import ones\n",
        "from numpy import vstack\n",
        "from numpy import asarray\n",
        "import numpy as np\n",
        "\n",
        "import os\n",
        "from keras.callbacks import ModelCheckpoint\n",
        "from keras.initializers import RandomNormal\n",
        "from numpy.random import random\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras import Model\n",
        "from numpy.random import randn\n",
        "from numpy.random import randint\n",
        "import time\n",
        "import matplotlib.pyplot as pyplot\n",
        "from keras.preprocessing.image import array_to_img\n",
        "from numpy import linspace\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "from nltk.tokenize import word_tokenize\n",
        "\n",
        "from numpy.random import randint\n",
        "import numpy as np"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "id": "Ojp1ZXRoODBd"
      },
      "outputs": [],
      "source": [
        "def create_sent_vector(sent):\n",
        "  result_array = np.empty((0, 300))\n",
        "\n",
        "  for word in word_tokenize(sent):\n",
        "    result_array = np.append(result_array, [model[word]], axis=0)\n",
        "  final = np.mean(result_array, axis=0).astype('float32')\n",
        "  return final"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "XdhNNHT_Oe0Q"
      },
      "outputs": [],
      "source": [
        "# User input to generate image from text\n",
        "def generate_random_word_vectors_from_dataset(n_samples, create_new_captions = True):\n",
        "\n",
        "  if create_new_captions:\n",
        "      e = create_sent_vector('A girl with yellow dress')\n",
        "      f = create_sent_vector('A boy with blue tshirt')\n",
        "      v = []\n",
        "      flag = True\n",
        "      for i in range(n_samples):\n",
        "        \n",
        "        if not flag:\n",
        "          v.append(f)\n",
        "          flag = True\n",
        "        elif flag:\n",
        "          v.append(e)\n",
        "          flag = False\n",
        "      return np.asarray(v), np.asarray([])\n",
        "  else:\n",
        "    ix = randint(0,len(captions), n_samples)\n",
        "    print(ix)\n",
        "    return  np.asarray(captions)[ix], np.asarray(labels)[ix]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "8d3lUH9kOfc0"
      },
      "outputs": [],
      "source": [
        "def generate_images(model, test_input):\n",
        "\n",
        "  predictions = model(test_input, training=False)\n",
        "\n",
        "  print(predictions.shape)\n",
        "  pyplot.figure(figsize=[15, 15])\n",
        "  for i in range(predictions.shape[0]):\n",
        "      pyplot.subplot(1, 9, i+1)\n",
        "      pyplot.imshow(array_to_img(predictions.numpy()[i]))\n",
        "      pyplot.axis('off')\n",
        "\n",
        "  pyplot.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "ywb3vO9hOhdB"
      },
      "outputs": [],
      "source": [
        "def interpolate_points(p1, p2, n_steps=9):\n",
        "\tratios = linspace(0, 1, num=n_steps)\n",
        "\tvectors = list()\n",
        "\tfor ratio in ratios:\n",
        "\t\tv = (1.0 - ratio) * p1 + ratio * p2\n",
        "\t\tvectors.append(v)\n",
        "\treturn np.asarray(vectors)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "id": "Iq5_fhUGOjLe"
      },
      "outputs": [],
      "source": [
        "def generate_latent_points(latent_dim, n_samples, interpolate = False):\n",
        "    x_input  = tf.random.normal([n_samples, latent_dim])\n",
        "    text_captions, labels = generate_random_word_vectors_from_dataset(n_samples, create_new_captions=False)\n",
        "    if interpolate:\n",
        "      text_captions = interpolate_points(text_captions[0], text_captions[1])\n",
        "      x_input = interpolate_points(x_input[0], x_input[1])\n",
        "    for index, s in enumerate(labels.flat):\n",
        "      print(index, s)\n",
        "\n",
        "    return [x_input, text_captions]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X9hvWrMwOlFA",
        "outputId": "14e56105-3247-40c5-c4ed-fe6b17575f35"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "WARNING:tensorflow:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
          ]
        }
      ],
      "source": [
        "#gen_model = tf.keras.models.load_model(path + 'Final/Bird/Weights/bird_model_weights_780.h5') #bird\n",
        "gen_model = tf.keras.models.load_model(path + 'Final/Weights/flickr8K_model_weights_840.h5') #flickr"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 43,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 310
        },
        "id": "kgrUkIqBOnMc",
        "outputId": "b49575a0-7ac5-4a38-b138-0c44aa1e468b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[1626]\n",
            "0 3644142276_caed26029e.jpg\n",
            "(1, 64, 64, 3)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x1080 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[5934]\n",
            "0 2837808847_5407af1986.jpg\n",
            "(1, 64, 64, 3)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x1080 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "for i in range(2):\n",
        "  generate_images(gen_model, generate_latent_points(100, 1))"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#FID calculation\n",
        "\n",
        "import numpy\n",
        "from numpy import cov\n",
        "from numpy import trace\n",
        "from numpy import iscomplexobj\n",
        "from numpy import asarray\n",
        "from numpy.random import randint\n",
        "from scipy.linalg import sqrtm\n",
        "from keras.applications.inception_v3 import InceptionV3\n",
        "from keras.applications.inception_v3 import preprocess_input\n",
        "from keras.datasets.mnist import load_data\n",
        "from skimage.transform import resize\n",
        "from PIL import Image\n",
        "from numpy import asarray\n",
        "\n",
        "# scale an array of images to a new size\n",
        "def scale_images(images, new_shape):\n",
        "\timages_list = list()\n",
        "\tfor image in images:\n",
        "\t\t# resize with nearest neighbor interpolation\n",
        "\t\tnew_image = resize(image, new_shape, 0)\n",
        "\t\t# store\n",
        "\t\timages_list.append(new_image)\n",
        "\treturn asarray(images_list)\n",
        " \n",
        "# calculate frechet inception distance\n",
        "def calculate_fid(model, images1, images2):\n",
        "\t# calculate activations\n",
        "\tact1 = model.predict(images1)\n",
        "\tact2 = model.predict(images2)\n",
        "\t# calculate mean and covariance statistics\n",
        "\tmu1, sigma1 = act1.mean(axis=0), cov(act1, rowvar=False)\n",
        "\tmu2, sigma2 = act2.mean(axis=0), cov(act2, rowvar=False)\n",
        "\t# calculate sum squared difference between means\n",
        "\tssdiff = numpy.sum((mu1 - mu2)**2.0)\n",
        "\t# calculate sqrt of product between cov\n",
        "\tcovmean = sqrtm(sigma1.dot(sigma2))\n",
        "\t# check and correct imaginary numbers from sqrt\n",
        "\tif iscomplexobj(covmean):\n",
        "\t\tcovmean = covmean.real\n",
        "\t# calculate score\n",
        "\tfid = ssdiff + trace(sigma1 + sigma2 - 2.0 * covmean)\n",
        "\treturn fid\n",
        " \n",
        "# prepare the inception v3 model\n",
        "model = InceptionV3(include_top=False, pooling='avg', input_shape=(299,299,3))\n",
        "# define two fake collections of images\n",
        "# load the image\n",
        "image = Image.open('path_to_generated_image')\n",
        "# convert image to numpy array\n",
        "images1 = asarray(image)\n",
        "images1=numpy.resize(images1, (10,32,32,3))\n",
        "images1 = images1.reshape((10,32,32,3))\n",
        "image2 = Image.open('path_to_real_image')\n",
        "# convert image to numpy array\n",
        "images2 = asarray(image2)\n",
        "# images2 = randint(0, 255, 10*32*32*3)\n",
        "images2=numpy.resize(images2, (10,32,32,3))\n",
        "images2 = images2.reshape((10,32,32,3))\n",
        "print('Prepared', images1.shape, images2.shape)\n",
        "# convert integer to floating point values\n",
        "images1 = images1.astype('float32')\n",
        "images2 = images2.astype('float32')\n",
        "# resize images\n",
        "images1 = scale_images(images1, (299,299,3))\n",
        "images2 = scale_images(images2, (299,299,3))\n",
        "print('Scaled', images1.shape, images2.shape)\n",
        "# pre-process images\n",
        "images1 = preprocess_input(images1)\n",
        "images2 = preprocess_input(images2)\n",
        "# fid between images1 and images1\n",
        "fid = calculate_fid(model, images1, images1)\n",
        "print('FID (same): %.3f' % fid)\n",
        "# fid between images1 and images2\n",
        "fid = calculate_fid(model, images1, images2)\n",
        "print('FID (different): %.3f' % fid)"
      ],
      "metadata": {
        "id": "z3P8iCNW0Jtk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "86ab8f7d-c749-4f25-a9e6-a343dc58bcad"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/inception_v3/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "87916544/87910968 [==============================] - 0s 0us/step\n",
            "87924736/87910968 [==============================] - 0s 0us/step\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "FileNotFoundError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-39-6970e63f6ec4>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     48\u001b[0m \u001b[0;31m# define two fake collections of images\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     49\u001b[0m \u001b[0;31m# load the image\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 50\u001b[0;31m \u001b[0mimage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mImage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'/content/drive/MyDrive/Project/Final/Bird/for_fid/birdddd.jpg'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     51\u001b[0m \u001b[0;31m# convert image to numpy array\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     52\u001b[0m \u001b[0mimages1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0masarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimage\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/PIL/Image.py\u001b[0m in \u001b[0;36mopen\u001b[0;34m(fp, mode)\u001b[0m\n\u001b[1;32m   2841\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2842\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfilename\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2843\u001b[0;31m         \u001b[0mfp\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbuiltins\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"rb\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2844\u001b[0m         \u001b[0mexclusive_fp\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2845\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: '/content/drive/MyDrive/Project/Final/Bird/for_fid/birdddd.jpg'"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "ThBRvyZjFY5Z"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "background_execution": "on",
      "collapsed_sections": [],
      "machine_shape": "hm",
      "name": "CSE_676_Text2Image_final.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}