improve_of_AC_DCGAN.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MN-30Z-ubc7x",
        "outputId": "fa9b7fd4-9f8c-4cba-d9e7-a2aa35307599"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/mydrive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/mydrive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# example of fitting an auxiliary classifier gan (ac-gan) on fashion mnsit\n",
        "from numpy import zeros\n",
        "from numpy import ones\n",
        "from numpy import expand_dims\n",
        "from numpy.random import randn\n",
        "from numpy.random import randint\n",
        "\n",
        "from keras.optimizers import Adam\n",
        "from keras.models import Model\n",
        "from keras.layers import Input\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Reshape\n",
        "from keras.layers import Flatten\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import Conv2DTranspose\n",
        "from keras.layers import LeakyReLU\n",
        "from keras.layers import BatchNormalization\n",
        "from keras.layers import Dropout\n",
        "from keras.layers import Embedding\n",
        "from keras.layers import Activation\n",
        "from keras.layers import Concatenate\n",
        "from keras.initializers import RandomNormal\n",
        "from matplotlib import pyplot\n",
        " "
      ],
      "metadata": {
        "id": "Eyd2Jw0gD4WV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EDfzPsRLcCJ7",
        "outputId": "98a85d08-de52-4d13-d1f4-a75a74ccb0be"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.9.2\n"
          ]
        }
      ],
      "source": [
        "\n",
        "%matplotlib inline\n",
        "from google.colab import files\n",
        "import os\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tqdm import tqdm\n",
        "from IPython import display\n",
        "import cv2\n",
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split \n",
        "from sklearn.metrics import confusion_matrix \n",
        "print(tf.__version__)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qtvwhmASbvnj",
        "outputId": "4ae80498-a990-4310-8a29-02b34843838e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1005\n"
          ]
        }
      ],
      "source": [
        "\n",
        "imagepaths = []\n",
        "\n",
        "for root, dirs, files in os.walk(\"mydrive/My Drive/DCGAN/\", topdown=False): \n",
        "  for name in files:\n",
        "    path = os.path.join(root, name)\n",
        "    if path.endswith(\"png\"): \n",
        "      imagepaths.append(path)\n",
        "\n",
        "print(len(imagepaths)) "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mV_embp-cWX0",
        "outputId": "b65dd10f-ccf4-4501-ffe3-7317e70eea89"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Images loaded:  1005\n",
            "Labels loaded:  1005\n",
            "4 mydrive/My Drive/DCGAN/Class 4/11.png\n"
          ]
        }
      ],
      "source": [
        "X = [] \n",
        "y = [] \n",
        "\n",
        "\n",
        "for path in imagepaths:\n",
        "  img = cv2.imread(path) \n",
        "  img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) \n",
        "  img = cv2.resize(img, (28, 28))\n",
        "  X.append(img)\n",
        "  \n",
        "\n",
        "  category = path.split(\"/\")[3]\n",
        "  label = int(category.split(\" \")[1]) \n",
        "  y.append(label)\n",
        "\n",
        "\n",
        "X = np.array(X, dtype=\"uint8\")\n",
        "#X = X.reshape(len(imagepaths), 232, 424, 3) # \n",
        "y = np.array(y)\n",
        "\n",
        "print(\"Images loaded: \", len(X))\n",
        "print(\"Labels loaded: \", len(y))\n",
        "\n",
        "print(y[0], imagepaths[0]) # Debugging"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "trainX, trainy=X,y\n",
        "X = expand_dims(trainX, axis=-1)\n",
        "\t# convert from ints to floats\n",
        "X = X.astype('float32')\n",
        "\t# scale from [0,255] to [-1,1]\n",
        "X = (X - 127.5) / 127.5\n",
        "print(X.shape, trainy.shape)\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jdiODsWTKvtA",
        "outputId": "bea55d21-11af-4a56-9756-182bb2215839"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(1005, 28, 28, 1) (1005,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Aq5QqS_4efg_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 578
        },
        "outputId": "f3d6e836-473c-4861-8a29-d489e6f99b2a"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 25 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# We plot first 25 images of training dataset\n",
        "plt.figure(figsize =(10, 10))\n",
        "for i in range(25):\n",
        "    plt.subplot(5, 5, i + 1)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.grid(False)\n",
        "    plt.imshow(x_train[i], cmap = plt.cm.binary)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "def define_discriminator(in_shape=(28,28,1), n_classes=5):\n",
        "\n",
        "\tinit = RandomNormal(stddev=0.02)\n",
        "\tin_image = Input(shape=in_shape)\n",
        "\t# downsample to 14x14\n",
        "\tfe = Conv2D(32, (3,3), strides=(2,2), padding='same', kernel_initializer=init)(in_image)\n",
        "\tfe = LeakyReLU(alpha=0.2)(fe)\n",
        "\tfe = Dropout(0.5)(fe)\n",
        "\t# normal\n",
        "\tfe = Conv2D(64, (3,3), padding='same', kernel_initializer=init)(fe)\n",
        "\tfe = BatchNormalization()(fe)\n",
        "\tfe = LeakyReLU(alpha=0.2)(fe)\n",
        "\tfe = Dropout(0.5)(fe)\n",
        "\t# downsample to 7x7\n",
        "\tfe = Conv2D(128, (3,3), strides=(2,2), padding='same', kernel_initializer=init)(fe)\n",
        "\tfe = BatchNormalization()(fe)\n",
        "\tfe = LeakyReLU(alpha=0.2)(fe)\n",
        "\tfe = Dropout(0.5)(fe)\n",
        "\t# normal\n",
        "\tfe = Conv2D(256, (3,3), padding='same', kernel_initializer=init)(fe)\n",
        "\tfe = BatchNormalization()(fe)\n",
        "\tfe = LeakyReLU(alpha=0.2)(fe)\n",
        "\tfe = Dropout(0.5)(fe)\n",
        "\tfe = Flatten()(fe)\n",
        "\t# real/fake output\n",
        "\tout1 = Dense(1, activation='sigmoid')(fe)\n",
        "\t# class label output\n",
        "\tout2 = Dense(n_classes, activation='softmax')(fe)\n",
        "\tmodel = Model(in_image, [out1, out2])\n",
        "\topt = Adam(learning_rate=0.0002, beta_1=0.5)\n",
        "\tmodel.compile(loss=['binary_crossentropy', 'sparse_categorical_crossentropy'], optimizer=opt, metrics=['binary_accuracy'])\n",
        "\treturn model\n"
      ],
      "metadata": {
        "id": "OeLmEKKWgxem"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "def define_generator(latent_dim, n_classes=10):\n",
        "\tinit = RandomNormal(stddev=0.02)\n",
        "\tin_label = Input(shape=(1,))\n",
        "\t# embedding for categorical input\n",
        "\tli = Embedding(n_classes, 50)(in_label)\n",
        "\tn_nodes = 7 * 7\n",
        "\tli = Dense(n_nodes, kernel_initializer=init)(li)\n",
        "\t# reshape to additional channel\n",
        "\tli = Reshape((7, 7, 1))(li)\n",
        "\tin_lat = Input(shape=(latent_dim,))\n",
        "\t# foundation for 7x7 image\n",
        "\tn_nodes = 384 * 7 * 7\n",
        "\tgen = Dense(n_nodes, kernel_initializer=init)(in_lat)\n",
        "\tgen = Activation('relu')(gen)\n",
        "\tgen = Reshape((7, 7, 384))(gen)\n",
        "\t# merge image gen and label input\n",
        "\tmerge = Concatenate()([gen, li])\n",
        "\t# upsample to 14x14\n",
        "\tgen = Conv2DTranspose(192, (5,5), strides=(2,2), padding='same', kernel_initializer=init)(merge)\n",
        "\tgen = BatchNormalization()(gen)\n",
        "\tgen = Activation('relu')(gen)\n",
        "\t# upsample to 28x28\n",
        "\tgen = Conv2DTranspose(1, (5,5), strides=(2,2), padding='same', kernel_initializer=init)(gen)\n",
        "\tout_layer = Activation('tanh')(gen)\n",
        "\tmodel = Model([in_lat, in_label], out_layer)\n",
        "\treturn model\n"
      ],
      "metadata": {
        "id": "p_vZxDOtg-6s"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# ترکیب مولد و تفکیک کننده\n",
        "def define_gan(g_model, d_model):\n",
        "\tfor layer in d_model.layers:\n",
        "\t\tif not isinstance(layer, BatchNormalization):\n",
        "\t\t\tlayer.trainable = False\n",
        "\tgan_output = d_model(g_model.output)\n",
        "\t# define gan model as taking noise and label and outputting real/fake and label outputs\n",
        "\tmodel = Model(g_model.input, gan_output)\n",
        "\topt = Adam(learning_rate=0.0002, beta_1=0.5)\n",
        "\tmodel.compile(loss=['binary_crossentropy', 'sparse_categorical_crossentropy'], optimizer=opt, metrics=['binary_accuracy'])\n",
        "\treturn model\n"
      ],
      "metadata": {
        "id": "Vj4bm9fRhHjs"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from numpy.random import random\n",
        "def generate_real_samples(dataset, n_samples):\n",
        "\timages, labels = dataset\n",
        "\t\n",
        "\tix = randint(0, images.shape[0], n_samples)\n",
        "\tX, labels = images[ix], labels[ix]\n",
        "\ty = ones((n_samples, 1))\n",
        "\ty_smooth=y - 0.2 + (random(y.shape) * 0.3)\n",
        "\treturn [X, labels], y_smooth"
      ],
      "metadata": {
        "id": "1GW42r9IiabH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def generate_latent_points(latent_dim, n_samples, n_classes=5):\n",
        "\t# generate points in the latent space\n",
        "\tx_input = randn(latent_dim * n_samples)\n",
        "\tz_input = x_input.reshape(n_samples, latent_dim)\n",
        "\tlabels = randint(0, n_classes, n_samples)\n",
        "\treturn [z_input, labels]\n"
      ],
      "metadata": {
        "id": "P1Ou-Vuqigl_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def generate_fake_samples(generator, latent_dim, n_samples):\n",
        "\t# generate points in latent space\n",
        "\tz_input, labels_input = generate_latent_points(latent_dim, n_samples)\n",
        "\timages = generator.predict([z_input, labels_input])\n",
        "\ty = zeros((n_samples, 1))\n",
        "\treturn [images, labels_input], y"
      ],
      "metadata": {
        "id": "yXRY_kpcisah"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def summarize_performance(step, g_model, latent_dim, n_samples=100):\n",
        "\t[X, _], _ = generate_fake_samples(g_model, latent_dim, n_samples)\n",
        "\tX = (X + 1) / 2.0\n",
        "\tfor i in range(100):\n",
        "\n",
        "\t\tpyplot.subplot(10, 10, 1 + i)\n",
        "\t\tpyplot.axis('off')\n",
        "\t\tpyplot.imshow(X[i, :, :, 0], cmap='gray_r')\n",
        "\tfilename1 = 'generated_plot_%04d.png' % (step+1)\n",
        "\tpyplot.savefig(\"mydrive/My Drive/improved/\"+filename1)\n",
        "\tpyplot.close()\n",
        "\tfilename2 = 'model_%04d.h5' % (step+1)\n",
        "\tg_model.save(\"mydrive/My Drive/improved/\"+filename2)\n",
        "\tprint('>Saved: %s and %s' % (filename1, filename2))\n"
      ],
      "metadata": {
        "id": "tgsgm9Bni1Ag"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def train(g_model, d_model, gan_model, dataset, latent_dim, n_epochs=50, n_batch=64):\n",
        "\td_loss=[]\n",
        "\td_loss_labels=[]\n",
        "\td_acc=[]\n",
        "\tg_loss=[]\n",
        "\tg_acc=[]\n",
        "\td_acc_labels=[]\n",
        "\tbat_per_epo = int(dataset[0].shape[0] / n_batch)\n",
        "\t# calculate the number of training iterations\n",
        "\tn_steps = bat_per_epo * n_epochs\n",
        "\thalf_batch = int(n_batch / 2)\n",
        "\tfor i in range(n_steps):\n",
        "\n",
        "\t\t[X_real, labels_real], y_real = generate_real_samples(dataset, half_batch)\n",
        "\t\t# آموزش مدل تفکیک کننده با داده های واقعی نویزی\n",
        "\t\t_,d_r1,d_r2,d_acc_r1,d_acc_r2 = d_model.train_on_batch(X_real, [noisy_labels(0.05,y_real), labels_real])\n",
        "\t\n",
        "\t\t[X_fake, labels_fake], y_fake = generate_fake_samples(g_model, latent_dim, half_batch)\n",
        "\t\t#   آموزش مدل تفکیک کننده با داده های فیک نویزی\n",
        "\t\t_,d_f,d_f2,d_acc_f1,d_acc_f2 = d_model.train_on_batch(X_fake, [noisy_labels(0.05,y_fake), labels_fake])\n",
        "\t\td_loss.append((d_f+d_r1)/2)\n",
        "\t\td_loss_labels.append((d_f2+d_r2)/2)\n",
        "\t\td_acc.append((d_acc_f1+d_acc_r1)/2)\n",
        "\t\td_acc_labels.append((d_acc_f2+d_acc_r2)/2)\n",
        "\t\t[z_input, z_labels] = generate_latent_points(latent_dim, n_batch)\n",
        "\t\ty_gan = ones((n_batch, 1))\n",
        "\t\t# update the generator via the discriminator's error\n",
        "\t\t_,g_1,g_2,g_acc_r1,g_acc_r2 = gan_model.train_on_batch([z_input, z_labels], [y_gan, z_labels])\n",
        "\t\tg_loss.append((g_1+g_2)/2)\n",
        "\t\tg_acc.append((g_acc_r1+g_acc_r2)/2)\n",
        "\t\t# summarize loss on this batch\n",
        "\t\t\n",
        "\t\t# evaluate the model performance every 'epoch'\n",
        "\t\tif (i+1) % (bat_per_epo * 10) == 0:\n",
        "\t\t\tprint('>%d, dr[%.3f,%.3f], df[%.3f,%.3f], g[%.3f,%.3f]' % (i+1, d_r1,d_r2, d_f,d_f2, g_1,g_2))\n",
        "\t\t\tprint(g_acc_r1,g_acc_r2)\n",
        "\t\t\tsummarize_performance(i, g_model, latent_dim)\n",
        "\treturn d_loss,d_loss_labels,d_acc,d_acc_labels,g_acc,g_loss\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "Aao4UUyIFBtA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "from numpy.random import choice\n",
        "def noisy_labels(p_flip, y):\n",
        " # determine the number of labels to flip\n",
        "  n_select = int(p_flip * len(y))\n",
        "  flip_ix = choice([i for i in range(len(y))], size=n_select)\n",
        "  for i in flip_ix:\n",
        "    y[i][0]= 1 - y[i][0]\n",
        "    if(y[i][0]<0.1):\n",
        "      y[i][0]=0\n",
        "\t\t\n",
        "  return y"
      ],
      "metadata": {
        "id": "iA50n6fBKMaJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "latent_dim = 100\n",
        "discriminator = define_discriminator()\n",
        "generator = define_generator(latent_dim)\n",
        "gan_model = define_gan(generator, discriminator)\n",
        "dataset = [X, trainy]\n",
        "# train\n",
        "d_loss,d_loss_labels,d_acc,d_acc_labels,g_acc,g_loss=train(generator, discriminator, gan_model, dataset, latent_dim)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yIXd49P6YcQ1",
        "outputId": "1838b6b1-c482-4023-a396-51e218e1fea5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 353ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            ">150, dr[0.260,0.069], df[0.088,0.159], g[0.160,0.168]\n",
            "0.953125 0.20000000298023224\n",
            "4/4 [==============================] - 0s 34ms/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "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"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved: generated_plot_0150.png and model_0150.h5\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            ">300, dr[0.506,0.054], df[0.253,0.003], g[1.691,0.030]\n",
            "0.21875 0.22187499701976776\n",
            "4/4 [==============================] - 0s 9ms/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "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"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved: generated_plot_0300.png and model_0300.h5\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 75ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 72ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 109ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            ">450, dr[0.908,0.032], df[0.860,0.248], g[2.067,0.159]\n",
            "0.125 0.18125000596046448\n",
            "4/4 [==============================] - 0s 12ms/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "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"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved: generated_plot_0450.png and model_0450.h5\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            ">600, dr[0.877,0.092], df[0.582,0.081], g[1.557,0.142]\n",
            "0.234375 0.23749999701976776\n",
            "4/4 [==============================] - 0s 9ms/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "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"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved: generated_plot_0600.png and model_0600.h5\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 17ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 16ms/step\n",
            "1/1 [==============================] - 0s 12ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            "1/1 [==============================] - 0s 14ms/step\n",
            "1/1 [==============================] - 0s 18ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 13ms/step\n",
            ">750, dr[0.755,0.009], df[0.632,0.072], g[1.759,0.038]\n",
            "0.09375 0.21250002086162567\n",
            "4/4 [==============================] - 0s 9ms/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "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"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved: generated_plot_0750.png and model_0750.h5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(d_loss)\n",
        "plt.plot(d_loss_labels)\n",
        "plt.title('Discriminator improved- loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('iteration')\n",
        "plt.legend(['Class loss', 'label(fake or real) loss'], loc='center right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-5n21tXhWjRj",
        "outputId": "2b0a64fd-aba8-4fd6-ea53-106098d10855"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(d_acc)\n",
        "plt.plot(d_acc_labels)\n",
        "plt.title('Discriminator improved- Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('iteration')\n",
        "plt.legend(['Class accuracy', 'label(fake or real) accuracy'], loc='center right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0U4hZOeI6hcL",
        "outputId": "9202b16b-3e44-418e-89e9-1ad88dbe6481"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(g_acc)\n",
        "plt.plot(g_loss)\n",
        "plt.title('Generator improved ')\n",
        "plt.ylabel(' ')\n",
        "plt.xlabel('iteration')\n",
        "plt.legend(['Accuracy', 'loss'], loc='center right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "id": "SXa4K0Ys6h7b",
        "outputId": "04de046f-6932-49a2-de5d-707798242df6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard",
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}