nb_ex6_1_ae_fc_mnist-cpu.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6장. 케라스2로 구현하는 AE(오토인코더)\n",
    "## 6.2 완전 연결 계층을 이용한 AE 구현\n",
    "필기체 숫자 즉 MNIST를 이용해 AE을 구현"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set to use CPU\n",
    "import os\n",
    "os.environ['CUDA_VISIBLE_DEVICES'] = '-1'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.2.1 완전 연결 계층 AE 모델링\n",
    "1. AE를 위한 딥러닝 모델을 만듭니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import layers, models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AE(models.Model):\n",
    "    def __init__(self, x_nodes=784, z_dim=36):\n",
    "        x_shape = (x_nodes,)\n",
    "        x = layers.Input(shape=x_shape)\n",
    "        z = layers.Dense(z_dim, activation='relu')(x)\n",
    "        y = layers.Dense(x_nodes, activation='sigmoid')(z)\n",
    "\n",
    "        super().__init__(x, y)\n",
    "\n",
    "        self.x = x\n",
    "        self.z = z\n",
    "        self.z_dim = z_dim\n",
    "\n",
    "        # Encoder, Decoder ??\n",
    "\n",
    "        self.compile(optimizer='adadelta', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "    def Encoder(self):\n",
    "        return models.Model(self.x, self.z)\n",
    "\n",
    "    def Decoder(self):\n",
    "        z_shape = (self.z_dim,)\n",
    "        z = layers.Input(shape=z_shape)\n",
    "        y_layer = self.layers[-1]\n",
    "        y = y_layer(z)\n",
    "        return models.Model(z, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.2.2 데이터 준비 \n",
    "2. 사용할 MNIST 데이터를 케라스의 서브패키지로부터 불러옵니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 784)\n",
      "(10000, 784)\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import mnist\n",
    "import numpy as np\n",
    "(X_train, _), (X_test, _) = mnist.load_data()\n",
    "\n",
    "X_train = X_train.astype('float32') / 255.\n",
    "X_test = X_test.astype('float32') / 255.\n",
    "X_train = X_train.reshape((len(X_train), np.prod(X_train.shape[1:])))\n",
    "X_test = X_test.reshape((len(X_test), np.prod(X_test.shape[1:])))\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.2.3 학습 효과 분석 \n",
    "3. 학습 효과를 그래프로 분석하기 위해 필요한 함수들과 패키지를 불러옵니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keraspp.skeras import plot_loss, plot_acc\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.2.4 완전 연결 계층 AE 동작 확인 \n",
    "4. 완전 연결 계층 AE가 어떻게 동작했는지 확인합시다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_ae(autoencoder):\n",
    "    encoder = autoencoder.Encoder()\n",
    "    decoder = autoencoder.Decoder()\n",
    "\n",
    "    encoded_imgs = encoder.predict(X_test)\n",
    "    decoded_imgs = decoder.predict(encoded_imgs)\n",
    "\n",
    "    n = 10\n",
    "    plt.figure(figsize=(20, 6))\n",
    "    for i in range(n):\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1)\n",
    "        plt.imshow(X_test[i].reshape(28, 28))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1 + n)\n",
    "        plt.stem(encoded_imgs[i].reshape(-1))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1 + n + n)\n",
    "        plt.imshow(decoded_imgs[i].reshape(28, 28))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.2.5 학습 및 성능 평가 \n",
    "5. 학습과 성능 평가에 사용할 main() 함수를 만듭니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def main():\n",
    "    x_nodes = 784\n",
    "    z_dim = 36\n",
    "\n",
    "    autoencoder = AE(x_nodes, z_dim)\n",
    "\n",
    "    history = autoencoder.fit(X_train, X_train,\n",
    "                              epochs=10,\n",
    "                              batch_size=256,\n",
    "                              shuffle=True,\n",
    "                              validation_data=(X_test, X_test))\n",
    "\n",
    "    plot_acc(history, 'Accuracy')\n",
    "    plt.show()\n",
    "    plot_loss(history, 'Loss')\n",
    "    plt.show()\n",
    "\n",
    "    show_ae(autoencoder)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-10-21 21:35:54.001437: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1\n",
      "2021-10-21 21:35:54.054257: E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected\n",
      "2021-10-21 21:35:54.054453: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (SJKIM-Home): /proc/driver/nvidia/version does not exist\n",
      "2021-10-21 21:35:54.056091: I tensorflow/core/platform/cpu_feature_guard.cc:143] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA\n",
      "2021-10-21 21:35:54.070529: I tensorflow/core/platform/profile_utils/cpu_utils.cc:102] CPU Frequency: 3599995000 Hz\n",
      "2021-10-21 21:35:54.072769: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55ddcd8c28c0 initialized for platform Host (this does not guarantee that XLA will be used). Devices:\n",
      "2021-10-21 21:35:54.072830: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6957 - accuracy: 0.0011 - val_loss: 0.6957 - val_accuracy: 0.0014\n",
      "Epoch 2/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6955 - accuracy: 0.0011 - val_loss: 0.6955 - val_accuracy: 0.0015\n",
      "Epoch 3/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6952 - accuracy: 0.0011 - val_loss: 0.6953 - val_accuracy: 0.0015\n",
      "Epoch 4/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6950 - accuracy: 0.0011 - val_loss: 0.6950 - val_accuracy: 0.0016\n",
      "Epoch 5/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6948 - accuracy: 0.0011 - val_loss: 0.6948 - val_accuracy: 0.0015\n",
      "Epoch 6/10\n",
      "235/235 [==============================] - 2s 8ms/step - loss: 0.6946 - accuracy: 0.0011 - val_loss: 0.6946 - val_accuracy: 0.0015\n",
      "Epoch 7/10\n",
      "235/235 [==============================] - 2s 7ms/step - loss: 0.6944 - accuracy: 0.0011 - val_loss: 0.6944 - val_accuracy: 0.0015\n",
      "Epoch 8/10\n",
      "235/235 [==============================] - 2s 7ms/step - loss: 0.6942 - accuracy: 0.0010 - val_loss: 0.6942 - val_accuracy: 0.0014\n",
      "Epoch 9/10\n",
      "235/235 [==============================] - 2s 7ms/step - loss: 0.6940 - accuracy: 0.0010 - val_loss: 0.6940 - val_accuracy: 0.0014\n",
      "Epoch 10/10\n",
      "235/235 [==============================] - 2s 7ms/step - loss: 0.6938 - accuracy: 0.0010 - val_loss: 0.6938 - val_accuracy: 0.0014\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54617 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49845 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44221 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44284 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 50640 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 46384 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 47480 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 51221 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54869 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 46020 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 48320 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54868 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 52628 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 51060 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54617 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49845 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44221 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44284 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 50640 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 46384 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 47480 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 51221 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54869 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 46020 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 48320 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54868 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 52628 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 51060 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49552 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49892 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44050 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49552 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49892 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44050 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "main()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "### 6.2.6 전체 코드"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 784)\n",
      "(10000, 784)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-10-21 21:47:42.248122: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1\n",
      "2021-10-21 21:47:42.329818: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.330309: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0 with properties: \n",
      "pciBusID: 0000:01:00.0 name: NVIDIA GeForce GTX 1050 computeCapability: 6.1\n",
      "coreClock: 1.455GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s\n",
      "2021-10-21 21:47:42.334011: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1\n",
      "2021-10-21 21:47:42.392417: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10\n",
      "2021-10-21 21:47:42.419715: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10\n",
      "2021-10-21 21:47:42.427511: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10\n",
      "2021-10-21 21:47:42.470964: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10\n",
      "2021-10-21 21:47:42.477217: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10\n",
      "2021-10-21 21:47:42.556294: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7\n",
      "2021-10-21 21:47:42.557203: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.558098: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.558390: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1703] Adding visible gpu devices: 0\n",
      "2021-10-21 21:47:42.559494: I tensorflow/core/platform/cpu_feature_guard.cc:143] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA\n",
      "2021-10-21 21:47:42.569907: I tensorflow/core/platform/profile_utils/cpu_utils.cc:102] CPU Frequency: 3599995000 Hz\n",
      "2021-10-21 21:47:42.572167: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5588282fb630 initialized for platform Host (this does not guarantee that XLA will be used). Devices:\n",
      "2021-10-21 21:47:42.572187: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version\n",
      "2021-10-21 21:47:42.573383: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.573621: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0 with properties: \n",
      "pciBusID: 0000:01:00.0 name: NVIDIA GeForce GTX 1050 computeCapability: 6.1\n",
      "coreClock: 1.455GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s\n",
      "2021-10-21 21:47:42.573698: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1\n",
      "2021-10-21 21:47:42.573725: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10\n",
      "2021-10-21 21:47:42.573747: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10\n",
      "2021-10-21 21:47:42.573767: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10\n",
      "2021-10-21 21:47:42.573786: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10\n",
      "2021-10-21 21:47:42.573805: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10\n",
      "2021-10-21 21:47:42.573825: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7\n",
      "2021-10-21 21:47:42.574548: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.575491: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.575680: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1703] Adding visible gpu devices: 0\n",
      "2021-10-21 21:47:42.576353: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1\n",
      "2021-10-21 21:47:42.801974: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
      "2021-10-21 21:47:42.802001: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1108]      0 \n",
      "2021-10-21 21:47:42.802094: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1121] 0:   N \n",
      "2021-10-21 21:47:42.803042: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.803144: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1330] Could not identify NUMA node of platform GPU id 0, defaulting to 0.  Your kernel may not have been built with NUMA support.\n",
      "2021-10-21 21:47:42.803762: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.804518: E tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node\n",
      "Your kernel may have been built without NUMA support.\n",
      "2021-10-21 21:47:42.805005: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1247] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 1421 MB memory) -> physical GPU (device: 0, name: NVIDIA GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1)\n",
      "2021-10-21 21:47:42.809854: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55882c1e1e10 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2021-10-21 21:47:42.809879: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): NVIDIA GeForce GTX 1050, Compute Capability 6.1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-10-21 21:47:46.051815: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6939 - accuracy: 8.5000e-04 - val_loss: 0.6938 - val_accuracy: 6.0000e-04\n",
      "Epoch 2/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6937 - accuracy: 8.8333e-04 - val_loss: 0.6936 - val_accuracy: 6.0000e-04\n",
      "Epoch 3/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6935 - accuracy: 9.1667e-04 - val_loss: 0.6934 - val_accuracy: 6.0000e-04\n",
      "Epoch 4/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6933 - accuracy: 9.5000e-04 - val_loss: 0.6933 - val_accuracy: 6.0000e-04\n",
      "Epoch 5/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6931 - accuracy: 0.0010 - val_loss: 0.6931 - val_accuracy: 6.0000e-04\n",
      "Epoch 6/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6930 - accuracy: 0.0011 - val_loss: 0.6929 - val_accuracy: 7.0000e-04\n",
      "Epoch 7/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6928 - accuracy: 0.0011 - val_loss: 0.6927 - val_accuracy: 7.0000e-04\n",
      "Epoch 8/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6926 - accuracy: 0.0011 - val_loss: 0.6925 - val_accuracy: 7.0000e-04\n",
      "Epoch 9/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6924 - accuracy: 0.0012 - val_loss: 0.6923 - val_accuracy: 7.0000e-04\n",
      "Epoch 10/10\n",
      "235/235 [==============================] - 1s 6ms/step - loss: 0.6922 - accuracy: 0.0012 - val_loss: 0.6922 - val_accuracy: 7.0000e-04\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54617 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49845 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44221 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44284 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 50640 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 46384 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 47480 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 51221 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54869 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 46020 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 48320 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 54868 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 52628 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 51060 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54617 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49845 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44221 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44284 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 50640 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 46384 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 47480 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 51221 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54869 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 46020 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 48320 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 54868 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 52628 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 51060 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49552 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 49892 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 44050 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49552 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 49892 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n",
      "/home/sjkim/anaconda3/envs/keras-gpu/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 44050 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "###########################\n",
    "# AE 모델링\n",
    "###########################\n",
    "from keras import layers, models  # (Input, Dense), (Model)\n",
    "\n",
    "class AE(models.Model):\n",
    "    def __init__(self, x_nodes=784, z_dim=36):\n",
    "        x_shape = (x_nodes,)\n",
    "        x = layers.Input(shape=x_shape)\n",
    "        z = layers.Dense(z_dim, activation='relu')(x)\n",
    "        y = layers.Dense(x_nodes, activation='sigmoid')(z)\n",
    "\n",
    "        super().__init__(x, y)\n",
    "\n",
    "        self.x = x\n",
    "        self.z = z\n",
    "        self.z_dim = z_dim\n",
    "\n",
    "        # Encoder, Decoder ??\n",
    "\n",
    "        self.compile(optimizer='adadelta', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "    def Encoder(self):\n",
    "        return models.Model(self.x, self.z)\n",
    "\n",
    "    def Decoder(self):\n",
    "        z_shape = (self.z_dim,)\n",
    "        z = layers.Input(shape=z_shape)\n",
    "        y_layer = self.layers[-1]\n",
    "        y = y_layer(z)\n",
    "        return models.Model(z, y)\n",
    "\n",
    "\n",
    "###########################\n",
    "# 데이터 준비\n",
    "###########################\n",
    "from keras.datasets import mnist\n",
    "import numpy as np\n",
    "(X_train, _), (X_test, _) = mnist.load_data()\n",
    "\n",
    "X_train = X_train.astype('float32') / 255.\n",
    "X_test = X_test.astype('float32') / 255.\n",
    "X_train = X_train.reshape((len(X_train), np.prod(X_train.shape[1:])))\n",
    "X_test = X_test.reshape((len(X_test), np.prod(X_test.shape[1:])))\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "\n",
    "\n",
    "###########################\n",
    "# 학습 효과 분석\n",
    "###########################\n",
    "from keraspp.skeras import plot_loss, plot_acc\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "###########################\n",
    "# AE 동작 확인\n",
    "###########################\n",
    "def show_ae(autoencoder):\n",
    "    encoder = autoencoder.Encoder()\n",
    "    decoder = autoencoder.Decoder()\n",
    "\n",
    "    encoded_imgs = encoder.predict(X_test)\n",
    "    decoded_imgs = decoder.predict(encoded_imgs)\n",
    "\n",
    "    n = 10\n",
    "    plt.figure(figsize=(20, 6))\n",
    "    for i in range(n):\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1)\n",
    "        plt.imshow(X_test[i].reshape(28, 28))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1 + n)\n",
    "        plt.stem(encoded_imgs[i].reshape(-1))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "        ax = plt.subplot(3, n, i + 1 + n + n)\n",
    "        plt.imshow(decoded_imgs[i].reshape(28, 28))\n",
    "        plt.gray()\n",
    "        ax.get_xaxis().set_visible(False)\n",
    "        ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "###########################\n",
    "# 학습\n",
    "###########################\n",
    "def main():\n",
    "    x_nodes = 784\n",
    "    z_dim = 36\n",
    "\n",
    "    autoencoder = AE(x_nodes, z_dim)\n",
    "\n",
    "    history = autoencoder.fit(X_train, X_train,\n",
    "                              epochs=10,\n",
    "                              batch_size=256,\n",
    "                              shuffle=True,\n",
    "                              validation_data=(X_test, X_test))\n",
    "\n",
    "    plot_acc(history, '(a) 학습 경과에 따른 정확도 변화 추이')\n",
    "    plt.show()\n",
    "    plot_loss(history, '(b) 학습 경과에 따른 손실값 변화 추이')\n",
    "    plt.show()\n",
    "\n",
    "    show_ae(autoencoder)\n",
    "    plt.show()\n",
    "\n",
    "main()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "keras-gpu",
   "language": "python",
   "name": "keras-gpu"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}