ch17/ch17_part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9y222HgfM-OA"
   },
   "source": [
    "# 머신 러닝 교과서 3판"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-5SXCZX_M-OD"
   },
   "source": [
    "# 17장 - 새로운 데이터 합성을 위한 생성적 적대 신경망 (1/2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "hx1MAB0FbouH"
   },
   "source": [
    "**아래 링크를 통해 이 노트북을 주피터 노트북 뷰어(nbviewer.jupyter.org)로 보거나 구글 코랩(colab.research.google.com)에서 실행할 수 있습니다.**\n",
    "\n",
    "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
    "  <td>\n",
    "    <a target=\"_blank\" href=\"https://nbviewer.jupyter.org/github/rickiepark/python-machine-learning-book-3rd-edition/blob/master/ch17/ch17_part1.ipynb\"><img src=\"https://jupyter.org/assets/main-logo.svg\" width=\"28\" />주피터 노트북 뷰어로 보기</a>\n",
    "  </td>\n",
    "  <td>\n",
    "    <a target=\"_blank\" href=\"https://colab.research.google.com/github/rickiepark/python-machine-learning-book-3rd-edition/blob/master/ch17/ch17_part1.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />구글 코랩(Colab)에서 실행하기</a>\n",
    "  </td>\n",
    "</table>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4wzfKSEhbouH"
   },
   "source": [
    "### 목차"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1vc8_zr4bouH"
   },
   "source": [
    "- 생성적 적대 신경망 소개\n",
    "    - 오토인코더\n",
    "    - 새로운 데이터 합성을 위한 생성 모델\n",
    "    - GAN으로 새로운 샘플 생성하기\n",
    "    - GAN의 생성자와 판별자 손실 함수 이해하기\n",
    "- 밑바닥부터 GAN 모델 구현하기\n",
    "    - 구글 코랩에서 GAN 모델 훈련하기\n",
    "    - 생성자와 판별자 신경망 구현하기\n",
    "    - 훈련 데이터셋 정의하기\n",
    "    - GAN 모델 훈련하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "2AfJgihdM-OE"
   },
   "outputs": [],
   "source": [
    "from IPython.display import Image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "41bk3eM5M-OE"
   },
   "source": [
    "# 생성적 적대 신경망 소개\n",
    "\n",
    "## 오토인코더"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 470
    },
    "id": "phbH6zGQM-OE",
    "outputId": "b1d09579-bc30-41c7-f2a2-31884ea1b97e"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQ2\" width=\"500\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 2,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQ2', width=500)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Cuw0DGR2M-OF"
   },
   "source": [
    "## 새로운 데이터 합성을 위한 생성 모델"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 437
    },
    "id": "ziTpZ3G5M-OF",
    "outputId": "df8f3c08-9031-4df8-80ab-23eb751b8701"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQH\" width=\"700\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 3,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQH', width=700)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "U1DsTJvLM-OF"
   },
   "source": [
    "## GAN으로 새로운 샘플 생성하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 337
    },
    "id": "2DijKimVM-OF",
    "outputId": "f2f95fce-9401-4adf-8c88-7e71294b4f6f"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQ7\" width=\"700\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 4,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQ7', width=700)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iamOayQfM-OF"
   },
   "source": [
    "## GAN의 생성자와 판별자 손실 함수 이해하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 516
    },
    "id": "P0SEKNGpM-OG",
    "outputId": "adeb9885-752b-4d78-fa55-e9f131409444"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQF\" width=\"800\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQF', width=800)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mgleEEF1M-OG"
   },
   "source": [
    "# 밑바닥부터 GAN 모델 구현하기"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "a1UBoKm_M-OG"
   },
   "source": [
    "## 구글 코랩에서 GAN 모델 훈련하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 507
    },
    "id": "sBdI0g22M-OG",
    "outputId": "f83c926f-ab99-45c5-ec4f-57acdaec9f9a"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQb\" width=\"700\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 6,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQb', width=700)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 526
    },
    "id": "Mf0pXqraM-OH",
    "outputId": "3e72f9d2-9668-4c09-f200-766994028877"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQN\" width=\"600\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 7,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQN', width=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 533
    },
    "id": "Nd67sNNIM-OH",
    "outputId": "201db3ea-722d-4c50-b686-5443a433a6c9"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQA\" width=\"600\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 8,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQA', width=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "_khN3hi1xxQ-",
    "outputId": "0720637b-1802-433b-cdb9-f2a3eafc8caa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.4.0\n",
      "GPU 여부: True\n",
      "/device:GPU:0\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "print(tf.__version__)\n",
    "\n",
    "print(\"GPU 여부:\", len(tf.config.list_physical_devices('GPU')) > 0)\n",
    "\n",
    "if tf.config.list_physical_devices('GPU'):\n",
    "    device_name = tf.test.gpu_device_name()\n",
    "else:\n",
    "    device_name = 'cpu:0'\n",
    "    \n",
    "print(device_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ceNFHOHHNVwz",
    "outputId": "f7d4997b-ea54-41f5-ace1-5d810f4e86b0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 10,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.config.experimental.get_memory_usage('GPU:0')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5sNcfuwHPQMs",
    "outputId": "ff3dfa5d-73c7-412c-aed8-ab6c950e3df3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Thu Jan  7 06:00:50 2021       \n",
      "+-----------------------------------------------------------------------------+\n",
      "| NVIDIA-SMI 460.27.04    Driver Version: 418.67       CUDA Version: 10.1     |\n",
      "|-------------------------------+----------------------+----------------------+\n",
      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
      "|                               |                      |               MIG M. |\n",
      "|===============================+======================+======================|\n",
      "|   0  Tesla P4            Off  | 00000000:00:04.0 Off |                    0 |\n",
      "| N/A   37C    P0    23W /  75W |    209MiB /  7611MiB |      0%      Default |\n",
      "|                               |                      |                 ERR! |\n",
      "+-------------------------------+----------------------+----------------------+\n",
      "                                                                               \n",
      "+-----------------------------------------------------------------------------+\n",
      "| Processes:                                                                  |\n",
      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
      "|        ID   ID                                                   Usage      |\n",
      "|=============================================================================|\n",
      "|  No running processes found                                                 |\n",
      "+-----------------------------------------------------------------------------+\n"
     ]
    }
   ],
   "source": [
    "!nvidia-smi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "4_yH3L4wTOLb",
    "outputId": "86aface1-6da2-4297-8b63-76273e4c96bc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No LSB modules are available.\n",
      "Distributor ID:\tUbuntu\n",
      "Description:\tUbuntu 18.04.5 LTS\n",
      "Release:\t18.04\n",
      "Codename:\tbionic\n"
     ]
    }
   ],
   "source": [
    "!lsb_release -a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "ehoy80rqwNnl"
   },
   "outputs": [],
   "source": [
    "#from google.colab import drive\n",
    "#drive.mount('/content/drive/')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jHHcQfM2M-OI"
   },
   "source": [
    "## 생성자와 판별자 신경망 구현하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 473
    },
    "id": "ig_rZtXqM-OI",
    "outputId": "1f16f27c-e6e9-43e6-e0b9-8288bf970091"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQp\" width=\"600\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 14,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQp', width=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 359
    },
    "id": "uc4FQzczM-OJ",
    "outputId": "4ff17041-dee5-48bd-f405-4f1f03a87995"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://git.io/JLAQh\" width=\"600\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 15,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(url='https://git.io/JLAQh', width=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "FRh8E66HwXRL"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_datasets as tfds\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "id": "tOL56FmXwoL-"
   },
   "outputs": [],
   "source": [
    "## 생성자 함수를 정의합니다:\n",
    "def make_generator_network(\n",
    "        num_hidden_layers=1,\n",
    "        num_hidden_units=100,\n",
    "        num_output_units=784):\n",
    "    model = tf.keras.Sequential()\n",
    "    for i in range(num_hidden_layers):\n",
    "        model.add(\n",
    "            tf.keras.layers.Dense(\n",
    "                units=num_hidden_units, \n",
    "                use_bias=False)\n",
    "            )\n",
    "        model.add(tf.keras.layers.LeakyReLU())\n",
    "        \n",
    "    model.add(tf.keras.layers.Dense(\n",
    "        units=num_output_units, activation='tanh'))\n",
    "    return model\n",
    "\n",
    "## 판별자 함수를 정의합니다:\n",
    "def make_discriminator_network(\n",
    "        num_hidden_layers=1,\n",
    "        num_hidden_units=100,\n",
    "        num_output_units=1):\n",
    "    model = tf.keras.Sequential()\n",
    "    for i in range(num_hidden_layers):\n",
    "        model.add(tf.keras.layers.Dense(units=num_hidden_units))\n",
    "        model.add(tf.keras.layers.LeakyReLU())\n",
    "        model.add(tf.keras.layers.Dropout(rate=0.5))\n",
    "        \n",
    "    model.add(\n",
    "        tf.keras.layers.Dense(\n",
    "            units=num_output_units, \n",
    "            activation=None)\n",
    "        )\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "9Z4JQb9MxD5p",
    "outputId": "ec093e68-9514-42af-ab7e-67b75519201e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 100)               2000      \n",
      "_________________________________________________________________\n",
      "leaky_re_lu (LeakyReLU)      (None, 100)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 784)               79184     \n",
      "=================================================================\n",
      "Total params: 81,184\n",
      "Trainable params: 81,184\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "image_size = (28, 28)\n",
    "z_size = 20\n",
    "mode_z = 'uniform'  # 'uniform' vs. 'normal'\n",
    "gen_hidden_layers = 1\n",
    "gen_hidden_size = 100\n",
    "disc_hidden_layers = 1\n",
    "disc_hidden_size = 100\n",
    "\n",
    "tf.random.set_seed(1)\n",
    "\n",
    "gen_model = make_generator_network(\n",
    "    num_hidden_layers=gen_hidden_layers, \n",
    "    num_hidden_units=gen_hidden_size,\n",
    "    num_output_units=np.prod(image_size))\n",
    "\n",
    "gen_model.build(input_shape=(None, z_size))\n",
    "gen_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "QAhbOUZHxN1b",
    "outputId": "64d2f5a3-1a94-4972-ef67-47186391b629"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_2 (Dense)              (None, 100)               78500     \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_1 (LeakyReLU)    (None, 100)               0         \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 100)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 101       \n",
      "=================================================================\n",
      "Total params: 78,601\n",
      "Trainable params: 78,601\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "disc_model = make_discriminator_network(\n",
    "    num_hidden_layers=disc_hidden_layers,\n",
    "    num_hidden_units=disc_hidden_size)\n",
    "\n",
    "disc_model.build(input_shape=(None, np.prod(image_size)))\n",
    "disc_model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "edHUQPQxM-OK"
   },
   "source": [
    "## 훈련 데이터셋 정의하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 290,
     "referenced_widgets": [
      "cb1a8a21d70841b4a2ba27ace07ccb11",
      "1dc2c5d3f42e4e51920b91c3098ba5c5",
      "f11d2daa89a149ad92ad5362f03de0b5",
      "d78d899c3e9c426fb049561894daf335",
      "527d1f8586a449df90bbdd38fd02c29f",
      "3d34dd3e0f09486294c41b372fd63a04",
      "bc37964978a747e883e94134035d847f",
      "bf41d9cedb7548868389eeab2134a90c"
     ]
    },
    "id": "ApQ1ICJixf2w",
    "outputId": "d7bc5247-1b52-4d9d-d224-6f7a59c302b2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mDownloading and preparing dataset mnist/3.0.1 (download: 11.06 MiB, generated: 21.00 MiB, total: 32.06 MiB) to /root/tensorflow_datasets/mnist/3.0.1...\u001b[0m\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Dataset mnist is hosted on GCS. It will automatically be downloaded to your\n",
      "local data directory. If you'd instead prefer to read directly from our public\n",
      "GCS bucket (recommended if you're running on GCP), you can instead pass\n",
      "`try_gcs=True` to `tfds.load` or set `data_dir=gs://tfds-data/datasets`.\n",
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb1a8a21d70841b4a2ba27ace07ccb11",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Dl Completed...', max=4.0, style=ProgressStyle(descriptio…"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1mDataset mnist downloaded and prepared to /root/tensorflow_datasets/mnist/3.0.1. Subsequent calls will reuse this data.\u001b[0m\n",
      "전처리 전:  \n",
      "dtype:  <dtype: 'uint8'>  최소: 0 최대: 255\n",
      "전처리 후:  \n",
      "dtype:  <dtype: 'float32'>  최소: -0.8737728595733643 최대: 0.9460210800170898\n"
     ]
    }
   ],
   "source": [
    "mnist_bldr = tfds.builder('mnist')\n",
    "mnist_bldr.download_and_prepare()\n",
    "mnist = mnist_bldr.as_dataset(shuffle_files=False)\n",
    "\n",
    "def preprocess(ex, mode='uniform'):\n",
    "    image = ex['image']\n",
    "    image = tf.image.convert_image_dtype(image, tf.float32)\n",
    "    image = tf.reshape(image, [-1])\n",
    "    image = image*2 - 1.0\n",
    "    if mode == 'uniform':\n",
    "        input_z = tf.random.uniform(\n",
    "            shape=(z_size,), minval=-1.0, maxval=1.0)\n",
    "    elif mode == 'normal':\n",
    "        input_z = tf.random.normal(shape=(z_size,))\n",
    "    return input_z, image\n",
    "\n",
    "\n",
    "\n",
    "mnist_trainset = mnist['train']\n",
    "\n",
    "print('전처리 전:  ')\n",
    "example = next(iter(mnist_trainset))['image']\n",
    "print('dtype: ', example.dtype, ' 최소: {} 최대: {}'.format(np.min(example), np.max(example)))\n",
    "\n",
    "mnist_trainset = mnist_trainset.map(preprocess)\n",
    "\n",
    "print('전처리 후:  ')\n",
    "example = next(iter(mnist_trainset))[0]\n",
    "print('dtype: ', example.dtype, ' 최소: {} 최대: {}'.format(np.min(example), np.max(example)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HT57CGAz0RDr"
   },
   "source": [
    " * **데이터 흐름을 단계별로 밟아보기**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "kdAXXUtryFGs",
    "outputId": "f9a3aeea-2512-45d1-b3db-7eb22683f0df"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input-z -- 크기: (32, 20)\n",
      "input-real -- 크기: (32, 784)\n",
      "생성자 출력 -- 크기: (32, 784)\n",
      "판별자 (진짜) -- 크기: (32, 1)\n",
      "판별자 (가짜) -- 크기: (32, 1)\n"
     ]
    }
   ],
   "source": [
    "mnist_trainset = mnist_trainset.batch(32, drop_remainder=True)\n",
    "input_z, input_real = next(iter(mnist_trainset))\n",
    "print('input-z -- 크기:', input_z.shape)\n",
    "print('input-real -- 크기:', input_real.shape)\n",
    "\n",
    "g_output = gen_model(input_z)\n",
    "print('생성자 출력 -- 크기:', g_output.shape)\n",
    "\n",
    "d_logits_real = disc_model(input_real)\n",
    "d_logits_fake = disc_model(g_output)\n",
    "print('판별자 (진짜) -- 크기:', d_logits_real.shape)\n",
    "print('판별자 (가짜) -- 크기:', d_logits_fake.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Nb2H-Ac3M-OK"
   },
   "source": [
    "## GAN 모델 훈련하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "-c9xVvjJySZk",
    "outputId": "f868cd67-9c7e-4edb-859a-e3a32b4d32e7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "생성자 손실: 0.7355\n",
      "판별자 손실: Real 1.5073 Fake 0.6583\n"
     ]
    }
   ],
   "source": [
    "loss_fn = tf.keras.losses.BinaryCrossentropy(from_logits=True)\n",
    "\n",
    "## 생성자 손실\n",
    "g_labels_real = tf.ones_like(d_logits_fake)\n",
    "g_loss = loss_fn(y_true=g_labels_real, y_pred=d_logits_fake)\n",
    "print('생성자 손실: {:.4f}'.format(g_loss))\n",
    "\n",
    "## 판별자 손실\n",
    "d_labels_real = tf.ones_like(d_logits_real)\n",
    "d_labels_fake = tf.zeros_like(d_logits_fake)\n",
    "\n",
    "d_loss_real = loss_fn(y_true=d_labels_real, y_pred=d_logits_real)\n",
    "d_loss_fake = loss_fn(y_true=d_labels_fake, y_pred=d_logits_fake)\n",
    "print('판별자 손실: 진짜 {:.4f} 가짜 {:.4f}'\n",
    "      .format(d_loss_real.numpy(), d_loss_fake.numpy()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jtxdTVyF0KCF"
   },
   "source": [
    " * **최종 훈련**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yeFKLGNfAF5J",
    "outputId": "1a7d8f74-c692-41b3-be1b-c6bb6ec0093f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "에포크 001 | 시간 0.42 min | 평균 손실 >> 생성자/판별자 2.8665/0.2895 [판별자-진짜: 0.0334 판별자-가짜: 0.2561]\n",
      "에포크 002 | 시간 0.81 min | 평균 손실 >> 생성자/판별자 5.2182/0.3529 [판별자-진짜: 0.1154 판별자-가짜: 0.2375]\n",
      "에포크 003 | 시간 1.20 min | 평균 손실 >> 생성자/판별자 3.0334/0.7355 [판별자-진짜: 0.3271 판별자-가짜: 0.4084]\n",
      "에포크 004 | 시간 1.60 min | 평균 손실 >> 생성자/판별자 2.2255/0.8048 [판별자-진짜: 0.4230 판별자-가짜: 0.3818]\n",
      "에포크 005 | 시간 1.99 min | 평균 손실 >> 생성자/판별자 2.1613/0.8179 [판별자-진짜: 0.4376 판별자-가짜: 0.3803]\n",
      "에포크 006 | 시간 2.38 min | 평균 손실 >> 생성자/판별자 1.8650/0.8739 [판별자-진짜: 0.4802 판별자-가짜: 0.3936]\n",
      "에포크 007 | 시간 2.76 min | 평균 손실 >> 생성자/판별자 1.5680/0.9967 [판별자-진짜: 0.5377 판별자-가짜: 0.4590]\n",
      "에포크 008 | 시간 3.14 min | 평균 손실 >> 생성자/판별자 1.4271/1.0091 [판별자-진짜: 0.5556 판별자-가짜: 0.4535]\n",
      "에포크 009 | 시간 3.53 min | 평균 손실 >> 생성자/판별자 1.4514/1.0203 [판별자-진짜: 0.5551 판별자-가짜: 0.4652]\n",
      "에포크 010 | 시간 3.92 min | 평균 손실 >> 생성자/판별자 1.3668/1.0605 [판별자-진짜: 0.5688 판별자-가짜: 0.4917]\n",
      "에포크 011 | 시간 4.30 min | 평균 손실 >> 생성자/판별자 1.4133/1.0507 [판별자-진짜: 0.5584 판별자-가짜: 0.4923]\n",
      "에포크 012 | 시간 4.68 min | 평균 손실 >> 생성자/판별자 1.3400/1.0857 [판별자-진짜: 0.5758 판별자-가짜: 0.5098]\n",
      "에포크 013 | 시간 5.07 min | 평균 손실 >> 생성자/판별자 1.2172/1.1498 [판별자-진짜: 0.5989 판별자-가짜: 0.5509]\n",
      "에포크 014 | 시간 5.45 min | 평균 손실 >> 생성자/판별자 1.2716/1.1640 [판별자-진짜: 0.5968 판별자-가짜: 0.5672]\n",
      "에포크 015 | 시간 5.83 min | 평균 손실 >> 생성자/판별자 1.1962/1.1786 [판별자-진짜: 0.6091 판별자-가짜: 0.5695]\n",
      "에포크 016 | 시간 6.22 min | 평균 손실 >> 생성자/판별자 1.1588/1.1767 [판별자-진짜: 0.6054 판별자-가짜: 0.5713]\n",
      "에포크 017 | 시간 6.62 min | 평균 손실 >> 생성자/판별자 1.1514/1.1734 [판별자-진짜: 0.6109 판별자-가짜: 0.5625]\n",
      "에포크 018 | 시간 7.01 min | 평균 손실 >> 생성자/판별자 1.2319/1.1455 [판별자-진짜: 0.5881 판별자-가짜: 0.5574]\n",
      "에포크 019 | 시간 7.39 min | 평균 손실 >> 생성자/판별자 1.1900/1.1483 [판별자-진짜: 0.5962 판별자-가짜: 0.5521]\n",
      "에포크 020 | 시간 7.77 min | 평균 손실 >> 생성자/판별자 1.1225/1.1948 [판별자-진짜: 0.6147 판별자-가짜: 0.5801]\n",
      "에포크 021 | 시간 8.14 min | 평균 손실 >> 생성자/판별자 1.1512/1.2045 [판별자-진짜: 0.6120 판별자-가짜: 0.5925]\n",
      "에포크 022 | 시간 8.52 min | 평균 손실 >> 생성자/판별자 1.0407/1.2206 [판별자-진짜: 0.6249 판별자-가짜: 0.5957]\n",
      "에포크 023 | 시간 8.90 min | 평균 손실 >> 생성자/판별자 1.0844/1.2081 [판별자-진짜: 0.6198 판별자-가짜: 0.5883]\n",
      "에포크 024 | 시간 9.28 min | 평균 손실 >> 생성자/판별자 1.1668/1.1909 [판별자-진짜: 0.6055 판별자-가짜: 0.5855]\n",
      "에포크 025 | 시간 9.65 min | 평균 손실 >> 생성자/판별자 1.0661/1.2190 [판별자-진짜: 0.6226 판별자-가짜: 0.5964]\n",
      "에포크 026 | 시간 10.02 min | 평균 손실 >> 생성자/판별자 1.0513/1.2232 [판별자-진짜: 0.6226 판별자-가짜: 0.6006]\n",
      "에포크 027 | 시간 10.41 min | 평균 손실 >> 생성자/판별자 1.0923/1.2147 [판별자-진짜: 0.6161 판별자-가짜: 0.5986]\n",
      "에포크 028 | 시간 10.78 min | 평균 손실 >> 생성자/판별자 1.0421/1.2480 [판별자-진짜: 0.6305 판별자-가짜: 0.6175]\n",
      "에포크 029 | 시간 11.16 min | 평균 손실 >> 생성자/판별자 1.0610/1.2395 [판별자-진짜: 0.6258 판별자-가짜: 0.6137]\n",
      "에포크 030 | 시간 11.56 min | 평균 손실 >> 생성자/판별자 1.0231/1.2459 [판별자-진짜: 0.6314 판별자-가짜: 0.6145]\n",
      "에포크 031 | 시간 11.98 min | 평균 손실 >> 생성자/판별자 1.0308/1.2368 [판별자-진짜: 0.6287 판별자-가짜: 0.6082]\n",
      "에포크 032 | 시간 12.37 min | 평균 손실 >> 생성자/판별자 1.0272/1.2536 [판별자-진짜: 0.6307 판별자-가짜: 0.6229]\n",
      "에포크 033 | 시간 12.75 min | 평균 손실 >> 생성자/판별자 1.0038/1.2614 [판별자-진짜: 0.6357 판별자-가짜: 0.6257]\n",
      "에포크 034 | 시간 13.14 min | 평균 손실 >> 생성자/판별자 0.9990/1.2590 [판별자-진짜: 0.6374 판별자-가짜: 0.6216]\n",
      "에포크 035 | 시간 13.52 min | 평균 손실 >> 생성자/판별자 1.0228/1.2793 [판별자-진짜: 0.6411 판별자-가짜: 0.6383]\n",
      "에포크 036 | 시간 13.90 min | 평균 손실 >> 생성자/판별자 0.9612/1.2889 [판별자-진짜: 0.6498 판별자-가짜: 0.6391]\n",
      "에포크 037 | 시간 14.28 min | 평균 손실 >> 생성자/판별자 0.9817/1.2900 [판별자-진짜: 0.6484 판별자-가짜: 0.6416]\n",
      "에포크 038 | 시간 14.66 min | 평균 손실 >> 생성자/판별자 0.9961/1.2786 [판별자-진짜: 0.6443 판별자-가짜: 0.6343]\n",
      "에포크 039 | 시간 15.04 min | 평균 손실 >> 생성자/판별자 0.9923/1.2694 [판별자-진짜: 0.6391 판별자-가짜: 0.6303]\n",
      "에포크 040 | 시간 15.42 min | 평균 손실 >> 생성자/판별자 0.9697/1.2814 [판별자-진짜: 0.6474 판별자-가짜: 0.6339]\n",
      "에포크 041 | 시간 15.81 min | 평균 손실 >> 생성자/판별자 1.0157/1.2776 [판별자-진짜: 0.6379 판별자-가짜: 0.6396]\n",
      "에포크 042 | 시간 16.19 min | 평균 손실 >> 생성자/판별자 0.9567/1.2866 [판별자-진짜: 0.6471 판별자-가짜: 0.6395]\n",
      "에포크 043 | 시간 16.57 min | 평균 손실 >> 생성자/판별자 0.8861/1.3179 [판별자-진짜: 0.6640 판별자-가짜: 0.6540]\n",
      "에포크 044 | 시간 16.97 min | 평균 손실 >> 생성자/판별자 1.0273/1.2787 [판별자-진짜: 0.6403 판별자-가짜: 0.6384]\n",
      "에포크 045 | 시간 17.37 min | 평균 손실 >> 생성자/판별자 1.0467/1.2610 [판별자-진짜: 0.6338 판별자-가짜: 0.6272]\n",
      "에포크 046 | 시간 17.75 min | 평균 손실 >> 생성자/판별자 0.9496/1.2831 [판별자-진짜: 0.6488 판별자-가짜: 0.6343]\n",
      "에포크 047 | 시간 18.13 min | 평균 손실 >> 생성자/판별자 0.9138/1.3107 [판별자-진짜: 0.6581 판별자-가짜: 0.6525]\n",
      "에포크 048 | 시간 18.52 min | 평균 손실 >> 생성자/판별자 0.9374/1.3018 [판별자-진짜: 0.6532 판별자-가짜: 0.6486]\n",
      "에포크 049 | 시간 18.91 min | 평균 손실 >> 생성자/판별자 0.9828/1.2856 [판별자-진짜: 0.6436 판별자-가짜: 0.6420]\n",
      "에포크 050 | 시간 19.29 min | 평균 손실 >> 생성자/판별자 0.9295/1.2975 [판별자-진짜: 0.6511 판별자-가짜: 0.6464]\n",
      "에포크 051 | 시간 19.68 min | 평균 손실 >> 생성자/판별자 0.9180/1.3116 [판별자-진짜: 0.6593 판별자-가짜: 0.6523]\n",
      "에포크 052 | 시간 20.06 min | 평균 손실 >> 생성자/판별자 0.9478/1.3045 [판별자-진짜: 0.6559 판별자-가짜: 0.6485]\n",
      "에포크 053 | 시간 20.44 min | 평균 손실 >> 생성자/판별자 0.9977/1.2845 [판별자-진짜: 0.6422 판별자-가짜: 0.6423]\n",
      "에포크 054 | 시간 20.83 min | 평균 손실 >> 생성자/판별자 0.9760/1.2877 [판별자-진짜: 0.6507 판별자-가짜: 0.6370]\n",
      "에포크 055 | 시간 21.21 min | 평균 손실 >> 생성자/판별자 0.9236/1.3009 [판별자-진짜: 0.6540 판별자-가짜: 0.6469]\n",
      "에포크 056 | 시간 21.59 min | 평균 손실 >> 생성자/판별자 0.9435/1.2951 [판별자-진짜: 0.6522 판별자-가짜: 0.6429]\n",
      "에포크 057 | 시간 21.98 min | 평균 손실 >> 생성자/판별자 0.9623/1.2953 [판별자-진짜: 0.6506 판별자-가짜: 0.6447]\n",
      "에포크 058 | 시간 22.39 min | 평균 손실 >> 생성자/판별자 0.9485/1.3007 [판별자-진짜: 0.6544 판별자-가짜: 0.6462]\n",
      "에포크 059 | 시간 22.77 min | 평균 손실 >> 생성자/판별자 0.9488/1.2967 [판별자-진짜: 0.6503 판별자-가짜: 0.6464]\n",
      "에포크 060 | 시간 23.15 min | 평균 손실 >> 생성자/판별자 0.9225/1.3088 [판별자-진짜: 0.6570 판별자-가짜: 0.6518]\n",
      "에포크 061 | 시간 23.53 min | 평균 손실 >> 생성자/판별자 0.9267/1.3077 [판별자-진짜: 0.6521 판별자-가짜: 0.6556]\n",
      "에포크 062 | 시간 23.92 min | 평균 손실 >> 생성자/판별자 0.9224/1.3105 [판별자-진짜: 0.6592 판별자-가짜: 0.6512]\n",
      "에포크 063 | 시간 24.30 min | 평균 손실 >> 생성자/판별자 0.9276/1.3111 [판별자-진짜: 0.6577 판별자-가짜: 0.6533]\n",
      "에포크 064 | 시간 24.68 min | 평균 손실 >> 생성자/판별자 0.9249/1.3098 [판별자-진짜: 0.6584 판별자-가짜: 0.6514]\n",
      "에포크 065 | 시간 25.07 min | 평균 손실 >> 생성자/판별자 0.9001/1.3109 [판별자-진짜: 0.6590 판별자-가짜: 0.6519]\n",
      "에포크 066 | 시간 25.44 min | 평균 손실 >> 생성자/판별자 0.9435/1.3136 [판별자-진짜: 0.6584 판별자-가짜: 0.6552]\n",
      "에포크 067 | 시간 25.81 min | 평균 손실 >> 생성자/판별자 0.9537/1.3059 [판별자-진짜: 0.6543 판별자-가짜: 0.6516]\n",
      "에포크 068 | 시간 26.19 min | 평균 손실 >> 생성자/판별자 0.8768/1.3154 [판별자-진짜: 0.6617 판별자-가짜: 0.6537]\n",
      "에포크 069 | 시간 26.56 min | 평균 손실 >> 생성자/판별자 0.8964/1.3219 [판별자-진짜: 0.6613 판별자-가짜: 0.6606]\n",
      "에포크 070 | 시간 26.93 min | 평균 손실 >> 생성자/판별자 0.9685/1.2944 [판별자-진짜: 0.6485 판별자-가짜: 0.6458]\n",
      "에포크 071 | 시간 27.32 min | 평균 손실 >> 생성자/판별자 0.9237/1.3043 [판별자-진짜: 0.6560 판별자-가짜: 0.6483]\n",
      "에포크 072 | 시간 27.71 min | 평균 손실 >> 생성자/판별자 0.9225/1.3055 [판별자-진짜: 0.6558 판별자-가짜: 0.6497]\n",
      "에포크 073 | 시간 28.10 min | 평균 손실 >> 생성자/판별자 0.9260/1.3028 [판별자-진짜: 0.6548 판별자-가짜: 0.6480]\n",
      "에포크 074 | 시간 28.48 min | 평균 손실 >> 생성자/판별자 0.9449/1.3015 [판별자-진짜: 0.6538 판별자-가짜: 0.6477]\n",
      "에포크 075 | 시간 28.86 min | 평균 손실 >> 생성자/판별자 0.9181/1.3067 [판별자-진짜: 0.6554 판별자-가짜: 0.6513]\n",
      "에포크 076 | 시간 29.23 min | 평균 손실 >> 생성자/판별자 0.9343/1.3056 [판별자-진짜: 0.6538 판별자-가짜: 0.6518]\n",
      "에포크 077 | 시간 29.61 min | 평균 손실 >> 생성자/판별자 0.9451/1.3091 [판별자-진짜: 0.6543 판별자-가짜: 0.6548]\n",
      "에포크 078 | 시간 29.98 min | 평균 손실 >> 생성자/판별자 0.9007/1.3105 [판별자-진짜: 0.6576 판별자-가짜: 0.6529]\n",
      "에포크 079 | 시간 30.36 min | 평균 손실 >> 생성자/판별자 0.9416/1.3068 [판별자-진짜: 0.6575 판별자-가짜: 0.6493]\n",
      "에포크 080 | 시간 30.73 min | 평균 손실 >> 생성자/판별자 0.9404/1.3020 [판별자-진짜: 0.6546 판별자-가짜: 0.6473]\n",
      "에포크 081 | 시간 31.11 min | 평균 손실 >> 생성자/판별자 0.9294/1.3046 [판별자-진짜: 0.6571 판별자-가짜: 0.6475]\n",
      "에포크 082 | 시간 31.48 min | 평균 손실 >> 생성자/판별자 0.9248/1.3172 [판별자-진짜: 0.6601 판별자-가짜: 0.6571]\n",
      "에포크 083 | 시간 31.86 min | 평균 손실 >> 생성자/판별자 0.9282/1.3183 [판별자-진짜: 0.6583 판별자-가짜: 0.6600]\n",
      "에포크 084 | 시간 32.23 min | 평균 손실 >> 생성자/판별자 0.8921/1.3230 [판별자-진짜: 0.6645 판별자-가짜: 0.6585]\n",
      "에포크 085 | 시간 32.63 min | 평균 손실 >> 생성자/판별자 0.9641/1.3053 [판별자-진짜: 0.6531 판별자-가짜: 0.6523]\n",
      "에포크 086 | 시간 33.01 min | 평균 손실 >> 생성자/판별자 0.9347/1.3058 [판별자-진짜: 0.6564 판별자-가짜: 0.6494]\n",
      "에포크 087 | 시간 33.39 min | 평균 손실 >> 생성자/판별자 0.8623/1.3313 [판별자-진짜: 0.6685 판별자-가짜: 0.6628]\n",
      "에포크 088 | 시간 33.76 min | 평균 손실 >> 생성자/판별자 0.9660/1.3011 [판별자-진짜: 0.6484 판별자-가짜: 0.6527]\n",
      "에포크 089 | 시간 34.14 min | 평균 손실 >> 생성자/판별자 0.9310/1.3067 [판별자-진짜: 0.6560 판별자-가짜: 0.6507]\n",
      "에포크 090 | 시간 34.52 min | 평균 손실 >> 생성자/판별자 0.9038/1.3144 [판별자-진짜: 0.6596 판별자-가짜: 0.6548]\n",
      "에포크 091 | 시간 34.89 min | 평균 손실 >> 생성자/판별자 0.9309/1.3168 [판별자-진짜: 0.6593 판별자-가짜: 0.6576]\n",
      "에포크 092 | 시간 35.27 min | 평균 손실 >> 생성자/판별자 0.9120/1.3118 [판별자-진짜: 0.6584 판별자-가짜: 0.6534]\n",
      "에포크 093 | 시간 35.64 min | 평균 손실 >> 생성자/판별자 0.9094/1.3180 [판별자-진짜: 0.6606 판별자-가짜: 0.6574]\n",
      "에포크 094 | 시간 36.02 min | 평균 손실 >> 생성자/판별자 0.9522/1.3056 [판별자-진짜: 0.6545 판별자-가짜: 0.6511]\n",
      "에포크 095 | 시간 36.39 min | 평균 손실 >> 생성자/판별자 0.9239/1.3115 [판별자-진짜: 0.6588 판별자-가짜: 0.6528]\n",
      "에포크 096 | 시간 36.77 min | 평균 손실 >> 생성자/판별자 0.8989/1.3177 [판별자-진짜: 0.6617 판별자-가짜: 0.6560]\n",
      "에포크 097 | 시간 37.15 min | 평균 손실 >> 생성자/판별자 0.9082/1.3182 [판별자-진짜: 0.6613 판별자-가짜: 0.6569]\n",
      "에포크 098 | 시간 37.54 min | 평균 손실 >> 생성자/판별자 0.9274/1.3184 [판별자-진짜: 0.6603 판별자-가짜: 0.6582]\n",
      "에포크 099 | 시간 37.92 min | 평균 손실 >> 생성자/판별자 0.9192/1.3157 [판별자-진짜: 0.6596 판별자-가짜: 0.6561]\n",
      "에포크 100 | 시간 38.29 min | 평균 손실 >> 생성자/판별자 0.9237/1.3179 [판별자-진짜: 0.6615 판별자-가짜: 0.6564]\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "\n",
    "num_epochs = 100\n",
    "batch_size = 64\n",
    "image_size = (28, 28)\n",
    "z_size = 20\n",
    "mode_z = 'uniform'\n",
    "gen_hidden_layers = 1\n",
    "gen_hidden_size = 100\n",
    "disc_hidden_layers = 1\n",
    "disc_hidden_size = 100\n",
    "\n",
    "tf.random.set_seed(1)\n",
    "np.random.seed(1)\n",
    "\n",
    "\n",
    "if mode_z == 'uniform':\n",
    "    fixed_z = tf.random.uniform(\n",
    "        shape=(batch_size, z_size),\n",
    "        minval=-1, maxval=1)\n",
    "elif mode_z == 'normal':\n",
    "    fixed_z = tf.random.normal(\n",
    "        shape=(batch_size, z_size))\n",
    "\n",
    "\n",
    "def create_samples(g_model, input_z):\n",
    "    g_output = g_model(input_z, training=False)\n",
    "    images = tf.reshape(g_output, (batch_size, *image_size))    \n",
    "    return (images+1)/2.0\n",
    "\n",
    "## 데이터셋 준비\n",
    "mnist_trainset = mnist['train']\n",
    "mnist_trainset = mnist_trainset.map(\n",
    "    lambda ex: preprocess(ex, mode=mode_z))\n",
    "\n",
    "mnist_trainset = mnist_trainset.shuffle(10000)\n",
    "mnist_trainset = mnist_trainset.batch(\n",
    "    batch_size, drop_remainder=True)\n",
    "\n",
    "## 모델 준비\n",
    "with tf.device(device_name):\n",
    "    gen_model = make_generator_network(\n",
    "        num_hidden_layers=gen_hidden_layers, \n",
    "        num_hidden_units=gen_hidden_size,\n",
    "        num_output_units=np.prod(image_size))\n",
    "    gen_model.build(input_shape=(None, z_size))\n",
    "\n",
    "    disc_model = make_discriminator_network(\n",
    "        num_hidden_layers=disc_hidden_layers,\n",
    "        num_hidden_units=disc_hidden_size)\n",
    "    disc_model.build(input_shape=(None, np.prod(image_size)))\n",
    "\n",
    "## 손실 함수와 옵티마이저:\n",
    "loss_fn = tf.keras.losses.BinaryCrossentropy(from_logits=True)\n",
    "g_optimizer = tf.keras.optimizers.Adam()\n",
    "d_optimizer = tf.keras.optimizers.Adam()\n",
    "\n",
    "all_losses = []\n",
    "all_d_vals = []\n",
    "epoch_samples = []\n",
    "\n",
    "start_time = time.time()\n",
    "for epoch in range(1, num_epochs+1):\n",
    "    epoch_losses, epoch_d_vals = [], []\n",
    "    for i,(input_z,input_real) in enumerate(mnist_trainset):\n",
    "        \n",
    "        ## 생성자 손실을 계산합니다\n",
    "        with tf.GradientTape() as g_tape:\n",
    "            g_output = gen_model(input_z)\n",
    "            d_logits_fake = disc_model(g_output, training=True)\n",
    "            labels_real = tf.ones_like(d_logits_fake)\n",
    "            g_loss = loss_fn(y_true=labels_real, y_pred=d_logits_fake)\n",
    "            \n",
    "        # g_loss의 그래디언트를 계산합니다\n",
    "        g_grads = g_tape.gradient(g_loss, gen_model.trainable_variables)\n",
    "        \n",
    "        # 최적화: 그래디언트를 적용합니다\n",
    "        g_optimizer.apply_gradients(\n",
    "            grads_and_vars=zip(g_grads, gen_model.trainable_variables))\n",
    "\n",
    "        ## 판별자 손실을 계산합니다\n",
    "        with tf.GradientTape() as d_tape:\n",
    "            d_logits_real = disc_model(input_real, training=True)\n",
    "\n",
    "            d_labels_real = tf.ones_like(d_logits_real)\n",
    "            \n",
    "            d_loss_real = loss_fn(\n",
    "                y_true=d_labels_real, y_pred=d_logits_real)\n",
    "\n",
    "            d_logits_fake = disc_model(g_output, training=True)\n",
    "            d_labels_fake = tf.zeros_like(d_logits_fake)\n",
    "\n",
    "            d_loss_fake = loss_fn(\n",
    "                y_true=d_labels_fake, y_pred=d_logits_fake)\n",
    "\n",
    "            d_loss = d_loss_real + d_loss_fake\n",
    "\n",
    "        ## d_loss의 그래디언트를 계산합니다\n",
    "        d_grads = d_tape.gradient(d_loss, disc_model.trainable_variables)\n",
    "        \n",
    "        ## 최적화: 그래디언트를 적용합니다\n",
    "        d_optimizer.apply_gradients(\n",
    "            grads_and_vars=zip(d_grads, disc_model.trainable_variables))\n",
    "                           \n",
    "        epoch_losses.append(\n",
    "            (g_loss.numpy(), d_loss.numpy(), \n",
    "             d_loss_real.numpy(), d_loss_fake.numpy()))\n",
    "        \n",
    "        d_probs_real = tf.reduce_mean(tf.sigmoid(d_logits_real))\n",
    "        d_probs_fake = tf.reduce_mean(tf.sigmoid(d_logits_fake))\n",
    "        epoch_d_vals.append((d_probs_real.numpy(), d_probs_fake.numpy()))        \n",
    "    all_losses.append(epoch_losses)\n",
    "    all_d_vals.append(epoch_d_vals)\n",
    "    print(\n",
    "        '에포크 {:03d} | 시간 {:.2f} min | 평균 손실 >>'\n",
    "        ' 생성자/판별자 {:.4f}/{:.4f} [판별자-진짜: {:.4f} 판별자-가짜: {:.4f}]'\n",
    "        .format(\n",
    "            epoch, (time.time() - start_time)/60, \n",
    "            *list(np.mean(all_losses[-1], axis=0))))\n",
    "    epoch_samples.append(\n",
    "        create_samples(gen_model, fixed_z).numpy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 413
    },
    "id": "TQyQ8deLaHmw",
    "outputId": "e75e7f10-148b-45a4-880d-bb151e3b82ee"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import itertools\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize=(16, 6))\n",
    "\n",
    "## 손실 그래프\n",
    "ax = fig.add_subplot(1, 2, 1)\n",
    "g_losses = [item[0] for item in itertools.chain(*all_losses)]\n",
    "d_losses = [item[1]/2.0 for item in itertools.chain(*all_losses)]\n",
    "plt.plot(g_losses, label='Generator loss', alpha=0.95)\n",
    "plt.plot(d_losses, label='Discriminator loss', alpha=0.95)\n",
    "plt.legend(fontsize=20)\n",
    "ax.set_xlabel('Iteration', size=15)\n",
    "ax.set_ylabel('Loss', size=15)\n",
    "\n",
    "epochs = np.arange(1, 101)\n",
    "epoch2iter = lambda e: e*len(all_losses[-1])\n",
    "epoch_ticks = [1, 20, 40, 60, 80, 100]\n",
    "newpos = [epoch2iter(e) for e in epoch_ticks]\n",
    "ax2 = ax.twiny()\n",
    "ax2.set_xticks(newpos)\n",
    "ax2.set_xticklabels(epoch_ticks)\n",
    "ax2.xaxis.set_ticks_position('bottom')\n",
    "ax2.xaxis.set_label_position('bottom')\n",
    "ax2.spines['bottom'].set_position(('outward', 60))\n",
    "ax2.set_xlabel('Epoch', size=15)\n",
    "ax2.set_xlim(ax.get_xlim())\n",
    "ax.tick_params(axis='both', which='major', labelsize=15)\n",
    "ax2.tick_params(axis='both', which='major', labelsize=15)\n",
    "\n",
    "## 판별자의 출력\n",
    "ax = fig.add_subplot(1, 2, 2)\n",
    "d_vals_real = [item[0] for item in itertools.chain(*all_d_vals)]\n",
    "d_vals_fake = [item[1] for item in itertools.chain(*all_d_vals)]\n",
    "plt.plot(d_vals_real, alpha=0.75, label=r'Real: $D(\\mathbf{x})$')\n",
    "plt.plot(d_vals_fake, alpha=0.75, label=r'Fake: $D(G(\\mathbf{z}))$')\n",
    "plt.legend(fontsize=20)\n",
    "ax.set_xlabel('Iteration', size=15)\n",
    "ax.set_ylabel('Discriminator output', size=15)\n",
    "\n",
    "ax2 = ax.twiny()\n",
    "ax2.set_xticks(newpos)\n",
    "ax2.set_xticklabels(epoch_ticks)\n",
    "ax2.xaxis.set_ticks_position('bottom')\n",
    "ax2.xaxis.set_label_position('bottom')\n",
    "ax2.spines['bottom'].set_position(('outward', 60))\n",
    "ax2.set_xlabel('Epoch', size=15)\n",
    "ax2.set_xlim(ax.get_xlim())\n",
    "ax.tick_params(axis='both', which='major', labelsize=15)\n",
    "ax2.tick_params(axis='both', which='major', labelsize=15)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 783
    },
    "id": "iENdPX_gPoJ7",
    "outputId": "ef511e68-944c-4dc8-f719-aa59ea8d1084"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x1008 with 30 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "selected_epochs = [1, 2, 4, 10, 50, 100]\n",
    "fig = plt.figure(figsize=(10, 14))\n",
    "for i,e in enumerate(selected_epochs):\n",
    "    for j in range(5):\n",
    "        ax = fig.add_subplot(6, 5, i*5+j+1)\n",
    "        ax.set_xticks([])\n",
    "        ax.set_yticks([])\n",
    "        if j == 0:\n",
    "            ax.text(\n",
    "                -0.06, 0.5, 'Epoch {}'.format(e),\n",
    "                rotation=90, size=18, color='red',\n",
    "                horizontalalignment='right',\n",
    "                verticalalignment='center', \n",
    "                transform=ax.transAxes)\n",
    "        \n",
    "        image = epoch_samples[e-1][j]\n",
    "        ax.imshow(image, cmap='gray_r')\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "ch17-basic-GAN.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "1dc2c5d3f42e4e51920b91c3098ba5c5": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "3d34dd3e0f09486294c41b372fd63a04": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "527d1f8586a449df90bbdd38fd02c29f": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": "initial"
     }
    },
    "bc37964978a747e883e94134035d847f": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "bf41d9cedb7548868389eeab2134a90c": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "cb1a8a21d70841b4a2ba27ace07ccb11": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_f11d2daa89a149ad92ad5362f03de0b5",
       "IPY_MODEL_d78d899c3e9c426fb049561894daf335"
      ],
      "layout": "IPY_MODEL_1dc2c5d3f42e4e51920b91c3098ba5c5"
     }
    },
    "d78d899c3e9c426fb049561894daf335": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_bf41d9cedb7548868389eeab2134a90c",
      "placeholder": "​",
      "style": "IPY_MODEL_bc37964978a747e883e94134035d847f",
      "value": " 4/4 [00:00&lt;00:00,  5.35 file/s]"
     }
    },
    "f11d2daa89a149ad92ad5362f03de0b5": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "Dl Completed...: 100%",
      "description_tooltip": null,
      "layout": "IPY_MODEL_3d34dd3e0f09486294c41b372fd63a04",
      "max": 4,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_527d1f8586a449df90bbdd38fd02c29f",
      "value": 4
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}