Colab 3 - Crear una Red Neuronal Artificial en TensorFlow 2.0.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://colab.research.google.com/github/joanby/tensorflow2/blob/master/Colab%203%20-%20Crear%20una%20Red%20Neuronal%20Artificial%20en%20TensorFlow%202.0.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "sJr4nMgyb9oS"
   },
   "source": [
    "\n",
    "<p align=\"center\">\n",
    "  <img src=\"https://storage.googleapis.com/kaggle-datasets-images/2243/3791/9384af51de8baa77f6320901f53bd26b/dataset-cover.png\" />\n",
    "  Imagen cortesía de: https://www.kaggle.com/\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "vVF9I9v6Zipb"
   },
   "source": [
    "## Paso 1: Instalar las dependencias y configurar el entorno de GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 496
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 36722,
     "status": "ok",
     "timestamp": 1564588185169,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "WS6fShx3Za4O",
    "outputId": "992d8abe-0f1c-4d37-966b-da81502a91b8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow-gpu==2.0.0.alpha0\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/1a/66/32cffad095253219d53f6b6c2a436637bbe45ac4e7be0244557210dc3918/tensorflow_gpu-2.0.0a0-cp36-cp36m-manylinux1_x86_64.whl (332.1MB)\n",
      "\u001b[K     |████████████████████████████████| 332.1MB 93kB/s \n",
      "\u001b[?25hRequirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.15.0)\n",
      "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.12.0)\n",
      "Requirement already satisfied: keras-applications>=1.0.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.0.8)\n",
      "Requirement already satisfied: google-pasta>=0.1.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (0.1.7)\n",
      "Requirement already satisfied: numpy<2.0,>=1.14.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.16.4)\n",
      "Requirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (0.8.0)\n",
      "Requirement already satisfied: gast>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (0.2.2)\n",
      "Collecting tf-estimator-nightly<1.14.0.dev2019030116,>=1.14.0.dev2019030115 (from tensorflow-gpu==2.0.0.alpha0)\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/13/82/f16063b4eed210dc2ab057930ac1da4fbe1e91b7b051a6c8370b401e6ae7/tf_estimator_nightly-1.14.0.dev2019030115-py2.py3-none-any.whl (411kB)\n",
      "\u001b[K     |████████████████████████████████| 419kB 54.1MB/s \n",
      "\u001b[?25hRequirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.1.0)\n",
      "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (0.33.4)\n",
      "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (1.1.0)\n",
      "Collecting tb-nightly<1.14.0a20190302,>=1.14.0a20190301 (from tensorflow-gpu==2.0.0.alpha0)\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/a9/51/aa1d756644bf4624c03844115e4ac4058eff77acd786b26315f051a4b195/tb_nightly-1.14.0a20190301-py3-none-any.whl (3.0MB)\n",
      "\u001b[K     |████████████████████████████████| 3.0MB 44.6MB/s \n",
      "\u001b[?25hRequirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (0.7.1)\n",
      "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0.alpha0) (3.7.1)\n",
      "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.6->tensorflow-gpu==2.0.0.alpha0) (2.8.0)\n",
      "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.14.0a20190302,>=1.14.0a20190301->tensorflow-gpu==2.0.0.alpha0) (0.15.5)\n",
      "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.14.0a20190302,>=1.14.0a20190301->tensorflow-gpu==2.0.0.alpha0) (3.1.1)\n",
      "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tensorflow-gpu==2.0.0.alpha0) (41.0.1)\n",
      "Installing collected packages: tf-estimator-nightly, tb-nightly, tensorflow-gpu\n",
      "Successfully installed tb-nightly-1.14.0a20190301 tensorflow-gpu-2.0.0a0 tf-estimator-nightly-1.14.0.dev2019030115\n"
     ]
    }
   ],
   "source": [
    "#!pip install tensorflow-gpu==2.3.0\n",
    "%tensorflow_version 2.x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "DjlnjPnjWYFw"
   },
   "source": [
    "## Paso 2: Importar las dependencias necesarias para el proyecto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "yt0-hrch6rZw"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import datetime\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.datasets import fashion_mnist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 728,
     "status": "ok",
     "timestamp": 1564588206066,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "_vv9AXUnZW78",
    "outputId": "a373495f-a620-462e-f10b-815bbeb81e42"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2.0.0-alpha0'"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "K2OiUS-kWkJU"
   },
   "source": [
    "## Paso 3: Pre procesado de datos\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "DdfoFiEEXYj1"
   },
   "source": [
    "### Cargar el dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 153
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 2705,
     "status": "ok",
     "timestamp": 1564588211057,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "-lCgz6UC8pKT",
    "outputId": "026f965e-3f9b-4fa0-f237-0dc0f8a09735"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
      "32768/29515 [=================================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
      "26427392/26421880 [==============================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
      "8192/5148 [===============================================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
      "4423680/4422102 [==============================] - 0s 0us/step\n"
     ]
    }
   ],
   "source": [
    "#Cargar el dataset Fashion Mnist \n",
    "(X_train, y_train), (X_test, y_test) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "AYxeEHzDXdSs"
   },
   "source": [
    "### Normalizar las imágenes\n",
    "\n",
    "Se divide cada imagen en los conjunto de entrenamiento y  de testing entre el valor máximo de cada uno de los píxeles (255).\n",
    "\n",
    "De este modo, cada píxel se hallará en el rango [0, 1]. Al normalizar las imágenes, nos aseguramos que nuestro modelo de RNA entrenará más rápidamente."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zvWzsB3G9IU8"
   },
   "outputs": [],
   "source": [
    "X_train = X_train / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lo--rpqo9ZtA"
   },
   "outputs": [],
   "source": [
    "X_test = X_test / 255.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "uBacLmGIX0Es"
   },
   "source": [
    "### Redimensionar el dataset\n",
    "\n",
    "Como vamos a utilizar una red neuronal totalmente conectada, vamos a redimensionar los subconjuntos de entrenamiento y testing a formato de vector en lugar de en formato de matriz."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "2Tao7pom-grn"
   },
   "outputs": [],
   "source": [
    "#Como cada imagen tiene 28x28 píxeles, usamos la función reshape en todo el dataset de entrenamiento para convertirlo \n",
    "# en vectores de tamaño [-1 (todos los elementos), anchura * altura]\n",
    "X_train = X_train.reshape(-1, 28*28)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 966,
     "status": "ok",
     "timestamp": 1564588234369,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "t9MbMrg9-kr_",
    "outputId": "a64ef84d-adbb-45f6-cc7b-e27aeab9c0ea"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 784)"
      ]
     },
     "execution_count": 10,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Y_duQGtbCTTL"
   },
   "outputs": [],
   "source": [
    "#Redimensionamos el conjunto de testing del mismo modo\n",
    "X_test = X_test.reshape(-1, 28*28)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "x5aDsaYSYmXD"
   },
   "source": [
    "## Paso 4: Construir la Red Neuronal Artificial"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "l30aZ6-GYtUP"
   },
   "source": [
    "### Definir el modelo\n",
    "\n",
    "Simplemente se define un objeto de modelo Sequential."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xmfogzmn9kqv"
   },
   "outputs": [],
   "source": [
    "model = tf.keras.models.Sequential()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "TNzLOAK5Y-mR"
   },
   "source": [
    "### Añadir la primera capa totalmente conectada (capa Densa)\n",
    "\n",
    "Hyper-parametros de la capa:\n",
    "- número de unidades/neuronas: 128\n",
    "- función de activación: ReLU\n",
    "- input_shape: (784, )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GBsfDyGE-FX5"
   },
   "outputs": [],
   "source": [
    "model.add(tf.keras.layers.Dense(units=128, activation='relu', input_shape=(784, )))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Vwqx1wZUa1rH"
   },
   "source": [
    "### Añadir una capa de Dropout \n",
    "\n",
    "Dropout es una técnica de Regularization donde aleatoriamente se asignan a ciertas neuronas de la red el valor cero. De este modo, mientras se entrena, estas neuronas no actualizarán sus valores. Al tener cierto porcentaje de neuronas sin actualizar, el proceso de entrenamiento toma más tiempo pero por contra tenemos menos posibilidades de sufrir overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "tAmpLPlr-pOX"
   },
   "outputs": [],
   "source": [
    "model.add(tf.keras.layers.Dropout(0.2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "BGqvyDvNbzwN"
   },
   "source": [
    "### Añadir la segunda capa (capa de salida)\n",
    "\n",
    "- unidades: número de clases (10 en el caso del Fashion MNIST)\n",
    "- función de activación: 'softmax' (probabilidades de cada clase)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "OmkUuF9Y-3mG"
   },
   "outputs": [],
   "source": [
    "model.add(tf.keras.layers.Dense(units=10, activation='softmax'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "2rRsMjsvcOua"
   },
   "source": [
    "### Compilar el modelo\n",
    "\n",
    "- Optimizer: Adam\n",
    "- Loss: Sparse softmax (categorical) crossentropy "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "nbW3xeRK_CrN"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 255
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 631,
     "status": "ok",
     "timestamp": 1564588582610,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "8dQOL_EtChrN",
    "outputId": "a839ebcc-3021-4b90-9598-3e4746c339fe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 128)               100480    \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 101,770\n",
      "Trainable params: 101,770\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "9kxIIFU1cany"
   },
   "source": [
    "### Entrenar el modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 26920,
     "status": "ok",
     "timestamp": 1564588612289,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "s-_oLiE0_3A2",
    "outputId": "71872c57-618f-42c1-ab22-2bedc29abcc8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "60000/60000 [==============================] - 6s 100us/sample - loss: 0.5289 - sparse_categorical_accuracy: 0.8131\n",
      "Epoch 2/5\n",
      "60000/60000 [==============================] - 5s 83us/sample - loss: 0.4007 - sparse_categorical_accuracy: 0.8553\n",
      "Epoch 3/5\n",
      "60000/60000 [==============================] - 5s 85us/sample - loss: 0.3667 - sparse_categorical_accuracy: 0.8661\n",
      "Epoch 4/5\n",
      "60000/60000 [==============================] - 5s 83us/sample - loss: 0.3471 - sparse_categorical_accuracy: 0.8724\n",
      "Epoch 5/5\n",
      "60000/60000 [==============================] - 5s 83us/sample - loss: 0.3303 - sparse_categorical_accuracy: 0.8780\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7f8404344518>"
      ]
     },
     "execution_count": 18,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, epochs=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "mj23nxmtcrhd"
   },
   "source": [
    "### Evaluación del modelo y predicción"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1041,
     "status": "ok",
     "timestamp": 1564588614802,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "-nQCioOmAL7i",
    "outputId": "4e4abac6-7e3b-49db-b9bf-7fcf3267fd2c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 1s 58us/sample - loss: 0.3563 - sparse_categorical_accuracy: 0.8708\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_accuracy = model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 686,
     "status": "ok",
     "timestamp": 1564588618623,
     "user": {
      "displayName": "Juan Gabriel Gomila Salas",
      "photoUrl": "https://lh4.googleusercontent.com/-RJ38mzk8Fog/AAAAAAAAAAI/AAAAAAAAA_g/VV8PcNMAUrk/s64/photo.jpg",
      "userId": "03106941341701838274"
     },
     "user_tz": -120
    },
    "id": "Ozv2YVlxcx1h",
    "outputId": "5ec94881-fd81-4779-ad24-a1b3664ba96b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy: 0.8708000183105469\n"
     ]
    }
   ],
   "source": [
    "print(\"Test accuracy: {}\".format(test_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Noi53-uq9yhl"
   },
   "source": [
    "## Paso 5 : Guardar el modelo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eMFm-Z9I99R5"
   },
   "source": [
    "### Guardar la arquitectura (topoligía) de la red neuronal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "pT7pmXWO9xxM"
   },
   "outputs": [],
   "source": [
    "model_json = model.to_json()\n",
    "with open(\"fashion_model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5UDk8L4A-CQX"
   },
   "source": [
    "### Guardar los pesos de la red neuronal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "R0ZOVcC498Mp"
   },
   "outputs": [],
   "source": [
    "model.save_weights(\"fashion_model.h5\")"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "Colab 3 - Crear una Red Neuronal Artificial en TensorFlow 2.0.ipynb",
   "provenance": [
    {
     "file_id": "1tj02zPuUkkbLob1oo63f9TliYj5DQHA3",
     "timestamp": 1564587360109
    }
   ]
  },
  "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}