CIFAR10-CNN-tipoResNET.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%config InlineBackend.figure_format = 'retina'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras import backend as K\n",
    "import tensorflow as tf\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth=True\n",
    "sess = tf.Session(config=config)\n",
    "K.set_session(sess)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy.random import seed\n",
    "seed(1)\n",
    "from tensorflow import set_random_seed\n",
    "set_random_seed(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Keras modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.utils import to_categorical\n",
    "from keras.callbacks import ModelCheckpoint, EarlyStopping\n",
    "from keras.models import Model\n",
    "from keras.layers.core import Dense, Flatten, Activation\n",
    "from keras.layers import Conv2D, AveragePooling2D, BatchNormalization, Dropout, MaxPooling2D, Input,Add"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cargo datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.datasets import cifar10\n",
    "((x, y), (x_test,y_test)) = cifar10.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_valid, y_train, y_valid = train_test_split(x, y, test_size = 0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Paso a float"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = x_train.astype('float32')\n",
    "x_valid = x_valid.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_valid /= 255\n",
    "x_test /= 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_categorical = to_categorical(y_train)\n",
    "y_val_categorical = to_categorical(y_valid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Red neuronal: CNN tipo RESNET"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_2 (InputLayer)            (None, 32, 32, 3)    0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_53 (Conv2D)              (None, 32, 32, 16)   448         input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_53 (BatchNo (None, 32, 32, 16)   64          conv2d_53[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_54 (Conv2D)              (None, 32, 32, 16)   2320        batch_normalization_53[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_54 (BatchNo (None, 32, 32, 16)   64          conv2d_54[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_49 (Activation)      (None, 32, 32, 16)   0           batch_normalization_54[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_55 (Conv2D)              (None, 32, 32, 16)   2320        activation_49[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_55 (BatchNo (None, 32, 32, 16)   64          conv2d_55[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_50 (Activation)      (None, 32, 32, 16)   0           batch_normalization_55[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_56 (Conv2D)              (None, 32, 32, 16)   2320        activation_50[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_56 (BatchNo (None, 32, 32, 16)   64          conv2d_56[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_17 (Add)                    (None, 32, 32, 16)   0           batch_normalization_53[0][0]     \n",
      "                                                                 batch_normalization_56[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_51 (Activation)      (None, 32, 32, 16)   0           add_17[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2D)  (None, 16, 16, 16)   0           activation_51[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_57 (Conv2D)              (None, 16, 16, 16)   2320        max_pooling2d_5[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_57 (BatchNo (None, 16, 16, 16)   64          conv2d_57[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_52 (Activation)      (None, 16, 16, 16)   0           batch_normalization_57[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_58 (Conv2D)              (None, 16, 16, 16)   2320        activation_52[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_58 (BatchNo (None, 16, 16, 16)   64          conv2d_58[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_53 (Activation)      (None, 16, 16, 16)   0           batch_normalization_58[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_59 (Conv2D)              (None, 16, 16, 16)   2320        activation_53[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_59 (BatchNo (None, 16, 16, 16)   64          conv2d_59[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_18 (Add)                    (None, 16, 16, 16)   0           max_pooling2d_5[0][0]            \n",
      "                                                                 batch_normalization_59[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_54 (Activation)      (None, 16, 16, 16)   0           add_18[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_60 (Conv2D)              (None, 16, 16, 16)   2320        activation_54[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_60 (BatchNo (None, 16, 16, 16)   64          conv2d_60[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_55 (Activation)      (None, 16, 16, 16)   0           batch_normalization_60[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_61 (Conv2D)              (None, 16, 16, 16)   2320        activation_55[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_61 (BatchNo (None, 16, 16, 16)   64          conv2d_61[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_56 (Activation)      (None, 16, 16, 16)   0           batch_normalization_61[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_62 (Conv2D)              (None, 16, 16, 16)   2320        activation_56[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_62 (BatchNo (None, 16, 16, 16)   64          conv2d_62[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_19 (Add)                    (None, 16, 16, 16)   0           activation_54[0][0]              \n",
      "                                                                 batch_normalization_62[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_57 (Activation)      (None, 16, 16, 16)   0           add_19[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_63 (Conv2D)              (None, 16, 16, 16)   2320        activation_57[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_63 (BatchNo (None, 16, 16, 16)   64          conv2d_63[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_58 (Activation)      (None, 16, 16, 16)   0           batch_normalization_63[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_64 (Conv2D)              (None, 16, 16, 16)   2320        activation_58[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_64 (BatchNo (None, 16, 16, 16)   64          conv2d_64[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_59 (Activation)      (None, 16, 16, 16)   0           batch_normalization_64[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_65 (Conv2D)              (None, 16, 16, 16)   2320        activation_59[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_65 (BatchNo (None, 16, 16, 16)   64          conv2d_65[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_20 (Add)                    (None, 16, 16, 16)   0           activation_57[0][0]              \n",
      "                                                                 batch_normalization_65[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_60 (Activation)      (None, 16, 16, 16)   0           add_20[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_6 (MaxPooling2D)  (None, 8, 8, 16)     0           activation_60[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_67 (Conv2D)              (None, 8, 8, 32)     4640        max_pooling2d_6[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_67 (BatchNo (None, 8, 8, 32)     128         conv2d_67[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_61 (Activation)      (None, 8, 8, 32)     0           batch_normalization_67[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_68 (Conv2D)              (None, 8, 8, 32)     9248        activation_61[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_68 (BatchNo (None, 8, 8, 32)     128         conv2d_68[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_62 (Activation)      (None, 8, 8, 32)     0           batch_normalization_68[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_66 (Conv2D)              (None, 8, 8, 32)     544         max_pooling2d_6[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_69 (Conv2D)              (None, 8, 8, 32)     9248        activation_62[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_66 (BatchNo (None, 8, 8, 32)     128         conv2d_66[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_69 (BatchNo (None, 8, 8, 32)     128         conv2d_69[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_21 (Add)                    (None, 8, 8, 32)     0           batch_normalization_66[0][0]     \n",
      "                                                                 batch_normalization_69[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_63 (Activation)      (None, 8, 8, 32)     0           add_21[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_70 (Conv2D)              (None, 8, 8, 32)     9248        activation_63[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_70 (BatchNo (None, 8, 8, 32)     128         conv2d_70[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_64 (Activation)      (None, 8, 8, 32)     0           batch_normalization_70[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_71 (Conv2D)              (None, 8, 8, 32)     9248        activation_64[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_71 (BatchNo (None, 8, 8, 32)     128         conv2d_71[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_65 (Activation)      (None, 8, 8, 32)     0           batch_normalization_71[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_72 (Conv2D)              (None, 8, 8, 32)     9248        activation_65[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_72 (BatchNo (None, 8, 8, 32)     128         conv2d_72[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_22 (Add)                    (None, 8, 8, 32)     0           activation_63[0][0]              \n",
      "                                                                 batch_normalization_72[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_66 (Activation)      (None, 8, 8, 32)     0           add_22[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_73 (Conv2D)              (None, 8, 8, 32)     9248        activation_66[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_73 (BatchNo (None, 8, 8, 32)     128         conv2d_73[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_67 (Activation)      (None, 8, 8, 32)     0           batch_normalization_73[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_74 (Conv2D)              (None, 8, 8, 32)     9248        activation_67[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_74 (BatchNo (None, 8, 8, 32)     128         conv2d_74[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_68 (Activation)      (None, 8, 8, 32)     0           batch_normalization_74[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_75 (Conv2D)              (None, 8, 8, 32)     9248        activation_68[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_75 (BatchNo (None, 8, 8, 32)     128         conv2d_75[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_23 (Add)                    (None, 8, 8, 32)     0           activation_66[0][0]              \n",
      "                                                                 batch_normalization_75[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_69 (Activation)      (None, 8, 8, 32)     0           add_23[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_76 (Conv2D)              (None, 8, 8, 32)     9248        activation_69[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_76 (BatchNo (None, 8, 8, 32)     128         conv2d_76[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_70 (Activation)      (None, 8, 8, 32)     0           batch_normalization_76[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_77 (Conv2D)              (None, 8, 8, 32)     9248        activation_70[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_77 (BatchNo (None, 8, 8, 32)     128         conv2d_77[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_71 (Activation)      (None, 8, 8, 32)     0           batch_normalization_77[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_78 (Conv2D)              (None, 8, 8, 32)     9248        activation_71[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_78 (BatchNo (None, 8, 8, 32)     128         conv2d_78[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_24 (Add)                    (None, 8, 8, 32)     0           activation_69[0][0]              \n",
      "                                                                 batch_normalization_78[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_72 (Activation)      (None, 8, 8, 32)     0           add_24[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2D)  (None, 4, 4, 32)     0           activation_72[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_80 (Conv2D)              (None, 4, 4, 64)     18496       max_pooling2d_7[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_80 (BatchNo (None, 4, 4, 64)     256         conv2d_80[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_73 (Activation)      (None, 4, 4, 64)     0           batch_normalization_80[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_81 (Conv2D)              (None, 4, 4, 64)     36928       activation_73[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_81 (BatchNo (None, 4, 4, 64)     256         conv2d_81[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_74 (Activation)      (None, 4, 4, 64)     0           batch_normalization_81[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_79 (Conv2D)              (None, 4, 4, 64)     2112        max_pooling2d_7[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_82 (Conv2D)              (None, 4, 4, 64)     36928       activation_74[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_79 (BatchNo (None, 4, 4, 64)     256         conv2d_79[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_82 (BatchNo (None, 4, 4, 64)     256         conv2d_82[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_25 (Add)                    (None, 4, 4, 64)     0           batch_normalization_79[0][0]     \n",
      "                                                                 batch_normalization_82[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_75 (Activation)      (None, 4, 4, 64)     0           add_25[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_83 (Conv2D)              (None, 4, 4, 64)     36928       activation_75[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_83 (BatchNo (None, 4, 4, 64)     256         conv2d_83[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_76 (Activation)      (None, 4, 4, 64)     0           batch_normalization_83[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_84 (Conv2D)              (None, 4, 4, 64)     36928       activation_76[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_84 (BatchNo (None, 4, 4, 64)     256         conv2d_84[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_77 (Activation)      (None, 4, 4, 64)     0           batch_normalization_84[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_85 (Conv2D)              (None, 4, 4, 64)     36928       activation_77[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_85 (BatchNo (None, 4, 4, 64)     256         conv2d_85[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_26 (Add)                    (None, 4, 4, 64)     0           activation_75[0][0]              \n",
      "                                                                 batch_normalization_85[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_78 (Activation)      (None, 4, 4, 64)     0           add_26[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_86 (Conv2D)              (None, 4, 4, 64)     36928       activation_78[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_86 (BatchNo (None, 4, 4, 64)     256         conv2d_86[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_79 (Activation)      (None, 4, 4, 64)     0           batch_normalization_86[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_87 (Conv2D)              (None, 4, 4, 64)     36928       activation_79[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_87 (BatchNo (None, 4, 4, 64)     256         conv2d_87[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_80 (Activation)      (None, 4, 4, 64)     0           batch_normalization_87[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_88 (Conv2D)              (None, 4, 4, 64)     36928       activation_80[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_88 (BatchNo (None, 4, 4, 64)     256         conv2d_88[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_27 (Add)                    (None, 4, 4, 64)     0           activation_78[0][0]              \n",
      "                                                                 batch_normalization_88[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_81 (Activation)      (None, 4, 4, 64)     0           add_27[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_89 (Conv2D)              (None, 4, 4, 64)     36928       activation_81[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_89 (BatchNo (None, 4, 4, 64)     256         conv2d_89[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_82 (Activation)      (None, 4, 4, 64)     0           batch_normalization_89[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_90 (Conv2D)              (None, 4, 4, 64)     36928       activation_82[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_90 (BatchNo (None, 4, 4, 64)     256         conv2d_90[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_83 (Activation)      (None, 4, 4, 64)     0           batch_normalization_90[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_91 (Conv2D)              (None, 4, 4, 64)     36928       activation_83[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_91 (BatchNo (None, 4, 4, 64)     256         conv2d_91[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_28 (Add)                    (None, 4, 4, 64)     0           activation_81[0][0]              \n",
      "                                                                 batch_normalization_91[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_84 (Activation)      (None, 4, 4, 64)     0           add_28[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2D)  (None, 2, 2, 64)     0           activation_84[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_93 (Conv2D)              (None, 2, 2, 128)    73856       max_pooling2d_8[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_93 (BatchNo (None, 2, 2, 128)    512         conv2d_93[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_85 (Activation)      (None, 2, 2, 128)    0           batch_normalization_93[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_94 (Conv2D)              (None, 2, 2, 128)    147584      activation_85[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_94 (BatchNo (None, 2, 2, 128)    512         conv2d_94[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_86 (Activation)      (None, 2, 2, 128)    0           batch_normalization_94[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_92 (Conv2D)              (None, 2, 2, 128)    8320        max_pooling2d_8[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_95 (Conv2D)              (None, 2, 2, 128)    147584      activation_86[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_92 (BatchNo (None, 2, 2, 128)    512         conv2d_92[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_95 (BatchNo (None, 2, 2, 128)    512         conv2d_95[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_29 (Add)                    (None, 2, 2, 128)    0           batch_normalization_92[0][0]     \n",
      "                                                                 batch_normalization_95[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_87 (Activation)      (None, 2, 2, 128)    0           add_29[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_96 (Conv2D)              (None, 2, 2, 128)    147584      activation_87[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_96 (BatchNo (None, 2, 2, 128)    512         conv2d_96[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_88 (Activation)      (None, 2, 2, 128)    0           batch_normalization_96[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_97 (Conv2D)              (None, 2, 2, 128)    147584      activation_88[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_97 (BatchNo (None, 2, 2, 128)    512         conv2d_97[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_89 (Activation)      (None, 2, 2, 128)    0           batch_normalization_97[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_98 (Conv2D)              (None, 2, 2, 128)    147584      activation_89[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_98 (BatchNo (None, 2, 2, 128)    512         conv2d_98[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "add_30 (Add)                    (None, 2, 2, 128)    0           activation_87[0][0]              \n",
      "                                                                 batch_normalization_98[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_90 (Activation)      (None, 2, 2, 128)    0           add_30[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_99 (Conv2D)              (None, 2, 2, 128)    147584      activation_90[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_99 (BatchNo (None, 2, 2, 128)    512         conv2d_99[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_91 (Activation)      (None, 2, 2, 128)    0           batch_normalization_99[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_100 (Conv2D)             (None, 2, 2, 128)    147584      activation_91[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_100 (BatchN (None, 2, 2, 128)    512         conv2d_100[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_92 (Activation)      (None, 2, 2, 128)    0           batch_normalization_100[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_101 (Conv2D)             (None, 2, 2, 128)    147584      activation_92[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_101 (BatchN (None, 2, 2, 128)    512         conv2d_101[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "add_31 (Add)                    (None, 2, 2, 128)    0           activation_90[0][0]              \n",
      "                                                                 batch_normalization_101[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "activation_93 (Activation)      (None, 2, 2, 128)    0           add_31[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_102 (Conv2D)             (None, 2, 2, 128)    147584      activation_93[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_102 (BatchN (None, 2, 2, 128)    512         conv2d_102[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_94 (Activation)      (None, 2, 2, 128)    0           batch_normalization_102[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_103 (Conv2D)             (None, 2, 2, 128)    147584      activation_94[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_103 (BatchN (None, 2, 2, 128)    512         conv2d_103[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_95 (Activation)      (None, 2, 2, 128)    0           batch_normalization_103[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_104 (Conv2D)             (None, 2, 2, 128)    147584      activation_95[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_104 (BatchN (None, 2, 2, 128)    512         conv2d_104[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "add_32 (Add)                    (None, 2, 2, 128)    0           activation_93[0][0]              \n",
      "                                                                 batch_normalization_104[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "activation_96 (Activation)      (None, 2, 2, 128)    0           add_32[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_2 (AveragePoo (None, 1, 1, 128)    0           activation_96[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "flatten_1 (Flatten)             (None, 128)          0           average_pooling2d_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "dense_4 (Dense)                 (None, 256)          33024       flatten_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_3 (Dropout)             (None, 256)          0           dense_4[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_5 (Dense)                 (None, 256)          65792       dropout_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_4 (Dropout)             (None, 256)          0           dense_5[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_6 (Dense)                 (None, 10)           2570        dropout_4[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 2,381,482\n",
      "Trainable params: 2,375,242\n",
      "Non-trainable params: 6,240\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "net_input = Input(shape=(32,32,3))\n",
    "\n",
    "def res_block_first(input_layer,filters=16):\n",
    "    \n",
    "    #Primera capa\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(input_layer)\n",
    "    aux_net=BatchNormalization()(net)\n",
    "    \n",
    "    #conexión directa\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(aux_net)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    \n",
    "    #Shortcut\n",
    "    net=Add()([aux_net,net])\n",
    "    \n",
    "    net=Activation(\"relu\")(net)\n",
    "    net=MaxPooling2D()(net)\n",
    "    \n",
    "    return net\n",
    "\n",
    "def res_block(input_layer,filters=16):\n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(input_layer)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    \n",
    "    #Shortcut\n",
    "    net=Add()([input_layer,net])\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    return net\n",
    "\n",
    "def res_block_dec(input_layer,filters=32):\n",
    "    \n",
    "    input_layer=MaxPooling2D()(input_layer)\n",
    "    net=Conv2D(filters=filters, kernel_size=(1, 1), padding='same')(input_layer)\n",
    "    aux_net=BatchNormalization()(net)\n",
    "    \n",
    "    #conexión directa\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(input_layer)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    net=Conv2D(filters=filters, kernel_size=(3, 3), padding='same')(net)\n",
    "    net=BatchNormalization()(net)\n",
    "    net=Add()([aux_net,net])\n",
    "    net=Activation(\"relu\")(net)\n",
    "    \n",
    "    return net\n",
    "\n",
    "net=res_block_first(net_input,filters=16)\n",
    "net=res_block(net,filters=16)\n",
    "net=res_block(net,filters=16)\n",
    "net=res_block(net,filters=16)\n",
    "net=res_block_dec(net,filters=32)\n",
    "net=res_block(net,filters=32)\n",
    "net=res_block(net,filters=32)\n",
    "net=res_block(net,filters=32)\n",
    "net=res_block_dec(net,filters=64)\n",
    "net=res_block(net,filters=64)\n",
    "net=res_block(net,filters=64)\n",
    "net=res_block(net,filters=64)\n",
    "net=res_block_dec(net,filters=128)\n",
    "net=res_block(net,filters=128)\n",
    "net=res_block(net,filters=128)\n",
    "net=res_block(net,filters=128)\n",
    "\n",
    "net = AveragePooling2D() (net)\n",
    "\n",
    "net=Flatten()(net)\n",
    "\n",
    "net=Dense(units=256, activation='relu')(net)\n",
    "net=Dropout(0.1)(net)\n",
    "net=Dense(units=256, activation='relu')(net)\n",
    "net=Dropout(0.1)(net)\n",
    "net=Dense(units=10, activation = 'softmax')(net)\n",
    "\n",
    "model=Model(net_input,net)\n",
    "\n",
    "model.compile(loss='categorical_crossentropy', optimizer='Adam', metrics=['accuracy'])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Callbacks\n",
    "checkpoint_file='checkpoint.mnist-cnn-tipo-RESNET.hdf5'\n",
    "earlystop = EarlyStopping(monitor='val_acc', min_delta=0, patience=5, verbose=2)\n",
    "checkpointer = ModelCheckpoint(monitor=\"val_acc\",filepath=checkpoint_file, verbose=1, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(45000, 32, 32, 3)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(45000, 10)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_categorical.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 45000 samples, validate on 5000 samples\n",
      "Epoch 1/100\n",
      "45000/45000 [==============================] - 35s 773us/step - loss: 1.6192 - acc: 0.4005 - val_loss: 1.8315 - val_acc: 0.4130\n",
      "\n",
      "Epoch 00001: val_acc improved from -inf to 0.41300, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 2/100\n",
      "45000/45000 [==============================] - 25s 559us/step - loss: 1.1946 - acc: 0.5761 - val_loss: 1.2874 - val_acc: 0.5496\n",
      "\n",
      "Epoch 00002: val_acc improved from 0.41300 to 0.54960, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 3/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.9906 - acc: 0.6530 - val_loss: 1.0689 - val_acc: 0.6338\n",
      "\n",
      "Epoch 00003: val_acc improved from 0.54960 to 0.63380, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 4/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.8460 - acc: 0.7067 - val_loss: 0.9587 - val_acc: 0.6816\n",
      "\n",
      "Epoch 00004: val_acc improved from 0.63380 to 0.68160, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 5/100\n",
      "45000/45000 [==============================] - 25s 556us/step - loss: 0.7531 - acc: 0.7401 - val_loss: 1.2374 - val_acc: 0.6140\n",
      "\n",
      "Epoch 00005: val_acc did not improve from 0.68160\n",
      "Epoch 6/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.6642 - acc: 0.7732 - val_loss: 0.8965 - val_acc: 0.7094\n",
      "\n",
      "Epoch 00006: val_acc improved from 0.68160 to 0.70940, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 7/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.6046 - acc: 0.7934 - val_loss: 0.7788 - val_acc: 0.7400\n",
      "\n",
      "Epoch 00007: val_acc improved from 0.70940 to 0.74000, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 8/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.5427 - acc: 0.8150 - val_loss: 0.8827 - val_acc: 0.7174\n",
      "\n",
      "Epoch 00008: val_acc did not improve from 0.74000\n",
      "Epoch 9/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.4999 - acc: 0.8291 - val_loss: 1.2692 - val_acc: 0.6332\n",
      "\n",
      "Epoch 00009: val_acc did not improve from 0.74000\n",
      "Epoch 10/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.4507 - acc: 0.8470 - val_loss: 0.7069 - val_acc: 0.7620\n",
      "\n",
      "Epoch 00010: val_acc improved from 0.74000 to 0.76200, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 11/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.4055 - acc: 0.8606 - val_loss: 0.7997 - val_acc: 0.7468\n",
      "\n",
      "Epoch 00011: val_acc did not improve from 0.76200\n",
      "Epoch 12/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.3648 - acc: 0.8749 - val_loss: 0.9355 - val_acc: 0.7268\n",
      "\n",
      "Epoch 00012: val_acc did not improve from 0.76200\n",
      "Epoch 13/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.3287 - acc: 0.8879 - val_loss: 0.8276 - val_acc: 0.7676\n",
      "\n",
      "Epoch 00013: val_acc improved from 0.76200 to 0.76760, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 14/100\n",
      "45000/45000 [==============================] - 25s 559us/step - loss: 0.3244 - acc: 0.8890 - val_loss: 0.7968 - val_acc: 0.7638\n",
      "\n",
      "Epoch 00014: val_acc did not improve from 0.76760\n",
      "Epoch 15/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.2688 - acc: 0.9072 - val_loss: 0.8572 - val_acc: 0.7642\n",
      "\n",
      "Epoch 00015: val_acc did not improve from 0.76760\n",
      "Epoch 16/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.2439 - acc: 0.9162 - val_loss: 0.8708 - val_acc: 0.7630\n",
      "\n",
      "Epoch 00016: val_acc did not improve from 0.76760\n",
      "Epoch 17/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.2335 - acc: 0.9193 - val_loss: 0.8053 - val_acc: 0.7756\n",
      "\n",
      "Epoch 00017: val_acc improved from 0.76760 to 0.77560, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 18/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.1989 - acc: 0.9315 - val_loss: 1.1628 - val_acc: 0.7336\n",
      "\n",
      "Epoch 00018: val_acc did not improve from 0.77560\n",
      "Epoch 19/100\n",
      "45000/45000 [==============================] - 25s 556us/step - loss: 0.1859 - acc: 0.9355 - val_loss: 0.8352 - val_acc: 0.7852\n",
      "\n",
      "Epoch 00019: val_acc improved from 0.77560 to 0.78520, saving model to checkpoint.mnist-cnn-tipo-RESNET.hdf5\n",
      "Epoch 20/100\n",
      "45000/45000 [==============================] - 25s 559us/step - loss: 0.1714 - acc: 0.9411 - val_loss: 0.9531 - val_acc: 0.7596\n",
      "\n",
      "Epoch 00020: val_acc did not improve from 0.78520\n",
      "Epoch 21/100\n",
      "45000/45000 [==============================] - 25s 557us/step - loss: 0.1640 - acc: 0.9431 - val_loss: 0.9314 - val_acc: 0.7768\n",
      "\n",
      "Epoch 00021: val_acc did not improve from 0.78520\n",
      "Epoch 22/100\n",
      "45000/45000 [==============================] - 25s 562us/step - loss: 0.1476 - acc: 0.9493 - val_loss: 0.9679 - val_acc: 0.7686\n",
      "\n",
      "Epoch 00022: val_acc did not improve from 0.78520\n",
      "Epoch 23/100\n",
      "45000/45000 [==============================] - 25s 556us/step - loss: 0.1412 - acc: 0.9512 - val_loss: 1.1865 - val_acc: 0.7596\n",
      "\n",
      "Epoch 00023: val_acc did not improve from 0.78520\n",
      "Epoch 24/100\n",
      "45000/45000 [==============================] - 25s 558us/step - loss: 0.1360 - acc: 0.9546 - val_loss: 1.0739 - val_acc: 0.7552\n",
      "\n",
      "Epoch 00024: val_acc did not improve from 0.78520\n",
      "Epoch 00024: early stopping\n"
     ]
    }
   ],
   "source": [
    "batch_size = 64\n",
    "epocs = 100\n",
    "\n",
    "history = model.fit(x_train, \n",
    "        y_train_categorical,\n",
    "        epochs=epocs, batch_size=batch_size, \n",
    "        verbose=1, \n",
    "        validation_data = (x_valid, y_val_categorical),\n",
    "        callbacks=[checkpointer, earlystop],\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 304,
       "width": 1049
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, (ax1, ax2) = plt.subplots(1, 2, sharex=True, figsize=(18,5))\n",
    "ax1.plot(history.history['loss'], label=\"loss\")\n",
    "ax1.plot(history.history['val_loss'], label=\"val_loss\")\n",
    "ax1.legend()\n",
    "\n",
    "ax2.plot(history.history['acc'], label=\"acc\")\n",
    "ax2.plot(history.history['val_acc'], label=\"val_acc\")\n",
    "ax2.legend()\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights('checkpoint.mnist-cnn.hdf5')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (tensorflow-gpu)",
   "language": "python",
   "name": "tensorflow-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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}