Making new model from Convbase.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python2.7/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.applications import VGG16\n",
    "\n",
    "conv_base = VGG16(weights='imagenet', include_top=False, input_shape=(150,150,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import models\n",
    "from keras import layers\n",
    "\n",
    "model = models.Sequential()\n",
    "model.add(conv_base)\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dense(1, activation='sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "vgg16 (Model)                (None, 512, 4, 4)         14714688  \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 8192)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 256)               2097408   \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 16,812,353\n",
      "Trainable params: 16,812,353\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(model.trainable_weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(model.trainable_weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2000 images belonging to 2 classes.\n",
      "Found 1000 images belonging to 2 classes.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING (theano.tensor.blas): We did not find a dynamic library in the library_dir of the library we use for blas. If you use ATLAS, make sure to compile it with dynamics library.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-9-f056c399d6d7>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     39\u001b[0m     \u001b[0mvalidation_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidation_generator\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     40\u001b[0m     \u001b[0mvalidation_steps\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m50\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 41\u001b[0;31m     verbose=2)\n\u001b[0m",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/keras/legacy/interfaces.pyc\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     89\u001b[0m                 warnings.warn('Update your `' + object_name +\n\u001b[1;32m     90\u001b[0m                               '` call to the Keras 2 API: ' + signature, stacklevel=2)\n\u001b[0;32m---> 91\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     92\u001b[0m         \u001b[0mwrapper\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_original_function\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     93\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mwrapper\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)\u001b[0m\n\u001b[1;32m   1413\u001b[0m             \u001b[0muse_multiprocessing\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0muse_multiprocessing\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1414\u001b[0m             \u001b[0mshuffle\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mshuffle\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1415\u001b[0;31m             initial_epoch=initial_epoch)\n\u001b[0m\u001b[1;32m   1416\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1417\u001b[0m     \u001b[0;34m@\u001b[0m\u001b[0minterfaces\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlegacy_generator_methods_support\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/keras/engine/training_generator.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(model, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)\u001b[0m\n\u001b[1;32m    211\u001b[0m                 outs = model.train_on_batch(x, y,\n\u001b[1;32m    212\u001b[0m                                             \u001b[0msample_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 213\u001b[0;31m                                             class_weight=class_weight)\n\u001b[0m\u001b[1;32m    214\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    215\u001b[0m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mto_list\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mtrain_on_batch\u001b[0;34m(self, x, y, sample_weight, class_weight)\u001b[0m\n\u001b[1;32m   1213\u001b[0m             \u001b[0mins\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0msample_weights\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1214\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_make_train_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1215\u001b[0;31m         \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1216\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0munpack_singleton\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1217\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/keras/backend/theano_backend.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   1271\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1272\u001b[0m         \u001b[0;32massert\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1273\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfunction\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1274\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1275\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python2.7/dist-packages/theano/compile/function_module.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    901\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    902\u001b[0m             \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 903\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0moutput_subset\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0mNone\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    904\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput_subset\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moutput_subset\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    905\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras import optimizers\n",
    "\n",
    "train_dir = '/home/geist/Desktop/projects/kaggle/cats_dogs/cats_and_dogs_small/train'\n",
    "validation_dir ='/home/geist/Desktop/projects/kaggle/cats_dogs/cats_and_dogs_small/validation'\n",
    "\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1./255,\n",
    "    rotation_range=40,\n",
    "    width_shift_range=0.2,\n",
    "    height_shift_range=0.2,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    fill_mode='nearest')\n",
    "\n",
    "test_datagen = ImageDataGenerator(rescale=1./255)\n",
    "\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    train_dir,\n",
    "    target_size=(150,150),\n",
    "    batch_size=20,\n",
    "    class_mode='binary')\n",
    "\n",
    "validation_generator = test_datagen.flow_from_directory(\n",
    "    validation_dir,\n",
    "    target_size=(150,150),\n",
    "    batch_size=20,\n",
    "    class_mode='binary')\n",
    "\n",
    "model.compile(loss='binary_crossentropy',\n",
    "             optimizer= optimizers.RMSprop(lr=2e-5),\n",
    "             metrics=['acc'])\n",
    "\n",
    "history = model.fit_generator(\n",
    "    train_generator,\n",
    "    steps_per_epoch=100,\n",
    "    epochs=30,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=50,\n",
    "    verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_4 (InputLayer)         (None, 150, 150, 3)       0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 150, 150, 64)      1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 150, 150, 64)      36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 75, 75, 64)        0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 75, 75, 128)       73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 75, 75, 128)       147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 37, 37, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 37, 37, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 37, 37, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 37, 37, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 18, 18, 256)       0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 18, 18, 512)       1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 18, 18, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 18, 18, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 9, 9, 512)         0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 4, 4, 512)         0         \n",
      "=================================================================\n",
      "Total params: 14,714,688\n",
      "Trainable params: 0\n",
      "Non-trainable params: 14,714,688\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "conv_base.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base.trainable = True\n",
    "\n",
    "set_trainable = False\n",
    "for layer in conv_base.layers:\n",
    "    if layer.name == 'block5_conv1':\n",
    "        set_trainable = True\n",
    "    if set_trainable:\n",
    "        layer.trainable = True\n",
    "    else:\n",
    "        layer.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " - 18s - loss: 0.2867 - acc: 0.8725 - val_loss: 0.2248 - val_acc: 0.9120\n",
      "Epoch 2/100\n",
      " - 17s - loss: 0.2606 - acc: 0.8905 - val_loss: 0.2133 - val_acc: 0.9140\n",
      "Epoch 3/100\n",
      " - 17s - loss: 0.2293 - acc: 0.9015 - val_loss: 0.2191 - val_acc: 0.9090\n",
      "Epoch 4/100\n",
      " - 17s - loss: 0.2264 - acc: 0.9030 - val_loss: 0.1946 - val_acc: 0.9260\n",
      "Epoch 5/100\n",
      " - 17s - loss: 0.1974 - acc: 0.9240 - val_loss: 0.2105 - val_acc: 0.9180\n",
      "Epoch 6/100\n",
      " - 17s - loss: 0.2000 - acc: 0.9240 - val_loss: 0.2088 - val_acc: 0.9170\n",
      "Epoch 7/100\n",
      " - 17s - loss: 0.1764 - acc: 0.9290 - val_loss: 0.2000 - val_acc: 0.9200\n",
      "Epoch 8/100\n",
      " - 17s - loss: 0.1732 - acc: 0.9270 - val_loss: 0.1928 - val_acc: 0.9260\n",
      "Epoch 9/100\n",
      " - 17s - loss: 0.1635 - acc: 0.9370 - val_loss: 0.2052 - val_acc: 0.9240\n",
      "Epoch 10/100\n",
      " - 17s - loss: 0.1621 - acc: 0.9335 - val_loss: 0.1887 - val_acc: 0.9210\n",
      "Epoch 11/100\n",
      " - 17s - loss: 0.1381 - acc: 0.9415 - val_loss: 0.1888 - val_acc: 0.9270\n",
      "Epoch 12/100\n",
      " - 17s - loss: 0.1358 - acc: 0.9460 - val_loss: 0.2281 - val_acc: 0.9200\n",
      "Epoch 13/100\n",
      " - 17s - loss: 0.1361 - acc: 0.9425 - val_loss: 0.1804 - val_acc: 0.9310\n",
      "Epoch 14/100\n",
      " - 17s - loss: 0.1374 - acc: 0.9395 - val_loss: 0.1788 - val_acc: 0.9350\n",
      "Epoch 15/100\n",
      " - 17s - loss: 0.1184 - acc: 0.9565 - val_loss: 0.1849 - val_acc: 0.9270\n",
      "Epoch 16/100\n",
      " - 17s - loss: 0.1306 - acc: 0.9490 - val_loss: 0.1859 - val_acc: 0.9280\n",
      "Epoch 17/100\n",
      " - 17s - loss: 0.1190 - acc: 0.9520 - val_loss: 0.1926 - val_acc: 0.9300\n",
      "Epoch 18/100\n",
      " - 17s - loss: 0.0992 - acc: 0.9640 - val_loss: 0.1752 - val_acc: 0.9380\n",
      "Epoch 19/100\n",
      " - 17s - loss: 0.1042 - acc: 0.9575 - val_loss: 0.2109 - val_acc: 0.9180\n",
      "Epoch 20/100\n",
      " - 17s - loss: 0.0935 - acc: 0.9595 - val_loss: 0.2503 - val_acc: 0.9100\n",
      "Epoch 21/100\n",
      " - 17s - loss: 0.0908 - acc: 0.9645 - val_loss: 0.1872 - val_acc: 0.9270\n",
      "Epoch 22/100\n",
      " - 17s - loss: 0.0810 - acc: 0.9685 - val_loss: 0.2418 - val_acc: 0.9220\n",
      "Epoch 23/100\n",
      " - 17s - loss: 0.0977 - acc: 0.9655 - val_loss: 0.1829 - val_acc: 0.9350\n",
      "Epoch 24/100\n",
      " - 17s - loss: 0.0668 - acc: 0.9740 - val_loss: 0.1855 - val_acc: 0.9310\n",
      "Epoch 25/100\n",
      " - 17s - loss: 0.0806 - acc: 0.9680 - val_loss: 0.1860 - val_acc: 0.9340\n",
      "Epoch 26/100\n",
      " - 17s - loss: 0.0886 - acc: 0.9650 - val_loss: 0.1789 - val_acc: 0.9350\n",
      "Epoch 27/100\n",
      " - 17s - loss: 0.0707 - acc: 0.9750 - val_loss: 0.1699 - val_acc: 0.9350\n",
      "Epoch 28/100\n",
      " - 17s - loss: 0.0715 - acc: 0.9745 - val_loss: 0.2194 - val_acc: 0.9290\n",
      "Epoch 29/100\n",
      " - 17s - loss: 0.0724 - acc: 0.9730 - val_loss: 0.2399 - val_acc: 0.9230\n",
      "Epoch 30/100\n",
      " - 17s - loss: 0.0711 - acc: 0.9735 - val_loss: 0.2140 - val_acc: 0.9280\n",
      "Epoch 31/100\n",
      " - 17s - loss: 0.0645 - acc: 0.9755 - val_loss: 0.2337 - val_acc: 0.9270\n",
      "Epoch 32/100\n",
      " - 17s - loss: 0.0637 - acc: 0.9755 - val_loss: 0.2334 - val_acc: 0.9280\n",
      "Epoch 33/100\n",
      " - 17s - loss: 0.0620 - acc: 0.9755 - val_loss: 0.1799 - val_acc: 0.9440\n",
      "Epoch 34/100\n",
      " - 17s - loss: 0.0584 - acc: 0.9770 - val_loss: 0.1870 - val_acc: 0.9330\n",
      "Epoch 35/100\n",
      " - 17s - loss: 0.0551 - acc: 0.9795 - val_loss: 0.2149 - val_acc: 0.9370\n",
      "Epoch 36/100\n",
      " - 17s - loss: 0.0557 - acc: 0.9815 - val_loss: 0.3004 - val_acc: 0.9200\n",
      "Epoch 37/100\n",
      " - 17s - loss: 0.0578 - acc: 0.9775 - val_loss: 0.1754 - val_acc: 0.9420\n",
      "Epoch 38/100\n",
      " - 17s - loss: 0.0524 - acc: 0.9795 - val_loss: 0.2296 - val_acc: 0.9280\n",
      "Epoch 39/100\n",
      " - 17s - loss: 0.0525 - acc: 0.9825 - val_loss: 0.2078 - val_acc: 0.9310\n",
      "Epoch 40/100\n",
      " - 17s - loss: 0.0580 - acc: 0.9805 - val_loss: 0.2237 - val_acc: 0.9270\n",
      "Epoch 41/100\n",
      " - 17s - loss: 0.0537 - acc: 0.9780 - val_loss: 0.2215 - val_acc: 0.9310\n",
      "Epoch 42/100\n",
      " - 17s - loss: 0.0519 - acc: 0.9780 - val_loss: 0.2679 - val_acc: 0.9280\n",
      "Epoch 43/100\n",
      " - 17s - loss: 0.0457 - acc: 0.9860 - val_loss: 0.2112 - val_acc: 0.9380\n",
      "Epoch 44/100\n",
      " - 17s - loss: 0.0610 - acc: 0.9780 - val_loss: 0.1846 - val_acc: 0.9430\n",
      "Epoch 45/100\n",
      " - 17s - loss: 0.0482 - acc: 0.9830 - val_loss: 0.1861 - val_acc: 0.9350\n",
      "Epoch 46/100\n",
      " - 17s - loss: 0.0317 - acc: 0.9895 - val_loss: 0.2035 - val_acc: 0.9440\n",
      "Epoch 47/100\n",
      " - 17s - loss: 0.0389 - acc: 0.9845 - val_loss: 0.2700 - val_acc: 0.9260\n",
      "Epoch 48/100\n",
      " - 17s - loss: 0.0275 - acc: 0.9895 - val_loss: 0.3032 - val_acc: 0.9280\n",
      "Epoch 49/100\n",
      " - 17s - loss: 0.0513 - acc: 0.9810 - val_loss: 0.2192 - val_acc: 0.9330\n",
      "Epoch 50/100\n",
      " - 17s - loss: 0.0380 - acc: 0.9880 - val_loss: 0.3634 - val_acc: 0.9130\n",
      "Epoch 51/100\n",
      " - 17s - loss: 0.0340 - acc: 0.9875 - val_loss: 0.2825 - val_acc: 0.9270\n",
      "Epoch 52/100\n",
      " - 17s - loss: 0.0299 - acc: 0.9875 - val_loss: 0.2514 - val_acc: 0.9340\n",
      "Epoch 53/100\n",
      " - 17s - loss: 0.0398 - acc: 0.9875 - val_loss: 0.2753 - val_acc: 0.9330\n",
      "Epoch 54/100\n",
      " - 17s - loss: 0.0295 - acc: 0.9875 - val_loss: 0.2740 - val_acc: 0.9330\n",
      "Epoch 55/100\n",
      " - 17s - loss: 0.0385 - acc: 0.9875 - val_loss: 0.2536 - val_acc: 0.9350\n",
      "Epoch 56/100\n",
      " - 17s - loss: 0.0351 - acc: 0.9870 - val_loss: 0.3843 - val_acc: 0.9130\n",
      "Epoch 57/100\n",
      " - 17s - loss: 0.0351 - acc: 0.9885 - val_loss: 0.2173 - val_acc: 0.9450\n",
      "Epoch 58/100\n",
      " - 17s - loss: 0.0421 - acc: 0.9850 - val_loss: 0.2105 - val_acc: 0.9450\n",
      "Epoch 59/100\n",
      " - 17s - loss: 0.0426 - acc: 0.9885 - val_loss: 0.2507 - val_acc: 0.9310\n",
      "Epoch 60/100\n",
      " - 17s - loss: 0.0316 - acc: 0.9885 - val_loss: 0.2277 - val_acc: 0.9380\n",
      "Epoch 61/100\n",
      " - 17s - loss: 0.0407 - acc: 0.9875 - val_loss: 0.2122 - val_acc: 0.9390\n",
      "Epoch 62/100\n",
      " - 17s - loss: 0.0236 - acc: 0.9930 - val_loss: 0.2325 - val_acc: 0.9400\n",
      "Epoch 63/100\n",
      " - 17s - loss: 0.0446 - acc: 0.9865 - val_loss: 0.3178 - val_acc: 0.9210\n",
      "Epoch 64/100\n",
      " - 17s - loss: 0.0176 - acc: 0.9945 - val_loss: 0.2195 - val_acc: 0.9410\n",
      "Epoch 65/100\n",
      " - 17s - loss: 0.0358 - acc: 0.9880 - val_loss: 0.2391 - val_acc: 0.9420\n",
      "Epoch 66/100\n",
      " - 17s - loss: 0.0287 - acc: 0.9890 - val_loss: 0.2142 - val_acc: 0.9430\n",
      "Epoch 67/100\n",
      " - 17s - loss: 0.0306 - acc: 0.9920 - val_loss: 0.2564 - val_acc: 0.9350\n",
      "Epoch 68/100\n",
      " - 17s - loss: 0.0248 - acc: 0.9920 - val_loss: 0.2371 - val_acc: 0.9460\n",
      "Epoch 69/100\n",
      " - 17s - loss: 0.0305 - acc: 0.9900 - val_loss: 0.4270 - val_acc: 0.9050\n",
      "Epoch 70/100\n",
      " - 17s - loss: 0.0377 - acc: 0.9855 - val_loss: 0.3713 - val_acc: 0.9140\n",
      "Epoch 71/100\n",
      " - 17s - loss: 0.0229 - acc: 0.9920 - val_loss: 0.2522 - val_acc: 0.9330\n",
      "Epoch 72/100\n",
      " - 17s - loss: 0.0271 - acc: 0.9900 - val_loss: 0.4651 - val_acc: 0.9120\n",
      "Epoch 73/100\n",
      " - 17s - loss: 0.0286 - acc: 0.9910 - val_loss: 0.2159 - val_acc: 0.9450\n",
      "Epoch 74/100\n",
      " - 17s - loss: 0.0269 - acc: 0.9885 - val_loss: 0.2225 - val_acc: 0.9470\n",
      "Epoch 75/100\n",
      " - 17s - loss: 0.0229 - acc: 0.9930 - val_loss: 0.2866 - val_acc: 0.9350\n",
      "Epoch 76/100\n",
      " - 17s - loss: 0.0339 - acc: 0.9880 - val_loss: 0.2645 - val_acc: 0.9360\n",
      "Epoch 77/100\n",
      " - 17s - loss: 0.0248 - acc: 0.9935 - val_loss: 0.2509 - val_acc: 0.9370\n",
      "Epoch 78/100\n",
      " - 17s - loss: 0.0243 - acc: 0.9905 - val_loss: 0.2924 - val_acc: 0.9340\n",
      "Epoch 79/100\n",
      " - 17s - loss: 0.0189 - acc: 0.9920 - val_loss: 0.2724 - val_acc: 0.9380\n",
      "Epoch 80/100\n",
      " - 17s - loss: 0.0231 - acc: 0.9905 - val_loss: 0.2329 - val_acc: 0.9420\n",
      "Epoch 81/100\n",
      " - 17s - loss: 0.0238 - acc: 0.9930 - val_loss: 0.2440 - val_acc: 0.9430\n",
      "Epoch 82/100\n",
      " - 17s - loss: 0.0193 - acc: 0.9930 - val_loss: 0.2646 - val_acc: 0.9350\n",
      "Epoch 83/100\n",
      " - 17s - loss: 0.0171 - acc: 0.9930 - val_loss: 0.2408 - val_acc: 0.9420\n",
      "Epoch 84/100\n",
      " - 17s - loss: 0.0193 - acc: 0.9930 - val_loss: 0.2497 - val_acc: 0.9440\n",
      "Epoch 85/100\n",
      " - 17s - loss: 0.0187 - acc: 0.9950 - val_loss: 0.2915 - val_acc: 0.9370\n",
      "Epoch 86/100\n",
      " - 17s - loss: 0.0240 - acc: 0.9925 - val_loss: 0.2693 - val_acc: 0.9400\n",
      "Epoch 87/100\n",
      " - 17s - loss: 0.0224 - acc: 0.9920 - val_loss: 0.3217 - val_acc: 0.9330\n",
      "Epoch 88/100\n",
      " - 17s - loss: 0.0174 - acc: 0.9940 - val_loss: 0.2600 - val_acc: 0.9410\n",
      "Epoch 89/100\n",
      " - 17s - loss: 0.0172 - acc: 0.9945 - val_loss: 0.2992 - val_acc: 0.9400\n",
      "Epoch 90/100\n",
      " - 17s - loss: 0.0184 - acc: 0.9935 - val_loss: 0.2787 - val_acc: 0.9380\n",
      "Epoch 91/100\n",
      " - 17s - loss: 0.0193 - acc: 0.9945 - val_loss: 0.2920 - val_acc: 0.9360\n",
      "Epoch 92/100\n",
      " - 17s - loss: 0.0170 - acc: 0.9950 - val_loss: 0.3267 - val_acc: 0.9340\n",
      "Epoch 93/100\n",
      " - 17s - loss: 0.0171 - acc: 0.9925 - val_loss: 0.2580 - val_acc: 0.9430\n",
      "Epoch 94/100\n",
      " - 17s - loss: 0.0198 - acc: 0.9940 - val_loss: 0.3074 - val_acc: 0.9350\n",
      "Epoch 95/100\n",
      " - 17s - loss: 0.0149 - acc: 0.9950 - val_loss: 0.3009 - val_acc: 0.9320\n",
      "Epoch 96/100\n",
      " - 17s - loss: 0.0144 - acc: 0.9945 - val_loss: 0.4172 - val_acc: 0.9190\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 97/100\n",
      " - 17s - loss: 0.0167 - acc: 0.9925 - val_loss: 0.2953 - val_acc: 0.9370\n",
      "Epoch 98/100\n",
      " - 17s - loss: 0.0162 - acc: 0.9935 - val_loss: 0.2878 - val_acc: 0.9340\n",
      "Epoch 99/100\n",
      " - 17s - loss: 0.0145 - acc: 0.9930 - val_loss: 0.3249 - val_acc: 0.9400\n",
      "Epoch 100/100\n",
      " - 17s - loss: 0.0220 - acc: 0.9915 - val_loss: 0.4674 - val_acc: 0.9200\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss='binary_crossentropy', \n",
    "             optimizer= optimizers.RMSprop(lr=1e-5),\n",
    "             metrics=['acc'])\n",
    "\n",
    "history = model.fit_generator(\n",
    "    train_generator,\n",
    "    steps_per_epoch=100,\n",
    "    epochs=100,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=50,\n",
    "    verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa71071e850>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa7106e3810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "acc = history.history['acc']\n",
    "val_acc = history.history['val_acc']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs = range(len(acc))\n",
    "\n",
    "plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEICAYAAACzliQjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzt3Xl8VNXZwPHfQ1jCKqsbgYBKZQsgBgQRQaWIYMEFRAUL1oqvaKutG0pbEaVaa7X4igtaFRAV6kqVt1YRCy4oQQFlC4uAYTOAQcIihDzvH+dOGCazJZlkkpnn+/nMh7n3nrn33JnwzJnnnHuuqCrGGGOSQ7V4V8AYY0zFsaBvjDFJxIK+McYkEQv6xhiTRCzoG2NMErGgb4wxScSCfhISkRQRyReRlrEsG08icpqIxHz8sYj0E5GNfstrRKR3NGVLcaznROSe0r7emGhUj3cFTGQiku+3WAf4CTjiLd+gqjNLsj9VPQLUi3XZZKCqp8diPyLya2Ckqvb12/evY7FvY8KxoF8FqGpR0PVakr9W1Q9ClReR6qpaUBF1MyYS+3usXCy9kwBE5AERmSUir4jIXmCkiPQUkUUikici20TkcRGp4ZWvLiIqIq285Ze87f8nIntF5DMRaV3Sst72i0QkW0T2iMj/isgnIjI6RL2jqeMNIrJORH4Qkcf9XpsiIo+JyC4RWQ8MCPP+/EFEXg1YN0VEHvWe/1pEVnnns95rhYfaV46I9PWe1xGRGV7dVgBnBjnuBm+/K0RksLc+A3gC6O2lznb6vbcT/F7/P9657xKRt0TkpGjem5K8z776iMgHIrJbRLaLyJ1+x/mj9578KCJZInJysFSaiHzs+5y993OBd5zdwB9EpI2IzPfOZaf3vh3n9/p07xxzve2TRSTVq3M7v3Inich+EWkS6nxNBKpqjyr0ADYC/QLWPQAcAn6B+yKvDXQDzsL9mjsFyAZu9spXBxRo5S2/BOwEMoEawCzgpVKUPR7YCwzxtv0eOAyMDnEu0dTxbeA4oBWw23fuwM3ACiANaAIscH/OQY9zCpAP1PXb9/dAprf8C6+MAOcDB4BO3rZ+wEa/feUAfb3njwAfAY2AdGBlQNkrgJO8z+Rqrw4neNt+DXwUUM+XgAne8/5eHbsAqcCTwIfRvDclfJ+PA3YAtwC1gAZAd2/b3cAyoI13Dl2AxsBpge818LHvc/bOrQC4EUjB/T3+DLgAqOn9nXwCPOJ3Pt9472ddr3wvb9tUYJLfcW4D3oz3/8Oq/Ih7BexRwg8sdND/MMLrbgf+6T0PFsif9is7GPimFGV/BSz02ybANkIE/Sjr2MNv+xvA7d7zBbg0l2/bwMBAFLDvRcDV3vOLgOwwZd8BbvKehwv6m/0/C2Csf9kg+/0GGOQ9jxT0pwF/9tvWANePkxbpvSnh+3wNkBWi3HpffQPWRxP0N0Sow1Bgsfe8N7AdSAlSrhfwLSDe8lLgslj/v0qmh6V3Esd3/gsi0lZE3vV+rv8ITASahnn9dr/n+wnfeRuq7Mn+9VD3vzQn1E6irGNUxwI2hakvwMvAVd7zq4Gizm8RuVhEPvfSG3m4Vna498rnpHB1EJHRIrLMS1HkAW2j3C+48yvan6r+CPwANPcrE9VnFuF9bgGsC1GHFrjAXxqBf48nishsEdni1eHFgDpsVDdo4Biq+gnuV8M5ItIRaAm8W8o6GSynn0gChys+g2tZnqaqDYA/4Vre5WkbriUKgIgIxwapQGWp4zZcsPCJNKR0FtBPRNJw6aeXvTrWBl4DHsSlXhoC/4myHttD1UFETgGewqU4mnj7Xe2330jDS7fiUka+/dXHpZG2RFGvQOHe5++AU0O8LtS2fV6d6vitOzGgTOD5/QU36izDq8PogDqki0hKiHpMB0bifpXMVtWfQpQzUbCgn7jqA3uAfV5H2A0VcMx3gK4i8gsRqY7LEzcrpzrOBm4VkeZep95d4Qqr6g5cCuIFYI2qrvU21cLlmXOBIyJyMS73HG0d7hGRhuKuY7jZb1s9XODLxX3//RrX0vfZAaT5d6gGeAW4TkQ6iUgt3JfSQlUN+cspjHDv8xygpYjcLCI1RaSBiHT3tj0HPCAip4rTRUQa477stuMGDKSIyBj8vqDC1GEfsEdEWuBSTD6fAbuAP4vrHK8tIr38ts/ApYOuxn0BmDKwoJ+4bgNG4TpWn8G1dMuVF1iHA4/i/hOfCnyFa+HFuo5PAfOAr4HFuNZ6JC/jcvQv+9U5D/gd8CauM3Qo7ssrGvfifnFsBP4Pv4CkqsuBx4EvvDJtgc/9Xvs+sBbYISL+aRrf6/+NS8O86b2+JTAiynoFCvk+q+oe4OfA5biO42ygj7f5r8BbuPf5R1ynaqqXtrseuAfXqX9awLkFcy/QHfflMwd43a8OBcDFQDtcq38z7nPwbd+I+5wPqeqnJTx3E8DXOWJMzHk/17cCQ1V1YbzrY6ouEZmO6xyeEO+6VHV2cZaJKREZgPu5fhA35K8A19o1plS8/pEhQEa865IIIqZ3ROR5EfleRL4JsV28izDWichyEenqt22UiKz1HqNiWXFTaZ0DbMD97B8AXGIdb6a0RORB3LUCf1bVzfGuTyKImN4RkXNxF5VMV9WOQbYPBH6DGyd9FjBZVc/yOnyycBfxKLAEOFNVf4jtKRhjjIlWxJa+qi7AdXCFMgT3haCqughoKO5y8QuB91V1txfo3yfMpfLGGGPKXyxy+s059kKMHG9dqPVhNW3aVFu1ahWDahljTPJYsmTJTlUNN0QaiE3QD3YRi4ZZX3wHbpzvGICWLVuSlZUVg2oZY0zyEJFIV6UDsRmnn8OxVyWm4YbphVpfjKpOVdVMVc1s1iziF5UxxphSikXQnwP80hvF0wPYo6rbgPeA/iLSSEQa4eYzeS8GxzPGGFNKEdM7IvIK0BdoKiI5uCvragCo6tPAXNzInXW4SZ+u9bbtFpH7cVdLAkxU1XAdwsYYY8pZxKCvqldF2K7ATSG2PQ88X7qqGWOMiTWbe8cYY5KIBX1jjImBmTOhVSuoVg2aNnWPatXcupkzI7264ljQN8aYAOEC+NixxbeJwDXXwKZNoAq7drmHqlt3zTWujP++4vXFYEHfGJMwomlthyoTbQB/6qni28Ath+Lb5r+vUF8M5f0FUOmmVs7MzFS7OMsYE62ZM2H8eBc4RUIHX9+2cGUqizp1YOpUGFGCOyiIyBJVzYxUzlr6xpgqIVgL3b9VDtG1tit7wAfYv999kZUHC/rGmLjyD+YlzZlD1QjipbG5nCaStqBvjImpwCAeLpceGMxLmzOvLMSbcaxJE/fwX1dSLVvGpk6BLOgbk2QidWSWtOMzUhD3dVBWtZZ6YAAXgfR0uPFG969I8W0zZrjz2bnTPVTdumDlw30x1KkDkyaV03lZR64xicHXobl5MzRu7Nbt3u1ajAMHwty5kTs7/VWljk9/vvr6AuquXcXPIbDM7t3F37NJk0rWkVoW/p9daY8dbUeuBX1jqphgAQJgzBjXAZiMfEE8PT14wIxFUK3sLOgbk4Bmziwe3KtaS7wsgrXQEzWIl5QN2TSmiguWSx85snhrPhECvi+nXdKceWEhbNxoAb8kYnHnLGNMFELl3IM9D8xD+zo9q4pQ/QHWUo8/a+kbE2PRDE0MdTl+RY1qCTYyJdxoknDlw7XI/UeuWEu9crCgb0wphLqgqLIMTQw3zDBU8A01zDBS+cDn/kF8xAi3bMG98rCOXGOiFO0cLxUpsB6lmbPFJAbryDUmBnwt+pLM8VIR6tSBl14q3iK3gG8isY5cY4hu7Hs8An00HZ8W5E1JWNA3SS9w7Ltv6oDyCPKRrgSN51WhJjlY0DdJyz9HH6isAd+GJprKyoK+SSqx7Iz1v/TfN7dNIl/mbxKDBX2TNALTOKUJ+JHmeDGmsrOgb5LG+PFlm5DMAr1JBFEN2RSRASKyRkTWici4INvTRWSeiCwXkY9EJM1v28MiskJEVonI4yKlvaWAMeFFunlHsNx9MMHmNn/pJbu4yCSGiEFfRFKAKcBFQHvgKhFpH1DsEWC6qnYCJgIPeq89G+gFdAI6At2APjGrvTEeX+om3M07IrGx7yYZRNPS7w6sU9UNqnoIeBUYElCmPTDPez7fb7sCqUBNoBZQA9hR1kqb5BOqFe8TLHUTzY2w/Wd39AV3mzrAJLJogn5z4Du/5Rxvnb9lwOXe80uB+iLSRFU/w30JbPMe76nqqsADiMgYEckSkazc3NySnoNJQIGTlv3qV8Fb8b4JzaJN3fjzn1PGgrtJFtF05AbLwQe2nW4HnhCR0cACYAtQICKnAe0AX47/fRE5V1UXHLMz1anAVHBz70RffZOIAkfZBJtWuKxTDqenu0BvTLKJpqWfA7TwW04DtvoXUNWtqnqZqp4BjPfW7cG1+hepar6q5gP/B/SISc1NwvG17oPdKCSWyvOm08ZUdtEE/cVAGxFpLSI1gSuBOf4FRKSpiPj2dTfwvPd8M9BHRKqLSA1cJ26x9I5JXqEmNCurUPPBW8esSXYRg76qFgA3A+/hAvZsVV0hIhNFZLBXrC+wRkSygRMAXzvqNWA98DUu779MVf8V21MwVU15z1zpy9UHmw/ecvcm2dl8+qZCBbuxdzRq1IAGDYrfRjCQzSdvkpXNp28qlbLk69PT4YUXgt/VKfAOURbwjQnPpmEw5a60rftQrXbfWHpjTMlZS9+Um9K07q3D1ZjyZS19E1OlmbrYZq40puJY0DcxU5qpiy3QG1OxLOibMgt3B6pQbJSNMfFhOX1TKmW5qMry9cbEj7X0TYmV9g5U1ro3Jv6spW+iZqNxjKn6rKVvolKasfbWSWtM5WNB34RlnbTGJBZL75iQ/G9BGImlcYypGizom2JKmru3WSyNqTosvWOOUZLcvaVxjKl6rKVvgNK17i3gG1P1WEvfWOvemCRiLX3D+PHWujcmWVhL37B5c/jt1ro3JnFYSz+J+fL44aZRsNa9MYnFWvpJxHeh1ebN0Lgx7N0Lhw4FL2ute2MSkwX9JBHYWbtrV+iyNn2CMYnLgn6CK+k0CiLuAitjTGKyoJ/ASjNJWsuW5VcfY0z8RdWRKyIDRGSNiKwTkXFBtqeLyDwRWS4iH4lImt+2liLyHxFZJSIrRaRV7Kpvwol2KKZPnTourWOMSVwRg76IpABTgIuA9sBVItI+oNgjwHRV7QRMBB702zYd+KuqtgO6A9/HouImskhDMWvUgCZNXErHRukYkxyiael3B9ap6gZVPQS8CgwJKNMemOc9n+/b7n05VFfV9wFUNV9VS9D2NGURLlWTng4vvAA7d0JhoU2UZkyyiCboNwe+81vO8db5WwZc7j2/FKgvIk2AnwF5IvKGiHwlIn/1fjmYcuQbf79p09Epj33q1IGXXrIgb0yyiiboS5B1gZfz3A70EZGvgD7AFqAA11Hc29veDTgFGF3sACJjRCRLRLJyc3Ojr70pJnAOfFWb694Yc1Q0o3dygBZ+y2nAVv8CqroVuAxAROoBl6vqHhHJAb5S1Q3etreAHsA/Al4/FZgKkJmZGeVttk0wwTpvVV3At6GYxphoWvqLgTYi0lpEagJXAnP8C4hIUxHx7etu4Hm/1zYSkWbe8vnAyrJX2wTyT+kEE6lT1xiTHCIGfVUtAG4G3gNWAbNVdYWITBSRwV6xvsAaEckGTgAmea89gkvtzBORr3GpomdjfhZJLprbGtr4e2MMgGi42bbiIDMzU7OysuJdjSolXAsfbB4dY5KBiCxR1cxI5WyWzQQQLnVjnbfGGH82DUMCaNkyeEvfOm+NMYGspV+FRRqPb1MqGGMCWdCvomw8vjGmNCy9U0XZeHxjTGlYS7+KCtV5a+PxjTHhWNCvYiLd19bG4xtjwrH0ThUS6aYo1nlrjInEWvpVSLiboljnrTEmGtbSr0JC5evtvrbGmGhZS78KsDy+MSZWrKVfyVke3xgTS9bSr+Qsj2+MiSVr6Vdylsc3xsSStfQrKcvjG2PKg7X0KyHL4xtjyou19Cshy+MbY8qLtfQrIcvjG2PKi7X0K6FQ+XrL4xtjysqCfiU0aZLL2/uzPL4xJhYs6FdCI0a4vH16ukvpWB7fGBMrltOvRGbOdJ24mze7VM6kSRbojTGxZUG/kggcprlpk1sGC/zGmNix9E4lEWyY5v79br0xxsRKVEFfRAaIyBoRWSci44JsTxeReSKyXEQ+EpG0gO0NRGSLiDwRq4onGrv9oTGmIkQM+iKSAkwBLgLaA1eJSPuAYo8A01W1EzAReDBg+/3Af8te3cRj0y0YYypSNC397sA6Vd2gqoeAV4EhAWXaA/O85/P9t4vImcAJwH/KXt3E4svjb9oUfLsN0zTGxFo0Qb858J3fco63zt8y4HLv+aVAfRFpIiLVgL8Bd4Q7gIiMEZEsEcnKzc2NruYJwKZbMKbs9u51jaMHHoCPP4ZDh9z6/HxYtQry8uJbv8ommtE7EmRdYDLiduAJERkNLAC2AAXAWGCuqn4nEmw33s5UpwJTATIzM0MkOhKHb2hmqBa+TbdgKsr27XDPPXDXXXD66fGuTcmowuuvw623wpYtR9fXrg2pqfDDD265USN49FEYNcr93wLYtQs++wxWr3aPnTuhWTM44QSXUu3XD045JfSxDx+GGjXK79zKUzRBPwdo4becBmz1L6CqW4HLAESkHnC5qu4RkZ5AbxEZC9QDaopIvqoW6wxOFpFm0ATL45uKsXmzC25r10K9evD44/GuUXD5+S7VWc0vL7F4MfzhD/Cf/0DnzvDaa9CmDSxcCP/9Lxw86H4tn3wyPPssXHstvPwyDBkCb70F8+fDkSNuX8cf7wL+559Dbu7R9W3awMCBcMMN0K6dW5ebC+PGwQsvQLducM01cOWV0LRpxb4nZaKqYR+4L4YNQGugJi6V0yGgTFOgmvd8EjAxyH5GA09EOt6ZZ56piSw9XdW1UYI/6tRRfemleNfSJJoVK1T/539UJ09WXblSdc0a1RYtVI87TrVDB9Wf/SzeNSzu4EHVO+9UrVZN9aSTVH/9a9UZM1QHDXL/Vxo3Vn3sMdXDh8Pv58gR1SlTVOvVc69r00b1nntUFy5U3bWreNk1a1Qff1x14EDVWrXcay68UPWBB1QbNVKtXl312mtVO3d222rUUP3971V//LH83otoAFkaIb6qq3IUhWAgkA2sB8Z76yYCg73nQ4G1XpnngFpB9mFBX1VFQgf89HQL+Ca2CgtVn3xSNTXVBSff31q1aqpNm6p++aX7IgDVDRviXdujvvxStWNHV6+RI1WHDVOtX/9osP/zn0seZHfscF94hYXRv+b771Xvv1/1xBPdsc87z+3DZ9ky1euuc9tOPll19uyS7T+WYhr0K/KR6EE/VEs/PT3eNUsea9aorloV71qUXmGh6ocfqr7ySvhyO3eqDhni/r7691fdts0F9meeUf3Nb46+B6tXuzJPPVX+dY/kyBHVhx5yremTTlJ9992j2376SXXRovi0qH/6yf1aChXQP/tMtUsX9z4+8UTF1s3Hgn4l9dJLLoVjKZ34+Pxz12KsV0/1iy/iXZuSKSxU/eAD1XPOOfq3M29e8LJffKHasqVr3T/6qAum4fbbqpXqJZeUT72jlZOjev757ryGDSueeqnsDh92vwSaNlXds6fij29BvxJ76SXXshexlE5F+uor1YYNVVu3Vj3lFJcmWLEi3rWKrLBQ9d//Vu3Vy/2Pbd7c5ZxPPdXl4g8ePLbslCmqNWu6oB/tF9sNN7gvw0OHyuccwlm9WvXuu93nUbeu6vPPxy9FUlZffOE+oz/+MXLZfftUJ06M3f9/C/rGePbtU/30U9cCS0tT/fZb1fXrXfrg5JPdcmVUWKj69tuq3bq5/6ktWrjUwYEDbvt777n1Eya45YMHVUePdusGDnTpnWi9+aZ73X//G/vzUFXdvdsFt8svd78qMjJU+/RRPfNMLepjGDTIpd6qumHD3JfXtm2hy8yfr3raae7ca9WKzXlb0K9krHVfsQ4eVL3iCjfawpcKOfFE1ezso2W+/tptP+kkF1zjWdcePVzL/c47XQrqlVdcYAT3y2TqVJdXDnTlla5Vv2CB6tlnu/L33hs+nRPMnj0uj3733TE5JT1wwNVpwgTV3r3dvsG918OHu76Gc891qaq//lV169bYHLcyyM5WTUlRHTvWLS9bpnrLLapDh7oU2gUXuPfi1FNVZ81yI6jOP7/sv24s6FcilsevWIWFR1u8117rRnpMm6a6ZUvxskuXqnbq5MpecYXqunWu3ObNFZeXvesud3z/4Aiq7dqpTp8efkji1q2qDRq48rVrq/7zn6WvR+/eql27lv71WVmqo0a599N3HiKqmZmq48a5TtiSfhlVVTfe6N4D3xdxaqr7PDMyVM84w33m+/a5sk8/7cpMn162Y1rQr0RsxI7q2rXuZ2/PnsVHX6xbpzp+vCsTC3/5y9EWbzQOHXJjsGvWPPbzadxY9YcfYlOnUBYscIHx+uvd8q5dbiz6229HHyBnzHAjR5YsKVtdJk1y5719e8lfe+DA0XH/Awa4cfBvvunSOslo2zbXf3TqqaqPPBI+1XbkiPt/0bRpyVJygSzoVwK+lE6ocfki8a5h+Tp0yI1pvuUWN4qkbl2Xu7366qM/ZfPyVE8/XYvyuiNHup/D/p2TJfHmm+59HT685D+XV692Y9qfftqlHMB9GZSVLwcf6Mcfj3Yq791b9uOUVVaWO+cXXyz5ax9+2L32ww9jX6+qKj8/+i/u5cvdL4Nf/ar0x7OgH2fBUjrJ0tJ/9lnV9u2PXgxUrZrqmDGu9XP//W7d88+7/xCDBrk/9tdeU73ttmPfszp1XEC8//7oWtwvv+x+Rnfrprp/f9nP46KLXOvL9zO8pBYtcjn3lBTV//3f4tuvu869Nx9/XLZ6xsqRI+5q1YyMkqVhdu50LfxBg8qvbsngnntUf/vb0qfALOjHWVWbbmHFCtWbbnIdb2+95XLapbF6tQv2Xbu6PO60acembQoKXKdVnTpH8+5TphzdnpvrvjQmTXJfAv37uzINGqjefrsbc/7AA25I3FtvuRbykSPuPwy4zsHc3LK9Fz4LF7p9Tp4cXfnCQtdH8OCDqt27H633KaeoNmvmWn4+ixa57XfcEZu6xsorr7h6vfxy9K/5/e/dl9fXX5dfvZKBdeRWcVVluoXt290Y7WrVXCvZv96jRpWsxVxYqNqvn2v1hcsLb93qgiC41m6kP/Yvv3QjH4K9pzVquLHq4OZmCTbCpSzOOcflqkPtt7DQpUV+9zs3ft5Xr86d3Vj6H390w0XB9TX4XnPOOarHHx//+VoCHTniOmJPOy26MfsbNri+kOuuK/+6mfAs6MdZZe+8LSxUfe45d0FO9erusvzcXNca/ewzN3RQxI00iHZOltmz3TkGS2UEWrjQHbMkufu8PPc4dMgF4Q8/dC3lXr3cMcvjgp65c905vfDCsesLC12ruG1bt71mTTcM8fnng48SuvBC1SZNXJD3jYl/+unY1zcW/vUvV7+pU91yYaHq4sXBv8hHjnSjhnJyKraOpjgL+nHgPxa/SZPio0FKktLJz1ft29d1ehYUxLaeO3aoDh7s6nTeeaEvDHnnHTcCoVEjN+b4gQdU//EPF7ADhzPu3etaumecEfv6xlNhoRsZ06qV6+T94guXmvFdHduliwuOkUap+NI5993n8ubt2kWeHTJeCgvdaJLmzd0oIt+59uhx7Bfr6tXuF+Kdd8avruYoC/oVLFjHbY0aLviX9IKsggJ3EYdvP7ffHrt6zp3r0gq1arlpaSN1Gq1b574YGjcu/qvllFPcVZUDBx7NYX/6aezqWlnMm+feM/9zP/5490upJF9wAwceTVG980751TcW5s8/eq5paa7xAaqvv360zDXXuL/5HTviVk3jx4J+BYtlOue229xr//5318IG17lZFocOuVQIuNEZpel0O3DApXreece1+ocNcx2nmZmu5Xr//WWrY2VWWKi6caMbZTRlikszldTixe79j8XVlxXhr391v+x++sn9KmnXzvWfHDrkrjqtVs39rZrKwYJ+BQvVcVuSsfiHD6v+7W/udTfd5ALD4cMuH1y9eugZFSP5+mv30xzclYKxGM5oSuf110s/Mire5sxxf0NPPuk6+VNTS3chlykfFvQrWFla+hs3uqtHfaM/Bg06Nt+bl+fGvTdvXrIx4+vXu442EZebL8sl+sYUFrqpGpo0cdce3HprvGtk/EUb9P3uOmnKYtIkdx9Pf3XquPX+9u519+J87jl3385TT4VWrWDiRMjIgDffdPfwrO539+LjjoNnnnE3f3700ejq88IL7kbXr78Od94J69fD0KFlOkWT5ETg4YfdTcVr1HB/V6bqEfcFUXlkZmZqVlZWvKtRKjNnwvjx7obTLVu6gD9ihNu2eTNcfDF8/fXR8k2aQO/e7nHJJXDKKeH3f/nl8N57sG4dnHhi6HLz50P//tCnD0yf7m4ObUys3H23u5n4734X75oYfyKyRFUzI5azoF824QK9z8GDLrBnZ7vWUceO0KGDC/LVSvBba906aN8eRo2CZ58NXiY7G3r0gJNOgk8/db8SjDGJL9qgXz1SARPazJkwZgzs3++WN21yy3Bs4L/5ZsjKcmmbIUNKf7zTToObboLHH4erroIVK1xLfvNmF+h794apUyElBf71Lwv4xpjirKVfBq1auUAfKD0dNm50z5991n0R/OEPcP/9ZT/m7t0u+P/wg1s+4wz3y2HRIli7FmrWhHnz4Jxzyn4sY0zVYS39CrB5c/j18+e7Vv6FF8KECbE5ZuPGrnX/2Weutd+x49Ft27e7Xx2R+gaMMcnLgn4ZtGwZvKXfsqVref/iF9CmDbz8sku5xMrFF7tHoHCdu8YYA0Q3ZFNEBojIGhFZJyLjgmxPF5F5IrJcRD4SkTRvfRcR+UxEVnjbhsf6BOIp1DDNG26Aiy5yQfj9913r3BhjKoOIQV9EUoApwEVAe+AqEWkfUOwRYLqqdgImAg966/cDv1TVDsAA4O8i0jBWlY+3ESNcp2rdukfX1avnxjLXrQsffOBG0RhjTGURTXqIy3CVAAAX8ElEQVSnO7BOVTcAiMirwBBgpV+Z9oBv1O584C0AVc32FVDVrSLyPdAMyCt71eNv1Sp3sdT+/XDrrS7Qb9gA+/bBX/7iOnqNMaYyiSboNwe+81vOAc4KKLMMuByYDFwK1BeRJqq6y1dARLoDNYH1ZapxHOXlwbvvwj//6dI2+/e7cfZ33QUPPhj59cYYE2/R5PQlyLrAcZ63A31E5CugD7AFKCjagchJwAzgWlUtLHYAkTEikiUiWbm5uVFXvqIUFrppE9q0gZEj4e23j47NLyx0KZ6ZM+NbR2OMiUY0QT8HaOG3nAZs9S+gqltV9TJVPQMY763bAyAiDYB3gT+o6qJgB1DVqaqaqaqZzZo1K8VplJ8VK9yFT9df7+ayCTZCZv9+d1WuMcZUdtEE/cVAGxFpLSI1gSuBOf4FRKSpiPj2dTfwvLe+JvAmrpP3n7GrdsVYtgzOPdcNy5wxAxYuhB07gpcNNWbfGGMqk4hBX1ULgJuB94BVwGxVXSEiE0VksFesL7BGRLKBEwDf3JJXAOcCo0VkqffoEuuTKA/Ll8MFF7ghmJ9+6tI6Im4MfjCh1htjTGVi0zAE8fnn7uKnWrXgo4/ctAc+gfPtgPtimDq1+ERrxhhTUaKdhsHm0/eowocfwoABLodfq5abRsEX8GfOdEMwr7kGatd20yKLuHl2LOAbY6qKpJ2G4cgRF+QXLnS5+6VLXV7++OPdlbY33giNGrmyga37Xbtc637GDAv2xpiqJenSO2vXwtNPu/lwtm934+xPPx06d4bzz3e5+9q1j31NNLNpGmNMPNksm0Fs2wZnnQX5+TBokEvVDBhQfP6cQJFm0zTGmKoiaYK+qpsI7cABNzKnbdvoXxtuNk1jjKlKkqYj9+WX3d2kJk0qWcCH6G96bowxlV1CtfR37oTFi+GLL1zHbIcOLkffsCH89rfQsyfcckvJ9+vrrI10L1xjjKnsEqYjd9Omo7Nairi7R337rZsbp3FjN/Pl0qUlb+UbY0xVkHTj9Fu2hL/9zY2t37MH1q2DnBw39XHbtjB5cskDvm9sfrVq7l+bVM0YU9UlTEs/1uzKW2NMVZJ0Lf1YGz/+2IAPNpumMabqs6Afgo3NN8YkIgv6IdhsmsaYRGRBPwQbm2+MSUQW9EMYMcJ12qan22yaxpjEkVAXZ8XaiBEW5I0xicVa+gFsbL4xJpFZS99P4Nj8TZvcMliL3xiTGKyl78fG5htjEp0FfT82Nt8Yk+gs6PuxsfnGmERnQd+Pjc03xiQ6C/p+bGy+MSbR2eidADY23xiTyKJq6YvIABFZIyLrRGRckO3pIjJPRJaLyEcikua3bZSIrPUeo2JZeWOMMSUTMeiLSAowBbgIaA9cJSLtA4o9AkxX1U7AROBB77WNgXuBs4DuwL0i0ih21Y8NuyDLGJMsomnpdwfWqeoGVT0EvAoMCSjTHpjnPZ/vt/1C4H1V3a2qPwDvAwPKXu3Y8V2QtWkTqB69IMsCvzEmEUUT9JsD3/kt53jr/C0DLveeXwrUF5EmUb4WERkjIlkikpWbmxtt3WPCLsgyxiSTaIK+BFkXeI/F24E+IvIV0AfYAhRE+VpUdaqqZqpqZrNmzaKoUuzYBVnGmGQSTdDPAVr4LacBW/0LqOpWVb1MVc8Axnvr9kTz2nizC7KMMckkmqC/GGgjIq1FpCZwJTDHv4CINBUR377uBp73nr8H9BeRRl4Hbn9vXaVhF2QZY5JJxKCvqgXAzbhgvQqYraorRGSiiAz2ivUF1ohINnACMMl77W7gftwXx2Jgoreu0rALsowxyURUi6XY4yozM1OzsrLiXQ1jjKlSRGSJqmZGKmfTMBhjTBKxoG+MMUkkaYO+XYVrjElGSTnhmt0W0RiTrJKypW9X4RpjklVSBn27CtcYk6ySMujbVbjGmGSVlEHfrsI1xiSrpAz6dhWuMSZZJeXoHbDbIhpjklNStvSNMSZZJW1L35iyOnz4MDk5ORw8eDDeVTFJJDU1lbS0NGrUqFGq11vQN6aUcnJyqF+/Pq1atUIk2P2CjIktVWXXrl3k5OTQunXrUu3D0jvGlNLBgwdp0qSJBXxTYUSEJk2alOnXpQV9Y8rAAr6paGX9m0uqoG+TrBljkl3SBH3fJGubNoHq0UnWLPCbilIejY5JkybRoUMHOnXqRJcuXfj888/LvtMQNm7cyMsvv1y0/OKLL3LzzTeXen8fffQRF198cbH1S5cuZe7cuSXe39atWxk6dGjEcgMHDiQvL6/E+08USRP0bZI1E0/l0ej47LPPeOedd/jyyy9Zvnw5H3zwAS1atIhdpQMEBv3yEi7oFxQUhHzdySefzGuvvRZx/3PnzqVhw4alrl9VlzRB3yZZM/FUHo2Obdu20bRpU2rVqgVA06ZNOfnkkwFo1aoV99xzDz179iQzM5Mvv/ySCy+8kFNPPZWnn34acCNB7rjjDjp27EhGRgazZs0Ku37cuHEsXLiQLl268NhjjwGudT1gwADatGnDnXfeWVS3//znP/Ts2ZOuXbsybNgw8vPzAfj3v/9N27ZtOeecc3jjjTeKndOhQ4f405/+xKxZs+jSpQuzZs1iwoQJjBkzhv79+/PLX/6SjRs30rt3b7p27UrXrl359NNPAfel1LFjR8D9CrnsssuC1q1Vq1bs3LmTjRs30q5dO66//no6dOhA//79OXDgAACLFy+mU6dO9OzZs+i9CJSfn88FF1xA165dycjI4O233y7aNn36dDp16kTnzp255pprANixYweXXnopnTt3pnPnzkX1rnCqWqkeZ555ppaH9HRV18Y69pGeXi6HM0lg5cqVUZcVCf73J1L64+/du1c7d+6sbdq00RtvvFE/+uijom3p6en65JNPqqrqrbfeqhkZGfrjjz/q999/r82aNVNV1ddee0379eunBQUFun37dm3RooVu3bo15Pr58+froEGDio7xwgsvaOvWrTUvL08PHDigLVu21M2bN2tubq727t1b8/PzVVX1oYce0vvuu08PHDigaWlpmp2drYWFhTps2LBj9ue/35tuuqlo+d5779WuXbvq/v37VVV13759euDAAVVVzc7OVl/M+Pbbb7VDhw5h6+Z7b3Jzc/Xbb7/VlJQU/eqrr1RVddiwYTpjxgxVVe3QoYN+8sknqqp61113Fe3X3+HDh3XPnj2qqpqbm6unnnqqFhYW6jfffKM/+9nPNDc3V1VVd+3apaqqV1xxhT722GOqqlpQUKB5eXmRP+QQgv3tAVkaRYxNmpa+TbJm4qk8ZnatV68eS5YsYerUqTRr1ozhw4fz4osvFm0fPHgwABkZGZx11lnUr1+fZs2akZqaSl5eHh9//DFXXXUVKSkpnHDCCfTp04fFixeHXB/MBRdcwHHHHUdqairt27dn06ZNLFq0iJUrV9KrVy+6dOnCtGnT2LRpE6tXr6Z169a0adMGEWHkyJFRn+vgwYOpXbs24C6Ku/7668nIyGDYsGGsXLky6roFat26NV26dAHgzDPPZOPGjeTl5bF3717OPvtsAK6++uqg+1dV7rnnHjp16kS/fv3YsmULO3bs4MMPP2To0KE0bdoUgMaNGwPw4YcfcuONNwKQkpLCcccdF/X5x1LSXJzlm2dn/HiX0mnZ0gV8m3/HVIRJk469WxvEptGRkpJC37596du3LxkZGUybNo3Ro0cDFKV9qlWrVvTct1xQUIBrHBYXan0w/vtNSUkp2u/Pf/5zXnnllWPKLl26tNTDDevWrVv0/LHHHuOEE05g2bJlFBYWkpqaGnXdIpU5cOBA1Oc/c+ZMcnNzWbJkCTVq1KBVq1YcPHgQVa3UQ3kTvqXvP2Ji/Hj3n6ywEDZutIBvKk55zOy6Zs0a1q5dW7S8dOlS0tPTo379ueeey6xZszhy5Ai5ubksWLCA7t27h1xfv3599u7dG3G/PXr04JNPPmHdunUA7N+/n+zsbNq2bcu3337L+vXrAYp9KfhEOs6ePXs46aSTqFatGjNmzODIkSNRn3M0GjVqRP369Vm0aBEAr776ash6HH/88dSoUYP58+cX/ZK44IILmD17Nrt27QJg9+7dReufeuopAI4cOcKPP/4Y03pHK6GDvg3TNJXJiBGusRGrRkd+fj6jRo2iffv2dOrUiZUrVzJhwoSoX3/ppZcWdTaef/75PPzww5x44okh13fq1Inq1avTuXPnoo7cYJo1a8aLL77IVVddRadOnejRowerV68mNTWVqVOnMmjQIM4555yQX1DnnXceK1euLOrIDTR27FimTZtGjx49yM7OPuZXQKz84x//YMyYMfTs2RNVDZqKGTFiBFlZWWRmZjJz5kzatm0LQIcOHRg/fjx9+vShc+fO/P73vwdg8uTJzJ8/n4yMDM4880xWrFgR83pHQ6L5KSMiA4DJQArwnKo+FLC9JTANaOiVGaeqc0WkBvAc0BWXSpquqg+GO1ZmZqZmZWWV5lyKadXKBfpA6enuP50xZbFq1SratWsX72qYcpCfn0+9evUAeOihh9i2bRuTJ0+Oc62OCva3JyJLVDUz0msj5vRFJAWYAvwcyAEWi8gcVfXvPfkDMFtVnxKR9sBcoBUwDKilqhkiUgdYKSKvqOrG6E6tbGyYpjGmNN59910efPBBCgoKSE9PP6aDvKqLpiO3O7BOVTcAiMirwBDAP+gr0MB7fhyw1W99XRGpDtQGDgEVlshq2TJ4S9/uhWuMCWf48OEMHz483tUoF9Hk9JsD3/kt53jr/E0ARopIDq6V/xtv/WvAPmAbsBl4RFV3Bx5ARMaISJaIZOXm5pbsDMKwYZrGGHOsaIJ+sLFHgR0BVwEvqmoaMBCYISLVcL8SjgAnA62B20TklGI7U52qqpmqmtmsWbMSnUA4di9cY4w5VjTpnRzAf0KPNI6mb3yuAwYAqOpnIpIKNAWuBv6tqoeB70XkEyAT2FDWikfL7oVrjDFHRdPSXwy0EZHWIlITuBKYE1BmM3ABgIi0A1KBXG/9+eLUBXoAq2NV+VBsCmVjjAkuYtBX1QLgZuA9YBVulM4KEZkoIoO9YrcB14vIMuAVYLQ3F8QUoB7wDe7L4wVVXV4O51HExuabZJKIUyuXZT9z5szhoYceClrONwQzlLy8PJ588smi5Winaq5qopqGQVXn4jpo/df9ye/5SqBXkNfl44ZtVphwsxlamsckEv+plWvVqsXOnTs5dOhQuR3PF/RDzUVTGQwePLhozqGS8gX9sWPHAtFP1VzVJNwVuTY238TDrbdC376xfdx6a/hjJuLUygBnnXXWMVer9u3blyVLlvDFF19w9tlnc8YZZ3D22WezZs2aYq/1//Xx7bff0rNnT7p168Yf//jHojKhpkQeN24c69evp0uXLtxxxx3HTNV88OBBrr32WjIyMjjjjDOYP39+0fFCTeHsb+LEiXTr1o2OHTsyZsyYovl91q1bR79+/ejcuTNdu3YtmqLi4YcfJiMjg86dOzNu3Lig+yy1aKbirMhHWadWtimUTUXxn972lltU+/SJ7eOWW8IfP1GnVn700Uf1T3/6k6qqbt26Vdu0aaOqqnv27NHDhw+rqur777+vl112marqMfXyn5b5F7/4hU6bNk1VVZ944gmtW7euqoaeEtl/ambVY6dqfuSRR3T06NGqqrpq1Spt0aKFHjhwIOwUzv580yurqo4cOVLnzJmjqqrdu3fXN954Q1VVDxw4oPv27dO5c+dqz549dd++fcVe61OWqZUTbpbN8prN0Jhw/v73ij+mb2rlhQsXMn/+fIYPH85DDz1UNMum/9TK+fn51K9fn/r165d6auUGDRoUq4Nv+mKgaPrivLy8oqmVwd0YpWfPnsdMrQwwcuRIpk6dWmyfV1xxBT//+c+57777mD17NsOGuQzxnj17GDVqFGvXrkVEOHz4cNj355NPPuH1118H4JprruGuu+4Cjk6JvGDBAqpVq1Y0JXI4H3/8Mb/5jbv8qG3btqSnp5OdnR3yPQi8g9n8+fN5+OGH2b9/P7t376ZDhw707duXLVu2cOmllwIUzRb6wQcfcO2111LHu8jINzVzrCRM0J858+i0yY0bQ+3asHu3TaFsElsiTq3cvHlzmjRpwvLly5k1axbPPPMMAH/84x8577zzePPNN9m4cSN9+/aNuK9gxws1JXI44d6TSFM4Hzx4kLFjx5KVlUWLFi2YMGFC0RTMoY5VnlMzJ0ROP3DEzq5dcOAAzJhhUyibxJWoUysDXHnllTz88MPs2bOHjIwMwLX0mzd3kwFEMxdOr169iqZFnuk3fC/UlMjhzu/cc88t2kd2djabN2/m9NNPj1gHoOgLpWnTpuTn5xd1Djdo0IC0tDTeeustAH766Sf2799P//79ef7559nvpSt8UzPHSkIEfbvpuUlGiTq1MsDQoUN59dVXueKKK4rW3Xnnndx999306tUrqjn0J0+ezJQpU+jWrRt79uwpWh9qSuQmTZrQq1cvOnbsyB133HHMvsaOHcuRI0fIyMgoukOZfws/nIYNGxbd6euSSy6hW7duRdtmzJjB448/TqdOnTj77LPZvn07AwYMYPDgwWRmZtKlSxceeeSRqI4TraimVq5IpZlauVo118IPJOLmLjemPNjUyiZeyjK1ckK09Mvj/qPGGJOIEiLo22yaxhgTnYQI+jabpomXypYeNYmvrH9zCTNk02bTNBUtNTWVXbt20aRJk3IdYmeMj6qya9euojH9pZEwQd+YipaWlkZOTg6xvPGPMZGkpqaSlpZW6tdb0DemlGrUqEHr1q3jXQ1jSiQhcvrGGGOiY0HfGGOSiAV9Y4xJIpXuilwRyQU2lWEXTYGdMapOVZGM5wzJed7JeM6QnOdd0nNOV9VmkQpVuqBfViKSFc2lyIkkGc8ZkvO8k/GcITnPu7zO2dI7xhiTRCzoG2NMEknEoF/8VjyJLxnPGZLzvJPxnCE5z7tczjnhcvrGGGNCS8SWvjHGmBAs6BtjTBJJmKAvIgNEZI2IrBORcfGuT3kRkRYiMl9EVonIChG5xVvfWETeF5G13r+N4l3XWBORFBH5SkTe8ZZbi8jn3jnPEpGa8a5jrIlIQxF5TURWe595z0T/rEXkd97f9jci8oqIpCbiZy0iz4vI9yLyjd+6oJ+tOI978W25iHQt7XETIuiLSAowBbgIaA9cJSLt41urclMA3Kaq7YAewE3euY4D5qlqG2Cet5xobgFW+S3/BXjMO+cfgOviUqvyNRn4t6q2BTrjzj9hP2sRaQ78FshU1Y5ACnAliflZvwgMCFgX6rO9CGjjPcYAT5X2oAkR9IHuwDpV3aCqh4BXgSFxrlO5UNVtqvql93wvLgg0x53vNK/YNOCS+NSwfIhIGjAIeM5bFuB84DWvSCKecwPgXOAfAKp6SFXzSPDPGjf7b20RqQ7UAbaRgJ+1qi4AdgesDvXZDgGmq7MIaCgiJ5XmuIkS9JsD3/kt53jrEpqItALOAD4HTlDVbeC+GIDj41ezcvF34E7Ad6v7JkCeqhZ4y4n4mZ8C5AIveGmt50SkLgn8WavqFuARYDMu2O8BlpD4n7VPqM82ZjEuUYJ+sNsWJfRYVBGpB7wO3KqqP8a7PuVJRC4GvlfVJf6rgxRNtM+8OtAVeEpVzwD2kUCpnGC8HPYQoDVwMlAXl9oIlGifdSQx+3tPlKCfA7TwW04DtsapLuVORGrgAv5MVX3DW73D93PP+/f7eNWvHPQCBovIRlzq7nxcy7+hlwKAxPzMc4AcVf3cW34N9yWQyJ91P+BbVc1V1cPAG8DZJP5n7RPqs41ZjEuUoL8YaOP18NfEdfzMiXOdyoWXy/4HsEpVH/XbNAcY5T0fBbxd0XUrL6p6t6qmqWor3Gf7oaqOAOYDQ71iCXXOAKq6HfhORE73Vl0ArCSBP2tcWqeHiNTx/tZ955zQn7WfUJ/tHOCX3iieHsAeXxqoxFQ1IR7AQCAbWA+Mj3d9yvE8z8H9rFsOLPUeA3E57nnAWu/fxvGuazmdf1/gHe/5KcAXwDrgn0CteNevHM63C5Dlfd5vAY0S/bMG7gNWA98AM4BaifhZA6/g+i0O41ry14X6bHHpnSlefPsaN7qpVMe1aRiMMSaJJEp6xxhjTBQs6BtjTBKxoG+MMUnEgr4xxiQRC/rGGJNELOgbY0wSsaBvjDFJ5P8BNZmkeF60bD0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa7101a07d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa710178050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def smooth_curve(points, factor=0.8):\n",
    "  smoothed_points = []\n",
    "  for point in points:\n",
    "    if smoothed_points:\n",
    "      previous = smoothed_points[-1]\n",
    "      smoothed_points.append(previous * factor + point * (1 - factor))\n",
    "    else:\n",
    "      smoothed_points.append(point)\n",
    "  return smoothed_points\n",
    "\n",
    "plt.plot(epochs,\n",
    "         smooth_curve(acc), 'bo', label='Smoothed training acc')\n",
    "plt.plot(epochs,\n",
    "         smooth_curve(val_acc), 'b', label='Smoothed validation acc')\n",
    "plt.title('Training and validation accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs,\n",
    "         smooth_curve(loss), 'bo', label='Smoothed training loss')\n",
    "plt.plot(epochs,\n",
    "         smooth_curve(val_loss), 'b', label='Smoothed validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1000 images belonging to 2 classes.\n",
      "('test acc:', 0.9189999949932098)\n"
     ]
    }
   ],
   "source": [
    "test_dir = '/home/geist/Desktop/projects/kaggle/cats_dogs/cats_and_dogs_small/test'\n",
    "\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "        test_dir,\n",
    "        target_size=(150,150),\n",
    "        batch_size=20,\n",
    "        class_mode='binary')\n",
    "\n",
    "test_loss, test_acc = model.evaluate_generator(test_generator, steps=50)\n",
    "print('test acc:', test_acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.15rc1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}