Attempt 2 NeuroLex Featureset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lucaslyon/anaconda3/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6\n",
      "  return f(*args, **kwds)\n",
      "/Users/lucaslyon/anaconda3/lib/python3.6/site-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": [
    "import librosa\n",
    "#import librosa.display\n",
    "import numpy as np\n",
    "#import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "#from matplotlib.pyplot import specgram\n",
    "import pandas as pd\n",
    "import os\n",
    "import keras\n",
    "from keras.preprocessing import sequence\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Embedding\n",
    "from keras.layers import LSTM\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
    "from keras.utils import to_categorical\n",
    "from keras.layers import Input, Flatten, Dropout, Activation\n",
    "from keras.layers import Conv1D, MaxPooling1D, AveragePooling1D\n",
    "from keras.models import Model\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "angryFiles = os.listdir('emotion_train_set/angry')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['.425_3.wav.icloud',\n",
       " '441_3.json',\n",
       " '.705_7.wav.icloud',\n",
       " '.91_2.wav.icloud',\n",
       " '210_3.json']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "angryFiles[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "disgustFiles = os.listdir('emotion_train_set/disgust')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fearFiles = os.listdir('emotion_train_set/fear')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "happyFiles = os.listdir('emotion_train_set/happy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "neutralFiles = os.listdir('emotion_train_set/neutral')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sadFiles = os.listdir('emotion_train_set/sad')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "surpriseFiles = os.listdir('emotion_train_set/surprise')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "angryFiles.sort()\n",
    "disgustFiles.sort()\n",
    "fearFiles.sort()\n",
    "happyFiles.sort()\n",
    "neutralFiles.sort()\n",
    "sadFiles.sort()\n",
    "surpriseFiles.sort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "angryJSONFiles = os.listdir('emotion_train_set/angry')\n",
    "angryJSONFiles.sort()\n",
    "angryData=[]\n",
    "for file in angryJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/angry/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            angryData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "disgustJSONFiles = os.listdir('emotion_train_set/disgust')\n",
    "disgustJSONFiles.sort()\n",
    "disgustData=[]\n",
    "for file in disgustJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/disgust/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            disgustData.append(d['features'])\n",
    "    except: pass\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fearJSONFiles = os.listdir('emotion_train_set/fear')\n",
    "fearJSONFiles.sort()\n",
    "fearData=[]\n",
    "for file in fearJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/fear/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            fearData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "happyJSONFiles = os.listdir('emotion_train_set/happy')\n",
    "happyJSONFiles.sort()\n",
    "happyData=[]\n",
    "for file in happyJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/happy/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            happyData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "neutralJSONFiles = os.listdir('emotion_train_set/neutral')\n",
    "neutralJSONFiles.sort()\n",
    "neutralData=[]\n",
    "for file in neutralJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/neutral/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            neutralData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sadJSONFiles = os.listdir('emotion_train_set/sad')\n",
    "sadJSONFiles.sort()\n",
    "sadData=[]\n",
    "for file in sadJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/sad/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            sadData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "surpriseJSONFiles = os.listdir('emotion_train_set/surprise')\n",
    "surpriseJSONFiles.sort()\n",
    "surpriseData=[]\n",
    "for file in surpriseJSONFiles:\n",
    "    try:\n",
    "        with open('emotion_train_set/surprise/'+file) as json_data:\n",
    "            d = json.load(json_data)\n",
    "            surpriseData.append(d['features'])\n",
    "    except: pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "angryDf = pd.DataFrame(angryData)\n",
    "disgustDf = pd.DataFrame(disgustData)\n",
    "fearDf = pd.DataFrame(fearData)\n",
    "happyDf = pd.DataFrame(happyData)\n",
    "neutralDf = pd.DataFrame(neutralData)\n",
    "sadDf = pd.DataFrame(sadData)\n",
    "surpriseDf = pd.DataFrame(surpriseData)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "angryDf['label'] = 'angry'\n",
    "disgustDf['label'] = 'disgust'\n",
    "fearDf['label'] = 'fear'\n",
    "happyDf['label'] = 'happy'\n",
    "neutralDf['label'] = 'neutral'\n",
    "sadDf['label'] = 'sad'\n",
    "surpriseDf['label'] = 'surprise'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "emotionsDf = pd.concat([angryDf, disgustDf, fearDf, happyDf, neutralDf, sadDf, surpriseDf])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.utils import shuffle\n",
    "shuffledEmotionsDf = shuffle(emotionsDf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "divider = np.random.rand(len(shuffledEmotionsDf)) < 0.85\n",
    "train = shuffledEmotionsDf[divider]\n",
    "test = shuffledEmotionsDf[~divider]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "trainfeatures = train.iloc[:, :-1]\n",
    "trainlabel = train.iloc[:, -1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "testfeatures = test.iloc[:, :-1]\n",
    "testlabel = test.iloc[:, -1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lucaslyon/anaconda3/lib/python3.6/site-packages/sklearn/preprocessing/label.py:111: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    }
   ],
   "source": [
    "from keras.utils import np_utils\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "X_train = np.array(trainfeatures)\n",
    "y_train = np.array(trainlabel)\n",
    "X_test = np.array(testfeatures)\n",
    "y_test = np.array(testlabel)\n",
    "\n",
    "lb = LabelEncoder()\n",
    "\n",
    "y_train = np_utils.to_categorical(lb.fit_transform(y_train))\n",
    "y_test = np_utils.to_categorical(lb.fit_transform(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5722, 272)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Conv1D(128, 5,padding='same',\n",
    "                 input_shape=(272,1)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Conv1D(128, 5,padding='same'))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(MaxPooling1D(pool_size=(8)))\n",
    "model.add(Conv1D(128, 5,padding='same',))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Conv1D(128, 5,padding='same',))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Conv1D(128, 5,padding='same',))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Conv1D(128, 5,padding='same',))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(7))\n",
    "model.add(Activation('softmax'))\n",
    "opt = keras.optimizers.rmsprop(lr=0.00001, decay=1e-6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv1d_1 (Conv1D)            (None, 272, 128)          768       \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 272, 128)          0         \n",
      "_________________________________________________________________\n",
      "conv1d_2 (Conv1D)            (None, 272, 128)          82048     \n",
      "_________________________________________________________________\n",
      "activation_2 (Activation)    (None, 272, 128)          0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 272, 128)          0         \n",
      "_________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1 (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_3 (Conv1D)            (None, 34, 128)           82048     \n",
      "_________________________________________________________________\n",
      "activation_3 (Activation)    (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_4 (Conv1D)            (None, 34, 128)           82048     \n",
      "_________________________________________________________________\n",
      "activation_4 (Activation)    (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_5 (Conv1D)            (None, 34, 128)           82048     \n",
      "_________________________________________________________________\n",
      "activation_5 (Activation)    (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "conv1d_6 (Conv1D)            (None, 34, 128)           82048     \n",
      "_________________________________________________________________\n",
      "activation_6 (Activation)    (None, 34, 128)           0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 4352)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 7)                 30471     \n",
      "_________________________________________________________________\n",
      "activation_7 (Activation)    (None, 7)                 0         \n",
      "=================================================================\n",
      "Total params: 441,479\n",
      "Trainable params: 441,479\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pad sequences\n"
     ]
    }
   ],
   "source": [
    "print('Pad sequences')\n",
    "x_traincnn =np.expand_dims(X_train, axis=2)\n",
    "x_testcnn= np.expand_dims(X_test, axis=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy', optimizer=opt,metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 5722 samples, validate on 996 samples\n",
      "Epoch 1/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 2.0725 - acc: 0.2253 - val_loss: 1.6651 - val_acc: 0.3022\n",
      "Epoch 2/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.8688 - acc: 0.2396 - val_loss: 1.6568 - val_acc: 0.3112\n",
      "Epoch 3/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.8010 - acc: 0.2386 - val_loss: 1.6681 - val_acc: 0.2912\n",
      "Epoch 4/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.7518 - acc: 0.2604 - val_loss: 1.6575 - val_acc: 0.3072\n",
      "Epoch 5/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.7246 - acc: 0.2723 - val_loss: 1.6494 - val_acc: 0.3253\n",
      "Epoch 6/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.7137 - acc: 0.2665 - val_loss: 1.6288 - val_acc: 0.3404\n",
      "Epoch 7/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.7034 - acc: 0.2812 - val_loss: 1.6276 - val_acc: 0.3434\n",
      "Epoch 8/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6959 - acc: 0.2878 - val_loss: 1.6249 - val_acc: 0.3303\n",
      "Epoch 9/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6854 - acc: 0.2931 - val_loss: 1.6281 - val_acc: 0.3293\n",
      "Epoch 10/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6821 - acc: 0.2936 - val_loss: 1.6294 - val_acc: 0.3353\n",
      "Epoch 11/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6730 - acc: 0.2910 - val_loss: 1.6225 - val_acc: 0.3424\n",
      "Epoch 12/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6722 - acc: 0.3016 - val_loss: 1.6094 - val_acc: 0.3414\n",
      "Epoch 13/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6631 - acc: 0.3044 - val_loss: 1.6137 - val_acc: 0.3444\n",
      "Epoch 14/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6622 - acc: 0.2901 - val_loss: 1.6113 - val_acc: 0.3574\n",
      "Epoch 15/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6511 - acc: 0.3165 - val_loss: 1.6070 - val_acc: 0.3474\n",
      "Epoch 16/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6531 - acc: 0.3161 - val_loss: 1.6096 - val_acc: 0.3474\n",
      "Epoch 17/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6508 - acc: 0.3106 - val_loss: 1.6116 - val_acc: 0.3504\n",
      "Epoch 18/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6530 - acc: 0.3093 - val_loss: 1.6116 - val_acc: 0.3464\n",
      "Epoch 19/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6513 - acc: 0.3071 - val_loss: 1.6140 - val_acc: 0.3454\n",
      "Epoch 20/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6434 - acc: 0.3193 - val_loss: 1.6076 - val_acc: 0.3394\n",
      "Epoch 21/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6430 - acc: 0.3196 - val_loss: 1.6065 - val_acc: 0.3594\n",
      "Epoch 22/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6377 - acc: 0.3182 - val_loss: 1.6135 - val_acc: 0.3394\n",
      "Epoch 23/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6407 - acc: 0.3244 - val_loss: 1.6100 - val_acc: 0.3514\n",
      "Epoch 24/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6371 - acc: 0.3261 - val_loss: 1.6055 - val_acc: 0.3584\n",
      "Epoch 25/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6378 - acc: 0.3158 - val_loss: 1.6027 - val_acc: 0.3454\n",
      "Epoch 26/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6312 - acc: 0.3224 - val_loss: 1.6015 - val_acc: 0.3504\n",
      "Epoch 27/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6326 - acc: 0.3251 - val_loss: 1.6058 - val_acc: 0.3424\n",
      "Epoch 28/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6352 - acc: 0.3219 - val_loss: 1.6135 - val_acc: 0.3384\n",
      "Epoch 29/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6275 - acc: 0.3196 - val_loss: 1.5982 - val_acc: 0.3514\n",
      "Epoch 30/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6257 - acc: 0.3298 - val_loss: 1.5990 - val_acc: 0.3484\n",
      "Epoch 31/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6277 - acc: 0.3334 - val_loss: 1.6254 - val_acc: 0.3514\n",
      "Epoch 32/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6223 - acc: 0.3352 - val_loss: 1.6070 - val_acc: 0.3444\n",
      "Epoch 33/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6232 - acc: 0.3266 - val_loss: 1.5978 - val_acc: 0.3564\n",
      "Epoch 34/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6193 - acc: 0.3368 - val_loss: 1.6100 - val_acc: 0.3474\n",
      "Epoch 35/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6152 - acc: 0.3340 - val_loss: 1.6007 - val_acc: 0.3544\n",
      "Epoch 36/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6196 - acc: 0.3376 - val_loss: 1.6011 - val_acc: 0.3504\n",
      "Epoch 37/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6147 - acc: 0.3348 - val_loss: 1.6175 - val_acc: 0.3353\n",
      "Epoch 38/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6140 - acc: 0.3371 - val_loss: 1.6024 - val_acc: 0.3514\n",
      "Epoch 39/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6163 - acc: 0.3266 - val_loss: 1.6070 - val_acc: 0.3524\n",
      "Epoch 40/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6113 - acc: 0.3397 - val_loss: 1.6193 - val_acc: 0.3444\n",
      "Epoch 41/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6115 - acc: 0.3380 - val_loss: 1.6093 - val_acc: 0.3444\n",
      "Epoch 42/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6055 - acc: 0.3403 - val_loss: 1.6084 - val_acc: 0.3444\n",
      "Epoch 43/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6102 - acc: 0.3354 - val_loss: 1.5985 - val_acc: 0.3494\n",
      "Epoch 44/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6107 - acc: 0.3368 - val_loss: 1.6070 - val_acc: 0.3474\n",
      "Epoch 45/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6085 - acc: 0.3369 - val_loss: 1.6086 - val_acc: 0.3373\n",
      "Epoch 46/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6078 - acc: 0.3362 - val_loss: 1.6105 - val_acc: 0.3484\n",
      "Epoch 47/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6006 - acc: 0.3425 - val_loss: 1.6036 - val_acc: 0.3333\n",
      "Epoch 48/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5984 - acc: 0.3392 - val_loss: 1.5982 - val_acc: 0.3414\n",
      "Epoch 49/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.6025 - acc: 0.3371 - val_loss: 1.6088 - val_acc: 0.3414\n",
      "Epoch 50/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5974 - acc: 0.3408 - val_loss: 1.6096 - val_acc: 0.3474\n",
      "Epoch 51/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5993 - acc: 0.3373 - val_loss: 1.6140 - val_acc: 0.3474\n",
      "Epoch 52/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5973 - acc: 0.3453 - val_loss: 1.6061 - val_acc: 0.3434\n",
      "Epoch 53/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5996 - acc: 0.3448 - val_loss: 1.6066 - val_acc: 0.3434\n",
      "Epoch 54/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5918 - acc: 0.3494 - val_loss: 1.6221 - val_acc: 0.3384\n",
      "Epoch 55/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5985 - acc: 0.3422 - val_loss: 1.6216 - val_acc: 0.3394\n",
      "Epoch 56/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5902 - acc: 0.3523 - val_loss: 1.6119 - val_acc: 0.3554\n",
      "Epoch 57/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5914 - acc: 0.3467 - val_loss: 1.6143 - val_acc: 0.3524\n",
      "Epoch 58/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5914 - acc: 0.3478 - val_loss: 1.6045 - val_acc: 0.3554\n",
      "Epoch 59/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5871 - acc: 0.3569 - val_loss: 1.6036 - val_acc: 0.3504\n",
      "Epoch 60/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5852 - acc: 0.3483 - val_loss: 1.6057 - val_acc: 0.3504\n",
      "Epoch 61/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5858 - acc: 0.3446 - val_loss: 1.6111 - val_acc: 0.3514\n",
      "Epoch 62/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5853 - acc: 0.3474 - val_loss: 1.6069 - val_acc: 0.3544\n",
      "Epoch 63/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5845 - acc: 0.3509 - val_loss: 1.5963 - val_acc: 0.3534\n",
      "Epoch 64/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5900 - acc: 0.3518 - val_loss: 1.6187 - val_acc: 0.3444\n",
      "Epoch 65/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5805 - acc: 0.3459 - val_loss: 1.6094 - val_acc: 0.3404\n",
      "Epoch 66/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5819 - acc: 0.3471 - val_loss: 1.6127 - val_acc: 0.3454\n",
      "Epoch 67/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5821 - acc: 0.3595 - val_loss: 1.6165 - val_acc: 0.3544\n",
      "Epoch 68/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5784 - acc: 0.3567 - val_loss: 1.6089 - val_acc: 0.3464\n",
      "Epoch 69/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5786 - acc: 0.3485 - val_loss: 1.6071 - val_acc: 0.3534\n",
      "Epoch 70/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5767 - acc: 0.3590 - val_loss: 1.6074 - val_acc: 0.3574\n",
      "Epoch 71/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5785 - acc: 0.3558 - val_loss: 1.6110 - val_acc: 0.3544\n",
      "Epoch 72/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5754 - acc: 0.3539 - val_loss: 1.6187 - val_acc: 0.3584\n",
      "Epoch 73/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5753 - acc: 0.3577 - val_loss: 1.6206 - val_acc: 0.3414\n",
      "Epoch 74/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5744 - acc: 0.3488 - val_loss: 1.6259 - val_acc: 0.3434\n",
      "Epoch 75/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5699 - acc: 0.3570 - val_loss: 1.6184 - val_acc: 0.3363\n",
      "Epoch 76/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5744 - acc: 0.3569 - val_loss: 1.6207 - val_acc: 0.3474\n",
      "Epoch 77/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5746 - acc: 0.3539 - val_loss: 1.6037 - val_acc: 0.3604\n",
      "Epoch 78/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5637 - acc: 0.3642 - val_loss: 1.6252 - val_acc: 0.3544\n",
      "Epoch 79/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5673 - acc: 0.3605 - val_loss: 1.6366 - val_acc: 0.3504\n",
      "Epoch 80/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5647 - acc: 0.3653 - val_loss: 1.6225 - val_acc: 0.3544\n",
      "Epoch 81/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5687 - acc: 0.3555 - val_loss: 1.6036 - val_acc: 0.3534\n",
      "Epoch 82/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5649 - acc: 0.3597 - val_loss: 1.6399 - val_acc: 0.3494\n",
      "Epoch 83/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5644 - acc: 0.3598 - val_loss: 1.6306 - val_acc: 0.3353\n",
      "Epoch 84/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5681 - acc: 0.3611 - val_loss: 1.6279 - val_acc: 0.3394\n",
      "Epoch 85/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5647 - acc: 0.3639 - val_loss: 1.6321 - val_acc: 0.3363\n",
      "Epoch 86/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5618 - acc: 0.3621 - val_loss: 1.6173 - val_acc: 0.3414\n",
      "Epoch 87/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5637 - acc: 0.3560 - val_loss: 1.6283 - val_acc: 0.3394\n",
      "Epoch 88/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5627 - acc: 0.3600 - val_loss: 1.6245 - val_acc: 0.3404\n",
      "Epoch 89/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5583 - acc: 0.3593 - val_loss: 1.6321 - val_acc: 0.3384\n",
      "Epoch 90/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5613 - acc: 0.3623 - val_loss: 1.6198 - val_acc: 0.3514\n",
      "Epoch 91/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5575 - acc: 0.3595 - val_loss: 1.6339 - val_acc: 0.3434\n",
      "Epoch 92/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5589 - acc: 0.3675 - val_loss: 1.6230 - val_acc: 0.3434\n",
      "Epoch 93/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5565 - acc: 0.3677 - val_loss: 1.6327 - val_acc: 0.3404\n",
      "Epoch 94/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5562 - acc: 0.3661 - val_loss: 1.6365 - val_acc: 0.3363\n",
      "Epoch 95/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5542 - acc: 0.3689 - val_loss: 1.6140 - val_acc: 0.3353\n",
      "Epoch 96/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5553 - acc: 0.3653 - val_loss: 1.6276 - val_acc: 0.3434\n",
      "Epoch 97/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5524 - acc: 0.3726 - val_loss: 1.6444 - val_acc: 0.3343\n",
      "Epoch 98/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5530 - acc: 0.3696 - val_loss: 1.6372 - val_acc: 0.3424\n",
      "Epoch 99/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5483 - acc: 0.3675 - val_loss: 1.6324 - val_acc: 0.3293\n",
      "Epoch 100/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5493 - acc: 0.3674 - val_loss: 1.6306 - val_acc: 0.3343\n",
      "Epoch 101/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5469 - acc: 0.3743 - val_loss: 1.6289 - val_acc: 0.3343\n",
      "Epoch 102/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5509 - acc: 0.3702 - val_loss: 1.6230 - val_acc: 0.3384\n",
      "Epoch 103/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5488 - acc: 0.3668 - val_loss: 1.6484 - val_acc: 0.3424\n",
      "Epoch 104/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5484 - acc: 0.3649 - val_loss: 1.6227 - val_acc: 0.3353\n",
      "Epoch 105/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5433 - acc: 0.3682 - val_loss: 1.6255 - val_acc: 0.3544\n",
      "Epoch 106/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5457 - acc: 0.3763 - val_loss: 1.6521 - val_acc: 0.3323\n",
      "Epoch 107/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5448 - acc: 0.3750 - val_loss: 1.6328 - val_acc: 0.3333\n",
      "Epoch 108/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5425 - acc: 0.3710 - val_loss: 1.6391 - val_acc: 0.3434\n",
      "Epoch 109/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5392 - acc: 0.3801 - val_loss: 1.6523 - val_acc: 0.3384\n",
      "Epoch 110/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5449 - acc: 0.3681 - val_loss: 1.6513 - val_acc: 0.3404\n",
      "Epoch 111/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5385 - acc: 0.3752 - val_loss: 1.6614 - val_acc: 0.3273\n",
      "Epoch 112/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5397 - acc: 0.3768 - val_loss: 1.6345 - val_acc: 0.3464\n",
      "Epoch 113/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5409 - acc: 0.3770 - val_loss: 1.6349 - val_acc: 0.3504\n",
      "Epoch 114/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5379 - acc: 0.3703 - val_loss: 1.6449 - val_acc: 0.3353\n",
      "Epoch 115/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5356 - acc: 0.3759 - val_loss: 1.6371 - val_acc: 0.3534\n",
      "Epoch 116/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5373 - acc: 0.3740 - val_loss: 1.6362 - val_acc: 0.3384\n",
      "Epoch 117/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5364 - acc: 0.3820 - val_loss: 1.6517 - val_acc: 0.3243\n",
      "Epoch 118/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5344 - acc: 0.3742 - val_loss: 1.6346 - val_acc: 0.3574\n",
      "Epoch 119/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5378 - acc: 0.3796 - val_loss: 1.6402 - val_acc: 0.3484\n",
      "Epoch 120/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5321 - acc: 0.3759 - val_loss: 1.6403 - val_acc: 0.3444\n",
      "Epoch 121/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5355 - acc: 0.3796 - val_loss: 1.6395 - val_acc: 0.3404\n",
      "Epoch 122/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5329 - acc: 0.3847 - val_loss: 1.6609 - val_acc: 0.3353\n",
      "Epoch 123/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5283 - acc: 0.3887 - val_loss: 1.6528 - val_acc: 0.3404\n",
      "Epoch 124/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5289 - acc: 0.3736 - val_loss: 1.6404 - val_acc: 0.3313\n",
      "Epoch 125/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5302 - acc: 0.3826 - val_loss: 1.6260 - val_acc: 0.3514\n",
      "Epoch 126/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5270 - acc: 0.3806 - val_loss: 1.6283 - val_acc: 0.3534\n",
      "Epoch 127/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5217 - acc: 0.3862 - val_loss: 1.6543 - val_acc: 0.3394\n",
      "Epoch 128/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5231 - acc: 0.3834 - val_loss: 1.6401 - val_acc: 0.3414\n",
      "Epoch 129/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5231 - acc: 0.3784 - val_loss: 1.6504 - val_acc: 0.3293\n",
      "Epoch 130/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5272 - acc: 0.3806 - val_loss: 1.6451 - val_acc: 0.3444\n",
      "Epoch 131/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5222 - acc: 0.3880 - val_loss: 1.6683 - val_acc: 0.3384\n",
      "Epoch 132/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5201 - acc: 0.3889 - val_loss: 1.6555 - val_acc: 0.3384\n",
      "Epoch 133/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5225 - acc: 0.3857 - val_loss: 1.6477 - val_acc: 0.3343\n",
      "Epoch 134/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5213 - acc: 0.3857 - val_loss: 1.6601 - val_acc: 0.3544\n",
      "Epoch 135/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5214 - acc: 0.3878 - val_loss: 1.6391 - val_acc: 0.3343\n",
      "Epoch 136/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5168 - acc: 0.3899 - val_loss: 1.6571 - val_acc: 0.3454\n",
      "Epoch 137/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5160 - acc: 0.3845 - val_loss: 1.6396 - val_acc: 0.3253\n",
      "Epoch 138/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5160 - acc: 0.3889 - val_loss: 1.6673 - val_acc: 0.3263\n",
      "Epoch 139/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5152 - acc: 0.3873 - val_loss: 1.6604 - val_acc: 0.3363\n",
      "Epoch 140/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5150 - acc: 0.3895 - val_loss: 1.6917 - val_acc: 0.3203\n",
      "Epoch 141/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5182 - acc: 0.3824 - val_loss: 1.6839 - val_acc: 0.3343\n",
      "Epoch 142/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5154 - acc: 0.3833 - val_loss: 1.6716 - val_acc: 0.3353\n",
      "Epoch 143/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5134 - acc: 0.3908 - val_loss: 1.6752 - val_acc: 0.3353\n",
      "Epoch 144/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5080 - acc: 0.3925 - val_loss: 1.6860 - val_acc: 0.3203\n",
      "Epoch 145/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5076 - acc: 0.3932 - val_loss: 1.7007 - val_acc: 0.3092\n",
      "Epoch 146/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5098 - acc: 0.3939 - val_loss: 1.6746 - val_acc: 0.3323\n",
      "Epoch 147/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5093 - acc: 0.3911 - val_loss: 1.6834 - val_acc: 0.3293\n",
      "Epoch 148/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5050 - acc: 0.3941 - val_loss: 1.6620 - val_acc: 0.3323\n",
      "Epoch 149/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5067 - acc: 0.3880 - val_loss: 1.6699 - val_acc: 0.3363\n",
      "Epoch 150/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5050 - acc: 0.3946 - val_loss: 1.6637 - val_acc: 0.3353\n",
      "Epoch 151/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5031 - acc: 0.3993 - val_loss: 1.6576 - val_acc: 0.3353\n",
      "Epoch 152/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5020 - acc: 0.3934 - val_loss: 1.6894 - val_acc: 0.3143\n",
      "Epoch 153/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4997 - acc: 0.3992 - val_loss: 1.6755 - val_acc: 0.3404\n",
      "Epoch 154/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5068 - acc: 0.3964 - val_loss: 1.6607 - val_acc: 0.3363\n",
      "Epoch 155/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5021 - acc: 0.3965 - val_loss: 1.6897 - val_acc: 0.3263\n",
      "Epoch 156/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5018 - acc: 0.3934 - val_loss: 1.6715 - val_acc: 0.3343\n",
      "Epoch 157/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5004 - acc: 0.3875 - val_loss: 1.6884 - val_acc: 0.3153\n",
      "Epoch 158/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5000 - acc: 0.4011 - val_loss: 1.6846 - val_acc: 0.3193\n",
      "Epoch 159/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.5021 - acc: 0.4025 - val_loss: 1.6942 - val_acc: 0.3143\n",
      "Epoch 160/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4955 - acc: 0.4048 - val_loss: 1.6844 - val_acc: 0.3303\n",
      "Epoch 161/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4943 - acc: 0.4023 - val_loss: 1.6752 - val_acc: 0.3303\n",
      "Epoch 162/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4976 - acc: 0.4027 - val_loss: 1.6840 - val_acc: 0.3323\n",
      "Epoch 163/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4937 - acc: 0.4067 - val_loss: 1.6976 - val_acc: 0.3193\n",
      "Epoch 164/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4939 - acc: 0.4067 - val_loss: 1.6844 - val_acc: 0.3253\n",
      "Epoch 165/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4985 - acc: 0.3976 - val_loss: 1.7258 - val_acc: 0.3092\n",
      "Epoch 166/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4880 - acc: 0.4095 - val_loss: 1.6660 - val_acc: 0.3434\n",
      "Epoch 167/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4888 - acc: 0.4056 - val_loss: 1.7185 - val_acc: 0.2972\n",
      "Epoch 168/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4869 - acc: 0.3999 - val_loss: 1.7072 - val_acc: 0.3143\n",
      "Epoch 169/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4876 - acc: 0.4046 - val_loss: 1.6895 - val_acc: 0.3223\n",
      "Epoch 170/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4884 - acc: 0.4082 - val_loss: 1.6759 - val_acc: 0.3313\n",
      "Epoch 171/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4840 - acc: 0.4086 - val_loss: 1.7034 - val_acc: 0.3122\n",
      "Epoch 172/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4883 - acc: 0.4002 - val_loss: 1.6962 - val_acc: 0.3283\n",
      "Epoch 173/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4795 - acc: 0.4154 - val_loss: 1.7080 - val_acc: 0.3143\n",
      "Epoch 174/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4774 - acc: 0.4096 - val_loss: 1.7190 - val_acc: 0.3052\n",
      "Epoch 175/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4771 - acc: 0.4144 - val_loss: 1.7165 - val_acc: 0.3072\n",
      "Epoch 176/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4845 - acc: 0.4075 - val_loss: 1.7180 - val_acc: 0.3082\n",
      "Epoch 177/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4772 - acc: 0.4144 - val_loss: 1.6953 - val_acc: 0.3213\n",
      "Epoch 178/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4784 - acc: 0.4114 - val_loss: 1.6949 - val_acc: 0.3173\n",
      "Epoch 179/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4833 - acc: 0.4055 - val_loss: 1.7174 - val_acc: 0.3032\n",
      "Epoch 180/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4773 - acc: 0.4109 - val_loss: 1.7007 - val_acc: 0.3112\n",
      "Epoch 181/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4709 - acc: 0.4098 - val_loss: 1.6974 - val_acc: 0.3122\n",
      "Epoch 182/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4755 - acc: 0.4110 - val_loss: 1.7187 - val_acc: 0.3012\n",
      "Epoch 183/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4718 - acc: 0.4182 - val_loss: 1.7268 - val_acc: 0.2912\n",
      "Epoch 184/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4770 - acc: 0.4096 - val_loss: 1.7602 - val_acc: 0.2721\n",
      "Epoch 185/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4734 - acc: 0.4117 - val_loss: 1.7239 - val_acc: 0.2972\n",
      "Epoch 186/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4745 - acc: 0.4133 - val_loss: 1.7606 - val_acc: 0.2801\n",
      "Epoch 187/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4741 - acc: 0.4098 - val_loss: 1.7008 - val_acc: 0.3143\n",
      "Epoch 188/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4691 - acc: 0.4096 - val_loss: 1.7235 - val_acc: 0.3012\n",
      "Epoch 189/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4689 - acc: 0.4102 - val_loss: 1.7055 - val_acc: 0.3082\n",
      "Epoch 190/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4662 - acc: 0.4161 - val_loss: 1.7148 - val_acc: 0.3012\n",
      "Epoch 191/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4690 - acc: 0.4154 - val_loss: 1.7191 - val_acc: 0.3173\n",
      "Epoch 192/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4648 - acc: 0.4154 - val_loss: 1.7422 - val_acc: 0.2992\n",
      "Epoch 193/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4659 - acc: 0.4165 - val_loss: 1.7347 - val_acc: 0.3042\n",
      "Epoch 194/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4640 - acc: 0.4186 - val_loss: 1.7155 - val_acc: 0.3082\n",
      "Epoch 195/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4627 - acc: 0.4243 - val_loss: 1.7212 - val_acc: 0.2952\n",
      "Epoch 196/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4603 - acc: 0.4264 - val_loss: 1.7172 - val_acc: 0.3052\n",
      "Epoch 197/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4600 - acc: 0.4173 - val_loss: 1.7275 - val_acc: 0.3012\n",
      "Epoch 198/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4642 - acc: 0.4191 - val_loss: 1.7202 - val_acc: 0.3102\n",
      "Epoch 199/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4618 - acc: 0.4207 - val_loss: 1.7348 - val_acc: 0.2982\n",
      "Epoch 200/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4567 - acc: 0.4217 - val_loss: 1.7056 - val_acc: 0.3042\n",
      "Epoch 201/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4558 - acc: 0.4166 - val_loss: 1.7235 - val_acc: 0.2982\n",
      "Epoch 202/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4557 - acc: 0.4245 - val_loss: 1.7026 - val_acc: 0.3052\n",
      "Epoch 203/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4641 - acc: 0.4151 - val_loss: 1.7531 - val_acc: 0.2871\n",
      "Epoch 204/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4598 - acc: 0.4193 - val_loss: 1.7834 - val_acc: 0.2671\n",
      "Epoch 205/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4559 - acc: 0.4289 - val_loss: 1.7713 - val_acc: 0.2631\n",
      "Epoch 206/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4564 - acc: 0.4273 - val_loss: 1.7679 - val_acc: 0.2841\n",
      "Epoch 207/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4527 - acc: 0.4278 - val_loss: 1.7760 - val_acc: 0.2671\n",
      "Epoch 208/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4564 - acc: 0.4228 - val_loss: 1.7471 - val_acc: 0.2882\n",
      "Epoch 209/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4446 - acc: 0.4303 - val_loss: 1.7623 - val_acc: 0.2741\n",
      "Epoch 210/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4519 - acc: 0.4191 - val_loss: 1.7425 - val_acc: 0.2972\n",
      "Epoch 211/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4455 - acc: 0.4313 - val_loss: 1.7543 - val_acc: 0.2992\n",
      "Epoch 212/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4470 - acc: 0.4257 - val_loss: 1.7461 - val_acc: 0.2882\n",
      "Epoch 213/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4473 - acc: 0.4264 - val_loss: 1.7415 - val_acc: 0.2851\n",
      "Epoch 214/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4460 - acc: 0.4290 - val_loss: 1.7741 - val_acc: 0.2661\n",
      "Epoch 215/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4419 - acc: 0.4245 - val_loss: 1.7558 - val_acc: 0.2861\n",
      "Epoch 216/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4430 - acc: 0.4359 - val_loss: 1.7674 - val_acc: 0.2711\n",
      "Epoch 217/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4427 - acc: 0.4327 - val_loss: 1.7800 - val_acc: 0.2821\n",
      "Epoch 218/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4403 - acc: 0.4313 - val_loss: 1.7591 - val_acc: 0.2882\n",
      "Epoch 219/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4445 - acc: 0.4180 - val_loss: 1.7768 - val_acc: 0.2761\n",
      "Epoch 220/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4408 - acc: 0.4271 - val_loss: 1.7875 - val_acc: 0.2691\n",
      "Epoch 221/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4387 - acc: 0.4339 - val_loss: 1.7380 - val_acc: 0.2992\n",
      "Epoch 222/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4360 - acc: 0.4327 - val_loss: 1.7625 - val_acc: 0.2841\n",
      "Epoch 223/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4339 - acc: 0.4280 - val_loss: 1.7558 - val_acc: 0.2912\n",
      "Epoch 224/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4357 - acc: 0.4280 - val_loss: 1.7775 - val_acc: 0.2751\n",
      "Epoch 225/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4387 - acc: 0.4315 - val_loss: 1.7943 - val_acc: 0.2651\n",
      "Epoch 226/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4294 - acc: 0.4395 - val_loss: 1.7740 - val_acc: 0.2751\n",
      "Epoch 227/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4298 - acc: 0.4343 - val_loss: 1.7802 - val_acc: 0.2741\n",
      "Epoch 228/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4363 - acc: 0.4315 - val_loss: 1.7818 - val_acc: 0.2681\n",
      "Epoch 229/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4375 - acc: 0.4343 - val_loss: 1.7431 - val_acc: 0.2922\n",
      "Epoch 230/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4213 - acc: 0.4446 - val_loss: 1.8391 - val_acc: 0.2440\n",
      "Epoch 231/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4254 - acc: 0.4339 - val_loss: 1.7850 - val_acc: 0.2631\n",
      "Epoch 232/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4217 - acc: 0.4355 - val_loss: 1.8798 - val_acc: 0.2329\n",
      "Epoch 233/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4259 - acc: 0.4303 - val_loss: 1.7627 - val_acc: 0.2821\n",
      "Epoch 234/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4204 - acc: 0.4434 - val_loss: 1.8187 - val_acc: 0.2671\n",
      "Epoch 235/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4246 - acc: 0.4416 - val_loss: 1.8061 - val_acc: 0.2641\n",
      "Epoch 236/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4179 - acc: 0.4415 - val_loss: 1.8310 - val_acc: 0.2550\n",
      "Epoch 237/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4177 - acc: 0.4451 - val_loss: 1.8111 - val_acc: 0.2681\n",
      "Epoch 238/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4229 - acc: 0.4383 - val_loss: 1.8489 - val_acc: 0.2520\n",
      "Epoch 239/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4201 - acc: 0.4357 - val_loss: 1.7865 - val_acc: 0.2751\n",
      "Epoch 240/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4164 - acc: 0.4444 - val_loss: 1.8534 - val_acc: 0.2460\n",
      "Epoch 241/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4173 - acc: 0.4380 - val_loss: 1.8137 - val_acc: 0.2671\n",
      "Epoch 242/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4111 - acc: 0.4460 - val_loss: 1.8028 - val_acc: 0.2751\n",
      "Epoch 243/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4173 - acc: 0.4456 - val_loss: 1.7831 - val_acc: 0.2691\n",
      "Epoch 244/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4141 - acc: 0.4436 - val_loss: 1.8747 - val_acc: 0.2309\n",
      "Epoch 245/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4133 - acc: 0.4474 - val_loss: 1.8140 - val_acc: 0.2570\n",
      "Epoch 246/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4156 - acc: 0.4504 - val_loss: 1.7548 - val_acc: 0.2952\n",
      "Epoch 247/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4096 - acc: 0.4444 - val_loss: 1.8410 - val_acc: 0.2600\n",
      "Epoch 248/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4090 - acc: 0.4415 - val_loss: 1.8126 - val_acc: 0.2530\n",
      "Epoch 249/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4081 - acc: 0.4462 - val_loss: 1.8163 - val_acc: 0.2641\n",
      "Epoch 250/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4082 - acc: 0.4460 - val_loss: 1.8106 - val_acc: 0.2661\n",
      "Epoch 251/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4049 - acc: 0.4448 - val_loss: 1.8097 - val_acc: 0.2641\n",
      "Epoch 252/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4006 - acc: 0.4525 - val_loss: 1.8186 - val_acc: 0.2550\n",
      "Epoch 253/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4100 - acc: 0.4493 - val_loss: 1.8211 - val_acc: 0.2620\n",
      "Epoch 254/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4016 - acc: 0.4476 - val_loss: 1.8191 - val_acc: 0.2580\n",
      "Epoch 255/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4022 - acc: 0.4477 - val_loss: 1.9145 - val_acc: 0.2118\n",
      "Epoch 256/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3992 - acc: 0.4518 - val_loss: 1.8477 - val_acc: 0.2560\n",
      "Epoch 257/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4009 - acc: 0.4490 - val_loss: 1.8528 - val_acc: 0.2600\n",
      "Epoch 258/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.4026 - acc: 0.4500 - val_loss: 1.8046 - val_acc: 0.2721\n",
      "Epoch 259/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3980 - acc: 0.4596 - val_loss: 1.8327 - val_acc: 0.2681\n",
      "Epoch 260/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3952 - acc: 0.4533 - val_loss: 1.8817 - val_acc: 0.2450\n",
      "Epoch 261/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3914 - acc: 0.4616 - val_loss: 1.7970 - val_acc: 0.2711\n",
      "Epoch 262/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3951 - acc: 0.4490 - val_loss: 1.8249 - val_acc: 0.2550\n",
      "Epoch 263/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3922 - acc: 0.4530 - val_loss: 1.8217 - val_acc: 0.2560\n",
      "Epoch 264/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3962 - acc: 0.4469 - val_loss: 1.9062 - val_acc: 0.2199\n",
      "Epoch 265/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3912 - acc: 0.4470 - val_loss: 1.8720 - val_acc: 0.2349\n",
      "Epoch 266/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3847 - acc: 0.4607 - val_loss: 1.8643 - val_acc: 0.2430\n",
      "Epoch 267/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3873 - acc: 0.4638 - val_loss: 1.8786 - val_acc: 0.2349\n",
      "Epoch 268/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3885 - acc: 0.4675 - val_loss: 1.8511 - val_acc: 0.2530\n",
      "Epoch 269/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3866 - acc: 0.4642 - val_loss: 1.9161 - val_acc: 0.2189\n",
      "Epoch 270/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3812 - acc: 0.4666 - val_loss: 1.8380 - val_acc: 0.2450\n",
      "Epoch 271/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3874 - acc: 0.4617 - val_loss: 1.8702 - val_acc: 0.2490\n",
      "Epoch 272/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3825 - acc: 0.4581 - val_loss: 1.8463 - val_acc: 0.2470\n",
      "Epoch 273/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3852 - acc: 0.4553 - val_loss: 1.8362 - val_acc: 0.2701\n",
      "Epoch 274/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3826 - acc: 0.4512 - val_loss: 1.9051 - val_acc: 0.2339\n",
      "Epoch 275/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3798 - acc: 0.4600 - val_loss: 1.9565 - val_acc: 0.2088\n",
      "Epoch 276/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3794 - acc: 0.4626 - val_loss: 1.8374 - val_acc: 0.2671\n",
      "Epoch 277/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3765 - acc: 0.4610 - val_loss: 1.9802 - val_acc: 0.2058\n",
      "Epoch 278/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3757 - acc: 0.4591 - val_loss: 1.9247 - val_acc: 0.2279\n",
      "Epoch 279/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3756 - acc: 0.4659 - val_loss: 1.9076 - val_acc: 0.2380\n",
      "Epoch 280/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3695 - acc: 0.4635 - val_loss: 1.9032 - val_acc: 0.2430\n",
      "Epoch 281/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3679 - acc: 0.4652 - val_loss: 1.9534 - val_acc: 0.2149\n",
      "Epoch 282/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3740 - acc: 0.4626 - val_loss: 1.8889 - val_acc: 0.2420\n",
      "Epoch 283/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3679 - acc: 0.4678 - val_loss: 1.8596 - val_acc: 0.2540\n",
      "Epoch 284/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3685 - acc: 0.4713 - val_loss: 1.9344 - val_acc: 0.2239\n",
      "Epoch 285/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3683 - acc: 0.4747 - val_loss: 1.8964 - val_acc: 0.2390\n",
      "Epoch 286/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3746 - acc: 0.4577 - val_loss: 1.9208 - val_acc: 0.2249\n",
      "Epoch 287/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3741 - acc: 0.4645 - val_loss: 1.9315 - val_acc: 0.2189\n",
      "Epoch 288/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3622 - acc: 0.4740 - val_loss: 1.9632 - val_acc: 0.2139\n",
      "Epoch 289/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3624 - acc: 0.4607 - val_loss: 1.9153 - val_acc: 0.2279\n",
      "Epoch 290/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3639 - acc: 0.4717 - val_loss: 2.0137 - val_acc: 0.2008\n",
      "Epoch 291/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3543 - acc: 0.4771 - val_loss: 1.9730 - val_acc: 0.2118\n",
      "Epoch 292/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3599 - acc: 0.4741 - val_loss: 1.9514 - val_acc: 0.2269\n",
      "Epoch 293/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3506 - acc: 0.4768 - val_loss: 1.9200 - val_acc: 0.2329\n",
      "Epoch 294/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3586 - acc: 0.4703 - val_loss: 1.8845 - val_acc: 0.2590\n",
      "Epoch 295/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3448 - acc: 0.4726 - val_loss: 2.0643 - val_acc: 0.1878\n",
      "Epoch 296/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3555 - acc: 0.4724 - val_loss: 1.9960 - val_acc: 0.2028\n",
      "Epoch 297/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3509 - acc: 0.4764 - val_loss: 1.9039 - val_acc: 0.2410\n",
      "Epoch 298/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3510 - acc: 0.4708 - val_loss: 1.9996 - val_acc: 0.2189\n",
      "Epoch 299/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3530 - acc: 0.4783 - val_loss: 1.9649 - val_acc: 0.2249\n",
      "Epoch 300/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3603 - acc: 0.4656 - val_loss: 1.8362 - val_acc: 0.2570\n",
      "Epoch 301/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3515 - acc: 0.4703 - val_loss: 1.9079 - val_acc: 0.2540\n",
      "Epoch 302/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3478 - acc: 0.4755 - val_loss: 2.0060 - val_acc: 0.2139\n",
      "Epoch 303/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3510 - acc: 0.4769 - val_loss: 2.0739 - val_acc: 0.1878\n",
      "Epoch 304/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3334 - acc: 0.4848 - val_loss: 1.9244 - val_acc: 0.2540\n",
      "Epoch 305/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3336 - acc: 0.4850 - val_loss: 2.0380 - val_acc: 0.1978\n",
      "Epoch 306/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3408 - acc: 0.4796 - val_loss: 1.9099 - val_acc: 0.2339\n",
      "Epoch 307/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3435 - acc: 0.4783 - val_loss: 2.0192 - val_acc: 0.1958\n",
      "Epoch 308/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3358 - acc: 0.4808 - val_loss: 1.9653 - val_acc: 0.2249\n",
      "Epoch 309/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3363 - acc: 0.4841 - val_loss: 1.9771 - val_acc: 0.2269\n",
      "Epoch 310/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3406 - acc: 0.4797 - val_loss: 1.9673 - val_acc: 0.2319\n",
      "Epoch 311/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3335 - acc: 0.4801 - val_loss: 1.9346 - val_acc: 0.2229\n",
      "Epoch 312/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3380 - acc: 0.4750 - val_loss: 1.9831 - val_acc: 0.2249\n",
      "Epoch 313/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3323 - acc: 0.4816 - val_loss: 2.0626 - val_acc: 0.1958\n",
      "Epoch 314/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3271 - acc: 0.4780 - val_loss: 1.9698 - val_acc: 0.2199\n",
      "Epoch 315/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3284 - acc: 0.4858 - val_loss: 2.0273 - val_acc: 0.2098\n",
      "Epoch 316/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3327 - acc: 0.4867 - val_loss: 1.9575 - val_acc: 0.2209\n",
      "Epoch 317/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3264 - acc: 0.4897 - val_loss: 1.9208 - val_acc: 0.2520\n",
      "Epoch 318/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3212 - acc: 0.4893 - val_loss: 2.0634 - val_acc: 0.1998\n",
      "Epoch 319/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3235 - acc: 0.4890 - val_loss: 1.9025 - val_acc: 0.2570\n",
      "Epoch 320/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3286 - acc: 0.4888 - val_loss: 2.0080 - val_acc: 0.2108\n",
      "Epoch 321/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3172 - acc: 0.4946 - val_loss: 2.0737 - val_acc: 0.2008\n",
      "Epoch 322/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3175 - acc: 0.4844 - val_loss: 1.9352 - val_acc: 0.2319\n",
      "Epoch 323/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3201 - acc: 0.4944 - val_loss: 2.0693 - val_acc: 0.1948\n",
      "Epoch 324/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3147 - acc: 0.4939 - val_loss: 1.9458 - val_acc: 0.2339\n",
      "Epoch 325/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3180 - acc: 0.4904 - val_loss: 2.1039 - val_acc: 0.1908\n",
      "Epoch 326/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3215 - acc: 0.4892 - val_loss: 1.9976 - val_acc: 0.2259\n",
      "Epoch 327/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3129 - acc: 0.4979 - val_loss: 2.0335 - val_acc: 0.2098\n",
      "Epoch 328/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3112 - acc: 0.4948 - val_loss: 1.9735 - val_acc: 0.2279\n",
      "Epoch 329/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3117 - acc: 0.4900 - val_loss: 2.0381 - val_acc: 0.2108\n",
      "Epoch 330/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3105 - acc: 0.4981 - val_loss: 2.0746 - val_acc: 0.1958\n",
      "Epoch 331/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3118 - acc: 0.5010 - val_loss: 2.0802 - val_acc: 0.1928\n",
      "Epoch 332/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3152 - acc: 0.4904 - val_loss: 2.0214 - val_acc: 0.2108\n",
      "Epoch 333/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3009 - acc: 0.4977 - val_loss: 2.1206 - val_acc: 0.1988\n",
      "Epoch 334/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3031 - acc: 0.5010 - val_loss: 2.0503 - val_acc: 0.2209\n",
      "Epoch 335/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3127 - acc: 0.4942 - val_loss: 2.0277 - val_acc: 0.2108\n",
      "Epoch 336/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2942 - acc: 0.5061 - val_loss: 2.0653 - val_acc: 0.2068\n",
      "Epoch 337/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3094 - acc: 0.4962 - val_loss: 2.0435 - val_acc: 0.2209\n",
      "Epoch 338/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2949 - acc: 0.4948 - val_loss: 2.0245 - val_acc: 0.2359\n",
      "Epoch 339/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3039 - acc: 0.4949 - val_loss: 2.0721 - val_acc: 0.2098\n",
      "Epoch 340/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2977 - acc: 0.5073 - val_loss: 2.0710 - val_acc: 0.1978\n",
      "Epoch 341/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2943 - acc: 0.5068 - val_loss: 2.0608 - val_acc: 0.2048\n",
      "Epoch 342/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2944 - acc: 0.5037 - val_loss: 2.1854 - val_acc: 0.1576\n",
      "Epoch 343/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.3031 - acc: 0.4977 - val_loss: 2.1522 - val_acc: 0.1928\n",
      "Epoch 344/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2951 - acc: 0.4970 - val_loss: 2.0520 - val_acc: 0.2098\n",
      "Epoch 345/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2976 - acc: 0.4986 - val_loss: 2.2020 - val_acc: 0.1596\n",
      "Epoch 346/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2889 - acc: 0.4986 - val_loss: 2.0765 - val_acc: 0.1988\n",
      "Epoch 347/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2885 - acc: 0.5044 - val_loss: 1.9880 - val_acc: 0.2329\n",
      "Epoch 348/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2894 - acc: 0.5059 - val_loss: 1.9949 - val_acc: 0.2229\n",
      "Epoch 349/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2916 - acc: 0.4998 - val_loss: 2.0960 - val_acc: 0.2129\n",
      "Epoch 350/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2925 - acc: 0.4997 - val_loss: 2.1125 - val_acc: 0.1978\n",
      "Epoch 351/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2888 - acc: 0.5112 - val_loss: 2.2970 - val_acc: 0.1456\n",
      "Epoch 352/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2869 - acc: 0.5058 - val_loss: 2.1526 - val_acc: 0.1797\n",
      "Epoch 353/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2838 - acc: 0.5042 - val_loss: 2.1142 - val_acc: 0.2048\n",
      "Epoch 354/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2781 - acc: 0.5122 - val_loss: 2.1506 - val_acc: 0.1918\n",
      "Epoch 355/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2879 - acc: 0.5031 - val_loss: 1.9962 - val_acc: 0.2309\n",
      "Epoch 356/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2781 - acc: 0.5103 - val_loss: 2.1634 - val_acc: 0.1918\n",
      "Epoch 357/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2763 - acc: 0.5066 - val_loss: 2.0797 - val_acc: 0.2018\n",
      "Epoch 358/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2764 - acc: 0.5168 - val_loss: 2.1845 - val_acc: 0.1807\n",
      "Epoch 359/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2790 - acc: 0.5093 - val_loss: 2.0592 - val_acc: 0.2078\n",
      "Epoch 360/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2691 - acc: 0.5124 - val_loss: 2.0358 - val_acc: 0.2259\n",
      "Epoch 361/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2786 - acc: 0.5052 - val_loss: 2.0617 - val_acc: 0.2199\n",
      "Epoch 362/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2629 - acc: 0.5150 - val_loss: 2.1489 - val_acc: 0.1968\n",
      "Epoch 363/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2685 - acc: 0.5161 - val_loss: 2.2095 - val_acc: 0.1827\n",
      "Epoch 364/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2627 - acc: 0.5156 - val_loss: 2.2335 - val_acc: 0.1596\n",
      "Epoch 365/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2667 - acc: 0.5241 - val_loss: 2.0553 - val_acc: 0.2209\n",
      "Epoch 366/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2602 - acc: 0.5150 - val_loss: 2.1227 - val_acc: 0.2018\n",
      "Epoch 367/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2626 - acc: 0.5138 - val_loss: 2.1042 - val_acc: 0.2068\n",
      "Epoch 368/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2624 - acc: 0.5150 - val_loss: 2.0664 - val_acc: 0.2169\n",
      "Epoch 369/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2645 - acc: 0.5225 - val_loss: 2.1456 - val_acc: 0.1978\n",
      "Epoch 370/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2645 - acc: 0.5171 - val_loss: 2.0704 - val_acc: 0.2219\n",
      "Epoch 371/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2576 - acc: 0.5170 - val_loss: 2.1207 - val_acc: 0.1998\n",
      "Epoch 372/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2579 - acc: 0.5117 - val_loss: 2.1637 - val_acc: 0.1888\n",
      "Epoch 373/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2480 - acc: 0.5229 - val_loss: 2.2355 - val_acc: 0.1747\n",
      "Epoch 374/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2528 - acc: 0.5229 - val_loss: 2.1930 - val_acc: 0.1928\n",
      "Epoch 375/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2672 - acc: 0.5126 - val_loss: 2.2517 - val_acc: 0.1767\n",
      "Epoch 376/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2601 - acc: 0.5184 - val_loss: 2.2316 - val_acc: 0.1647\n",
      "Epoch 377/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2514 - acc: 0.5170 - val_loss: 2.2709 - val_acc: 0.1606\n",
      "Epoch 378/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2492 - acc: 0.5243 - val_loss: 2.0457 - val_acc: 0.2209\n",
      "Epoch 379/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2417 - acc: 0.5245 - val_loss: 2.1679 - val_acc: 0.1878\n",
      "Epoch 380/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2398 - acc: 0.5238 - val_loss: 2.1899 - val_acc: 0.1958\n",
      "Epoch 381/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2464 - acc: 0.5241 - val_loss: 2.2949 - val_acc: 0.1637\n",
      "Epoch 382/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2383 - acc: 0.5287 - val_loss: 2.3319 - val_acc: 0.1486\n",
      "Epoch 383/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2446 - acc: 0.5114 - val_loss: 2.1850 - val_acc: 0.1978\n",
      "Epoch 384/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2387 - acc: 0.5280 - val_loss: 2.2268 - val_acc: 0.1667\n",
      "Epoch 385/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2326 - acc: 0.5287 - val_loss: 2.1581 - val_acc: 0.1867\n",
      "Epoch 386/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2297 - acc: 0.5306 - val_loss: 2.2243 - val_acc: 0.1847\n",
      "Epoch 387/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2351 - acc: 0.5355 - val_loss: 2.2140 - val_acc: 0.1777\n",
      "Epoch 388/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2341 - acc: 0.5294 - val_loss: 2.2837 - val_acc: 0.1637\n",
      "Epoch 389/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2386 - acc: 0.5211 - val_loss: 2.1240 - val_acc: 0.2149\n",
      "Epoch 390/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2388 - acc: 0.5253 - val_loss: 1.9927 - val_acc: 0.2359\n",
      "Epoch 391/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2377 - acc: 0.5281 - val_loss: 2.2661 - val_acc: 0.1627\n",
      "Epoch 392/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2301 - acc: 0.5337 - val_loss: 2.2002 - val_acc: 0.1787\n",
      "Epoch 393/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2364 - acc: 0.5262 - val_loss: 2.2665 - val_acc: 0.1657\n",
      "Epoch 394/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2248 - acc: 0.5341 - val_loss: 2.2645 - val_acc: 0.1627\n",
      "Epoch 395/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2262 - acc: 0.5264 - val_loss: 2.2826 - val_acc: 0.1737\n",
      "Epoch 396/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2206 - acc: 0.5364 - val_loss: 2.2725 - val_acc: 0.1616\n",
      "Epoch 397/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2181 - acc: 0.5409 - val_loss: 2.1943 - val_acc: 0.1837\n",
      "Epoch 398/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2151 - acc: 0.5327 - val_loss: 2.1761 - val_acc: 0.1787\n",
      "Epoch 399/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2303 - acc: 0.5301 - val_loss: 2.2575 - val_acc: 0.1707\n",
      "Epoch 400/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2188 - acc: 0.5344 - val_loss: 2.0960 - val_acc: 0.2179\n",
      "Epoch 401/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2164 - acc: 0.5334 - val_loss: 2.1463 - val_acc: 0.2159\n",
      "Epoch 402/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2171 - acc: 0.5383 - val_loss: 2.2596 - val_acc: 0.1847\n",
      "Epoch 403/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2103 - acc: 0.5370 - val_loss: 2.4490 - val_acc: 0.1426\n",
      "Epoch 404/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2108 - acc: 0.5397 - val_loss: 2.4181 - val_acc: 0.1456\n",
      "Epoch 405/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2166 - acc: 0.5411 - val_loss: 2.3085 - val_acc: 0.1677\n",
      "Epoch 406/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2088 - acc: 0.5414 - val_loss: 2.4558 - val_acc: 0.1396\n",
      "Epoch 407/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2141 - acc: 0.5414 - val_loss: 2.3378 - val_acc: 0.1677\n",
      "Epoch 408/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2060 - acc: 0.5379 - val_loss: 2.5619 - val_acc: 0.1225\n",
      "Epoch 409/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1986 - acc: 0.5578 - val_loss: 2.3636 - val_acc: 0.1596\n",
      "Epoch 410/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2114 - acc: 0.5351 - val_loss: 2.1875 - val_acc: 0.1908\n",
      "Epoch 411/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2037 - acc: 0.5400 - val_loss: 2.3818 - val_acc: 0.1456\n",
      "Epoch 412/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2017 - acc: 0.5402 - val_loss: 2.2696 - val_acc: 0.1797\n",
      "Epoch 413/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2099 - acc: 0.5323 - val_loss: 2.3625 - val_acc: 0.1647\n",
      "Epoch 414/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.2055 - acc: 0.5470 - val_loss: 2.1836 - val_acc: 0.1908\n",
      "Epoch 415/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1967 - acc: 0.5510 - val_loss: 2.3595 - val_acc: 0.1536\n",
      "Epoch 416/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1989 - acc: 0.5432 - val_loss: 2.2908 - val_acc: 0.1697\n",
      "Epoch 417/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1924 - acc: 0.5482 - val_loss: 2.4959 - val_acc: 0.1436\n",
      "Epoch 418/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1952 - acc: 0.5484 - val_loss: 2.4501 - val_acc: 0.1315\n",
      "Epoch 419/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1912 - acc: 0.5426 - val_loss: 2.3133 - val_acc: 0.1777\n",
      "Epoch 420/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1896 - acc: 0.5531 - val_loss: 2.3460 - val_acc: 0.1606\n",
      "Epoch 421/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1909 - acc: 0.5458 - val_loss: 2.2562 - val_acc: 0.1928\n",
      "Epoch 422/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1992 - acc: 0.5437 - val_loss: 2.2046 - val_acc: 0.1898\n",
      "Epoch 423/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1885 - acc: 0.5503 - val_loss: 2.3881 - val_acc: 0.1526\n",
      "Epoch 424/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1857 - acc: 0.5549 - val_loss: 2.6135 - val_acc: 0.1255\n",
      "Epoch 425/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1900 - acc: 0.5468 - val_loss: 2.2953 - val_acc: 0.1867\n",
      "Epoch 426/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1783 - acc: 0.5571 - val_loss: 2.4010 - val_acc: 0.1446\n",
      "Epoch 427/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1818 - acc: 0.5475 - val_loss: 2.3634 - val_acc: 0.1647\n",
      "Epoch 428/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1767 - acc: 0.5535 - val_loss: 2.4249 - val_acc: 0.1606\n",
      "Epoch 429/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1715 - acc: 0.5570 - val_loss: 2.4738 - val_acc: 0.1295\n",
      "Epoch 430/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1712 - acc: 0.5551 - val_loss: 2.5658 - val_acc: 0.1325\n",
      "Epoch 431/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1776 - acc: 0.5477 - val_loss: 2.4283 - val_acc: 0.1647\n",
      "Epoch 432/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1818 - acc: 0.5491 - val_loss: 2.3511 - val_acc: 0.1737\n",
      "Epoch 433/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1761 - acc: 0.5509 - val_loss: 2.3823 - val_acc: 0.1566\n",
      "Epoch 434/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1722 - acc: 0.5582 - val_loss: 2.2396 - val_acc: 0.1918\n",
      "Epoch 435/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1663 - acc: 0.5613 - val_loss: 2.3790 - val_acc: 0.1506\n",
      "Epoch 436/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1664 - acc: 0.5626 - val_loss: 2.3184 - val_acc: 0.1727\n",
      "Epoch 437/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1657 - acc: 0.5592 - val_loss: 2.2589 - val_acc: 0.1888\n",
      "Epoch 438/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1673 - acc: 0.5573 - val_loss: 2.4756 - val_acc: 0.1406\n",
      "Epoch 439/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1564 - acc: 0.5659 - val_loss: 2.5037 - val_acc: 0.1406\n",
      "Epoch 440/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1542 - acc: 0.5603 - val_loss: 2.2072 - val_acc: 0.1807\n",
      "Epoch 441/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1636 - acc: 0.5557 - val_loss: 2.5339 - val_acc: 0.1396\n",
      "Epoch 442/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1531 - acc: 0.5675 - val_loss: 2.4285 - val_acc: 0.1476\n",
      "Epoch 443/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1500 - acc: 0.5673 - val_loss: 2.2814 - val_acc: 0.1827\n",
      "Epoch 444/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1541 - acc: 0.5613 - val_loss: 2.5464 - val_acc: 0.1386\n",
      "Epoch 445/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1484 - acc: 0.5622 - val_loss: 2.4362 - val_acc: 0.1506\n",
      "Epoch 446/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1525 - acc: 0.5713 - val_loss: 2.6065 - val_acc: 0.1345\n",
      "Epoch 447/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1631 - acc: 0.5631 - val_loss: 2.3508 - val_acc: 0.1777\n",
      "Epoch 448/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1532 - acc: 0.5598 - val_loss: 2.3505 - val_acc: 0.1878\n",
      "Epoch 449/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1536 - acc: 0.5629 - val_loss: 2.2874 - val_acc: 0.1978\n",
      "Epoch 450/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1516 - acc: 0.5652 - val_loss: 2.3680 - val_acc: 0.1757\n",
      "Epoch 451/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1496 - acc: 0.5704 - val_loss: 2.2449 - val_acc: 0.1918\n",
      "Epoch 452/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1455 - acc: 0.5645 - val_loss: 2.6804 - val_acc: 0.1275\n",
      "Epoch 453/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1444 - acc: 0.5708 - val_loss: 2.4467 - val_acc: 0.1606\n",
      "Epoch 454/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1403 - acc: 0.5746 - val_loss: 2.4320 - val_acc: 0.1606\n",
      "Epoch 455/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1394 - acc: 0.5708 - val_loss: 2.4787 - val_acc: 0.1446\n",
      "Epoch 456/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1311 - acc: 0.5762 - val_loss: 2.4810 - val_acc: 0.1516\n",
      "Epoch 457/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1249 - acc: 0.5741 - val_loss: 2.3682 - val_acc: 0.1707\n",
      "Epoch 458/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1384 - acc: 0.5696 - val_loss: 2.7368 - val_acc: 0.1165\n",
      "Epoch 459/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1326 - acc: 0.5748 - val_loss: 2.3542 - val_acc: 0.1767\n",
      "Epoch 460/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1364 - acc: 0.5704 - val_loss: 2.4005 - val_acc: 0.1606\n",
      "Epoch 461/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1312 - acc: 0.5678 - val_loss: 2.4206 - val_acc: 0.1486\n",
      "Epoch 462/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1357 - acc: 0.5711 - val_loss: 2.5946 - val_acc: 0.1396\n",
      "Epoch 463/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1264 - acc: 0.5806 - val_loss: 2.5517 - val_acc: 0.1496\n",
      "Epoch 464/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1249 - acc: 0.5793 - val_loss: 2.4729 - val_acc: 0.1506\n",
      "Epoch 465/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1278 - acc: 0.5708 - val_loss: 2.3077 - val_acc: 0.1837\n",
      "Epoch 466/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1255 - acc: 0.5755 - val_loss: 2.3356 - val_acc: 0.1817\n",
      "Epoch 467/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1193 - acc: 0.5762 - val_loss: 2.4650 - val_acc: 0.1526\n",
      "Epoch 468/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1190 - acc: 0.5797 - val_loss: 2.5158 - val_acc: 0.1466\n",
      "Epoch 469/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1163 - acc: 0.5758 - val_loss: 2.3183 - val_acc: 0.1747\n",
      "Epoch 470/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1202 - acc: 0.5746 - val_loss: 2.5006 - val_acc: 0.1446\n",
      "Epoch 471/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1299 - acc: 0.5731 - val_loss: 2.4951 - val_acc: 0.1355\n",
      "Epoch 472/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1258 - acc: 0.5664 - val_loss: 2.6211 - val_acc: 0.1466\n",
      "Epoch 473/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1032 - acc: 0.5855 - val_loss: 2.5226 - val_acc: 0.1616\n",
      "Epoch 474/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1165 - acc: 0.5779 - val_loss: 2.5994 - val_acc: 0.1386\n",
      "Epoch 475/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1105 - acc: 0.5851 - val_loss: 2.4619 - val_acc: 0.1627\n",
      "Epoch 476/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1085 - acc: 0.5804 - val_loss: 2.4501 - val_acc: 0.1616\n",
      "Epoch 477/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1069 - acc: 0.5757 - val_loss: 2.4961 - val_acc: 0.1566\n",
      "Epoch 478/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1047 - acc: 0.5841 - val_loss: 2.6513 - val_acc: 0.1335\n",
      "Epoch 479/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0975 - acc: 0.5916 - val_loss: 2.3991 - val_acc: 0.1727\n",
      "Epoch 480/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1084 - acc: 0.5907 - val_loss: 2.4535 - val_acc: 0.1697\n",
      "Epoch 481/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1093 - acc: 0.5783 - val_loss: 2.4464 - val_acc: 0.1687\n",
      "Epoch 482/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1018 - acc: 0.5816 - val_loss: 2.6168 - val_acc: 0.1476\n",
      "Epoch 483/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1029 - acc: 0.5865 - val_loss: 2.4656 - val_acc: 0.1727\n",
      "Epoch 484/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0998 - acc: 0.5849 - val_loss: 2.8375 - val_acc: 0.1165\n",
      "Epoch 485/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1002 - acc: 0.5886 - val_loss: 2.5697 - val_acc: 0.1496\n",
      "Epoch 486/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0963 - acc: 0.5923 - val_loss: 2.4640 - val_acc: 0.1616\n",
      "Epoch 487/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0883 - acc: 0.5825 - val_loss: 2.3632 - val_acc: 0.1827\n",
      "Epoch 488/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0894 - acc: 0.5905 - val_loss: 2.4678 - val_acc: 0.1667\n",
      "Epoch 489/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.1018 - acc: 0.5855 - val_loss: 2.5651 - val_acc: 0.1416\n",
      "Epoch 490/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0859 - acc: 0.5883 - val_loss: 2.4193 - val_acc: 0.1697\n",
      "Epoch 491/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0783 - acc: 0.5959 - val_loss: 2.8119 - val_acc: 0.1225\n",
      "Epoch 492/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0926 - acc: 0.5909 - val_loss: 2.6078 - val_acc: 0.1406\n",
      "Epoch 493/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0815 - acc: 0.5975 - val_loss: 2.8379 - val_acc: 0.1185\n",
      "Epoch 494/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0843 - acc: 0.5897 - val_loss: 2.7909 - val_acc: 0.1355\n",
      "Epoch 495/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0835 - acc: 0.6022 - val_loss: 2.4780 - val_acc: 0.1687\n",
      "Epoch 496/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0848 - acc: 0.5886 - val_loss: 2.6338 - val_acc: 0.1406\n",
      "Epoch 497/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0797 - acc: 0.5965 - val_loss: 2.7146 - val_acc: 0.1386\n",
      "Epoch 498/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0862 - acc: 0.5947 - val_loss: 2.6532 - val_acc: 0.1496\n",
      "Epoch 499/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0716 - acc: 0.5945 - val_loss: 2.5341 - val_acc: 0.1546\n",
      "Epoch 500/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0774 - acc: 0.5921 - val_loss: 2.4446 - val_acc: 0.1807\n",
      "Epoch 501/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0759 - acc: 0.5973 - val_loss: 2.6963 - val_acc: 0.1255\n",
      "Epoch 502/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0716 - acc: 0.6073 - val_loss: 2.5175 - val_acc: 0.1606\n",
      "Epoch 503/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0771 - acc: 0.5872 - val_loss: 2.6346 - val_acc: 0.1436\n",
      "Epoch 504/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0715 - acc: 0.6085 - val_loss: 2.6468 - val_acc: 0.1416\n",
      "Epoch 505/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0716 - acc: 0.5954 - val_loss: 2.3880 - val_acc: 0.1797\n",
      "Epoch 506/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0587 - acc: 0.5982 - val_loss: 2.8458 - val_acc: 0.1245\n",
      "Epoch 507/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0670 - acc: 0.6001 - val_loss: 2.4357 - val_acc: 0.1637\n",
      "Epoch 508/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0570 - acc: 0.6045 - val_loss: 2.5171 - val_acc: 0.1616\n",
      "Epoch 509/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0524 - acc: 0.6122 - val_loss: 2.7662 - val_acc: 0.1265\n",
      "Epoch 510/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0608 - acc: 0.6029 - val_loss: 2.7030 - val_acc: 0.1295\n",
      "Epoch 511/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0586 - acc: 0.6008 - val_loss: 2.7120 - val_acc: 0.1335\n",
      "Epoch 512/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0644 - acc: 0.6075 - val_loss: 2.3837 - val_acc: 0.1747\n",
      "Epoch 513/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0413 - acc: 0.6162 - val_loss: 2.8175 - val_acc: 0.1195\n",
      "Epoch 514/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0491 - acc: 0.6064 - val_loss: 2.6392 - val_acc: 0.1386\n",
      "Epoch 515/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0504 - acc: 0.6099 - val_loss: 2.6171 - val_acc: 0.1647\n",
      "Epoch 516/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0546 - acc: 0.6043 - val_loss: 2.7009 - val_acc: 0.1436\n",
      "Epoch 517/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0529 - acc: 0.6059 - val_loss: 2.5525 - val_acc: 0.1596\n",
      "Epoch 518/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0480 - acc: 0.6038 - val_loss: 2.9176 - val_acc: 0.1124\n",
      "Epoch 519/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0345 - acc: 0.6211 - val_loss: 2.5525 - val_acc: 0.1556\n",
      "Epoch 520/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0412 - acc: 0.6059 - val_loss: 2.7635 - val_acc: 0.1265\n",
      "Epoch 521/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0359 - acc: 0.6194 - val_loss: 2.7848 - val_acc: 0.1305\n",
      "Epoch 522/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0393 - acc: 0.6251 - val_loss: 2.4612 - val_acc: 0.1717\n",
      "Epoch 523/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0305 - acc: 0.6204 - val_loss: 2.6607 - val_acc: 0.1386\n",
      "Epoch 524/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0334 - acc: 0.6164 - val_loss: 2.6645 - val_acc: 0.1476\n",
      "Epoch 525/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0346 - acc: 0.6153 - val_loss: 2.7190 - val_acc: 0.1466\n",
      "Epoch 526/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0329 - acc: 0.6167 - val_loss: 2.9539 - val_acc: 0.1205\n",
      "Epoch 527/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0468 - acc: 0.6077 - val_loss: 2.6231 - val_acc: 0.1546\n",
      "Epoch 528/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0376 - acc: 0.6078 - val_loss: 2.5751 - val_acc: 0.1667\n",
      "Epoch 529/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0402 - acc: 0.6208 - val_loss: 2.4832 - val_acc: 0.1777\n",
      "Epoch 530/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0241 - acc: 0.6169 - val_loss: 2.7374 - val_acc: 0.1396\n",
      "Epoch 531/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0319 - acc: 0.6187 - val_loss: 2.9662 - val_acc: 0.1175\n",
      "Epoch 532/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0276 - acc: 0.6117 - val_loss: 2.7985 - val_acc: 0.1325\n",
      "Epoch 533/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0314 - acc: 0.6201 - val_loss: 2.6719 - val_acc: 0.1526\n",
      "Epoch 534/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0192 - acc: 0.6227 - val_loss: 2.7912 - val_acc: 0.1255\n",
      "Epoch 535/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0173 - acc: 0.6183 - val_loss: 2.6931 - val_acc: 0.1526\n",
      "Epoch 536/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0095 - acc: 0.6230 - val_loss: 2.9941 - val_acc: 0.1124\n",
      "Epoch 537/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0212 - acc: 0.6150 - val_loss: 2.8642 - val_acc: 0.1165\n",
      "Epoch 538/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0150 - acc: 0.6267 - val_loss: 2.5620 - val_acc: 0.1496\n",
      "Epoch 539/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0157 - acc: 0.6288 - val_loss: 2.8201 - val_acc: 0.1325\n",
      "Epoch 540/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0052 - acc: 0.6285 - val_loss: 2.8119 - val_acc: 0.1396\n",
      "Epoch 541/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0038 - acc: 0.6312 - val_loss: 2.4540 - val_acc: 0.1647\n",
      "Epoch 542/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0167 - acc: 0.6264 - val_loss: 2.6959 - val_acc: 0.1436\n",
      "Epoch 543/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0084 - acc: 0.6278 - val_loss: 2.9116 - val_acc: 0.1215\n",
      "Epoch 544/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0034 - acc: 0.6323 - val_loss: 2.5891 - val_acc: 0.1606\n",
      "Epoch 545/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0071 - acc: 0.6312 - val_loss: 3.0655 - val_acc: 0.1175\n",
      "Epoch 546/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0070 - acc: 0.6293 - val_loss: 2.5068 - val_acc: 0.1647\n",
      "Epoch 547/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0053 - acc: 0.6241 - val_loss: 2.7730 - val_acc: 0.1345\n",
      "Epoch 548/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0088 - acc: 0.6183 - val_loss: 2.7835 - val_acc: 0.1426\n",
      "Epoch 549/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9927 - acc: 0.6290 - val_loss: 2.6752 - val_acc: 0.1406\n",
      "Epoch 550/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 1.0048 - acc: 0.6285 - val_loss: 2.6772 - val_acc: 0.1596\n",
      "Epoch 551/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9875 - acc: 0.6358 - val_loss: 2.6289 - val_acc: 0.1566\n",
      "Epoch 552/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9902 - acc: 0.6344 - val_loss: 2.8714 - val_acc: 0.1255\n",
      "Epoch 553/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9941 - acc: 0.6309 - val_loss: 2.9235 - val_acc: 0.1265\n",
      "Epoch 554/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9935 - acc: 0.6321 - val_loss: 2.6017 - val_acc: 0.1486\n",
      "Epoch 555/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9911 - acc: 0.6297 - val_loss: 2.7412 - val_acc: 0.1436\n",
      "Epoch 556/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9820 - acc: 0.6396 - val_loss: 3.2464 - val_acc: 0.0934\n",
      "Epoch 557/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9807 - acc: 0.6444 - val_loss: 3.0322 - val_acc: 0.1124\n",
      "Epoch 558/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9997 - acc: 0.6255 - val_loss: 2.6659 - val_acc: 0.1586\n",
      "Epoch 559/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9768 - acc: 0.6405 - val_loss: 2.8606 - val_acc: 0.1305\n",
      "Epoch 560/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9851 - acc: 0.6351 - val_loss: 2.7523 - val_acc: 0.1546\n",
      "Epoch 561/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9897 - acc: 0.6379 - val_loss: 2.9399 - val_acc: 0.1225\n",
      "Epoch 562/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9744 - acc: 0.6400 - val_loss: 2.6904 - val_acc: 0.1506\n",
      "Epoch 563/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9728 - acc: 0.6442 - val_loss: 2.7285 - val_acc: 0.1546\n",
      "Epoch 564/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9662 - acc: 0.6433 - val_loss: 2.7951 - val_acc: 0.1376\n",
      "Epoch 565/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9659 - acc: 0.6445 - val_loss: 2.7195 - val_acc: 0.1596\n",
      "Epoch 566/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9708 - acc: 0.6452 - val_loss: 2.8629 - val_acc: 0.1325\n",
      "Epoch 567/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9620 - acc: 0.6465 - val_loss: 2.9315 - val_acc: 0.1275\n",
      "Epoch 568/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9739 - acc: 0.6417 - val_loss: 2.7632 - val_acc: 0.1365\n",
      "Epoch 569/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9678 - acc: 0.6416 - val_loss: 2.7643 - val_acc: 0.1365\n",
      "Epoch 570/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9581 - acc: 0.6435 - val_loss: 2.8340 - val_acc: 0.1506\n",
      "Epoch 571/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9727 - acc: 0.6463 - val_loss: 2.7075 - val_acc: 0.1536\n",
      "Epoch 572/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9642 - acc: 0.6419 - val_loss: 2.7522 - val_acc: 0.1536\n",
      "Epoch 573/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9543 - acc: 0.6466 - val_loss: 2.7420 - val_acc: 0.1436\n",
      "Epoch 574/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9488 - acc: 0.6479 - val_loss: 2.9957 - val_acc: 0.1215\n",
      "Epoch 575/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9519 - acc: 0.6447 - val_loss: 2.7061 - val_acc: 0.1476\n",
      "Epoch 576/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9511 - acc: 0.6522 - val_loss: 3.3110 - val_acc: 0.1044\n",
      "Epoch 577/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9567 - acc: 0.6444 - val_loss: 2.9456 - val_acc: 0.1355\n",
      "Epoch 578/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9448 - acc: 0.6533 - val_loss: 2.6880 - val_acc: 0.1616\n",
      "Epoch 579/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9527 - acc: 0.6547 - val_loss: 3.2053 - val_acc: 0.1034\n",
      "Epoch 580/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9543 - acc: 0.6487 - val_loss: 2.7201 - val_acc: 0.1546\n",
      "Epoch 581/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9524 - acc: 0.6531 - val_loss: 2.7699 - val_acc: 0.1466\n",
      "Epoch 582/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9448 - acc: 0.6541 - val_loss: 2.7748 - val_acc: 0.1446\n",
      "Epoch 583/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9468 - acc: 0.6554 - val_loss: 2.5724 - val_acc: 0.1737\n",
      "Epoch 584/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9408 - acc: 0.6531 - val_loss: 3.1468 - val_acc: 0.1145\n",
      "Epoch 585/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9390 - acc: 0.6524 - val_loss: 3.0814 - val_acc: 0.1225\n",
      "Epoch 586/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9468 - acc: 0.6506 - val_loss: 2.6607 - val_acc: 0.1546\n",
      "Epoch 587/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9472 - acc: 0.6526 - val_loss: 3.0212 - val_acc: 0.1215\n",
      "Epoch 588/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9447 - acc: 0.6498 - val_loss: 3.0209 - val_acc: 0.1175\n",
      "Epoch 589/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9369 - acc: 0.6540 - val_loss: 2.6926 - val_acc: 0.1647\n",
      "Epoch 590/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9342 - acc: 0.6561 - val_loss: 2.7305 - val_acc: 0.1506\n",
      "Epoch 591/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9340 - acc: 0.6608 - val_loss: 2.6466 - val_acc: 0.1576\n",
      "Epoch 592/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9220 - acc: 0.6667 - val_loss: 3.2118 - val_acc: 0.1124\n",
      "Epoch 593/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9232 - acc: 0.6580 - val_loss: 2.9849 - val_acc: 0.1285\n",
      "Epoch 594/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9346 - acc: 0.6498 - val_loss: 2.7048 - val_acc: 0.1586\n",
      "Epoch 595/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9157 - acc: 0.6666 - val_loss: 2.9065 - val_acc: 0.1365\n",
      "Epoch 596/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9240 - acc: 0.6589 - val_loss: 2.9774 - val_acc: 0.1295\n",
      "Epoch 597/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9159 - acc: 0.6585 - val_loss: 2.6570 - val_acc: 0.1657\n",
      "Epoch 598/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9191 - acc: 0.6606 - val_loss: 2.7782 - val_acc: 0.1516\n",
      "Epoch 599/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9074 - acc: 0.6660 - val_loss: 3.4269 - val_acc: 0.1004\n",
      "Epoch 600/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9100 - acc: 0.6660 - val_loss: 2.9862 - val_acc: 0.1255\n",
      "Epoch 601/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9159 - acc: 0.6634 - val_loss: 2.8014 - val_acc: 0.1476\n",
      "Epoch 602/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9145 - acc: 0.6669 - val_loss: 2.9158 - val_acc: 0.1355\n",
      "Epoch 603/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9139 - acc: 0.6629 - val_loss: 2.7976 - val_acc: 0.1516\n",
      "Epoch 604/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9121 - acc: 0.6639 - val_loss: 3.0481 - val_acc: 0.1355\n",
      "Epoch 605/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9150 - acc: 0.6688 - val_loss: 2.8878 - val_acc: 0.1546\n",
      "Epoch 606/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9148 - acc: 0.6627 - val_loss: 2.5848 - val_acc: 0.1797\n",
      "Epoch 607/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9115 - acc: 0.6580 - val_loss: 2.3774 - val_acc: 0.2199\n",
      "Epoch 608/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9067 - acc: 0.6676 - val_loss: 2.8923 - val_acc: 0.1446\n",
      "Epoch 609/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9024 - acc: 0.6657 - val_loss: 2.9112 - val_acc: 0.1335\n",
      "Epoch 610/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9080 - acc: 0.6686 - val_loss: 3.1847 - val_acc: 0.1124\n",
      "Epoch 611/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9063 - acc: 0.6669 - val_loss: 3.0101 - val_acc: 0.1245\n",
      "Epoch 612/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8984 - acc: 0.6760 - val_loss: 2.7263 - val_acc: 0.1657\n",
      "Epoch 613/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8999 - acc: 0.6737 - val_loss: 3.0531 - val_acc: 0.1355\n",
      "Epoch 614/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8917 - acc: 0.6751 - val_loss: 2.7570 - val_acc: 0.1586\n",
      "Epoch 615/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8976 - acc: 0.6652 - val_loss: 2.9921 - val_acc: 0.1396\n",
      "Epoch 616/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.9047 - acc: 0.6603 - val_loss: 2.8291 - val_acc: 0.1556\n",
      "Epoch 617/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8828 - acc: 0.6816 - val_loss: 3.0464 - val_acc: 0.1315\n",
      "Epoch 618/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8946 - acc: 0.6765 - val_loss: 2.8465 - val_acc: 0.1576\n",
      "Epoch 619/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8901 - acc: 0.6786 - val_loss: 2.8698 - val_acc: 0.1446\n",
      "Epoch 620/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8825 - acc: 0.6805 - val_loss: 2.9394 - val_acc: 0.1345\n",
      "Epoch 621/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8700 - acc: 0.6805 - val_loss: 2.8439 - val_acc: 0.1496\n",
      "Epoch 622/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8788 - acc: 0.6770 - val_loss: 2.9662 - val_acc: 0.1496\n",
      "Epoch 623/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8825 - acc: 0.6770 - val_loss: 3.0016 - val_acc: 0.1255\n",
      "Epoch 624/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8915 - acc: 0.6672 - val_loss: 3.0080 - val_acc: 0.1345\n",
      "Epoch 625/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8685 - acc: 0.6826 - val_loss: 2.9899 - val_acc: 0.1315\n",
      "Epoch 626/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8775 - acc: 0.6765 - val_loss: 2.9243 - val_acc: 0.1376\n",
      "Epoch 627/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8829 - acc: 0.6744 - val_loss: 2.8981 - val_acc: 0.1486\n",
      "Epoch 628/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8654 - acc: 0.6847 - val_loss: 3.2876 - val_acc: 0.1104\n",
      "Epoch 629/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8735 - acc: 0.6791 - val_loss: 2.9946 - val_acc: 0.1396\n",
      "Epoch 630/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8725 - acc: 0.6882 - val_loss: 3.0241 - val_acc: 0.1335\n",
      "Epoch 631/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8736 - acc: 0.6833 - val_loss: 2.9437 - val_acc: 0.1335\n",
      "Epoch 632/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8523 - acc: 0.6933 - val_loss: 3.2098 - val_acc: 0.1145\n",
      "Epoch 633/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8720 - acc: 0.6802 - val_loss: 2.9108 - val_acc: 0.1426\n",
      "Epoch 634/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8541 - acc: 0.6889 - val_loss: 3.1446 - val_acc: 0.1185\n",
      "Epoch 635/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8637 - acc: 0.6866 - val_loss: 3.1369 - val_acc: 0.1205\n",
      "Epoch 636/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8641 - acc: 0.6861 - val_loss: 2.8206 - val_acc: 0.1627\n",
      "Epoch 637/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8560 - acc: 0.6896 - val_loss: 2.8677 - val_acc: 0.1486\n",
      "Epoch 638/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8619 - acc: 0.6900 - val_loss: 2.9326 - val_acc: 0.1556\n",
      "Epoch 639/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8558 - acc: 0.6882 - val_loss: 3.1440 - val_acc: 0.1235\n",
      "Epoch 640/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8519 - acc: 0.6861 - val_loss: 2.9290 - val_acc: 0.1556\n",
      "Epoch 641/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8664 - acc: 0.6875 - val_loss: 2.9042 - val_acc: 0.1526\n",
      "Epoch 642/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8469 - acc: 0.6947 - val_loss: 3.1306 - val_acc: 0.1335\n",
      "Epoch 643/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8674 - acc: 0.6825 - val_loss: 2.7858 - val_acc: 0.1657\n",
      "Epoch 644/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8516 - acc: 0.6905 - val_loss: 3.0889 - val_acc: 0.1335\n",
      "Epoch 645/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8512 - acc: 0.6889 - val_loss: 3.1856 - val_acc: 0.1285\n",
      "Epoch 646/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8489 - acc: 0.6938 - val_loss: 3.3275 - val_acc: 0.1114\n",
      "Epoch 647/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8457 - acc: 0.6886 - val_loss: 2.8813 - val_acc: 0.1466\n",
      "Epoch 648/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8531 - acc: 0.6900 - val_loss: 3.1072 - val_acc: 0.1456\n",
      "Epoch 649/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8402 - acc: 0.6977 - val_loss: 3.1487 - val_acc: 0.1295\n",
      "Epoch 650/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8415 - acc: 0.6942 - val_loss: 2.9107 - val_acc: 0.1576\n",
      "Epoch 651/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8407 - acc: 0.6873 - val_loss: 2.8882 - val_acc: 0.1576\n",
      "Epoch 652/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8342 - acc: 0.6973 - val_loss: 2.8352 - val_acc: 0.1496\n",
      "Epoch 653/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8398 - acc: 0.6924 - val_loss: 3.2480 - val_acc: 0.1124\n",
      "Epoch 654/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8343 - acc: 0.6936 - val_loss: 3.1085 - val_acc: 0.1335\n",
      "Epoch 655/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8274 - acc: 0.6975 - val_loss: 3.0812 - val_acc: 0.1275\n",
      "Epoch 656/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8374 - acc: 0.6931 - val_loss: 2.9379 - val_acc: 0.1486\n",
      "Epoch 657/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8294 - acc: 0.6947 - val_loss: 3.3981 - val_acc: 0.1165\n",
      "Epoch 658/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8242 - acc: 0.7046 - val_loss: 2.9949 - val_acc: 0.1446\n",
      "Epoch 659/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8315 - acc: 0.6984 - val_loss: 3.1145 - val_acc: 0.1396\n",
      "Epoch 660/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8168 - acc: 0.7074 - val_loss: 3.2328 - val_acc: 0.1255\n",
      "Epoch 661/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8244 - acc: 0.6998 - val_loss: 2.8322 - val_acc: 0.1637\n",
      "Epoch 662/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8161 - acc: 0.7022 - val_loss: 3.2025 - val_acc: 0.1245\n",
      "Epoch 663/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8190 - acc: 0.6954 - val_loss: 2.9595 - val_acc: 0.1647\n",
      "Epoch 664/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8257 - acc: 0.7033 - val_loss: 3.2534 - val_acc: 0.1245\n",
      "Epoch 665/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8157 - acc: 0.7017 - val_loss: 3.0986 - val_acc: 0.1345\n",
      "Epoch 666/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8268 - acc: 0.6982 - val_loss: 3.2007 - val_acc: 0.1215\n",
      "Epoch 667/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8002 - acc: 0.7199 - val_loss: 3.3447 - val_acc: 0.1305\n",
      "Epoch 668/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8156 - acc: 0.7015 - val_loss: 3.4097 - val_acc: 0.1215\n",
      "Epoch 669/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8072 - acc: 0.7008 - val_loss: 2.9143 - val_acc: 0.1596\n",
      "Epoch 670/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8085 - acc: 0.7095 - val_loss: 3.2419 - val_acc: 0.1335\n",
      "Epoch 671/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8106 - acc: 0.7064 - val_loss: 2.8271 - val_acc: 0.1546\n",
      "Epoch 672/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8054 - acc: 0.7053 - val_loss: 2.9437 - val_acc: 0.1647\n",
      "Epoch 673/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8119 - acc: 0.7017 - val_loss: 2.9639 - val_acc: 0.1556\n",
      "Epoch 674/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7934 - acc: 0.7164 - val_loss: 2.9623 - val_acc: 0.1526\n",
      "Epoch 675/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8057 - acc: 0.7109 - val_loss: 3.0940 - val_acc: 0.1416\n",
      "Epoch 676/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8027 - acc: 0.7113 - val_loss: 3.3710 - val_acc: 0.1175\n",
      "Epoch 677/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7945 - acc: 0.7115 - val_loss: 3.2242 - val_acc: 0.1325\n",
      "Epoch 678/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.8097 - acc: 0.7104 - val_loss: 2.9575 - val_acc: 0.1596\n",
      "Epoch 679/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7922 - acc: 0.7139 - val_loss: 3.0117 - val_acc: 0.1516\n",
      "Epoch 680/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7896 - acc: 0.7162 - val_loss: 3.3482 - val_acc: 0.1225\n",
      "Epoch 681/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7990 - acc: 0.7088 - val_loss: 3.0779 - val_acc: 0.1376\n",
      "Epoch 682/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7805 - acc: 0.7155 - val_loss: 3.1969 - val_acc: 0.1365\n",
      "Epoch 683/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7749 - acc: 0.7233 - val_loss: 2.7439 - val_acc: 0.1787\n",
      "Epoch 684/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7789 - acc: 0.7197 - val_loss: 3.2508 - val_acc: 0.1305\n",
      "Epoch 685/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7802 - acc: 0.7176 - val_loss: 3.0483 - val_acc: 0.1556\n",
      "Epoch 686/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7832 - acc: 0.7148 - val_loss: 3.5798 - val_acc: 0.1124\n",
      "Epoch 687/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7930 - acc: 0.7083 - val_loss: 3.4439 - val_acc: 0.1185\n",
      "Epoch 688/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7785 - acc: 0.7223 - val_loss: 3.3954 - val_acc: 0.1225\n",
      "Epoch 689/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7802 - acc: 0.7179 - val_loss: 3.4578 - val_acc: 0.1265\n",
      "Epoch 690/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7792 - acc: 0.7247 - val_loss: 3.0634 - val_acc: 0.1486\n",
      "Epoch 691/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7920 - acc: 0.7160 - val_loss: 3.0966 - val_acc: 0.1416\n",
      "Epoch 692/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7578 - acc: 0.7251 - val_loss: 3.1379 - val_acc: 0.1586\n",
      "Epoch 693/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7724 - acc: 0.7225 - val_loss: 3.0589 - val_acc: 0.1526\n",
      "Epoch 694/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7724 - acc: 0.7240 - val_loss: 3.2265 - val_acc: 0.1355\n",
      "Epoch 695/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7603 - acc: 0.7251 - val_loss: 3.6306 - val_acc: 0.1114\n",
      "Epoch 696/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7635 - acc: 0.7239 - val_loss: 3.1533 - val_acc: 0.1275\n",
      "Epoch 697/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7701 - acc: 0.7230 - val_loss: 3.2187 - val_acc: 0.1355\n",
      "Epoch 698/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7683 - acc: 0.7223 - val_loss: 3.6441 - val_acc: 0.0964\n",
      "Epoch 699/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7709 - acc: 0.7228 - val_loss: 3.2631 - val_acc: 0.1426\n",
      "Epoch 700/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7633 - acc: 0.7225 - val_loss: 2.9071 - val_acc: 0.1697\n",
      "Epoch 701/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7703 - acc: 0.7209 - val_loss: 3.1625 - val_acc: 0.1416\n",
      "Epoch 702/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7709 - acc: 0.7218 - val_loss: 3.3878 - val_acc: 0.1185\n",
      "Epoch 703/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7531 - acc: 0.7279 - val_loss: 3.2784 - val_acc: 0.1386\n",
      "Epoch 704/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7631 - acc: 0.7211 - val_loss: 4.1461 - val_acc: 0.0743\n",
      "Epoch 705/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7525 - acc: 0.7296 - val_loss: 3.0272 - val_acc: 0.1566\n",
      "Epoch 706/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7618 - acc: 0.7302 - val_loss: 3.5244 - val_acc: 0.1265\n",
      "Epoch 707/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7611 - acc: 0.7242 - val_loss: 3.1170 - val_acc: 0.1516\n",
      "Epoch 708/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7502 - acc: 0.7326 - val_loss: 3.4613 - val_acc: 0.1285\n",
      "Epoch 709/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7470 - acc: 0.7289 - val_loss: 3.3141 - val_acc: 0.1295\n",
      "Epoch 710/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7472 - acc: 0.7317 - val_loss: 3.2613 - val_acc: 0.1345\n",
      "Epoch 711/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7571 - acc: 0.7270 - val_loss: 3.1688 - val_acc: 0.1496\n",
      "Epoch 712/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7451 - acc: 0.7289 - val_loss: 3.3129 - val_acc: 0.1355\n",
      "Epoch 713/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7470 - acc: 0.7274 - val_loss: 3.7497 - val_acc: 0.1034\n",
      "Epoch 714/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7467 - acc: 0.7242 - val_loss: 3.1460 - val_acc: 0.1355\n",
      "Epoch 715/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7433 - acc: 0.7349 - val_loss: 3.6981 - val_acc: 0.0984\n",
      "Epoch 716/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7440 - acc: 0.7305 - val_loss: 3.3853 - val_acc: 0.1245\n",
      "Epoch 717/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7403 - acc: 0.7349 - val_loss: 3.1627 - val_acc: 0.1526\n",
      "Epoch 718/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7296 - acc: 0.7405 - val_loss: 3.3451 - val_acc: 0.1365\n",
      "Epoch 719/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7390 - acc: 0.7326 - val_loss: 3.2516 - val_acc: 0.1285\n",
      "Epoch 720/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7244 - acc: 0.7342 - val_loss: 3.4354 - val_acc: 0.1275\n",
      "Epoch 721/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7365 - acc: 0.7389 - val_loss: 3.0983 - val_acc: 0.1526\n",
      "Epoch 722/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7403 - acc: 0.7330 - val_loss: 3.0520 - val_acc: 0.1677\n",
      "Epoch 723/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7249 - acc: 0.7454 - val_loss: 3.1655 - val_acc: 0.1466\n",
      "Epoch 724/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7171 - acc: 0.7480 - val_loss: 3.4573 - val_acc: 0.1285\n",
      "Epoch 725/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7263 - acc: 0.7443 - val_loss: 3.3234 - val_acc: 0.1255\n",
      "Epoch 726/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7237 - acc: 0.7426 - val_loss: 3.7362 - val_acc: 0.1034\n",
      "Epoch 727/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7182 - acc: 0.7403 - val_loss: 3.2365 - val_acc: 0.1526\n",
      "Epoch 728/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7243 - acc: 0.7377 - val_loss: 3.5011 - val_acc: 0.1205\n",
      "Epoch 729/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7286 - acc: 0.7415 - val_loss: 3.4046 - val_acc: 0.1275\n",
      "Epoch 730/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7156 - acc: 0.7464 - val_loss: 3.4446 - val_acc: 0.1235\n",
      "Epoch 731/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7123 - acc: 0.7386 - val_loss: 3.1461 - val_acc: 0.1506\n",
      "Epoch 732/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7273 - acc: 0.7457 - val_loss: 3.4378 - val_acc: 0.1365\n",
      "Epoch 733/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7074 - acc: 0.7487 - val_loss: 3.3402 - val_acc: 0.1376\n",
      "Epoch 734/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7323 - acc: 0.7373 - val_loss: 3.2105 - val_acc: 0.1586\n",
      "Epoch 735/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7044 - acc: 0.7478 - val_loss: 3.5820 - val_acc: 0.1235\n",
      "Epoch 736/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7081 - acc: 0.7459 - val_loss: 3.4498 - val_acc: 0.1295\n",
      "Epoch 737/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7045 - acc: 0.7454 - val_loss: 3.5998 - val_acc: 0.1265\n",
      "Epoch 738/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7102 - acc: 0.7450 - val_loss: 3.1154 - val_acc: 0.1526\n",
      "Epoch 739/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7210 - acc: 0.7436 - val_loss: 3.4827 - val_acc: 0.1285\n",
      "Epoch 740/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6980 - acc: 0.7524 - val_loss: 3.2359 - val_acc: 0.1546\n",
      "Epoch 741/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7024 - acc: 0.7529 - val_loss: 3.2708 - val_acc: 0.1345\n",
      "Epoch 742/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7017 - acc: 0.7574 - val_loss: 3.3596 - val_acc: 0.1305\n",
      "Epoch 743/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7059 - acc: 0.7517 - val_loss: 3.4481 - val_acc: 0.1355\n",
      "Epoch 744/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6976 - acc: 0.7545 - val_loss: 3.5622 - val_acc: 0.1235\n",
      "Epoch 745/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7086 - acc: 0.7524 - val_loss: 3.2137 - val_acc: 0.1406\n",
      "Epoch 746/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6970 - acc: 0.7483 - val_loss: 3.2313 - val_acc: 0.1486\n",
      "Epoch 747/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6980 - acc: 0.7534 - val_loss: 3.5948 - val_acc: 0.1295\n",
      "Epoch 748/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6900 - acc: 0.7555 - val_loss: 3.0886 - val_acc: 0.1747\n",
      "Epoch 749/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.7016 - acc: 0.7476 - val_loss: 3.5497 - val_acc: 0.1255\n",
      "Epoch 750/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6798 - acc: 0.7609 - val_loss: 3.2765 - val_acc: 0.1526\n",
      "Epoch 751/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6815 - acc: 0.7616 - val_loss: 3.2376 - val_acc: 0.1406\n",
      "Epoch 752/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6882 - acc: 0.7557 - val_loss: 3.4682 - val_acc: 0.1305\n",
      "Epoch 753/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6882 - acc: 0.7494 - val_loss: 3.3450 - val_acc: 0.1396\n",
      "Epoch 754/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6822 - acc: 0.7592 - val_loss: 3.5812 - val_acc: 0.1225\n",
      "Epoch 755/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6882 - acc: 0.7552 - val_loss: 3.1447 - val_acc: 0.1606\n",
      "Epoch 756/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6707 - acc: 0.7625 - val_loss: 3.5539 - val_acc: 0.1155\n",
      "Epoch 757/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6707 - acc: 0.7651 - val_loss: 3.7661 - val_acc: 0.1124\n",
      "Epoch 758/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6809 - acc: 0.7641 - val_loss: 3.3674 - val_acc: 0.1446\n",
      "Epoch 759/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6836 - acc: 0.7592 - val_loss: 3.8487 - val_acc: 0.1084\n",
      "Epoch 760/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6828 - acc: 0.7580 - val_loss: 3.2854 - val_acc: 0.1506\n",
      "Epoch 761/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6916 - acc: 0.7482 - val_loss: 3.1828 - val_acc: 0.1596\n",
      "Epoch 762/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6699 - acc: 0.7607 - val_loss: 3.6347 - val_acc: 0.1255\n",
      "Epoch 763/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6670 - acc: 0.7635 - val_loss: 3.9281 - val_acc: 0.1104\n",
      "Epoch 764/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6736 - acc: 0.7611 - val_loss: 3.2933 - val_acc: 0.1456\n",
      "Epoch 765/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6673 - acc: 0.7642 - val_loss: 3.2357 - val_acc: 0.1586\n",
      "Epoch 766/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6567 - acc: 0.7677 - val_loss: 3.8529 - val_acc: 0.1195\n",
      "Epoch 767/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6571 - acc: 0.7714 - val_loss: 3.4702 - val_acc: 0.1396\n",
      "Epoch 768/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6501 - acc: 0.7730 - val_loss: 3.4094 - val_acc: 0.1315\n",
      "Epoch 769/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6738 - acc: 0.7595 - val_loss: 3.4460 - val_acc: 0.1335\n",
      "Epoch 770/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6508 - acc: 0.7698 - val_loss: 3.2443 - val_acc: 0.1466\n",
      "Epoch 771/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6639 - acc: 0.7616 - val_loss: 3.5417 - val_acc: 0.1335\n",
      "Epoch 772/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6580 - acc: 0.7667 - val_loss: 3.4070 - val_acc: 0.1365\n",
      "Epoch 773/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6627 - acc: 0.7651 - val_loss: 3.3237 - val_acc: 0.1486\n",
      "Epoch 774/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6645 - acc: 0.7623 - val_loss: 3.6083 - val_acc: 0.1315\n",
      "Epoch 775/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6693 - acc: 0.7648 - val_loss: 3.2492 - val_acc: 0.1576\n",
      "Epoch 776/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6586 - acc: 0.7739 - val_loss: 3.5569 - val_acc: 0.1275\n",
      "Epoch 777/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6574 - acc: 0.7670 - val_loss: 3.3230 - val_acc: 0.1466\n",
      "Epoch 778/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6609 - acc: 0.7616 - val_loss: 3.3305 - val_acc: 0.1536\n",
      "Epoch 779/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6572 - acc: 0.7700 - val_loss: 3.2992 - val_acc: 0.1386\n",
      "Epoch 780/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6564 - acc: 0.7723 - val_loss: 3.7980 - val_acc: 0.1094\n",
      "Epoch 781/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6516 - acc: 0.7700 - val_loss: 3.2927 - val_acc: 0.1627\n",
      "Epoch 782/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6354 - acc: 0.7787 - val_loss: 3.2913 - val_acc: 0.1536\n",
      "Epoch 783/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6503 - acc: 0.7642 - val_loss: 3.6635 - val_acc: 0.1315\n",
      "Epoch 784/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6428 - acc: 0.7747 - val_loss: 3.8074 - val_acc: 0.1165\n",
      "Epoch 785/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6376 - acc: 0.7780 - val_loss: 3.7634 - val_acc: 0.1245\n",
      "Epoch 786/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6358 - acc: 0.7747 - val_loss: 3.7379 - val_acc: 0.1145\n",
      "Epoch 787/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6377 - acc: 0.7767 - val_loss: 3.5981 - val_acc: 0.1305\n",
      "Epoch 788/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6289 - acc: 0.7842 - val_loss: 3.9275 - val_acc: 0.1074\n",
      "Epoch 789/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6358 - acc: 0.7761 - val_loss: 3.7545 - val_acc: 0.1265\n",
      "Epoch 790/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6422 - acc: 0.7723 - val_loss: 3.3952 - val_acc: 0.1386\n",
      "Epoch 791/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6365 - acc: 0.7768 - val_loss: 3.5757 - val_acc: 0.1315\n",
      "Epoch 792/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6340 - acc: 0.7704 - val_loss: 3.2785 - val_acc: 0.1566\n",
      "Epoch 793/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6368 - acc: 0.7772 - val_loss: 3.7148 - val_acc: 0.1195\n",
      "Epoch 794/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6265 - acc: 0.7767 - val_loss: 3.8900 - val_acc: 0.1135\n",
      "Epoch 795/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6292 - acc: 0.7768 - val_loss: 3.7258 - val_acc: 0.1325\n",
      "Epoch 796/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6396 - acc: 0.7801 - val_loss: 3.3957 - val_acc: 0.1466\n",
      "Epoch 797/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6279 - acc: 0.7826 - val_loss: 3.9236 - val_acc: 0.1135\n",
      "Epoch 798/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6288 - acc: 0.7782 - val_loss: 3.8441 - val_acc: 0.1205\n",
      "Epoch 799/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6178 - acc: 0.7808 - val_loss: 3.6131 - val_acc: 0.1285\n",
      "Epoch 800/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6159 - acc: 0.7854 - val_loss: 3.4028 - val_acc: 0.1506\n",
      "Epoch 801/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6223 - acc: 0.7829 - val_loss: 3.4784 - val_acc: 0.1546\n",
      "Epoch 802/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6226 - acc: 0.7782 - val_loss: 3.8086 - val_acc: 0.1295\n",
      "Epoch 803/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6146 - acc: 0.7793 - val_loss: 3.4479 - val_acc: 0.1446\n",
      "Epoch 804/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6231 - acc: 0.7864 - val_loss: 3.2791 - val_acc: 0.1616\n",
      "Epoch 805/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6119 - acc: 0.7836 - val_loss: 3.6236 - val_acc: 0.1315\n",
      "Epoch 806/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6091 - acc: 0.7847 - val_loss: 3.4756 - val_acc: 0.1416\n",
      "Epoch 807/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6168 - acc: 0.7852 - val_loss: 3.3349 - val_acc: 0.1566\n",
      "Epoch 808/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5925 - acc: 0.8002 - val_loss: 3.2773 - val_acc: 0.1586\n",
      "Epoch 809/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6119 - acc: 0.7845 - val_loss: 3.9628 - val_acc: 0.1084\n",
      "Epoch 810/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6089 - acc: 0.7835 - val_loss: 3.7564 - val_acc: 0.1215\n",
      "Epoch 811/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6171 - acc: 0.7821 - val_loss: 3.9951 - val_acc: 0.1094\n",
      "Epoch 812/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6043 - acc: 0.7887 - val_loss: 3.9632 - val_acc: 0.1155\n",
      "Epoch 813/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6097 - acc: 0.7814 - val_loss: 3.6679 - val_acc: 0.1365\n",
      "Epoch 814/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5909 - acc: 0.8004 - val_loss: 3.5244 - val_acc: 0.1516\n",
      "Epoch 815/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6011 - acc: 0.7905 - val_loss: 3.3937 - val_acc: 0.1456\n",
      "Epoch 816/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.6049 - acc: 0.7878 - val_loss: 3.3498 - val_acc: 0.1496\n",
      "Epoch 817/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5961 - acc: 0.7903 - val_loss: 3.9045 - val_acc: 0.1114\n",
      "Epoch 818/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5846 - acc: 0.8025 - val_loss: 3.6489 - val_acc: 0.1396\n",
      "Epoch 819/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5896 - acc: 0.7976 - val_loss: 3.5190 - val_acc: 0.1345\n",
      "Epoch 820/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5815 - acc: 0.8055 - val_loss: 3.6814 - val_acc: 0.1416\n",
      "Epoch 821/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5969 - acc: 0.7866 - val_loss: 3.8550 - val_acc: 0.1315\n",
      "Epoch 822/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5988 - acc: 0.7898 - val_loss: 3.6204 - val_acc: 0.1446\n",
      "Epoch 823/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5945 - acc: 0.7948 - val_loss: 3.8155 - val_acc: 0.1185\n",
      "Epoch 824/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5899 - acc: 0.7967 - val_loss: 3.6580 - val_acc: 0.1305\n",
      "Epoch 825/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5964 - acc: 0.7898 - val_loss: 4.0585 - val_acc: 0.1135\n",
      "Epoch 826/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5879 - acc: 0.7947 - val_loss: 3.7140 - val_acc: 0.1275\n",
      "Epoch 827/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5812 - acc: 0.7985 - val_loss: 3.2667 - val_acc: 0.1667\n",
      "Epoch 828/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5751 - acc: 0.8002 - val_loss: 3.5861 - val_acc: 0.1466\n",
      "Epoch 829/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5912 - acc: 0.7856 - val_loss: 3.5265 - val_acc: 0.1406\n",
      "Epoch 830/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5863 - acc: 0.7988 - val_loss: 3.6738 - val_acc: 0.1376\n",
      "Epoch 831/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5870 - acc: 0.7952 - val_loss: 3.6490 - val_acc: 0.1406\n",
      "Epoch 832/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5737 - acc: 0.8027 - val_loss: 3.4543 - val_acc: 0.1446\n",
      "Epoch 833/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5683 - acc: 0.8023 - val_loss: 3.6834 - val_acc: 0.1365\n",
      "Epoch 834/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5767 - acc: 0.7943 - val_loss: 3.5114 - val_acc: 0.1456\n",
      "Epoch 835/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5716 - acc: 0.8001 - val_loss: 3.7217 - val_acc: 0.1376\n",
      "Epoch 836/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5774 - acc: 0.7994 - val_loss: 3.6759 - val_acc: 0.1315\n",
      "Epoch 837/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5683 - acc: 0.8100 - val_loss: 3.1894 - val_acc: 0.1747\n",
      "Epoch 838/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5655 - acc: 0.7962 - val_loss: 3.6555 - val_acc: 0.1406\n",
      "Epoch 839/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5604 - acc: 0.8046 - val_loss: 3.6452 - val_acc: 0.1365\n",
      "Epoch 840/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5756 - acc: 0.7987 - val_loss: 3.7194 - val_acc: 0.1355\n",
      "Epoch 841/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5769 - acc: 0.8008 - val_loss: 3.5386 - val_acc: 0.1365\n",
      "Epoch 842/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5785 - acc: 0.7938 - val_loss: 3.8161 - val_acc: 0.1265\n",
      "Epoch 843/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5641 - acc: 0.8048 - val_loss: 3.2671 - val_acc: 0.1627\n",
      "Epoch 844/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5611 - acc: 0.8030 - val_loss: 3.3450 - val_acc: 0.1667\n",
      "Epoch 845/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5586 - acc: 0.7999 - val_loss: 4.0457 - val_acc: 0.1155\n",
      "Epoch 846/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5585 - acc: 0.8055 - val_loss: 3.8011 - val_acc: 0.1295\n",
      "Epoch 847/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5563 - acc: 0.8025 - val_loss: 3.6040 - val_acc: 0.1426\n",
      "Epoch 848/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5627 - acc: 0.8088 - val_loss: 3.9263 - val_acc: 0.1315\n",
      "Epoch 849/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5499 - acc: 0.8109 - val_loss: 3.8977 - val_acc: 0.1325\n",
      "Epoch 850/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5413 - acc: 0.8097 - val_loss: 4.1794 - val_acc: 0.1135\n",
      "Epoch 851/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5672 - acc: 0.7973 - val_loss: 3.8773 - val_acc: 0.1285\n",
      "Epoch 852/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5556 - acc: 0.8048 - val_loss: 4.1356 - val_acc: 0.1114\n",
      "Epoch 853/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5490 - acc: 0.8132 - val_loss: 3.8072 - val_acc: 0.1285\n",
      "Epoch 854/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5610 - acc: 0.7999 - val_loss: 4.1108 - val_acc: 0.1135\n",
      "Epoch 855/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5434 - acc: 0.8161 - val_loss: 3.4562 - val_acc: 0.1637\n",
      "Epoch 856/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5292 - acc: 0.8172 - val_loss: 3.2516 - val_acc: 0.1757\n",
      "Epoch 857/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5516 - acc: 0.8121 - val_loss: 3.5870 - val_acc: 0.1526\n",
      "Epoch 858/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5406 - acc: 0.8121 - val_loss: 3.3104 - val_acc: 0.1757\n",
      "Epoch 859/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5472 - acc: 0.8106 - val_loss: 3.7206 - val_acc: 0.1426\n",
      "Epoch 860/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5459 - acc: 0.8106 - val_loss: 3.7252 - val_acc: 0.1376\n",
      "Epoch 861/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5419 - acc: 0.8113 - val_loss: 3.9468 - val_acc: 0.1245\n",
      "Epoch 862/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5438 - acc: 0.8128 - val_loss: 3.6248 - val_acc: 0.1476\n",
      "Epoch 863/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5371 - acc: 0.8141 - val_loss: 3.8830 - val_acc: 0.1335\n",
      "Epoch 864/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5438 - acc: 0.8160 - val_loss: 3.5033 - val_acc: 0.1596\n",
      "Epoch 865/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5343 - acc: 0.8128 - val_loss: 3.6049 - val_acc: 0.1436\n",
      "Epoch 866/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5344 - acc: 0.8067 - val_loss: 3.6574 - val_acc: 0.1416\n",
      "Epoch 867/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5289 - acc: 0.8158 - val_loss: 3.8251 - val_acc: 0.1315\n",
      "Epoch 868/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5316 - acc: 0.8161 - val_loss: 3.9928 - val_acc: 0.1205\n",
      "Epoch 869/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5225 - acc: 0.8196 - val_loss: 3.4579 - val_acc: 0.1727\n",
      "Epoch 870/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5298 - acc: 0.8149 - val_loss: 4.0201 - val_acc: 0.1275\n",
      "Epoch 871/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5294 - acc: 0.8172 - val_loss: 3.7582 - val_acc: 0.1335\n",
      "Epoch 872/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5184 - acc: 0.8168 - val_loss: 4.0409 - val_acc: 0.1245\n",
      "Epoch 873/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5284 - acc: 0.8160 - val_loss: 3.7327 - val_acc: 0.1466\n",
      "Epoch 874/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5321 - acc: 0.8163 - val_loss: 3.4527 - val_acc: 0.1516\n",
      "Epoch 875/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5269 - acc: 0.8163 - val_loss: 3.8851 - val_acc: 0.1265\n",
      "Epoch 876/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5191 - acc: 0.8244 - val_loss: 3.8353 - val_acc: 0.1265\n",
      "Epoch 877/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5250 - acc: 0.8198 - val_loss: 4.0124 - val_acc: 0.1215\n",
      "Epoch 878/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5219 - acc: 0.8161 - val_loss: 3.8197 - val_acc: 0.1335\n",
      "Epoch 879/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5171 - acc: 0.8163 - val_loss: 3.9128 - val_acc: 0.1305\n",
      "Epoch 880/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5133 - acc: 0.8228 - val_loss: 3.7185 - val_acc: 0.1516\n",
      "Epoch 881/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5200 - acc: 0.8251 - val_loss: 3.8384 - val_acc: 0.1345\n",
      "Epoch 882/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5131 - acc: 0.8270 - val_loss: 3.8286 - val_acc: 0.1315\n",
      "Epoch 883/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5064 - acc: 0.8266 - val_loss: 3.7239 - val_acc: 0.1396\n",
      "Epoch 884/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5169 - acc: 0.8224 - val_loss: 4.1489 - val_acc: 0.1155\n",
      "Epoch 885/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5033 - acc: 0.8277 - val_loss: 3.3819 - val_acc: 0.1747\n",
      "Epoch 886/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5136 - acc: 0.8196 - val_loss: 3.2325 - val_acc: 0.1847\n",
      "Epoch 887/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5030 - acc: 0.8228 - val_loss: 3.9209 - val_acc: 0.1295\n",
      "Epoch 888/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5041 - acc: 0.8252 - val_loss: 4.3262 - val_acc: 0.1064\n",
      "Epoch 889/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5155 - acc: 0.8149 - val_loss: 3.2925 - val_acc: 0.1847\n",
      "Epoch 890/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5085 - acc: 0.8207 - val_loss: 3.8846 - val_acc: 0.1396\n",
      "Epoch 891/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4964 - acc: 0.8317 - val_loss: 3.4476 - val_acc: 0.1687\n",
      "Epoch 892/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5032 - acc: 0.8307 - val_loss: 3.8087 - val_acc: 0.1386\n",
      "Epoch 893/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5045 - acc: 0.8259 - val_loss: 3.7586 - val_acc: 0.1456\n",
      "Epoch 894/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4934 - acc: 0.8350 - val_loss: 3.8364 - val_acc: 0.1436\n",
      "Epoch 895/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5054 - acc: 0.8252 - val_loss: 3.7710 - val_acc: 0.1355\n",
      "Epoch 896/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5060 - acc: 0.8242 - val_loss: 4.0373 - val_acc: 0.1295\n",
      "Epoch 897/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.5077 - acc: 0.8280 - val_loss: 3.7270 - val_acc: 0.1376\n",
      "Epoch 898/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4870 - acc: 0.8350 - val_loss: 3.8008 - val_acc: 0.1396\n",
      "Epoch 899/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4927 - acc: 0.8352 - val_loss: 3.8335 - val_acc: 0.1386\n",
      "Epoch 900/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4912 - acc: 0.8321 - val_loss: 3.7087 - val_acc: 0.1496\n",
      "Epoch 901/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4915 - acc: 0.8319 - val_loss: 4.3507 - val_acc: 0.1124\n",
      "Epoch 902/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4964 - acc: 0.8270 - val_loss: 4.0914 - val_acc: 0.1265\n",
      "Epoch 903/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4924 - acc: 0.8331 - val_loss: 3.7743 - val_acc: 0.1546\n",
      "Epoch 904/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4905 - acc: 0.8296 - val_loss: 3.8708 - val_acc: 0.1355\n",
      "Epoch 905/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4870 - acc: 0.8352 - val_loss: 3.5505 - val_acc: 0.1657\n",
      "Epoch 906/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4835 - acc: 0.8357 - val_loss: 3.8515 - val_acc: 0.1506\n",
      "Epoch 907/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4902 - acc: 0.8329 - val_loss: 3.8145 - val_acc: 0.1446\n",
      "Epoch 908/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4691 - acc: 0.8397 - val_loss: 3.6902 - val_acc: 0.1486\n",
      "Epoch 909/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4883 - acc: 0.8308 - val_loss: 3.9721 - val_acc: 0.1285\n",
      "Epoch 910/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4674 - acc: 0.8424 - val_loss: 3.8995 - val_acc: 0.1325\n",
      "Epoch 911/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4849 - acc: 0.8355 - val_loss: 4.1818 - val_acc: 0.1225\n",
      "Epoch 912/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4744 - acc: 0.8394 - val_loss: 3.6928 - val_acc: 0.1496\n",
      "Epoch 913/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4753 - acc: 0.8319 - val_loss: 4.1224 - val_acc: 0.1175\n",
      "Epoch 914/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4860 - acc: 0.8368 - val_loss: 3.7306 - val_acc: 0.1406\n",
      "Epoch 915/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4871 - acc: 0.8305 - val_loss: 4.4183 - val_acc: 0.1145\n",
      "Epoch 916/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4720 - acc: 0.8417 - val_loss: 3.8355 - val_acc: 0.1376\n",
      "Epoch 917/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4679 - acc: 0.8368 - val_loss: 3.8833 - val_acc: 0.1386\n",
      "Epoch 918/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4875 - acc: 0.8321 - val_loss: 4.1971 - val_acc: 0.1225\n",
      "Epoch 919/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4782 - acc: 0.8329 - val_loss: 3.9511 - val_acc: 0.1386\n",
      "Epoch 920/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4698 - acc: 0.8406 - val_loss: 3.8844 - val_acc: 0.1436\n",
      "Epoch 921/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4655 - acc: 0.8390 - val_loss: 3.5925 - val_acc: 0.1687\n",
      "Epoch 922/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4645 - acc: 0.8420 - val_loss: 3.9024 - val_acc: 0.1426\n",
      "Epoch 923/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4762 - acc: 0.8341 - val_loss: 3.8725 - val_acc: 0.1396\n",
      "Epoch 924/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4608 - acc: 0.8464 - val_loss: 3.8381 - val_acc: 0.1486\n",
      "Epoch 925/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4657 - acc: 0.8422 - val_loss: 3.9702 - val_acc: 0.1396\n",
      "Epoch 926/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4641 - acc: 0.8434 - val_loss: 3.7919 - val_acc: 0.1506\n",
      "Epoch 927/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4678 - acc: 0.8427 - val_loss: 3.6695 - val_acc: 0.1526\n",
      "Epoch 928/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4523 - acc: 0.8452 - val_loss: 4.2651 - val_acc: 0.1245\n",
      "Epoch 929/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4603 - acc: 0.8453 - val_loss: 4.0135 - val_acc: 0.1325\n",
      "Epoch 930/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4654 - acc: 0.8396 - val_loss: 4.2141 - val_acc: 0.1215\n",
      "Epoch 931/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4562 - acc: 0.8460 - val_loss: 3.7965 - val_acc: 0.1446\n",
      "Epoch 932/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4559 - acc: 0.8460 - val_loss: 4.0088 - val_acc: 0.1355\n",
      "Epoch 933/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4403 - acc: 0.8523 - val_loss: 3.8252 - val_acc: 0.1436\n",
      "Epoch 934/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4377 - acc: 0.8527 - val_loss: 4.0089 - val_acc: 0.1416\n",
      "Epoch 935/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4567 - acc: 0.8427 - val_loss: 4.3514 - val_acc: 0.1195\n",
      "Epoch 936/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4626 - acc: 0.8390 - val_loss: 4.0435 - val_acc: 0.1275\n",
      "Epoch 937/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4643 - acc: 0.8383 - val_loss: 3.8745 - val_acc: 0.1516\n",
      "Epoch 938/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4615 - acc: 0.8387 - val_loss: 3.9229 - val_acc: 0.1456\n",
      "Epoch 939/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4339 - acc: 0.8563 - val_loss: 4.4625 - val_acc: 0.1165\n",
      "Epoch 940/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4486 - acc: 0.8495 - val_loss: 3.7888 - val_acc: 0.1546\n",
      "Epoch 941/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4437 - acc: 0.8476 - val_loss: 4.0983 - val_acc: 0.1355\n",
      "Epoch 942/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4494 - acc: 0.8490 - val_loss: 3.7762 - val_acc: 0.1506\n",
      "Epoch 943/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4444 - acc: 0.8474 - val_loss: 3.8492 - val_acc: 0.1476\n",
      "Epoch 944/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4357 - acc: 0.8504 - val_loss: 3.8644 - val_acc: 0.1526\n",
      "Epoch 945/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4454 - acc: 0.8544 - val_loss: 4.0487 - val_acc: 0.1365\n",
      "Epoch 946/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4353 - acc: 0.8541 - val_loss: 4.2088 - val_acc: 0.1245\n",
      "Epoch 947/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4365 - acc: 0.8483 - val_loss: 4.1489 - val_acc: 0.1365\n",
      "Epoch 948/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4413 - acc: 0.8494 - val_loss: 3.9367 - val_acc: 0.1396\n",
      "Epoch 949/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4344 - acc: 0.8534 - val_loss: 3.9348 - val_acc: 0.1496\n",
      "Epoch 950/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4328 - acc: 0.8539 - val_loss: 3.8358 - val_acc: 0.1496\n",
      "Epoch 951/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4143 - acc: 0.8600 - val_loss: 3.8044 - val_acc: 0.1526\n",
      "Epoch 952/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4276 - acc: 0.8527 - val_loss: 4.2355 - val_acc: 0.1295\n",
      "Epoch 953/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4349 - acc: 0.8541 - val_loss: 3.9235 - val_acc: 0.1446\n",
      "Epoch 954/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4165 - acc: 0.8598 - val_loss: 4.3195 - val_acc: 0.1255\n",
      "Epoch 955/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4376 - acc: 0.8478 - val_loss: 4.0297 - val_acc: 0.1325\n",
      "Epoch 956/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4341 - acc: 0.8494 - val_loss: 4.3002 - val_acc: 0.1225\n",
      "Epoch 957/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4224 - acc: 0.8618 - val_loss: 3.6957 - val_acc: 0.1616\n",
      "Epoch 958/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4300 - acc: 0.8553 - val_loss: 3.8706 - val_acc: 0.1456\n",
      "Epoch 959/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4333 - acc: 0.8492 - val_loss: 4.2563 - val_acc: 0.1335\n",
      "Epoch 960/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4216 - acc: 0.8598 - val_loss: 4.1426 - val_acc: 0.1255\n",
      "Epoch 961/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4205 - acc: 0.8525 - val_loss: 4.3705 - val_acc: 0.1175\n",
      "Epoch 962/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4157 - acc: 0.8612 - val_loss: 4.1777 - val_acc: 0.1345\n",
      "Epoch 963/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4264 - acc: 0.8560 - val_loss: 4.1915 - val_acc: 0.1285\n",
      "Epoch 964/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4096 - acc: 0.8691 - val_loss: 3.9639 - val_acc: 0.1396\n",
      "Epoch 965/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4265 - acc: 0.8544 - val_loss: 4.2196 - val_acc: 0.1265\n",
      "Epoch 966/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4248 - acc: 0.8569 - val_loss: 3.9941 - val_acc: 0.1396\n",
      "Epoch 967/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4140 - acc: 0.8616 - val_loss: 4.0616 - val_acc: 0.1446\n",
      "Epoch 968/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4193 - acc: 0.8560 - val_loss: 4.5118 - val_acc: 0.1165\n",
      "Epoch 969/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4129 - acc: 0.8628 - val_loss: 4.2827 - val_acc: 0.1295\n",
      "Epoch 970/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4203 - acc: 0.8579 - val_loss: 3.9211 - val_acc: 0.1526\n",
      "Epoch 971/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4088 - acc: 0.8625 - val_loss: 3.8946 - val_acc: 0.1526\n",
      "Epoch 972/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4147 - acc: 0.8583 - val_loss: 4.1794 - val_acc: 0.1345\n",
      "Epoch 973/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4015 - acc: 0.8695 - val_loss: 4.1294 - val_acc: 0.1295\n",
      "Epoch 974/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4016 - acc: 0.8691 - val_loss: 4.4180 - val_acc: 0.1265\n",
      "Epoch 975/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4196 - acc: 0.8572 - val_loss: 4.1779 - val_acc: 0.1325\n",
      "Epoch 976/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4171 - acc: 0.8618 - val_loss: 3.7762 - val_acc: 0.1596\n",
      "Epoch 977/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4088 - acc: 0.8630 - val_loss: 4.3524 - val_acc: 0.1265\n",
      "Epoch 978/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4036 - acc: 0.8670 - val_loss: 3.9303 - val_acc: 0.1536\n",
      "Epoch 979/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4013 - acc: 0.8621 - val_loss: 3.8526 - val_acc: 0.1496\n",
      "Epoch 980/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4127 - acc: 0.8686 - val_loss: 4.8187 - val_acc: 0.1064\n",
      "Epoch 981/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3970 - acc: 0.8625 - val_loss: 4.1590 - val_acc: 0.1406\n",
      "Epoch 982/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4200 - acc: 0.8521 - val_loss: 4.2636 - val_acc: 0.1345\n",
      "Epoch 983/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4040 - acc: 0.8628 - val_loss: 4.4401 - val_acc: 0.1185\n",
      "Epoch 984/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3964 - acc: 0.8642 - val_loss: 4.2134 - val_acc: 0.1335\n",
      "Epoch 985/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4040 - acc: 0.8625 - val_loss: 4.4278 - val_acc: 0.1185\n",
      "Epoch 986/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4002 - acc: 0.8644 - val_loss: 4.0168 - val_acc: 0.1436\n",
      "Epoch 987/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.4023 - acc: 0.8633 - val_loss: 3.9670 - val_acc: 0.1496\n",
      "Epoch 988/1000\n",
      "5722/5722 [==============================] - 13s 2ms/step - loss: 0.3968 - acc: 0.8670 - val_loss: 4.6381 - val_acc: 0.1135\n",
      "Epoch 989/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3889 - acc: 0.8708 - val_loss: 4.4265 - val_acc: 0.1225\n",
      "Epoch 990/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3849 - acc: 0.8728 - val_loss: 3.8380 - val_acc: 0.1566\n",
      "Epoch 991/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3947 - acc: 0.8700 - val_loss: 4.2331 - val_acc: 0.1295\n",
      "Epoch 992/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3946 - acc: 0.8658 - val_loss: 4.1472 - val_acc: 0.1315\n",
      "Epoch 993/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3891 - acc: 0.8679 - val_loss: 4.3510 - val_acc: 0.1345\n",
      "Epoch 994/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3987 - acc: 0.8653 - val_loss: 4.1286 - val_acc: 0.1426\n",
      "Epoch 995/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3887 - acc: 0.8695 - val_loss: 4.4808 - val_acc: 0.1285\n",
      "Epoch 996/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3975 - acc: 0.8649 - val_loss: 4.4414 - val_acc: 0.1285\n",
      "Epoch 997/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3889 - acc: 0.8707 - val_loss: 4.1427 - val_acc: 0.1406\n",
      "Epoch 998/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3845 - acc: 0.8724 - val_loss: 4.5886 - val_acc: 0.1275\n",
      "Epoch 999/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3829 - acc: 0.8728 - val_loss: 4.5473 - val_acc: 0.1175\n",
      "Epoch 1000/1000\n",
      "5722/5722 [==============================] - 12s 2ms/step - loss: 0.3810 - acc: 0.8719 - val_loss: 4.5212 - val_acc: 0.1265\n"
     ]
    }
   ],
   "source": [
    "cnnhistory=model.fit(x_traincnn, y_train, batch_size=32, epochs=1000, validation_data=(x_testcnn, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "#sigmoid\n",
    "plt.plot(cnnhistory.history['acc'])\n",
    "plt.plot(cnnhistory.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pydot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "ename": "OSError",
     "evalue": "`pydot` failed to call GraphViz.Please install GraphViz (https://www.graphviz.org/) and ensure that its executables are in the $PATH.",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/pydot.py\u001b[0m in \u001b[0;36mcreate\u001b[0;34m(self, prog, format, encoding)\u001b[0m\n\u001b[1;32m   1860\u001b[0m                 \u001b[0mshell\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1861\u001b[0;31m                 stderr=subprocess.PIPE, stdout=subprocess.PIPE)\n\u001b[0m\u001b[1;32m   1862\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/subprocess.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, args, bufsize, executable, stdin, stdout, stderr, preexec_fn, close_fds, shell, cwd, env, universal_newlines, startupinfo, creationflags, restore_signals, start_new_session, pass_fds, encoding, errors)\u001b[0m\n\u001b[1;32m    708\u001b[0m                                 \u001b[0merrread\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merrwrite\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 709\u001b[0;31m                                 restore_signals, start_new_session)\n\u001b[0m\u001b[1;32m    710\u001b[0m         \u001b[0;32mexcept\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/subprocess.py\u001b[0m in \u001b[0;36m_execute_child\u001b[0;34m(self, args, executable, preexec_fn, close_fds, pass_fds, cwd, env, startupinfo, creationflags, shell, p2cread, p2cwrite, c2pread, c2pwrite, errread, errwrite, restore_signals, start_new_session)\u001b[0m\n\u001b[1;32m   1343\u001b[0m                             \u001b[0merr_msg\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;34m': '\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mrepr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr_filename\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1344\u001b[0;31m                     \u001b[0;32mraise\u001b[0m \u001b[0mchild_exception_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merrno_num\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merr_msg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merr_filename\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1345\u001b[0m                 \u001b[0;32mraise\u001b[0m \u001b[0mchild_exception_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr_msg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: 'dot': 'dot'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/keras/utils/vis_utils.py\u001b[0m in \u001b[0;36m_check_pydot\u001b[0;34m()\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0;31m# to check the pydot/graphviz installation.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m         \u001b[0mpydot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpydot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     27\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/pydot.py\u001b[0m in \u001b[0;36mcreate\u001b[0;34m(self, prog, format, encoding)\u001b[0m\n\u001b[1;32m   1866\u001b[0m                     prog=prog)\n\u001b[0;32m-> 1867\u001b[0;31m                 \u001b[0;32mraise\u001b[0m \u001b[0mOSError\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[0m\n\u001b[0m\u001b[1;32m   1868\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] \"dot\" not found in path.",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-3-d98aa1f73144>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'full_model.h5'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m#loading\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mplot_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mto_file\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'modelPrebuilt.png'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0mscore\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mevaluate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_testcnn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/keras/utils/vis_utils.py\u001b[0m in \u001b[0;36mplot_model\u001b[0;34m(model, to_file, show_shapes, show_layer_names, rankdir)\u001b[0m\n\u001b[1;32m    130\u001b[0m             \u001b[0;34m'LR'\u001b[0m \u001b[0mcreates\u001b[0m \u001b[0ma\u001b[0m \u001b[0mhorizontal\u001b[0m \u001b[0mplot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    131\u001b[0m     \"\"\"\n\u001b[0;32m--> 132\u001b[0;31m     \u001b[0mdot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel_to_dot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshow_shapes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshow_layer_names\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrankdir\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    133\u001b[0m     \u001b[0m_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mextension\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplitext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mto_file\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    134\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mextension\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/keras/utils/vis_utils.py\u001b[0m in \u001b[0;36mmodel_to_dot\u001b[0;34m(model, show_shapes, show_layer_names, rankdir)\u001b[0m\n\u001b[1;32m     53\u001b[0m     \u001b[0;32mfrom\u001b[0m \u001b[0;34m.\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodels\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mSequential\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     54\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 55\u001b[0;31m     \u001b[0m_check_pydot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     56\u001b[0m     \u001b[0mdot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpydot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     57\u001b[0m     \u001b[0mdot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'rankdir'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrankdir\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/keras/utils/vis_utils.py\u001b[0m in \u001b[0;36m_check_pydot\u001b[0;34m()\u001b[0m\n\u001b[1;32m     27\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     28\u001b[0m         raise OSError(\n\u001b[0;32m---> 29\u001b[0;31m             \u001b[0;34m'`pydot` failed to call GraphViz.'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     30\u001b[0m             \u001b[0;34m'Please install GraphViz (https://www.graphviz.org/) '\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m             'and ensure that its executables are in the $PATH.')\n",
      "\u001b[0;31mOSError\u001b[0m: `pydot` failed to call GraphViz.Please install GraphViz (https://www.graphviz.org/) and ensure that its executables are in the $PATH."
     ]
    }
   ],
   "source": [
    "from keras.utils import plot_model\n",
    "from keras.models import load_model\n",
    "import os\n",
    "model = load_model('full_model.h5') #loading\n",
    "plot_model(model, to_file='modelPrebuilt.png')\n",
    "\n",
    "score = model.evaluate(x_testcnn, y_test, verbose=0)\n",
    "print(\"test loss:\", score[0])\n",
    "print(\"test accuracy:\", score[1])\n",
    "from keras.models import load_model\n",
    "#model.save('full_model.h5') #saving\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}