keyword_classifier-LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/arc/fdp5/.local/lib/python3.6/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "# Import phase if every import is here it is to allow you to run this cell and know if you have any issue and not after a few hours of training\n",
    "\n",
    "import tensorflow\n",
    "from tensorflow import keras\n",
    "import os\n",
    "import numpy as np\n",
    "import librosa\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "from keras_preprocessing.sequence import pad_sequences\n",
    "from tensorflow.keras.callbacks import Callback\n",
    "from tensorflow.keras.layers import Dense, Dropout, Conv1D, Activation, Flatten, MaxPooling1D, BatchNormalization, LSTM, GaussianNoise\n",
    "from tensorflow.keras.models import Sequential\n",
    "\n",
    "\n",
    "from src.VoiceModule import VoiceModule\n",
    "from src.classifiers.KerasClassifier import KerasClassifier\n",
    "from prepare_data_utils import load_wav, preprare_wav, split_sample\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "down\n",
      "tree\n",
      "_background_noise_\n",
      "go\n",
      "house\n",
      "validation_list.txt\n",
      "eight\n",
      "up\n",
      "bed\n",
      "two\n",
      "dog\n",
      "no\n",
      "bird\n",
      "five\n",
      "marvin\n",
      "seven\n",
      "four\n",
      "visual\n",
      "happy\n",
      "nine\n",
      "data_speech_commands_v0.02.tar.gz\n",
      "off\n",
      "yes\n",
      "forward\n",
      "follow\n",
      "README.md\n",
      "cat\n",
      "three\n",
      "on\n",
      "right\n",
      "backward\n",
      "testing_list.txt\n",
      "sheila\n",
      "wow\n",
      ".DS_Store\n",
      "stop\n",
      "one\n",
      "zero\n",
      "six\n",
      "learn\n",
      "LICENSE\n",
      "left\n"
     ]
    }
   ],
   "source": [
    "# print environment\n",
    "folder_path = \"../../data/keywords\"\n",
    "for l in os.listdir(folder_path):\n",
    "    print(l)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Transform data method\n",
    "\n",
    "def mfcc(buffer, samplerate, dim):\n",
    "    a = librosa.feature.mfcc(y=buffer, sr=samplerate, n_mfcc=dim)\n",
    "    a = np.transpose(a)\n",
    "    return a\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create callback to later draw some graphs\n",
    "\n",
    "class PerformanceLogger(Callback):\n",
    "    def __init__(self):\n",
    "        # List of tuple (loss, accuracy)\n",
    "        self.training = []\n",
    "        self.testing = []\n",
    "    def on_epoch_end(self, epoch, logs=None):\n",
    "        self.training.append((logs[\"loss\"], logs[\"acc\"]))\n",
    "        self.testing.append((logs[\"val_loss\"], logs[\"val_acc\"]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create class\n",
    "class EmotionClassifierCnn(KerasClassifier):\n",
    "    def __init__(self, file_path=None):\n",
    "        super().__init__(file_path)\n",
    "\n",
    "    def predict(self, x):\n",
    "        x = np.expand_dims(np.array([x]), axis=2)\n",
    "        return self._model.predict(x)\n",
    "\n",
    "    def build(self):\n",
    "        model = Sequential()\n",
    "        # Adding noise to the training data\n",
    "        model.add(GaussianNoise(0.2))\n",
    "        \n",
    "        # feature Extraction layers\n",
    "        model.add(BatchNormalization())\n",
    "        model.add(LSTM(512))\n",
    "        # End of the neuron and classification part\n",
    "        model.add(Dropout(0.4))\n",
    "        model.add(Flatten())\n",
    "        model.add(BatchNormalization())\n",
    "        model.add(Dense(128))\n",
    "        model.add(Dense(35))\n",
    "        model.add(Activation('softmax'))\n",
    "        model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "        self._model = model\n",
    "        \n",
    "    def train(self, X, Y, batch_size=32, epoch=720, validation_data=None, callbacks=[]):\n",
    "        self._model.fit(X, Y, batch_size=batch_size, epochs=epoch, validation_data=validation_data, callbacks=callbacks)\n",
    "\n",
    "    def transform(self, x, samplerate):\n",
    "        to_process = mfcc(x, samplerate, 13)\n",
    "        to_process = pad_sequences([to_process], maxlen=50, padding='post')[0]\n",
    "        return to_process\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35\n"
     ]
    }
   ],
   "source": [
    "#Instanciate model\n",
    "digit_list = [\"zero\", \"one\", \"two\", \"three\", \"four\", \"five\", \"six\", \"seven\", \"eight\", \"nine\"]\n",
    "keyword_list = [\"backward\", \"bed\", \"bird\", \"cat\", \"dog\", \"down\", \"follow\", \"forward\",\n",
    "               \"go\", \"happy\", \"house\", \"learn\", \"left\", \"marvin\", \"no\", \"off\", \"on\",\n",
    "               \"right\", \"sheila\", \"stop\", \"tree\", \"up\", \"visual\", \"wow\", \"yes\"] + digit_list\n",
    "print(len(keyword_list))\n",
    "dimension_mfcc = 13\n",
    "samplerate = 16000\n",
    "nb_break=None\n",
    "sample_duration=1\n",
    "step=1\n",
    "\n",
    "logger = PerformanceLogger()\n",
    "\n",
    "cls = EmotionClassifierCnn()\n",
    "vm = VoiceModule(\"digit\", keyword_list, cls)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Doing backward\n",
      "Doing bed\n",
      "Doing bird\n",
      "Doing cat\n",
      "Doing dog\n",
      "Doing down\n",
      "Doing follow\n",
      "Doing forward\n",
      "Doing go\n",
      "Doing happy\n",
      "Doing house\n",
      "Doing learn\n",
      "Doing left\n",
      "Doing marvin\n",
      "Doing no\n",
      "Doing off\n",
      "Doing on\n",
      "Doing right\n",
      "Doing sheila\n",
      "Doing stop\n",
      "Doing tree\n",
      "Doing up\n",
      "Doing visual\n",
      "Doing wow\n",
      "Doing yes\n",
      "Doing zero\n",
      "Doing one\n",
      "Doing two\n",
      "Doing three\n",
      "Doing four\n",
      "Doing five\n",
      "Doing six\n",
      "Doing seven\n",
      "Doing eight\n",
      "Doing nine\n",
      "Done (95394, 50, 13)\n"
     ]
    }
   ],
   "source": [
    "# Prepare data for training\n",
    "data = {f: load_wav(f\"{folder_path}/{f}\", nb_break) for f in keyword_list}\n",
    "X, Y = preprare_wav(data, vm, sample_duration, step)\n",
    "print(\"Done\", X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_file = f\"x_{'_'.join(keyword_list)}-{vm._name}.npy\"\n",
    "Y_file = f\"y_{'_'.join(keyword_list)}.npy\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Saving the prepared input data \n",
    "np.save(X_file, X)\n",
    "np.save(Y_file, Y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data for training\n",
    "X, Y = np.load(X_file), np.load(Y_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(95394, 35, 13)\n",
      "76315 19079\n",
      "(35, 13)\n",
      "float32\n"
     ]
    }
   ],
   "source": [
    "\"\"\"train_size = 80 * len(X) // 100\n",
    "test_size = 20 * len(X) // 100\n",
    "#X = np.expand_dims(X, axis=3)\n",
    "X_train, X_test, = X[:train_size], X[train_size:]\n",
    "Y_train, Y_test = Y[:train_size], Y[train_size:]\"\"\"\n",
    "X = X.astype('float32')\n",
    "print(X.shape)\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=42)\n",
    "print(len(X_train), len(X_test))\n",
    "\n",
    "print(X[0].shape)\n",
    "print( X.dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 76315 samples, validate on 19079 samples\n",
      "Epoch 1/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 2.1451 - acc: 0.3962 - val_loss: 1.5260 - val_acc: 0.6240\n",
      "Epoch 2/100\n",
      "76315/76315 [==============================] - 108s 1ms/sample - loss: 0.8160 - acc: 0.7557 - val_loss: 0.6800 - val_acc: 0.8226\n",
      "Epoch 3/100\n",
      "76315/76315 [==============================] - 117s 2ms/sample - loss: 0.5239 - acc: 0.8406 - val_loss: 0.4688 - val_acc: 0.8637\n",
      "Epoch 4/100\n",
      "76315/76315 [==============================] - 121s 2ms/sample - loss: 0.4074 - acc: 0.8746 - val_loss: 0.4102 - val_acc: 0.8759\n",
      "Epoch 5/100\n",
      "76315/76315 [==============================] - 122s 2ms/sample - loss: 0.3319 - acc: 0.8979 - val_loss: 0.3578 - val_acc: 0.8928\n",
      "Epoch 6/100\n",
      "76315/76315 [==============================] - 117s 2ms/sample - loss: 0.2741 - acc: 0.9149 - val_loss: 0.3329 - val_acc: 0.9016\n",
      "Epoch 7/100\n",
      "76315/76315 [==============================] - 116s 2ms/sample - loss: 0.2404 - acc: 0.9255 - val_loss: 0.3193 - val_acc: 0.9074\n",
      "Epoch 8/100\n",
      "76315/76315 [==============================] - 116s 2ms/sample - loss: 0.2082 - acc: 0.9346 - val_loss: 0.2985 - val_acc: 0.9141\n",
      "Epoch 9/100\n",
      "76315/76315 [==============================] - 114s 1ms/sample - loss: 0.1772 - acc: 0.9445 - val_loss: 0.3201 - val_acc: 0.9132\n",
      "Epoch 10/100\n",
      "76315/76315 [==============================] - 117s 2ms/sample - loss: 0.1600 - acc: 0.9492 - val_loss: 0.3121 - val_acc: 0.9132\n",
      "Epoch 11/100\n",
      "76315/76315 [==============================] - 116s 2ms/sample - loss: 0.1445 - acc: 0.9541 - val_loss: 0.2890 - val_acc: 0.9216\n",
      "Epoch 12/100\n",
      "76315/76315 [==============================] - 116s 2ms/sample - loss: 0.1252 - acc: 0.9612 - val_loss: 0.2932 - val_acc: 0.9206\n",
      "Epoch 13/100\n",
      "76315/76315 [==============================] - 117s 2ms/sample - loss: 0.1117 - acc: 0.9645 - val_loss: 0.3041 - val_acc: 0.9209\n",
      "Epoch 14/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.1017 - acc: 0.9682 - val_loss: 0.3120 - val_acc: 0.9212\n",
      "Epoch 15/100\n",
      "76315/76315 [==============================] - 115s 2ms/sample - loss: 0.0859 - acc: 0.9726 - val_loss: 0.3059 - val_acc: 0.9247\n",
      "Epoch 16/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0811 - acc: 0.9743 - val_loss: 0.3137 - val_acc: 0.9244\n",
      "Epoch 17/100\n",
      "76315/76315 [==============================] - 103s 1ms/sample - loss: 0.0788 - acc: 0.9749 - val_loss: 0.3216 - val_acc: 0.9250\n",
      "Epoch 18/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0749 - acc: 0.9760 - val_loss: 0.3242 - val_acc: 0.9258\n",
      "Epoch 19/100\n",
      "76315/76315 [==============================] - 116s 2ms/sample - loss: 0.0686 - acc: 0.9781 - val_loss: 0.3189 - val_acc: 0.9263\n",
      "Epoch 20/100\n",
      "76315/76315 [==============================] - 114s 1ms/sample - loss: 0.0675 - acc: 0.9786 - val_loss: 0.3411 - val_acc: 0.9224\n",
      "Epoch 21/100\n",
      "76315/76315 [==============================] - 115s 2ms/sample - loss: 0.0620 - acc: 0.9804 - val_loss: 0.3379 - val_acc: 0.9241\n",
      "Epoch 22/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0549 - acc: 0.9826 - val_loss: 0.3316 - val_acc: 0.9266\n",
      "Epoch 23/100\n",
      "76315/76315 [==============================] - 115s 2ms/sample - loss: 0.0507 - acc: 0.9836 - val_loss: 0.3418 - val_acc: 0.9284\n",
      "Epoch 24/100\n",
      "76315/76315 [==============================] - 115s 2ms/sample - loss: 0.0493 - acc: 0.9843 - val_loss: 0.3395 - val_acc: 0.9270\n",
      "Epoch 25/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0466 - acc: 0.9853 - val_loss: 0.3415 - val_acc: 0.9281\n",
      "Epoch 26/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0456 - acc: 0.9857 - val_loss: 0.3632 - val_acc: 0.9245\n",
      "Epoch 27/100\n",
      "76315/76315 [==============================] - 114s 1ms/sample - loss: 0.0436 - acc: 0.9862 - val_loss: 0.3559 - val_acc: 0.9268\n",
      "Epoch 28/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0440 - acc: 0.9862 - val_loss: 0.3389 - val_acc: 0.9306\n",
      "Epoch 29/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0353 - acc: 0.9893 - val_loss: 0.3452 - val_acc: 0.9309\n",
      "Epoch 30/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0346 - acc: 0.9891 - val_loss: 0.3497 - val_acc: 0.9317\n",
      "Epoch 31/100\n",
      "76315/76315 [==============================] - 113s 1ms/sample - loss: 0.0404 - acc: 0.9873 - val_loss: 0.3803 - val_acc: 0.9269\n",
      "Epoch 32/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0407 - acc: 0.9870 - val_loss: 0.3405 - val_acc: 0.9345\n",
      "Epoch 33/100\n",
      "76315/76315 [==============================] - 101s 1ms/sample - loss: 0.0353 - acc: 0.9886 - val_loss: 0.3578 - val_acc: 0.9314\n",
      "Epoch 34/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0316 - acc: 0.9900 - val_loss: 0.3680 - val_acc: 0.9306\n",
      "Epoch 35/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0290 - acc: 0.9907 - val_loss: 0.3627 - val_acc: 0.9326\n",
      "Epoch 36/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0302 - acc: 0.9909 - val_loss: 0.3767 - val_acc: 0.9299\n",
      "Epoch 37/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0299 - acc: 0.9909 - val_loss: 0.3785 - val_acc: 0.9308\n",
      "Epoch 38/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0283 - acc: 0.9914 - val_loss: 0.3750 - val_acc: 0.9300\n",
      "Epoch 39/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0273 - acc: 0.9911 - val_loss: 0.3749 - val_acc: 0.9325\n",
      "Epoch 40/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0334 - acc: 0.9899 - val_loss: 0.3737 - val_acc: 0.9301\n",
      "Epoch 41/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0350 - acc: 0.9890 - val_loss: 0.3806 - val_acc: 0.9296\n",
      "Epoch 42/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0277 - acc: 0.9915 - val_loss: 0.3568 - val_acc: 0.9340\n",
      "Epoch 43/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0256 - acc: 0.9921 - val_loss: 0.3713 - val_acc: 0.9325\n",
      "Epoch 44/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0222 - acc: 0.9932 - val_loss: 0.3845 - val_acc: 0.9323\n",
      "Epoch 45/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0216 - acc: 0.9933 - val_loss: 0.3778 - val_acc: 0.9334\n",
      "Epoch 46/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0179 - acc: 0.9945 - val_loss: 0.3871 - val_acc: 0.9344\n",
      "Epoch 47/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0186 - acc: 0.9942 - val_loss: 0.3872 - val_acc: 0.9322\n",
      "Epoch 48/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0196 - acc: 0.9939 - val_loss: 0.3949 - val_acc: 0.9318\n",
      "Epoch 49/100\n",
      "76315/76315 [==============================] - 100s 1ms/sample - loss: 0.0214 - acc: 0.9938 - val_loss: 0.3853 - val_acc: 0.9324\n",
      "Epoch 50/100\n",
      "76315/76315 [==============================] - 108s 1ms/sample - loss: 0.0255 - acc: 0.9920 - val_loss: 0.4042 - val_acc: 0.9308\n",
      "Epoch 51/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0207 - acc: 0.9937 - val_loss: 0.3794 - val_acc: 0.9330\n",
      "Epoch 52/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0182 - acc: 0.9945 - val_loss: 0.3818 - val_acc: 0.9353\n",
      "Epoch 53/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0158 - acc: 0.9952 - val_loss: 0.3979 - val_acc: 0.9311\n",
      "Epoch 54/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0226 - acc: 0.9932 - val_loss: 0.3969 - val_acc: 0.9321\n",
      "Epoch 55/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0188 - acc: 0.9943 - val_loss: 0.3940 - val_acc: 0.9339\n",
      "Epoch 56/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0191 - acc: 0.9940 - val_loss: 0.3945 - val_acc: 0.9347\n",
      "Epoch 57/100\n",
      "76315/76315 [==============================] - 108s 1ms/sample - loss: 0.0128 - acc: 0.9961 - val_loss: 0.3866 - val_acc: 0.9356\n",
      "Epoch 58/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0159 - acc: 0.9953 - val_loss: 0.4085 - val_acc: 0.9346\n",
      "Epoch 59/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0189 - acc: 0.9941 - val_loss: 0.4252 - val_acc: 0.9313\n",
      "Epoch 60/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0204 - acc: 0.9933 - val_loss: 0.4153 - val_acc: 0.9321\n",
      "Epoch 61/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0254 - acc: 0.9923 - val_loss: 0.4048 - val_acc: 0.9318\n",
      "Epoch 62/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0210 - acc: 0.9940 - val_loss: 0.3721 - val_acc: 0.9361\n",
      "Epoch 63/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0155 - acc: 0.9951 - val_loss: 0.3836 - val_acc: 0.9355\n",
      "Epoch 64/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0158 - acc: 0.9954 - val_loss: 0.4250 - val_acc: 0.9322\n",
      "Epoch 65/100\n",
      "76315/76315 [==============================] - 105s 1ms/sample - loss: 0.0179 - acc: 0.9946 - val_loss: 0.4091 - val_acc: 0.9331\n",
      "Epoch 66/100\n",
      "76315/76315 [==============================] - 102s 1ms/sample - loss: 0.0182 - acc: 0.9945 - val_loss: 0.4055 - val_acc: 0.9342\n",
      "Epoch 67/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0130 - acc: 0.9959 - val_loss: 0.3958 - val_acc: 0.9371\n",
      "Epoch 68/100\n",
      "76315/76315 [==============================] - 107s 1ms/sample - loss: 0.0088 - acc: 0.9972 - val_loss: 0.4100 - val_acc: 0.9371\n",
      "Epoch 69/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0106 - acc: 0.9969 - val_loss: 0.4218 - val_acc: 0.9339\n",
      "Epoch 70/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0177 - acc: 0.9950 - val_loss: 0.4253 - val_acc: 0.9333\n",
      "Epoch 71/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0232 - acc: 0.9936 - val_loss: 0.4291 - val_acc: 0.9301\n",
      "Epoch 72/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0212 - acc: 0.9935 - val_loss: 0.4252 - val_acc: 0.9304\n",
      "Epoch 73/100\n",
      "76315/76315 [==============================] - 112s 1ms/sample - loss: 0.0159 - acc: 0.9948 - val_loss: 0.4261 - val_acc: 0.9337\n",
      "Epoch 74/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0138 - acc: 0.9959 - val_loss: 0.4247 - val_acc: 0.9347\n",
      "Epoch 75/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0118 - acc: 0.9962 - val_loss: 0.4447 - val_acc: 0.9326\n",
      "Epoch 76/100\n",
      "76315/76315 [==============================] - 108s 1ms/sample - loss: 0.0135 - acc: 0.9959 - val_loss: 0.4332 - val_acc: 0.9334\n",
      "Epoch 77/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0156 - acc: 0.9952 - val_loss: 0.4353 - val_acc: 0.9327\n",
      "Epoch 78/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0139 - acc: 0.9961 - val_loss: 0.3999 - val_acc: 0.9384\n",
      "Epoch 79/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0096 - acc: 0.9971 - val_loss: 0.4162 - val_acc: 0.9357\n",
      "Epoch 80/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0093 - acc: 0.9972 - val_loss: 0.4472 - val_acc: 0.9328\n",
      "Epoch 81/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0081 - acc: 0.9974 - val_loss: 0.4370 - val_acc: 0.9349\n",
      "Epoch 82/100\n",
      "76315/76315 [==============================] - 100s 1ms/sample - loss: 0.0087 - acc: 0.9974 - val_loss: 0.4220 - val_acc: 0.9370\n",
      "Epoch 83/100\n",
      "76315/76315 [==============================] - 105s 1ms/sample - loss: 0.0112 - acc: 0.9969 - val_loss: 0.4130 - val_acc: 0.9362\n",
      "Epoch 84/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0104 - acc: 0.9971 - val_loss: 0.4101 - val_acc: 0.9349\n",
      "Epoch 85/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0147 - acc: 0.9958 - val_loss: 0.4209 - val_acc: 0.9356\n",
      "Epoch 86/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0139 - acc: 0.9958 - val_loss: 0.4218 - val_acc: 0.9360\n",
      "Epoch 87/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0112 - acc: 0.9965 - val_loss: 0.4296 - val_acc: 0.9340\n",
      "Epoch 88/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0082 - acc: 0.9975 - val_loss: 0.4107 - val_acc: 0.9385\n",
      "Epoch 89/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0128 - acc: 0.9964 - val_loss: 0.4283 - val_acc: 0.9359\n",
      "Epoch 90/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0182 - acc: 0.9945 - val_loss: 0.4182 - val_acc: 0.9354\n",
      "Epoch 91/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0123 - acc: 0.9962 - val_loss: 0.4192 - val_acc: 0.9376\n",
      "Epoch 92/100\n",
      "76315/76315 [==============================] - 111s 1ms/sample - loss: 0.0098 - acc: 0.9971 - val_loss: 0.4201 - val_acc: 0.9386\n",
      "Epoch 93/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0112 - acc: 0.9966 - val_loss: 0.4201 - val_acc: 0.9367\n",
      "Epoch 94/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0115 - acc: 0.9965 - val_loss: 0.4281 - val_acc: 0.9355\n",
      "Epoch 95/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0129 - acc: 0.9963 - val_loss: 0.4215 - val_acc: 0.9355\n",
      "Epoch 96/100\n",
      "76315/76315 [==============================] - 109s 1ms/sample - loss: 0.0105 - acc: 0.9968 - val_loss: 0.4293 - val_acc: 0.9361\n",
      "Epoch 97/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0078 - acc: 0.9978 - val_loss: 0.4306 - val_acc: 0.9373\n",
      "Epoch 98/100\n",
      "76315/76315 [==============================] - 106s 1ms/sample - loss: 0.0074 - acc: 0.9979 - val_loss: 0.4315 - val_acc: 0.9367\n",
      "Epoch 99/100\n",
      "76315/76315 [==============================] - 100s 1ms/sample - loss: 0.0097 - acc: 0.9972 - val_loss: 0.4363 - val_acc: 0.9358\n",
      "Epoch 100/100\n",
      "76315/76315 [==============================] - 110s 1ms/sample - loss: 0.0111 - acc: 0.9967 - val_loss: 0.4529 - val_acc: 0.9347\n"
     ]
    }
   ],
   "source": [
    "vm.model.train(X_train, Y_train, batch_size=1024, epoch=100, validation_data=(X_test, Y_test), callbacks=[logger])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9f045d4828>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.plot(list(range(0, len(logger.training))), list(map(lambda x:x[1], logger.training)), label=\"training data\")\n",
    "plt.plot(list(range(0, len(logger.testing))), list(map(lambda x:x[1], logger.testing)), label=\"testing data\")\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}