HAR-02_model-training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Human Activity Recognition\n",
    "Credit to:\n",
    "* https://github.com/guillaume-chevalier/LSTM-Human-Activity-Recognition\n",
    "* https://github.com/cwi-dis/mobile-har-tutorial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 17:05:06.473370: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
      "To enable the following instructions: AVX2 AVX512F FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2024-02-10 17:05:07.575788: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "import scipy.io\n",
    "import sklearn\n",
    "import sklearn.metrics\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "#import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "#import matplotlib.colors as colors\n",
    "#import matplotlib.cm\n",
    "import pickle\n",
    "\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GPU and CPU settings\n",
    "If GPU is not available, comment out the bottom block."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "if True:\n",
    "    # Enter either 0 or 1 below to select the GPU to use\n",
    "    os.environ[\"CUDA_VISIBLE_DEVICES\"]=\"0\"\n",
    "\n",
    "    physical_devices = tf.config.experimental.list_physical_devices('GPU')\n",
    "    tf.config.set_visible_devices(physical_devices[0], 'GPU')\n",
    "    tf.config.experimental.set_memory_growth(physical_devices[0], True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Config your ConvLSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "## UNCOMMENT below line for segmenting the signal in overlapping windows of 90 samples with 50% overlap\n",
    "# segments, labels, subjects = segment_signal(df)\n",
    "\n",
    "## COMMENT below segments + labels files if you want to segment afresh . open a file, where you stored the pickled data\n",
    "segments = pickle.load(open('./data/segments_90_logo.p', 'rb'), encoding='latin1')\n",
    "labels = pickle.load(open('./data/labels_90_logo.p','rb'), encoding='latin1')\n",
    "subjects = pickle.load(open('./data/subjects_90_logo.p','rb'),encoding='latin1')\n",
    "subjects = subjects.astype(int)\n",
    "\n",
    "## dump information to that file (UNCOMMENT to save fresh segmentation!)\n",
    "# pickle.dump(segments, open( './data/segments_90_logo.p','wb'))\n",
    "# pickle.dump(labels, open( './data/labels_90_logo.p','wb'))\n",
    "# pickle.dump(subjects, open( './data/subjects_90_logo.p','wb'))\n",
    "\n",
    "# segments, labels, subjects = segment_signal(df)\n",
    "\n",
    "groups = np.array(subjects)\n",
    "\n",
    "## defining parameters for the input and network layers\n",
    "## we are treating each segmeent or chunk as a 2D image (90 X 3)\n",
    "numOfSamplesInWindow = segments.shape[1]\n",
    "numOfChannels = segments.shape[2]\n",
    "\n",
    "## reshaping the data for network input\n",
    "reshapedSegments = segments.reshape(segments.shape[0], numOfSamplesInWindow, numOfChannels, 1)\n",
    "## (observations, timesteps, features (x,y,z), channels)\n",
    "\n",
    "# categorically defining the classes of the activities\n",
    "labels = np.asarray(pd.get_dummies(labels),dtype = np.int8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(62476, 90, 6, 1)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reshapedSegments.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([10, 12, 1, 13, 6, 9, 3, 2, 14, 5], [8, 11, 4], [7])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lstPIds = sorted(set(subjects.astype(int)))\n",
    "np.random.seed(42)\n",
    "np.random.shuffle(lstPIds)\n",
    "splitTrain = .7\n",
    "splitVal = .2\n",
    "splitTest = .1\n",
    "splitTrainAmount = int(np.round(len(lstPIds)*splitTrain))\n",
    "splitValAmount = int(np.round(len(lstPIds)*splitVal)) + splitTrainAmount\n",
    "splitTestAmount = int(np.round(len(lstPIds)*splitTest)) + splitValAmount\n",
    "lstPIdsTrain = lstPIds[:splitTrainAmount]\n",
    "lstPIdsVal =  lstPIds[splitTrainAmount:splitValAmount]\n",
    "lstPIdsTest = lstPIds[splitValAmount:]\n",
    "lstPIdsTrain, lstPIdsVal, lstPIdsTest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "x_train = reshapedSegments[np.isin(subjects, lstPIdsTrain)]\n",
    "x_val = reshapedSegments[np.isin(subjects, lstPIdsVal)]\n",
    "x_test = reshapedSegments[np.isin(subjects, lstPIdsTrain)]\n",
    "\n",
    "y_train_one_hot = labels[np.isin(subjects, lstPIdsTrain)]\n",
    "y_val_one_hot = labels[np.isin(subjects, lstPIdsVal)]\n",
    "y_test_one_hot = labels[np.isin(subjects, lstPIdsTrain)]\n",
    "\n",
    "#x_train = np.expand_dims(x_train, 1)\n",
    "#x_val = np.expand_dims(x_val, 1)\n",
    "#x_test = np.expand_dims(x_test, 1)\n",
    "#\n",
    "#y_train_one_hot = np.expand_dims(y_train_one_hot, 1)\n",
    "#y_val_one_hot = np.expand_dims(y_val_one_hot, 1)\n",
    "#y_test_one_hot = np.expand_dims(y_test_one_hot, 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(45511, 90, 6, 1, 1)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.expand_dims(x_train,-1).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fc35b54a390>,\n",
       " <matplotlib.lines.Line2D at 0x7fc35b52e710>,\n",
       " <matplotlib.lines.Line2D at 0x7fc35b52ea50>,\n",
       " <matplotlib.lines.Line2D at 0x7fc35b52ee50>,\n",
       " <matplotlib.lines.Line2D at 0x7fc35b52f190>,\n",
       " <matplotlib.lines.Line2D at 0x7fc35b52f550>]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(x_train[0].reshape(90,-1))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(62476, 12)\n",
      "12\n"
     ]
    }
   ],
   "source": [
    "# number of epochs\n",
    "Epochs = 2\n",
    "\n",
    "batch_size = 32\n",
    "\n",
    "## ConvLSTM net hyperparameters\n",
    "numChannels = 1\n",
    "numFilters = 128 # number of filters in Conv2D layer\n",
    "# kernal size of the Conv2D layer\n",
    "kernalSize1 = 2\n",
    "# max pooling window size\n",
    "poolingWindowSz = 2\n",
    "# number of filters in fully connected layers\n",
    "numNueronsFCL1 = 128\n",
    "numNueronsFCL2 = 128\n",
    "\n",
    "# number of total classes\n",
    "numClasses = labels.shape[1]\n",
    "# dropout ratio for dropout layer\n",
    "dropOutRatio = 0.2\n",
    "\n",
    "## number of total clases\n",
    "numClasses = labels.shape[1]\n",
    "print(labels.shape)\n",
    "print(numClasses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 17:05:15.200876: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1635] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 22888 MB memory:  -> device: 0, name: Tesla V100-SXM2-32GB, pci bus id: 0000:62:00.0, compute capability: 7.0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input (InputLayer)          [(None, 90, 6, 1)]        0         \n",
      "                                                                 \n",
      " conv_lstm1d (ConvLSTM1D)    (None, 5, 128)            132608    \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 640)               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 128)               82048     \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 128)               16512     \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 12)                1548      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 232,716\n",
      "Trainable params: 232,716\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "\n",
    "inputs = tf.keras.layers.Input(shape=(numOfSamplesInWindow, numOfChannels, 1), name=\"input\")\n",
    "layer = tf.keras.layers.ConvLSTM1D(numFilters, kernalSize1)(inputs)\n",
    "layer = tf.keras.layers.Flatten()(layer)\n",
    "layer = tf.keras.layers.Dense(numNueronsFCL1, activation='relu')(layer)\n",
    "layer = tf.keras.layers.Dense(numNueronsFCL2, activation='relu')(layer)\n",
    "output = tf.keras.layers.Dense(numClasses, activation='softmax')(layer)\n",
    "\n",
    "model = tf.keras.Model(inputs=inputs, outputs=output)\n",
    "\n",
    "\n",
    "## Compiling the model to generate a model\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate = 0.001)\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer=optimizer,\n",
    "              metrics=['accuracy'])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train, evaluate & save your network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-10 17:05:19.826790: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:424] Loaded cuDNN version 8600\n",
      "2024-02-10 17:05:20.282150: I tensorflow/compiler/xla/service/service.cc:169] XLA service 0x7fc0c0008b60 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2024-02-10 17:05:20.282211: I tensorflow/compiler/xla/service/service.cc:177]   StreamExecutor device (0): Tesla V100-SXM2-32GB, Compute Capability 7.0\n",
      "2024-02-10 17:05:20.294616: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:269] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
      "2024-02-10 17:05:20.500601: I ./tensorflow/compiler/jit/device_compiler.h:180] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1423/1423 [==============================] - 755s 526ms/step - loss: 0.8266 - accuracy: 0.6990 - val_loss: 1.1959 - val_accuracy: 0.6133\n",
      "Epoch 2/2\n",
      "1423/1423 [==============================] - 746s 524ms/step - loss: 0.5211 - accuracy: 0.8062 - val_loss: 1.0496 - val_accuracy: 0.6453\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_train,\n",
    "                    y_train_one_hot,\n",
    "                    validation_data=(x_val, y_val_one_hot),\n",
    "                    epochs=Epochs,\n",
    "                    batch_size=batch_size,\n",
    "                    verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fc35b53b210>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAG2CAYAAACZEEfAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABDK0lEQVR4nO3deZyNdf/H8feZfWFmrLP42U2ZNMi+VGS5R+SOlCXLkGgZhGTJnoqkcou4uRktlnLfkRbEFEq2MFImWUOZUcmMIWPMXL8/5DTHbOeMM3Nmrl7Px+M8zPle2+dcJufd9/pe19diGIYhAAAAk3JzdQEAAACFibADAABMjbADAABMjbADAABMjbADAABMjbADAABMjbADAABMjbADAABMjbADAABMjbADAABMzaVhZ+vWrercubPCwsJksVi0Zs0am+WGYWjSpEkKDQ2Vr6+v2rVrp8OHD9usc+7cOfXu3VsBAQEKCgrSwIEDlZqaWoSfAgAAFGcuDTsXL15UvXr1NG/evByXz5w5U3PmzNGCBQu0c+dO+fv7KyoqSpcvX7au07t3b3333XfauHGjPvroI23dulWDBw8uqo8AAACKOUtxmQjUYrFo9erV6tKli6RrvTphYWF6+umnNWrUKElScnKygoODtXTpUvXs2VMJCQm67bbbtHv3bjVq1EiStH79enXs2FGnT59WWFiYqz4OAAAoJjxcXUBujh8/rsTERLVr187aFhgYqKZNm2r79u3q2bOntm/frqCgIGvQkaR27drJzc1NO3fuVNeuXXPcd1pamtLS0qzvMzMzde7cOZUrV04Wi6XwPhQAAHAawzB04cIFhYWFyc0t94tVxTbsJCYmSpKCg4Nt2oODg63LEhMTVbFiRZvlHh4eKlu2rHWdnEyfPl1Tp051csUAAMAVTp06pf/7v//LdXmxDTuFady4cRo5cqT1fXJysqpUqaJTp04pICDAhZUBAAB7paSkqHLlyipdunSe6xXbsBMSEiJJSkpKUmhoqLU9KSlJ9evXt65z9uxZm+2uXr2qc+fOWbfPibe3t7y9vbO1BwQEEHYAAChh8huCUmyfs1O9enWFhIQoLi7O2paSkqKdO3eqefPmkqTmzZvr/Pnz2rNnj3Wdzz77TJmZmWratGmR1wwAAIofl/bspKam6siRI9b3x48fV3x8vMqWLasqVapo+PDhev755xUeHq7q1atr4sSJCgsLs96xFRERoQ4dOmjQoEFasGCB0tPTNWTIEPXs2ZM7sQAAgCQXh52vv/5a99xzj/X99XE00dHRWrp0qUaPHq2LFy9q8ODBOn/+vO68806tX79ePj4+1m2WLVumIUOGqG3btnJzc1O3bt00Z86cIv8sAACgeCo2z9lxpZSUFAUGBio5OZkxOwBKnIyMDKWnp7u6DMDpPD095e7unutye7+/i+0AZQBA3gzDUGJios6fP+/qUoBCExQUpJCQkJt6Dh5hBwBKqOtBp2LFivLz8+OhqDAVwzB06dIl613XWe/MdhRhBwBKoIyMDGvQKVeunKvLAQqFr6+vJOns2bOqWLFinpe08lJsbz0HAOTu+hgdPz8/F1cCFK7rv+M3My6NsAMAJRiXrmB2zvgdJ+wAAABTI+wAAEqU1q1ba/jw4db31apV0+zZs/PcxmKxaM2aNTd9bGftB0WLsAMAKBKdO3dWhw4dclz2xRdfyGKx6JtvvnF4v7t379bgwYNvtjwbU6ZMsc7DmNWZM2d07733OvVYufnjjz9UtmxZlS9fXmlpaUVyTLMi7AAAisTAgQO1ceNGnT59Otuy2NhYNWrUSHXr1nV4vxUqVCiygdohISE5TiRdGP73v/+pTp06ql27tst7kwzD0NWrV11aw80g7AAAisR9992nChUqaOnSpTbtqampWrVqlQYOHKjffvtNvXr1UqVKleTn56fIyEitWLEiz/3eeBnr8OHDuvvuu+Xj46PbbrtNGzduzLbNmDFjdMstt8jPz081atTQxIkTrXf7LF26VFOnTtX+/ftlsVhksVisNd94GevAgQNq06aNfH19Va5cOQ0ePFipqanW5f3791eXLl00a9YshYaGqly5coqJibHrzqLFixerT58+6tOnjxYvXpxt+Xfffaf77rtPAQEBKl26tO666y4dPXrUunzJkiWqU6eOvL29FRoaqiFDhkiSTpw4IYvFovj4eOu658+fl8Vi0ebNmyVJmzdvlsVi0bp169SwYUN5e3vryy+/1NGjR3X//fcrODhYpUqVUuPGjbVp0yabutLS0jRmzBhVrlxZ3t7eqlWrlhYvXizDMFSrVi3NmjXLZv34+HhZLBabuTKdjefsAIBJGIahP9Izivy4vp7udt0x4+HhoX79+mnp0qUaP368dZtVq1YpIyNDvXr1Umpqqho2bKgxY8YoICBAH3/8sfr27auaNWuqSZMm+R4jMzNTDzzwgIKDg7Vz504lJyfbjO+5rnTp0lq6dKnCwsJ04MABDRo0SKVLl9bo0aPVo0cPffvtt1q/fr31izwwMDDbPi5evKioqCg1b95cu3fv1tmzZ/Xoo49qyJAhNoHu888/V2hoqD7//HMdOXJEPXr0UP369TVo0KBcP8fRo0e1fft2vf/++zIMQyNGjNCPP/6oqlWrSpJ++ukn3X333WrdurU+++wzBQQEaNu2bdbel/nz52vkyJGaMWOG7r33XiUnJ2vbtm35nr8bjR07VrNmzVKNGjVUpkwZnTp1Sh07dtQLL7wgb29vvfXWW+rcubMOHTqkKlWqSJL69eun7du3a86cOapXr56OHz+uX3/9VRaLRY888ohiY2M1atQo6zFiY2N19913q1atWg7XZy/CDgCYxB/pGbpt0oYiP+7B56Lk52Xf18kjjzyil19+WVu2bFHr1q0lXfuy69atmwIDAxUYGGjzRTh06FBt2LBB7733nl1hZ9OmTfr++++1YcMGhYWFSZJefPHFbONsJkyYYP25WrVqGjVqlFauXKnRo0fL19dXpUqVkoeHh0JCQnI91vLly3X58mW99dZb8vf3lyTNnTtXnTt31ksvvaTg4GBJUpkyZTR37ly5u7urdu3a6tSpk+Li4vIMO0uWLNG9996rMmXKSJKioqIUGxurKVOmSJLmzZunwMBArVy5Up6enpKkW265xbr9888/r6efflpPPfWUta1x48b5nr8bPffcc2rfvr31fdmyZVWvXj3r+2nTpmn16tVau3athgwZoh9++EHvvfeeNm7cqHbt2kmSatSoYV2/f//+mjRpknbt2qUmTZooPT1dy5cvz9bb42xcxgIAFJnatWurRYsWWrJkiSTpyJEj+uKLLzRw4EBJ154MPW3aNEVGRqps2bIqVaqUNmzYoJMnT9q1/4SEBFWuXNkadCSpefPm2dZ799131bJlS4WEhKhUqVKaMGGC3cfIeqx69epZg44ktWzZUpmZmTp06JC1rU6dOjZP/g0NDbVOgZCTjIwMvfnmm+rTp4+1rU+fPlq6dKkyMzMlXbv0c9ddd1mDTlZnz57Vzz//rLZt2zr0eXLSqFEjm/epqakaNWqUIiIiFBQUpFKlSikhIcF67uLj4+Xu7q5WrVrluL+wsDB16tTJ+vf/4YcfKi0tTQ899NBN15oXenYAwCR8Pd118LkolxzXEQMHDtTQoUM1b948xcbGqmbNmtYvx5dffln/+te/NHv2bEVGRsrf31/Dhw/XlStXnFbv9u3b1bt3b02dOlVRUVHWHpJXXnnFacfI6sZAYrFYrKElJxs2bNBPP/2kHj162LRnZGQoLi5O7du3t06jkJO8lkmSm9u1fg7DMKxtuY0hyhrkJGnUqFHauHGjZs2apVq1asnX11cPPvig9e8nv2NL0qOPPqq+ffvqtddeU2xsrHr06FHoA8zp2QEAk7BYLPLz8ijyl6NPuO3evbvc3Ny0fPlyvfXWW3rkkUes+9i2bZvuv/9+9enTR/Xq1VONGjX0ww8/2L3viIgInTp1SmfOnLG27dixw2adr776SlWrVtX48ePVqFEjhYeH68cff7RZx8vLSxkZeY9/ioiI0P79+3Xx4kVr27Zt2+Tm5qZbb73V7ppvtHjxYvXs2VPx8fE2r549e1oHKtetW1dffPFFjiGldOnSqlatmuLi4nLcf4UKFSTJ5hxlHaycl23btql///7q2rWrIiMjFRISohMnTliXR0ZGKjMzU1u2bMl1Hx07dpS/v7/mz5+v9evX65FHHrHr2DeDsAMAKFKlSpVSjx49NG7cOJ05c0b9+/e3LgsPD9fGjRv11VdfKSEhQY899piSkpLs3ne7du10yy23KDo6Wvv379cXX3yh8ePH26wTHh6ukydPauXKlTp69KjmzJmj1atX26xTrVo1HT9+XPHx8fr1119zfM5N79695ePjo+joaH377bf6/PPPNXToUPXt29c6XsdRv/zyiz788ENFR0fr9ttvt3n169dPa9as0blz5zRkyBClpKSoZ8+e+vrrr3X48GG9/fbb1stnU6ZM0SuvvKI5c+bo8OHD2rt3r15//XVJ13pfmjVrphkzZighIUFbtmyxGcOUl/DwcL3//vuKj4/X/v379fDDD9v0UlWrVk3R0dF65JFHtGbNGh0/flybN2/We++9Z13H3d1d/fv317hx4xQeHp7jZUZnI+wAAIrcwIED9fvvvysqKspmfM2ECRPUoEEDRUVFqXXr1goJCVGXLl3s3q+bm5tWr16tP/74Q02aNNGjjz6qF154wWadf/7znxoxYoSGDBmi+vXr66uvvtLEiRNt1unWrZs6dOige+65RxUqVMjx9nc/Pz9t2LBB586dU+PGjfXggw+qbdu2mjt3rmMnI4vrg51zGm/Ttm1b+fr66p133lG5cuX02WefKTU1Va1atVLDhg21aNEi6yWz6OhozZ49W2+88Ybq1Kmj++67T4cPH7bua8mSJbp69aoaNmyo4cOH6/nnn7ervldffVVlypRRixYt1LlzZ0VFRalBgwY268yfP18PPvignnzySdWuXVuDBg2y6f2Srv39X7lyRQMGDHD0FBWIxch60e5vKiUlRYGBgUpOTlZAQICrywGAfF2+fFnHjx9X9erV5ePj4+pyAId88cUXatu2rU6dOpVvL1hev+v2fn8zQBkAABSJtLQ0/fLLL5oyZYoeeuihAl/ucxSXsQAAQJFYsWKFqlatqvPnz2vmzJlFdlzCDgAAKBL9+/dXRkaG9uzZo0qVKhXZcQk7AADA1Ag7AADA1Ag7AADA1Ag7AADA1Ag7AADA1Ag7AADA1Ag7AIASrVq1apo9e7bd62/evFkWi0Xnz58vtJpQvBB2AABFwmKx5PmaMmVKgfa7e/duDR482O71W7RooTNnzigwMLBAxyuI2rVry9vbW4mJiUV2TPyFsAMAKBJnzpyxvmbPnq2AgACbtlGjRlnXNQxDV69etWu/FSpUkJ+fn911eHl5KSQkRBaLxeHPUBBffvml/vjjDz344IN68803i+SYeUlPT3d1CUWOsAMAKBIhISHWV2BgoCwWi/X9999/r9KlS2vdunVq2LChvL299eWXX+ro0aO6//77FRwcrFKlSqlx48batGmTzX5vvIxlsVj0n//8R127dpWfn5/Cw8O1du1a6/IbL2MtXbpUQUFB2rBhgyIiIlSqVCl16NBBZ86csW5z9epVDRs2TEFBQSpXrpzGjBmj6Ohou2ZkX7x4sR5++GH17dtXS5Ysybb89OnT6tWrl8qWLSt/f381atRIO3futC7/8MMP1bhxY/n4+Kh8+fLq2rWrzWdds2aNzf6CgoK0dOlSSdKJEydksVj07rvvqlWrVvLx8dGyZcv022+/qVevXqpUqZL8/PwUGRmZbWb3zMxMzZw5U7Vq1ZK3t7eqVKlinUG+TZs2GjJkiM36v/zyi7y8vBQXF5fvOSlqhB0AMAvDkK5cLPqXYTjtI4wdO1YzZsxQQkKC6tatq9TUVHXs2FFxcXHat2+fOnTooM6dO+vkyZN57mfq1Knq3r27vvnmG3Xs2FG9e/fWuXPncl3/0qVLmjVrlt5++21t3bpVJ0+etOlpeumll7Rs2TLFxsZq27ZtSklJyRYycnLhwgWtWrVKffr0Ufv27ZWcnKwvvvjCujw1NVWtWrXSTz/9pLVr12r//v0aPXq0MjMzJUkff/yxunbtqo4dO2rfvn2Ki4tTkyZN8j3ujcaOHaunnnpKCQkJioqK0uXLl9WwYUN9/PHH+vbbbzV48GD17dtXu3btsm4zbtw4zZgxQxMnTtTBgwe1fPly68Sdjz76qJYvX660tDTr+u+8844qVaqkNm3aOFxfYWPWcwAwi/RL0othRX/cZ3+WvPydsqvnnntO7du3t74vW7as6tWrZ30/bdo0rV69WmvXrs3Ws5BV//791atXL0nSiy++qDlz5mjXrl3q0KFDjuunp6drwYIFqlmzpiRpyJAheu6556zLX3/9dY0bN87aqzJ37lx98skn+X6elStXKjw8XHXq1JEk9ezZU4sXL9Zdd90lSVq+fLl++eUX7d69W2XLlpUk1apVy7r9Cy+8oJ49e2rq1KnWtqznw17Dhw/XAw88YNOWNcwNHTpUGzZs0HvvvacmTZrowoUL+te//qW5c+cqOjpaklSzZk3deeedkqQHHnhAQ4YM0QcffKDu3btLutZD1r9//yK7POgIenYAAMVGo0aNbN6npqZq1KhRioiIUFBQkEqVKqWEhIR8e3bq1q1r/dnf318BAQE6e/Zsruv7+flZg44khYaGWtdPTk5WUlKSTY+Ku7u7GjZsmO/nWbJkifr06WN936dPH61atUoXLlyQJMXHx+uOO+6wBp0bxcfHq23btvkeJz83nteMjAxNmzZNkZGRKlu2rEqVKqUNGzZYz2tCQoLS0tJyPbaPj4/NZbm9e/fq22+/Vf/+/W+61sJAzw4AmIWn37VeFlcc10n8/W17iEaNGqWNGzdq1qxZqlWrlnx9ffXggw/qypUreZfk6Wnz3mKxWC8N2bu+cZOX5w4ePKgdO3Zo165dGjNmjLU9IyNDK1eu1KBBg+Tr65vnPvJbnlOdOQ1AvvG8vvzyy/rXv/6l2bNnKzIyUv7+/ho+fLj1vOZ3XOnapaz69evr9OnTio2NVZs2bVS1atV8t3MFenYAwCwslmuXk4r6VYiXLbZt26b+/fura9euioyMVEhIiE6cOFFox8tJYGCggoODtXv3bmtbRkaG9u7dm+d2ixcv1t133639+/crPj7e+ho5cqQWL14s6VoPVHx8fK7jierWrZvngN8KFSrYDKQ+fPiwLl26lO9n2rZtm+6//3716dNH9erVU40aNfTDDz9Yl4eHh8vX1zfPY0dGRqpRo0ZatGiRli9frkceeSTf47oKYQcAUGyFh4fr/fffV3x8vPbv36+HH344zx6awjJ06FBNnz5dH3zwgQ4dOqSnnnpKv//+e67jU9LT0/X222+rV69euv32221ejz76qHbu3KnvvvtOvXr1UkhIiLp06aJt27bp2LFj+t///qft27dLkiZPnqwVK1Zo8uTJSkhI0IEDB/TSSy9Zj9OmTRvNnTtX+/bt09dff63HH388Wy9VTsLDw7Vx40Z99dVXSkhI0GOPPaakpCTrch8fH40ZM0ajR4/WW2+9paNHj2rHjh3WkHbdo48+qhkzZsgwDJu7xIobwg4AoNh69dVXVaZMGbVo0UKdO3dWVFSUGjRoUOR1jBkzRr169VK/fv3UvHlzlSpVSlFRUfLx8clx/bVr1+q3337LMQBEREQoIiJCixcvlpeXlz799FNVrFhRHTt2VGRkpGbMmCF3d3dJUuvWrbVq1SqtXbtW9evXV5s2bWzumHrllVdUuXJl3XXXXXr44Yc1atQou545NGHCBDVo0EBRUVFq3bq1NXBlNXHiRD399NOaNGmSIiIi1KNHj2zjnnr16iUPDw/16tUr13NRHFiMm70oaQIpKSkKDAxUcnKyAgICXF0OAOTr8uXLOn78uKpXr16sv2TMKjMzUxEREerevbumTZvm6nJc5sSJE6pZs6Z2795daCE0r991e7+/GaAMAEA+fvzxR3366adq1aqV0tLSNHfuXB0/flwPP/ywq0tzifT0dP3222+aMGGCmjVr5pLeNkdwGQsAgHy4ublp6dKlaty4sVq2bKkDBw5o06ZNioiIcHVpLrFt2zaFhoZq9+7dWrBggavLyRc9OwAA5KNy5cratm2bq8soNlq3bn3Tt+YXJXp2AACAqRF2AKAEK0n/dw0UhDN+xwk7AFACXX+Wij0PkANKsuu/4/Y8Pyg3jNkBgBLI3d1dQUFB1uee+Pn5FcsJGIGCMgxDly5d0tmzZxUUFGR99lBBEHYAoIQKCQmRpDwnuARKuqCgIOvvekERdgCghLJYLAoNDVXFihVznPwRKOk8PT1vqkfnOsIOAJRw7u7uTvlCAMyKAcoAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUinXYycjI0MSJE1W9enX5+vqqZs2amjZtmgzDsK5jGIYmTZqk0NBQ+fr6ql27djp8+LALqwYA4G8s46p0IUlK+k46tlk68F8py/e2K3i49Oj5eOmllzR//ny9+eabqlOnjr7++msNGDBAgYGBGjZsmCRp5syZmjNnjt58801Vr15dEydOVFRUlA4ePCgfHx8XfwIAAEq4zAzp0jnp4i/SpV+v/Xnxt+zvr//8x+/Z9xHeXvIJLPra/1Ssw85XX32l+++/X506dZIkVatWTStWrNCuXbskXevVmT17tiZMmKD7779fkvTWW28pODhYa9asUc+ePV1WOwAAxVJm5rVAYhNWfr32yinMXDonycGeGYub5FtW8q8g+ZeX0i8TdnLTokULLVy4UD/88INuueUW7d+/X19++aVeffVVSdLx48eVmJiodu3aWbcJDAxU06ZNtX379lzDTlpamtLS0qzvU1JSCveDAABQWDIzpcvn/wws1wPMjeEly/tLv0lGpoMHsUi+Zf4KL/7lJb/yub/3LSO5uRfGpy2QYh12xo4dq5SUFNWuXVvu7u7KyMjQCy+8oN69e0uSEhMTJUnBwcE22wUHB1uX5WT69OmaOnVq4RUOAEBBGcaf4SWHS0U5vv9NMjIcP45vmT8Dij3hpazkXqwjQ56KdeXvvfeeli1bpuXLl6tOnTqKj4/X8OHDFRYWpujo6ALvd9y4cRo5cqT1fUpKiipXruyMkgEAsGUYUlpK3r0tN4aXzHTHj+MTmHdgyfrer6zk7un8z1pMFeuw88wzz2js2LHWy1GRkZH68ccfNX36dEVHRyskJESSlJSUpNDQUOt2SUlJql+/fq779fb2lre3d6HWDgAwKcOQrqTm09tyw6WkjCuOH8c7QPIrZ2d4KSd5eDn/s5pEsQ47ly5dkpub7d3x7u7uysy8dq2xevXqCgkJUVxcnDXcpKSkaOfOnXriiSeKulwAQEl15WIu4SWX3peMtPz3eSOvUjcElj+DTNb3WcOMB/9T7izFOux07txZL7zwgqpUqaI6depo3759evXVV/XII49IkiwWi4YPH67nn39e4eHh1lvPw8LC1KVLF9cWDwBwnSuX8u9tybrs6h+OH8PTL/+xLtb35SVPX+d/TtilWIed119/XRMnTtSTTz6ps2fPKiwsTI899pgmTZpkXWf06NG6ePGiBg8erPPnz+vOO+/U+vXrecYOAJhJ+mU7BupmeZ9+0fFjePj8FVLyHbhbXvLyd/7nRKGwGIaLH2tYDKSkpCgwMFDJyckKCAhwdTkAYH5X07L0sNgxcPfKBceP4e5t351G1997+UsWi/M/KwqNvd/fxbpnBwBQQmSk5/5QupwuJaUV4Plmbp6OhRfv0oQXSCLsAAByknH12i3Qufa23PD+crLjx3DzyOVy0Y0Dd8tfu9vIJ5DwggIh7ADA34Ez5jfKj8XN/ofU+ZeXfIIILygShB0AKIlcMb9Rns98qXAtvNzwuBCgOCDsAEBxwPxGQKEh7ABAYSiq+Y18ghwILyV7fiOgoPitBwB7FLv5jf68rPQ3mt8IKCjCDoC/J+Y3Av42CDsAzMNV8xvl1RPD/EaAyxF2ABRfN85vlF94YX4jADkg7AAoOkU5v5Fdl46Y3wj4OyDsACi4q1fsf8Iu8xsBcBHCDoC/ODS/0W9SWkGmCGB+IwBFi7ADmFlRz2+U3xN2md8IgAsQdoCSJN/5jW4IL8xvBACEHcClimJ+I1n+7HHJ0rvC/EYA/kYIO4AzZZ3fyJ5LR8xvBACFjrAD5IX5jQCgxONfTPy9uGp+I5tLRzdcSmJ+IwAoVIQdlGzMbwQAyAdhB8VPUc1v5Eh48fRx/ucEABQJwg4KX3GZ3yhruGF+IwD42yDswHHMbwQAKEEIOyii+Y28svSwXH9g3Y1hJst75jcCADgJYceMmN8IAAArwk5JUGTzG5Wz7wm7zG8EAChBCDuukNv8RrmFl0Kb3yhreAliigAAgCkRdpyB+Y0AACi2CDs5ccX8Rnk9Yde/AvMbAQBQQISdrBa1k4zz10IM8xsBAGAKfNtm9ev3kneWQbfegX8Flbxuk2Z+IwAAii3CTlbd35FCqv3VC8P8RgAAlHiEnaxqtpYCAlxdBQAAcCJu1wEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKZG2AEAAKbm4eoCAABA4TAMQ+kZhq5kZCr9aqbSMzKV9uef19quLbtyvS3Lsms/G7pyNcO6j5zXM3LZ9q9j/O/JFirl7brIQdgBAKCAroeJ3IOC7Zf/9fWyrpO9Let6Rh7b/nWM7CEm0xpQioO09AzCDgAAOckaJnILCmk39Db81Wbk0GbbO3HlxqDy5/7Tbwgs2XosilmYsJe7m0We7hZ5urvJ28NNnu5u8vrzz+s/e7lbrG1e7m7y9Lj257WfLfJyd//zT9vlnnlsW8rHtXGDsAMAf2OGYehqZk5f6Nl7DLKGgrQbeiVyXs/IY9sbgkQuPRYlLUy4WWT9sve+IURkDRPZ29zkmTUo2ASMnIOIzXrWNkuOIeZ6m7ubxdWnyCUIOwBQiK6HCWdfvsjas5B9HEWWcRb59EpcuVpyw4TXDYHhWpslh7brP1sc39YaRP7aR04h5u8eJoo7wg6AEi1rmEi/aigtIyPHHoP0Gy5PXMkwcmi7PmAzw6ZnIadtr4eJ/Hol0jMyZRiuPkv2s1hkDQI3fsl7ebjn07NgyaW3wf5eidy3vfYnYQIFQdgBkCfDMJSRaViDwPUwkZ7lC952QGSGrly1vfvjikNhwsglYNj2SqRfzVRaSQ4T2XoWch9H4X1jr0S2IJLf5RHbno2ct7223MOdJ5LAfAg7gIvdGCauZNgGgOyXHjLyuNUzr8sWed8mej1MpGcJG2YJE9cHU+Y2jsL7hl6JnNfL+/JFXoEla88HYQIoeoQd/C1czeHSQl6DJQt6d0fWAZk5rvdnmLC59FHCwoR0bcyEt81dGBab3oLsPQeWPHoT8ut1+PPuj5x6NrIM4ry+D3c3iywWLnUA+AthB05hDRO59ibY0+tg5LFtwe/uSM/IVGYJDhO5DqrM+mXvYcmh7a/gYL1N1Kbtr22vh4ls217v0chyqylhAkBJQ9gpITIyc+6V+Kstr1s4HXs4Ve7b5j4gsySGiXzv0LAGglyeOXF926zPnHAgiNhcHvlzH57ubvIgTACAU9kVdho0aODQTi0Wi9auXatKlSoVqChXuZyeIeOP9HwvX9xMr0TugSXvB1uV1DCR43MjbO68KPq7OwgTAPD3YlfYiY+P19NPP61SpUrlu65hGJoxY4bS0tJuurii1uj5TXLz9nN1GXax77kRuQyWzCOIOOPuDsIEAKA4sfsy1jPPPKOKFSvate4rr7xS4IKKi/x7Fv4KCs64uyOnJ13+dZnENmwQJgAAsJ9dYef48eOqUKGC3Ts9ePCgwsLCClxUVj/99JPGjBmjdevW6dKlS6pVq5ZiY2PVqFEjSdd6kiZPnqxFixbp/PnzatmypebPn6/w8HCHj7Xz2bYqVyZInu6ECQAAzMKuBz5UrVrVoS//ypUry93dvcBFXff777+rZcuW8vT01Lp163Tw4EG98sorKlOmjHWdmTNnas6cOVqwYIF27twpf39/RUVF6fLlyw4fz9/bQ14ebgQdAABMxGIYBXvCx9WrV/Xvf/9bmzdvVkZGhlq2bKmYmBj5+Pg4rbixY8dq27Zt+uKLL3JcbhiGwsLC9PTTT2vUqFGSpOTkZAUHB2vp0qXq2bOnXcdJSUlRYGCgkpOTFRAQ4LT6AQBA4bH3+7vAj/IcNmyYVq9erXvuuUetWrXS8uXLNWDAgILuLkdr165Vo0aN9NBDD6lixYq64447tGjRIuvy48ePKzExUe3atbO2BQYGqmnTptq+fXuu+01LS1NKSorNCwAAmJPdA5RXr16trl27Wt9/+umnOnTokPVyVVRUlJo1a+bU4o4dO6b58+dr5MiRevbZZ7V7924NGzZMXl5eio6OVmJioiQpODjYZrvg4GDrspxMnz5dU6dOdWqtAACgeLK7Z2fJkiXq0qWLfv75Z0nXnr3z+OOPa/369frwww81evRoNW7c2KnFZWZmqkGDBnrxxRd1xx13aPDgwRo0aJAWLFhwU/sdN26ckpOTra9Tp045qWIAAFDc2B12PvzwQ/Xq1UutW7fW66+/roULFyogIEDjx4/XxIkTVblyZS1fvtypxYWGhuq2226zaYuIiNDJkyclSSEhIZKkpKQkm3WSkpKsy3Li7e2tgIAAmxcAADAnh8bs9OjRQ7t27dKBAwcUFRWlPn36aM+ePYqPj9e8efMcuj3dHi1bttShQ4ds2n744QdVrVpVklS9enWFhIQoLi7OujwlJUU7d+5U8+bNnVoLAAAomRweoBwUFKSFCxfq5ZdfVr9+/fTMM88U6DZve4wYMUI7duzQiy++qCNHjmj58uVauHChYmJiJF2blmL48OF6/vnntXbtWh04cED9+vVTWFiYunTpUig1AQCAksXusHPy5El1795dkZGR6t27t8LDw7Vnzx75+fmpXr16WrdundOLa9y4sVavXq0VK1bo9ttv17Rp0zR79mz17t3bus7o0aM1dOhQDR48WI0bN1ZqaqrWr1/v1FvgAQBAyWX3c3Zat26tkJAQ9e/fXxs2bNDRo0e1du1aSVJCQoIee+wxhYSE6L333ivUggsDz9kBAKDksff72+5bz7/++mvt379fNWvWVFRUlKpXr25dFhERoa1bt2rhwoU3VzUAAICT2R12GjZsqEmTJik6OlqbNm1SZGRktnUGDx7s1OIAAABult1jdt566y2lpaVpxIgR+umnn/Tvf/+7MOsCAABwCrt7dqpWrar//ve/hVkLAACA09nVs+Po3FEXLlwoUDEAAADOZlfYKVOmjM6ePWv3TitVqqRjx44VuCgAAABnsesylmEY+s9//qNSpUrZtdP09PSbKgoAAMBZ7Ao7VapU0aJFi+zeaUhIiDw9PQtcFAAAgLPYFXZOnDhRyGUAAAAUDofnxgIAAChJCDsAAMDUCDsAAMDUCDsAAMDUCDsAAMDUHA471apV03PPPaeTJ08WRj0AAABO5XDYGT58uN5//33VqFFD7du318qVK5WWllYYtQEAANy0AoWd+Ph47dq1SxERERo6dKhCQ0M1ZMgQ7d27tzBqBAAAKDCLYRjGzewgPT1db7zxhsaMGaP09HRFRkZq2LBhGjBggCwWi7PqLFQpKSkKDAxUcnKyAgICXF0OAACwg73f33Y9QTkn6enpWr16tWJjY7Vx40Y1a9ZMAwcO1OnTp/Xss89q06ZNWr58eUF3DwAA4BQOh529e/cqNjZWK1askJubm/r166fXXntNtWvXtq7TtWtXNW7c2KmFAgAAFITDYadx48Zq37695s+fry5duuQ44Wf16tXVs2dPpxQIAABwMxwOO8eOHVPVqlXzXMff31+xsbEFLgoAAMBZHL4b6+zZs9q5c2e29p07d+rrr792SlEAAADO4nDYiYmJ0alTp7K1//TTT4qJiXFKUQAAAM7icNg5ePCgGjRokK39jjvu0MGDB51SFAAAgLM4HHa8vb2VlJSUrf3MmTPy8CjwnewAAACFwuGw849//EPjxo1TcnKyte38+fN69tln1b59e6cWBwAAcLMc7oqZNWuW7r77blWtWlV33HGHJCk+Pl7BwcF6++23nV4gAADAzXA47FSqVEnffPONli1bpv3798vX11cDBgxQr169cnzmDgAAgCsVaJCNv7+/Bg8e7OxaAAAAnK7AI4oPHjyokydP6sqVKzbt//znP2+6KAAAAGcp0BOUu3btqgMHDshisej6pOnXZzjPyMhwboUAAAA3weG7sZ566ilVr15dZ8+elZ+fn7777jtt3bpVjRo10ubNmwuhRAAAgIJzuGdn+/bt+uyzz1S+fHm5ubnJzc1Nd955p6ZPn65hw4Zp3759hVEnAABAgTjcs5ORkaHSpUtLksqXL6+ff/5ZklS1alUdOnTIudUBAADcJId7dm6//Xbt379f1atXV9OmTTVz5kx5eXlp4cKFqlGjRmHUCAAAUGAOh50JEybo4sWLkqTnnntO9913n+666y6VK1dO7777rtMLBAAAuBkW4/rtVDfh3LlzKlOmjPWOrJImJSVFgYGBSk5OVkBAgKvLAQAAdrD3+9uhMTvp6eny8PDQt99+a9NetmzZEht0AACAuTkUdjw9PVWlShWepQMAAEoMh+/GGj9+vJ599lmdO3euMOoBAABwKocHKM+dO1dHjhxRWFiYqlatKn9/f5vle/fudVpxAAAAN8vhsNOlS5dCKAMAAKBwOOVurJKOu7EAACh5CuVuLAAAgJLG4ctYbm5ued5mzp1aAACgOHE47KxevdrmfXp6uvbt26c333xTU6dOdVphAAAAzuC0MTvLly/Xu+++qw8++MAZuytSjNkBAKDkKfIxO82aNVNcXJyzdgcAAOAUTgk7f/zxh+bMmaNKlSo5Y3cAAABO4/CYnRsn/DQMQxcuXJCfn5/eeecdpxYHAABwsxwOO6+99ppN2HFzc1OFChXUtGlTlSlTxqnFAQAA3CyHw07//v0LoQwAAIDC4fCYndjYWK1atSpb+6pVq/Tmm286pSgAAABncTjsTJ8+XeXLl8/WXrFiRb344otOKQoAAMBZHA47J0+eVPXq1bO1V61aVSdPnnRKUQAAAM7icNipWLGivvnmm2zt+/fvV7ly5ZxSFAAAgLM4HHZ69eqlYcOG6fPPP1dGRoYyMjL02Wef6amnnlLPnj0Lo0YAAIACc/hurGnTpunEiRNq27atPDyubZ6Zmal+/foxZgcAABQ7BZ4b6/Dhw4qPj5evr68iIyNVtWpVZ9dWZJgbCwCAksfe72+He3auCw8PV3h4eEE3BwAAKBIOj9np1q2bXnrppWztM2fO1EMPPeSUogAAAJzF4bCzdetWdezYMVv7vffeq61btzqlKAAAAGdxOOykpqbKy8srW7unp6dSUlKcUlRuZsyYIYvFouHDh1vbLl++rJiYGJUrV06lSpVSt27dlJSUVKh1AACAksPhsBMZGal33303W/vKlSt12223OaWonOzevVv//ve/VbduXZv2ESNG6MMPP9SqVau0ZcsW/fzzz3rggQcKrQ4AAFCyODxAeeLEiXrggQd09OhRtWnTRpIUFxenFStW5DhnljOkpqaqd+/eWrRokZ5//nlre3JyshYvXqzly5dba4mNjVVERIR27NihZs2aFUo9AACg5HC4Z6dz585as2aNjhw5oieffFJPP/20Tp8+rU2bNqlLly6FUKIUExOjTp06qV27djbte/bsUXp6uk177dq1VaVKFW3fvj3X/aWlpSklJcXmBQAAzKlAt5536tRJnTp1ytb+7bff6vbbb7/porJauXKl9u7dq927d2dblpiYKC8vLwUFBdm0BwcHKzExMdd9Tp8+XVOnTnVqnQAAoHhyuGfnRhcuXNDChQvVpEkT1atXzxk1WZ06dUpPPfWUli1bJh8fH6ftd9y4cUpOTra+Tp065bR9AwCA4qXAYWfr1q3q16+fQkNDNWvWLLVp00Y7duxwZm3as2ePzp49qwYNGsjDw0MeHh7asmWL5syZIw8PDwUHB+vKlSs6f/68zXZJSUkKCQnJdb/e3t4KCAiweQEAAHNy6DJWYmKili5dqsWLFyslJUXdu3dXWlqa1qxZUyh3YrVt21YHDhywaRswYIBq166tMWPGqHLlyvL09FRcXJy6desmSTp06JBOnjyp5s2bO70eAABQ8tgddjp37qytW7eqU6dOmj17tjp06CB3d3ctWLCg0IorXbp0tjFA/v7+KleunLV94MCBGjlypMqWLauAgAANHTpUzZs3504sAAAgyYGws27dOg0bNkxPPPFEsZoT67XXXpObm5u6deumtLQ0RUVF6Y033nB1WQAAoJiwe9bzHTt2aPHixXr33XcVERGhvn37qmfPngoNDdX+/fsL9YGChY1ZzwEAKHns/f62e4Bys2bNtGjRIp05c0aPPfaYVq5cqbCwMGVmZmrjxo26cOGCUwoHAABwJrt7dnJy6NAhLV68WG+//bbOnz+v9u3ba+3atc6sr0jQswMAQMnj9J6dnNx6662aOXOmTp8+rRUrVtzMrgAAAArFTfXsmAU9OwAAlDxF0rMDAABQ3BF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRF2AACAqRXrsDN9+nQ1btxYpUuXVsWKFdWlSxcdOnTIZp3Lly8rJiZG5cqVU6lSpdStWzclJSW5qGIAAFDcFOuws2XLFsXExGjHjh3auHGj0tPT9Y9//EMXL160rjNixAh9+OGHWrVqlbZs2aKff/5ZDzzwgAurBgAAxYnFMAzD1UXY65dfflHFihW1ZcsW3X333UpOTlaFChW0fPlyPfjgg5Kk77//XhEREdq+fbuaNWtm135TUlIUGBio5ORkBQQEFOZHAAAATmLv93ex7tm5UXJysiSpbNmykqQ9e/YoPT1d7dq1s65Tu3ZtValSRdu3b891P2lpaUpJSbF5AQAAcyoxYSczM1PDhw9Xy5Ytdfvtt0uSEhMT5eXlpaCgIJt1g4ODlZiYmOu+pk+frsDAQOurcuXKhVk6AABwoRITdmJiYvTtt99q5cqVN72vcePGKTk52fo6deqUEyoEAADFkYerC7DHkCFD9NFHH2nr1q36v//7P2t7SEiIrly5ovPnz9v07iQlJSkkJCTX/Xl7e8vb27swSwYAAMVEse7ZMQxDQ4YM0erVq/XZZ5+pevXqNssbNmwoT09PxcXFWdsOHTqkkydPqnnz5kVdLgAAKIaKdc9OTEyMli9frg8++EClS5e2jsMJDAyUr6+vAgMDNXDgQI0cOVJly5ZVQECAhg4dqubNm9t9JxYAADC3Yn3rucViybE9NjZW/fv3l3TtoYJPP/20VqxYobS0NEVFRemNN97I8zLWjbj1HACAksfe7+9iHXaKCmEHAICSx5TP2QEAAHAUYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJgaYQcAAJiaacLOvHnzVK1aNfn4+Khp06batWuXq0sCAADFgCnCzrvvvquRI0dq8uTJ2rt3r+rVq6eoqCidPXvW1aUBAAAXM0XYefXVVzVo0CANGDBAt912mxYsWCA/Pz8tWbLE1aUBAAAX83B1ATfrypUr2rNnj8aNG2dtc3NzU7t27bR9+/Yct0lLS1NaWpr1fXJysiQpJSWlcIsFAABOc/172zCMPNcr8WHn119/VUZGhoKDg23ag4OD9f333+e4zfTp0zV16tRs7ZUrVy6UGgEAQOG5cOGCAgMDc11e4sNOQYwbN04jR460vj9//ryqVq2qkydP5nmy4FwpKSmqXLmyTp06pYCAAFeX87fCuXcNzrvrcO5do7DPu2EYunDhgsLCwvJcr8SHnfLly8vd3V1JSUk27UlJSQoJCclxG29vb3l7e2drDwwM5D8CFwgICOC8uwjn3jU4767DuXeNwjzv9nRSlPgByl5eXmrYsKHi4uKsbZmZmYqLi1Pz5s1dWBkAACgOSnzPjiSNHDlS0dHRatSokZo0aaLZs2fr4sWLGjBggKtLAwAALmaKsNOjRw/98ssvmjRpkhITE1W/fn2tX78+26Dl3Hh7e2vy5Mk5XtpC4eG8uw7n3jU4767DuXeN4nLeLUZ+92sBAACUYCV+zA4AAEBeCDsAAMDUCDsAAMDUCDsAAMDU/jZhZ968eapWrZp8fHzUtGlT7dq1K8/1V61apdq1a8vHx0eRkZH65JNPiqhSc3HkvC9atEh33XWXypQpozJlyqhdu3b5/j0hd47+zl+3cuVKWSwWdenSpXALNClHz/v58+cVExOj0NBQeXt765ZbbuHfmwJw9LzPnj1bt956q3x9fVW5cmWNGDFCly9fLqJqzWPr1q3q3LmzwsLCZLFYtGbNmny32bx5sxo0aCBvb2/VqlVLS5cuLfQ6ZfwNrFy50vDy8jKWLFlifPfdd8agQYOMoKAgIykpKcf1t23bZri7uxszZ840Dh48aEyYMMHw9PQ0Dhw4UMSVl2yOnveHH37YmDdvnrFv3z4jISHB6N+/vxEYGGicPn26iCsv+Rw999cdP37cqFSpknHXXXcZ999/f9EUayKOnve0tDSjUaNGRseOHY0vv/zSOH78uLF582YjPj6+iCsv2Rw978uWLTO8vb2NZcuWGcePHzc2bNhghIaGGiNGjCjiyku+Tz75xBg/frzx/vvvG5KM1atX57n+sWPHDD8/P2PkyJHGwYMHjddff91wd3c31q9fX6h1/i3CTpMmTYyYmBjr+4yMDCMsLMyYPn16jut3797d6NSpk01b06ZNjccee6xQ6zQbR8/7ja5evWqULl3aePPNNwurRNMqyLm/evWq0aJFC+M///mPER0dTdgpAEfP+/z5840aNWoYV65cKaoSTcnR8x4TE2O0adPGpm3kyJFGy5YtC7VOs7Mn7IwePdqoU6eOTVuPHj2MqKioQqzMMEx/GevKlSvas2eP2rVrZ21zc3NTu3bttH379hy32b59u836khQVFZXr+siuIOf9RpcuXVJ6errKli1bWGWaUkHP/XPPPaeKFStq4MCBRVGm6RTkvK9du1bNmzdXTEyMgoODdfvtt+vFF19URkZGUZVd4hXkvLdo0UJ79uyxXuo6duyYPvnkE3Xs2LFIav47c9X3qymeoJyXX3/9VRkZGdmephwcHKzvv/8+x20SExNzXD8xMbHQ6jSbgpz3G40ZM0ZhYWHZ/sNA3gpy7r/88kstXrxY8fHxRVChORXkvB87dkyfffaZevfurU8++URHjhzRk08+qfT0dE2ePLkoyi7xCnLeH374Yf3666+68847ZRiGrl69qscff1zPPvtsUZT8t5bb92tKSor++OMP+fr6FspxTd+zg5JpxowZWrlypVavXi0fHx9Xl2NqFy5cUN++fbVo0SKVL1/e1eX8rWRmZqpixYpauHChGjZsqB49emj8+PFasGCBq0sztc2bN+vFF1/UG2+8ob179+r999/Xxx9/rGnTprm6NBQS0/fslC9fXu7u7kpKSrJpT0pKUkhISI7bhISEOLQ+sivIeb9u1qxZmjFjhjZt2qS6desWZpmm5Oi5P3r0qE6cOKHOnTtb2zIzMyVJHh4eOnTokGrWrFm4RZtAQX7nQ0ND5enpKXd3d2tbRESEEhMTdeXKFXl5eRVqzWZQkPM+ceJE9e3bV48++qgkKTIyUhcvXtTgwYM1fvx4ubnRD1BYcvt+DQgIKLReHelv0LPj5eWlhg0bKi4uztqWmZmpuLg4NW/ePMdtmjdvbrO+JG3cuDHX9ZFdQc67JM2cOVPTpk3T+vXr1ahRo6Io1XQcPfe1a9fWgQMHFB8fb33985//1D333KP4+HhVrly5KMsvsQryO9+yZUsdOXLEGi4l6YcfflBoaChBx04FOe+XLl3KFmiuB06D6SILlcu+Xwt1+HMxsXLlSsPb29tYunSpcfDgQWPw4MFGUFCQkZiYaBiGYfTt29cYO3asdf1t27YZHh4exqxZs4yEhARj8uTJ3HpeAI6e9xkzZhheXl7Gf//7X+PMmTPW14ULF1z1EUosR8/9jbgbq2AcPe8nT540SpcubQwZMsQ4dOiQ8dFHHxkVK1Y0nn/+eVd9hBLJ0fM+efJko3Tp0saKFSuMY8eOGZ9++qlRs2ZNo3v37q76CCXWhQsXjH379hn79u0zJBmvvvqqsW/fPuPHH380DMMwxo4da/Tt29e6/vVbz5955hkjISHBmDdvHreeO9Prr79uVKlSxfDy8jKaNGli7Nixw7qsVatWRnR0tM367733nnHLLbcYXl5eRp06dYyPP/64iCs2B0fOe9WqVQ1J2V6TJ08u+sJNwNHf+awIOwXn6Hn/6quvjKZNmxre3t5GjRo1jBdeeMG4evVqEVdd8jly3tPT040pU6YYNWvWNHx8fIzKlSsbTz75pPH7778XfeEl3Oeff57jv9vXz3d0dLTRqlWrbNvUr1/f8PLyMmrUqGHExsYWep0Ww6DPDgAAmJfpx+wAAIC/N8IOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAEiyWCxas2aNq8sAUAgIOwBcrn///rJYLNleHTp0cHVpAEzA9LOeAygZOnTooNjYWJs2b29vF1UDwEzo2QFQLHh7eyskJMTmVaZMGUnXLjHNnz9f9957r3x9fVWjRg3997//tdn+wIEDatOmjXx9fVWuXDkNHjxYqampNussWbJEderUkbe3t0JDQzVkyBCb5b/++qu6du0qPz8/hYeHa+3atdZlv//+u3r37q0KFSrI19dX4eHh2cIZgOKJsAOgRJg4caK6deum/fv3q3fv3urZs6cSEhIkSRcvXlRUVJTKlCmj3bt3a9WqVdq0aZNNmJk/f75iYmI0ePBgHThwQGvXrlWtWrVsjjF16lR1795d33zzjTp27KjevXvr3Llz1uMfPHhQ69atU0JCgubPn6/y5csX3QkAUHCFPtUoAOQjOjracHd3N/z9/W1eL7zwgmEYhiHJePzxx222adq0qfHEE08YhmEYCxcuNMqUKWOkpqZal3/88ceGm5ubkZiYaBiGYYSFhRnjx4/PtQZJxoQJE6zvU1NTDUnGunXrDMMwjM6dOxsDBgxwzgcGUKQYswOgWLjnnns0f/58m7ayZctaf27evLnNsubNmys+Pl6SlJCQoHr16snf39+6vGXLlsrMzNShQ4dksVj0888/q23btnnWULduXevP/v7+CggI0NmzZyVJTzzxhLp166a9e/fqH//4h7p06aIWLVoU6LMCKFqEHQDFgr+/f7bLSs7i6+tr13qenp427y0WizIzMyVJ9957r3788Ud98skn2rhxo9q2bauYmBjNmjXL6fUCcC7G7AAoEXbs2JHtfUREhCQpIiJC+/fv18WLF63Lt23bJjc3N916660qXbq0qlWrpri4uJuqoUKFCoqOjtY777yj2bNna+HChTe1PwBFg54dAMVCWlqaEhMTbdo8PDysg4BXrVqlRo0a6c4779SyZcu0a9cuLV68WJLUu3dvTZ48WdHR0ZoyZYp++eUXDR06VH379lVwcLAkacqUKXr88cdVsWJF3Xvvvbpw4YK2bdumoUOH2lXfpEmT1LBhQ9WpU0dpaWn66KOPrGELQPFG2AFQLKxfv16hoaE2bbfeequ+//57SdfulFq5cqWefPJJhYaGasWKFbrtttskSX5+ftqwYYOeeuopNW7cWH5+furWrZteffVV676io6N1+fJlvfbaaxo1apTKly+vBx980O76vLy8NG7cOJ04cUK+vr666667tHLlSid8cgCFzWIYhuHqIgAgLxaLRatXr1aXLl1cXQqAEogxOwAAwNQIOwAAwNQYswOg2ONqO4CbQc8OAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwNcIOAAAwtf8H+ME+tfNCIg0AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.array(history.history['val_accuracy'])*100, label=\"Validation Accuracy\")\n",
    "plt.plot(np.array(history.history['accuracy'])*100, label=\"Training Accuracy\")\n",
    "plt.ylabel(\"Accuracy [%]\")\n",
    "plt.xlabel(\"Epochs\")\n",
    "plt.ylim(0,100)\n",
    "plt.xlim(0,)\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "437/437 [==============================] - 33s 74ms/step\n",
      "\n",
      " Summary of the precision, recall, F1 score for each class:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.34      0.61      0.44       789\n",
      "           1       0.34      0.09      0.15       798\n",
      "           2       0.94      0.50      0.65       527\n",
      "           3       0.61      0.95      0.74      1019\n",
      "           4       0.85      0.99      0.91      1378\n",
      "           5       0.99      1.00      0.99      1667\n",
      "           6       0.66      0.57      0.61      1375\n",
      "           7       0.40      0.40      0.40       922\n",
      "           8       0.54      0.53      0.54      1722\n",
      "           9       0.65      0.68      0.67      1344\n",
      "          10       0.65      0.57      0.61      1447\n",
      "          11       0.54      0.40      0.46       986\n",
      "\n",
      "    accuracy                           0.65     13974\n",
      "   macro avg       0.63      0.61      0.60     13974\n",
      "weighted avg       0.65      0.65      0.63     13974\n",
      "\n",
      "\n",
      " Confusion matrix: \n",
      "[[ 482   69    0    1   81    2  146    3    0    3    0    2]\n",
      " [ 478   74    0    0   68    0  175    0    1    0    0    2]\n",
      " [   0    1  263  194    1    3    4   52    6    0    0    3]\n",
      " [   0    0    3  967    1    0    1   11   15   17    2    2]\n",
      " [   3    0    0    2 1363    2    8    0    0    0    0    0]\n",
      " [   1    0    0    2    1 1659    3    1    0    0    0    0]\n",
      " [ 432   70    0    0   85    0  784    0    0    1    1    2]\n",
      " [   0    0    9  114    2    0    6  370  168  196   19   38]\n",
      " [   0    1    0   18    0    0   21  167  920  211  280  104]\n",
      " [   0    0    0    7    1    0    7  219  157  918    7   28]\n",
      " [   0    0    1  114    0    3    6   28  298   13  821  163]\n",
      " [   2    0    3  171    3    2   18   74  131   50  136  396]]\n"
     ]
    }
   ],
   "source": [
    "# use model for inference to get test accuracy\n",
    "y_val_pred = model.predict(x_val)\n",
    "y_val_pred = np.argmax(y_val_pred, axis=1)\n",
    "y_val = np.argmax(y_val_one_hot, axis=1)\n",
    "print ('\\n Summary of the precision, recall, F1 score for each class:')\n",
    "print (sklearn.metrics.classification_report(y_val, y_val_pred))\n",
    "\n",
    "print ('\\n Confusion matrix: ')\n",
    "print (sklearn.metrics.confusion_matrix(y_val, y_val_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "#converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "#tflite_model = converter.convert()\n",
    "\n",
    "#with open('model.tflite', 'wb') as f:\n",
    "#    f.write(tflite_model)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}