fer2013_kaggle.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# Facial Emotion Recognition","metadata":{}},{"cell_type":"markdown","source":"## Import libraries","metadata":{}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport os\nimport tensorflow as tf\nfrom keras.preprocessing.image import ImageDataGenerator, load_img\nfrom keras.layers import Conv2D, Dense, BatchNormalization, Activation, Dropout, MaxPooling2D, Flatten\nfrom keras.optimizers import Adam, RMSprop, SGD\nfrom keras import regularizers\nfrom keras.callbacks import ModelCheckpoint, CSVLogger, TensorBoard, EarlyStopping, ReduceLROnPlateau\nimport datetime\nimport matplotlib.pyplot as plt\nfrom keras.utils import plot_model","metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","execution":{"iopub.status.busy":"2022-07-31T09:12:37.225230Z","iopub.execute_input":"2022-07-31T09:12:37.225493Z","iopub.status.idle":"2022-07-31T09:12:42.077436Z","shell.execute_reply.started":"2022-07-31T09:12:37.225467Z","shell.execute_reply":"2022-07-31T09:12:42.076538Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"code","source":"train_dir = '../input/fer2013/train/'\ntest_dir = '../input/fer2013/test/'\n\nimg_l, img_w = 48, 48\nnum_classes = 7\n\ndef count_exp(path, set_):\n    dict_ = {}\n    for x in os.listdir(path):\n        dir_ = path + x\n        dict_[x] = len(os.listdir(dir_))\n    df = pd.DataFrame(dict_, index=[set_])\n    return df\ntrain_count = count_exp(train_dir, 'train')\ntest_count = count_exp(test_dir, 'test')\nprint(train_count)\nprint(test_count)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:12:42.079108Z","iopub.execute_input":"2022-07-31T09:12:42.079461Z","iopub.status.idle":"2022-07-31T09:12:47.061907Z","shell.execute_reply.started":"2022-07-31T09:12:42.079425Z","shell.execute_reply":"2022-07-31T09:12:47.061101Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"       surprise  fear  angry  neutral   sad  disgust  happy\ntrain      3171  4097   3995     4965  4830      436   7215\n      surprise  fear  angry  neutral   sad  disgust  happy\ntest       831  1024    958     1233  1247      111   1774\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Graphical representation of training data","metadata":{}},{"cell_type":"code","source":"train_count.transpose().plot(kind='bar', figsize=(12,10))","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:12:50.403447Z","iopub.execute_input":"2022-07-31T09:12:50.403788Z","iopub.status.idle":"2022-07-31T09:12:50.617827Z","shell.execute_reply.started":"2022-07-31T09:12:50.403754Z","shell.execute_reply":"2022-07-31T09:12:50.617093Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"<matplotlib.axes._subplots.AxesSubplot at 0x7fe57a158610>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 864x720 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"### Graphical representation of testing data ","metadata":{}},{"cell_type":"code","source":"test_count.transpose().plot(kind='bar',figsize=(12, 10))","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:12:53.852159Z","iopub.execute_input":"2022-07-31T09:12:53.852511Z","iopub.status.idle":"2022-07-31T09:12:54.015359Z","shell.execute_reply.started":"2022-07-31T09:12:53.852476Z","shell.execute_reply":"2022-07-31T09:12:54.014497Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":"<matplotlib.axes._subplots.AxesSubplot at 0x7fe57a35ad10>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 864x720 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"### Display sample images","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(14,22))\ni = 1\nfor x in os.listdir(train_dir):\n    img = load_img((train_dir + x +'/'+ os.listdir(train_dir + x)[1]))\n    plt.subplot(1,7,i)\n    plt.imshow(img)\n    plt.title(x)\n    plt.axis('off')\n    i += 1\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:12:56.279807Z","iopub.execute_input":"2022-07-31T09:12:56.280137Z","iopub.status.idle":"2022-07-31T09:12:56.698396Z","shell.execute_reply.started":"2022-07-31T09:12:56.280105Z","shell.execute_reply":"2022-07-31T09:12:56.697472Z"},"trusted":true},"execution_count":6,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x1584 with 7 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAxsAAAB9CAYAAADKm34TAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOy9eZRsWVbe952IzMgYcnpj1Xs1Vw9FUdVT9Wqg1RZYAhqBwJKxLQsEWsLWgGwZS0ZLIFkD0hLIluVlWauFhbEtEIMAIauRGKwW0DS06FFdDfRQ3V3VXXO9evWmnCIjMjLi+o/I34kvdt7I97orXr2qenevlSsybtx77rnnnr3P3t8eTiqKQhVVVFFFFVVUUUUVVVRRRfOm2o3uQEUVVVRRRRVVVFFFFVX06qTK2KiooooqqqiiiiqqqKKKrgtVxkZFFVVUUUUVVVRRRRVVdF2oMjYqqqiiiiqqqKKKKqqooutClbFRUUUVVVRRRRVVVFFFFV0XqoyNiiqqqKKKKqqooooqqui60E1rbKSUfn9K6TM3uh83K6WU7kspPZxS2kopfc+N7k9FFVX00lNKqUgpvfZG96Oickop/UZK6U/f6H682iml9GMppb9b6SWvXEopPZ5S+rob3Y+XK920xkZRFL9VFMV9N7ofNzH9FUm/URTFSlEU/+hGd6aiiiq6dkop/ccppadvdD8qqujVRDdaL6kU5oquF73ijY2U0sJLcU1Fc6e7JH1yng2mMb3i5/TNThV/vjqoeo8VVVRRRRVJLwNjI6X0fSmlZw7CaT6TUvpaXIp2zhSKdmB9f19K6Xcl7aSUFg6O/dWU0qdSSpdTSv80pdT06w+uOSfpn5a0eagfB8drKaXvTyk9llK6mFL6uZTS8ZduhF59lFL6dUl/QNK7UkrbByFV/yCl9GRK6fmU0j9JKbUOzj2WUvrFlNILB+/1F1NKt1tbv5FS+sGU0r+X1JV07415qpuDjBe2DnjtPz04/qdSSu8/eI+XU0pfSCl9o113T0rpNw+u+9WU0j9OKf3kwW93H4TT/NcppScl/XpK6ZdSSv9duPfvppT+6Ev6wK8COpCNf/lg/DZSSj9rsvGbU0ofTyldSSn9dkrpjXbdVIiThXp0JP2KpLMH/LudUjqbUvqBlNLPp5R+MqW0KelPpZS+IqX0gYP2n0spvSul1HjJB+EmpBlr65HvI6X09SmlRw7mybskpRv4CK9aSim9JaX0sYN387OSpnQVO2+WXtJKKf34gaz9dErpr4TrSnn34P+TB+volZTSpZTSbx3oOT8h6U5J/+aAp//KSzQcryZ6c5Sz16jD/L2U0ocPrvuFdKBj2tr4Z1NKzx7w7Pce/HZrSqmbUjphbb314D6LL/2jH0031NhIKd0n6S9IeltRFCuSvkHS49d4+bdJ+sOS1oui2D849icO2niNpNdL+ut2/q2SjmuMqP/ZL6If3yPpj0r6GklnJV2W9I+v9RkrOkxFUfxBSb8l6S8URbEs6c9r/L7eLOm1km6T9DcPTq9J+qcav7c7Je1Keldo8js1fqcrkp643v2/yekxSb9f0pqkvy3pJ1NKZw5++0pJn5F0UtLfl/R/p5RQVn5a0oclnZD0Axq/s0hfI+l+jfnvxyV9Bz+klN6k8bz45fk+zk1Df0zSH5J0j6Q3amwIPCTp/5H05zR+Lz8i6V+nlJaOaqgoih1J3yjp2aIolg/+nj34+Y9I+nlJ65J+StJQ0l/SeE68XdLXSvpv5vxsFQU6Yk2b+T5SSicl/UuN182TGvP6O17qvr/a6cC4e7ekn9BYJ/kXkv6zkvOO0kv+lqS7NQbXvl4mK6+BvlfS05JOSbpF0l+TVBRF8Z2SnpT0LQc8/fe/2Ger6LCc1bXpMH9S0n+lsY65LymGlv8BSa+T9E5J359S+rqiKM5J+o2De0LfIelniqIYzO2J5kQ32rMxlLQk6ctTSotFUTxeFMVj13jtPyqK4qmiKHbt2LsOjl2S9IMaGyTQSNLfKoqiH665Wj/+nKT/sSiKp4ui6GusKP3nqQoRmAsdKKN/RtJfKoriUlEUW5J+SNIfl6SiKC4WRfEvi6LoHvz2gxorpU4/VhTFJ4ui2H85MtmriYqi+BdFUTxbFMWoKIqflfQ5SV9x8PMTRVH8aFEUQ42NhTOSbkkp3SnpbZL+ZlEUe0VRvF/Svy5p/geKotg54M9fkPS6lNLrDn77Tkk/WxTF3vV8vlcx/aOD93ZJ0r/R2LD/M5J+pCiKDxVFMSyK4scl9SV91Yu4zweKonj3wfzYLYriPxRF8cED3nxcY4Mm8m9F86fSNe0q7+ObJH2qKIqfP5Cj/1DSuRvS+1c3fZWkRUn/sCiKQVEUPy/pIyXnHaWX/DFJP1QUxeWiKJ7WYeX0KBpoLJvvOrj/bxVFUXzpj1OR0SE5e406zE8URfGJAyDnb0j6Yymluv3+tw/Wxt/T2HBBt82g3MH536axEfuyoxtqbBRF8aikv6ixAn8+pfQzKaWz13j5U1c59oTGViL0QlEUvS+hH3dJ+lcHLscrkj6tsRC45Rr7WdHRdEpSW9J/sDH+/w6OK6XUTin9SErpiYPQjN+UtB4YsWwuVHQdKKX0J9Mk7OaKpAc1RkElU0yKouge/LusMR9esmPSVfj3wLD/OUnfkcZ5OC9bIfoKIVcauxq/l7skfS/v8uB93qFpufnF0tR7TSm9/iBs4NwB//6QJvOloutEs9a0q7yPs5rmwUKVbL0edFbSM0HBP+SRv4peMvWu9MW9p/9F0qOS3pNS+nxK6fu/mM5XdCQdkrNfgg7zhMbG6Mkjfmce/ILGxigero2iKD48p2eZK91oz4aKovjpoij+I40XvkLS/yxpR2MFFLq17NKSY3fY/3dKeta+H2m5z+iHNH7J31gUxbr9NYuieOao9iq6ZrqgsVvxARvftYPwKmns8r1P0lcWRbEq6asPjnsscYXKvASUUrpL0o9q7No/URTFuqRP6Opx3c9JOp5Scp6+o+S8+B5/XOPQyK+V1C2K4gNfUscrmkVPSfrBINvaRVH884Pfu5oth2fxXDz+f0h6RNLrDvj3r6nKA3hJaMaadtT7eE7Glwde5zI+rejF0XOSbrMQU2msrxyiI/SS5yTdbqfG9zSTd4ui2CqK4nuLorhX0rdI+h/IBVG1ll4PuhYdJuquA411o1m/PytJBwD6z2m8Tn6nXsaA3A3P2Ugp/cGDGOGexkrnUNLHJX1TSul4SulWja37a6H/NqV0+0FyzV+T9LMvsh+S9E8k/eCBoqWU0qmU0h+51mes6GgqimKksQL7v6WUTktSSum2lNI3HJyyovH7uHLwXv/WjelpRZI6Gi9GL0hSSum7NPZsHElFUTwh6aOSfiCl1EgpvV3jRe5q131A4/DH/1UvYyH6CqYflfTdKaWvTGPqpJT+cEpp5eD3j0v69pRSPaX0hzTt+n9e0omU0tpV7rEiaVPSdkrpyzTOz6roOtMRa9pR7+OXJD2QUvrWgzDh71E50FfRi6MPaByX/z1pXNzmWzUJRc10Fb3k5yT91TROPr5NYwDIaSbvpnFRiNceGDubB23S7vOqiqzMm65Fh/mOlNKXHwByf0fSzx+EI0N/48BD8oCk79K0bvvPNM4N+U8k/eT1eIB50I32bCxJ+p80tuDOSTqtsZHwE5J+R+NkqPfoGo0GjZNQ3yPp8wd/f/fo06/aD0n63zWOL39PSmlL0gc1ToStaH70fRq7dT944Gb8VY2RAGkcN9zS+N18UOMQq4puABVF8SmNFf8PaLwovUHSv7/Gy/+ExgmpFzXmy5/VOD/gavTPDu7zshWir1QqiuKjGudtvEvjwheParxoQf+9xkbhFY3f37vt2kck/XNJnz8IwZoVevWXJX27pC2NjZtrleUVvTiatabNfB9FUVyQ9F8cXHdR44TUa+Xviq6RDvLOvlVjXrss6b+U9P+WnHqUXvJ3NE7y/oLG6+XPa1qezuRdjd/rr0ra1liW/3BRFL9x8Nvfk/TXD3j6L7+Ix6xoQteiw/yEpB/T+D03NTb0nd6nsXz+NUn/oCiK9/BDURT/XmNQ7mMHeVgvS0qvlryglNLjkv50URS/eqP7UlFFFR1NaVzu8ZGiKI70VKWU/qSkP3sQSlBRRRVVVFGglNKfl/THi6Koii+8wiil9BuSfrIoiv+r5Le7NTYoF4tJ1dWyNn5d0k+XtfFyoRvt2aiooopuAkopvS2l9Jo0ruf+hzQukfruq1zT1rgs5//5UvSxoooqquiVQCmlMymldxzI0/s0zgv4Vze6XxW99JRSepukh/Qy9xpXxkZFFVX0UtCtGtcE39a4TOOfL4ri4VknH+TsvKBxuNZPvxQdrKiiiip6hVBD47LFW5J+XeOqRD98Q3tU0UtOKaUf1zgk7i8elNV92dKrJoyqoooqqqiiiiqqqKKKKnp5UeXZqKiiiiqqqKKKKqqoooquC1XGRkUVVVRRRRVVVFFFFVV0XWjhqB/b7XYhSc1mU5L00EMPSZLOnj3L75Kk+++/X5L0pje9SZJ07NgxSdLS0pIkqV4fb5S4u7srSbp06ZIkaWtrHGLW64039m40GpKkW28dl/ZeW1ubOg4NBgNJUr8/rvQ2Go3GD7MwfhxCw/b396fuPxwOp45DfKdd9trhOP3jd75fuXJl6jleeOGFqf4wDidOnJAk3XbbbZKk1772tZKkWm1s6z355JOSpMcee0ySdPnyZUnS933f981t86tGo5Hj5ab3EionPyelpIWFBdXrdXU6HdXr9amxTilpcXFR9Xpd7XZbKSUNBgONRqP8Lvz3hYUFNZtNpZTU7/dVFEU+d39/X7VaTSsrK1pYWNDS0pJSSvk+jNlwOFRKKc8x3g39XF5eVkpJ3e540+pWq5WfqVar5fv7df7cw+FQw+FQ29vbGg6H+Tnoy/7+fv7k+UajkTY2NrS9va1PfOIT2t3d1d7enoqi0N7e3lze5Zvf/ObC+0G/GJ+FhQXVajU1Gg0NBgM9/fTT2tvbm3qXjLdfV6/Xp8Z3NBrl8V1YWNDa2loe21qtpsXFRUljHtzf39fW1paGw6EGg8FUn/b391Wv17W2tqaiKHT58mXt7+/nMWW8vR/c3+dFURTqdDp66KGH1G638/Xc8+LFixoOh7mPzBvvjz8fbfd6PQ2Hwzwei4uLU3Pfw0xTSvrIRz4yl/f4jne8o5Ck5eXx3pXIEGTd3XffLWkiY7a3t/N4OzH/6DPz+Z577pEkfc3XjIvTfPmXf7mkyTyHb/jOuDM+HOc9cz6fyFLuyxxDNvI7/dnZ2ZGkzI/nz5+XJH3uc5+TJD3xxHgDZdYGZCIytdVqSZIefPDBqX4hM3lPyFrWDvrHuF24MN4n6/3vf/9c3uPb3/72YjQaaWdnJ8skJ/++sLCgkydPam1tTW95y1u0tramW265RY1GQ81mM/NZrVabmvfIV5eDZfdxPnb+5tP5YGlpKfNxURR5bUZOjkYjjUaj/F79PH5zOTgajVSr1bS2tqZ6va7RaKThcKhut6t+v6/nnntO29vbeuyxx7SxsaEvfOEL2tvby+1ub29rMBio2+1qOBzmec18hI+Zl8yvRx99dG5r5Dvf+c5CmsxxxhjeWF1dlSSdOnVKknTy5Mmp4+vr65IO8xSfzNlZvMZ3KN4/XueyU5qMSdR/OM6ciGtmPJ93znHuBw9HfYnv586NN8/+2Mc+Jkl6/PHHJU1kG8/J92//9m+XJH3/9483MD9+/Phc3uV9991XMDasiUVRZF2Dvqyvr2thYSH/vru7m9eT0WiUnw/eg/+63W7mgYWFBR0/flzSWG8bDAZTc7VWq2WeKooi6x6S8tqDbGKtXVlZUUpJvV5vSoeiD7wP5oM0fkfLy8uq1Wra39/X/v5+lkm+Vrr8WFhYyDLDdSbuy3hxT/q9tbWVn380GuX332g0VKvV9MlPfrL0PR5pbPAwLICdTkfS4QWOScR5TGYm5+c//3lJ0mc/+1lJ0vPPPz/1O+3QPg91+vTpqU8WEph6ZWVlqp+8lPg9Mg2/R6aKRgm/o/yz8MFULFybm5uSJooB7fEc9JNxeeMb3yhJ+vqv/3pJE6MEYcQ43CgqM0ZYrFBC49hGpYxncYHpwpFJ3G6386LFgsU5MD6LogsQFkvmnN+nVqtlxmKu7u/vTxkbg8FAKaX8O/em33t7exqNRrk9N3JgMFeWYTwXCNEgmAfR78XFxal7MFfpK8pL2bt0pZ7+0a5fU6vVtLS0pGazqVOnTqnRaGSjkXHa39/PykS/3z9knEG8f97b7u5uqRHAgsinGx+DwUCf/OQntbS0pNOnT2txcTGPw+nTp7W/v5+NDowNKPI47fsiwnFX5qLsmBehPPtYSxPeZ17z6QuLX8d7QAbfddddkqS3ve1tkqQ777xz6nwoAjHcP943fkYjJSosfGd8aZd+8txnzpyZ+j0qRKwNzGtkLEYXgBTtocBgzMRxpP15v0f4B16s1WpTcsfndb1e14kTJ3T8+HGdPn1aKysrmU+RM67ESxPeGA6HqtVqU2CCv1Oui2uyz3NporT3+/08z1NKea1yecZzuaJCP10Wj0YjdbtdFUWhXq+ner2en6vVamkwGGh/f1/tdlv9fl8rKyva3t5Wt9vN8oLn9LXZ5QMKIWN+PYgxZ05hsKJMArLecsstkiZ6CKBrlBWMPRSNEJc5fh00y9iIsskBMm8HnvM1zc+LvBtlROz/rLnlgJvfP+qJGLQc/53f+R1J0rPPPitpMs7zINYL1jpf21kjWfc2NjamACdfQ30twAhxmbW/v69ut3toTZUmxgPzCd5zXvU5jw5BH5aXl7MR5HOevnNfAGB4FANqYWEhjwNGDSAA74VnRfbQD+dJ1uyFhYVDIN8Xsz4eaWxU9OqhWWht/G3W/1znf66kRXL0xf/8/hHJ9nvGe0WBGNus1+uHUDy/rsyQmdX/+IxlyvCsMYr9mjdF5Zh+IYAQZqAb8TniO4jP4c+LB2txcVFLS0v5zwVk9GixoLjgdEKwxXlxLf1ydMq9NSAzcdxnzdGydzRLkXEFrqKKIl1NLvknfNLpdNRsNjMSCEXjQdKUR8ORSf8sk49x8XejId7P5bAjsGX3wBh3uYMhhZLi90HpWVpa0nA4VLvd1v7+vjqdjoqimAIcUWjiWLj8ul6GRkWvHnJvgKRDa6Ar18zVMk+7r1PR0+jnEY3hIKN0eK3zNTr2x+/nAEbkRX8myA0GB0aJtuBajEuXAVFn4Pcyj4p7M3w8vN9laz50TcYGFn60kHH/cBxLHxf5hz/8YUkTdxroFCgKD8dD0Q7oFR4EBBAeAMKRHnjgganjoFkxHCoiB1F5oH2eB1c+Fjcu/eeee27q94i+YsGC/tEfxoX7P/XUU5Kk9773vZKkt771rZIm7lmfDC8lzVKqnAHcs3HU98h8THh+p13eOUyPggzR5t7eXl4UfeGB0WGs3d1dpZQy8udM4X3o9Xr5vHq9ns93hKNsEY7MFp8PBd2VjnkukI1GIyvbIH7cE4Sx1+tpY2Mjex2KopjyDIHo+PvCQ+BhSAsLC1peXlaz2dTq6qqazeaUq5V2i6JQo9HICEyv18suYu7jKBPoi88Ff9/+fxTKeFd6vV6WQ4wDLmtpOlQyGmfMgTIqMzbiAjMPYt4TdkTbs0I4YlgMzxk9IncfhF/h4eA8CKSfdmKIxywPRvSsRA9UHM+robTcj1AU+o3sZq3gHdM+shdZideY82OIB2FqGxsbU+3Nk3yR5TkZN6jRaKjT6eiuu+7SiRMndOzYsRzq6GOHYgF/dzodpZS0s7OjwWCgzc3NzNcoCD7/XfbEEI6Uxp7FxcVFHT9+fMozyHURVHEP1nA4zHxfZpwMBgNduXIlh08RqlKr1bL86Pf7ajab2t3d1cbGRjY24LOlpaW8BgwGgymPrdO8PVTSxHPB3OITxJ1PPBnxHUdPAjwWEf4oR+KYR96EZl0PXS2iI/J+DCeP940ekPgc8T6xXy6fpYl+RX9ihMu8CN5hTWftwbjn+VgbfQ13iqBBXAs4fzAYZG8e85bndN5zTwO/u04KLxJmzPVlHqrRaJSBP3i62Wxm+UGbUZ+RJkBgDLtDt3LwwmWDywDX8zzK4SiqPBsVHSIXGmXKoDPdrPMcQfDfo9LmKMBRKLILuIgeOnNAHh4SrXCp3ENR1q+jDIXYj/j9elLsF+OHAtLv96diyGf1yY1CBDHG1+LiolqtlprNZvZq4FrmXoQ3tNtt1Wo1dTqdqVASD6uJBmgcq4jaer+5DqSU4547A3KM4jML9Y3Hr2V8Zx2rqKI4r1wZ4TveweXlZS0vL0+FldJGnOuScnjpxsaG+v1+jgsndwL+QpFyY6Ner6vVak0hpO12W41GI/O3K2BlMhAwAQUHmeL3ATxwvsWIgi+5B/fE8Gg0Glm5cplMezx/lP8VL1Y0i6KOUAaElOkvnid4lD7ixnwEx2if8zgew3P9PNbDMmOR4wBtsT0HCAEcyryXsT/SRAe6WpQHv8dQ8bhGX21NPdLY4OGw+LHkIzrEcXIWPvWpT0maJE59y7d8i6QJIsBgEHJBLgSoFSgUli/nEY/L5zPPPDPVzxg7GVE3jkeLjufBk0LS4qOPPippkpsRLXNySEA8iEMmxwSPS3xeUDie4+mnn57qJ7/faCozNHxM4+LoC5YnHDsjlRkJoAIg4e7WXFpaysqvpOzJwIpnMfIY9IWFhRzWg7fE0Y2yPng/eRY8Ba5UI5BA6KOrFablntGl+2IpoiYxtGJvb0+DwSCjNj5uZQYZPOO5IIuLi1pZWVGr1dJtt92mVqultbU1NRqNnMBKHPbOzo5Go1FO2l5aWlKv19O5c+fU6/V05coV7e/vZ2WFsCs3hnxe8Ww+55g3i4uLWltbywrLcDjUxYsXNRgM1Ov1tLS0pFtuuUVLS0tZMdrc3JzKvYEXo6t4FkXBPy+iXWRCREtnFbeIaCLvF5mDDOI6PCd4PmgH7zIeDNqJcddQRDcjOspzxLWhbGHz9nleZChea9aEmBCPbCRHA88FzxOfm/tFD9G8aHFxcerdRC/6YDDInoRTp07ptttu09raWvbWIkOYn3xfXV3VYDDQs88+q42NDf3u7/6utra2pgosjEajPD5ugMMrLs+azaYWFxe1vLyspaWlnIOFMcK18CTyFf7s9/taWFjQ6dOn1W63ddttt00VbllcXMzJrb1eTy+88EKWk/ArfNhoNPJ7O3XqlLa3t/XCCy9M5cGR8+EyF0K+zpvIxcDbxpykr9HbF/OlIoIceSjqIZwX85/KgBb/hOJxvJjwYEwQnwW0Xc1L5HmP3v94Pc/BZ/Ta0q8YSUIkzLzI30cEs6RJXieGtK/T7lUEgHOZyDqPR4H2/Pkw4Gmf/CjXPeBRHx/G0QsZwVOLi4tZhuNd9HnkXn28nuhShFRzD/I7MB7ie3Edz0FIPlutVubBqxkYU+/lms6q6FVBX4zSO0vAlblonYGO8mCUtRljGlk0Y2WXqHT5NZ4fwKJXVl3K7++/u1U/y2szS9D7c3gbLrzmSS5gXFHHQPMQL/rFeMbnc6PQw6kc9Wy1Wmq1WlPJ4Y7IeNJcvV7XysqKGo2GdnZ2tLi4mA0N+oYR58JcUjaa4oLt4+s5JJ4g78QzgZjyzHHRiX8VVTQPKptrKSW1Wi212+0cnujhSH6+83C/39fFixd15coVXbx4Udvb29rc3JxKpo8KjgM8LlvpC8oHCgm/e65FSikrXPSl1+tlBLXX62UPCffx8D83UgAZYngH8qXdbmcQogzMISm1rOhERRWV0SwgMSrorEVlhmsEVrkmXi9NlHQ36pz34jrDb14Nin66/GCdY/0lJJJwLF/T4Sl42Cu60V/3uAC8lQHIZeuj6wzcM+Z+XI2ONDawpIiLBb2ahSrx+90H8begbHg4EIzRgr/jjjskHbZ0qWKFh4PjtBPLRdJetNxjDGWcRBcvXpQkfeELX5A0yTHB08J1IB48F88ZYzljbgsWJ+OFx4P+E7MIOongvhE0yygo8wZEI8ErN7kiGK/lM6VJdSIWN/KDsMqZK7Qdk5xAsZk7MKbHR/b7fXW73RzmU6/Xc7gP4TagjL7YSdPlClk0IV+cOQ8hQAm860WxgoXnSKDUg6ZAPp7StEeKMcNYIFeDijnNZjMLxaIopsJChsOhrly5oqIodPr0aY1GI7VaLfX7fW1ubmp3d1ePP/649vb2svJA3CceKNBaPCX+nAhY+ri4uKjV1dXsuSDHxJEiqlNduHAhl8R1QeqejYia0oZ/nzfF6kvISHif47MQefrFM8M3yOrYfze6pcOu/nh+RGFnobLMp1j61g1S/4Sip4OFERmKNz2WvmV8YgWcuMbwO/l/rFHI4HkRnlUHPFzhhrfOnDmjU6dOZbTe1wZky2g0zjnr9/t66qmntLGxoUceeUQ7Ozu53CRt4rlot9uq1+taXV3NnsfFxUW1220tLi5qfX19qoBCWe4DYVqufPB9Y2Mjew6LotBjjz2mhYUFdTqdnPjdbDZ1zz33qN1u6/jx46rVxiVw9/b28hpK7gbzZWVlRUtLSxoMBtre3tbu7q52d3e1s7OT5RY8u7+/r+3t7UNrx7x5kxwNeIm5wpxk7kVE2hFzJ1cu/fxZXsoIkMV3NSuXM54Xr48ejqiHRf0ojmusNhorzvHJ+Li3vKw9+ofeQ4TKvIi1n+fES0dffA3D8+C/uWxzIA69JnpIvLKghwt7aCF85TkWlPf3MvKsTbVaLfMIXkh0Jc6h3Lx7b3ku92Sgk3g1KfeOxut5Xs9n9VxUz/nw62MO06H3Mp/XW9GrjaISUobuS9NW/CxE7ai2ytAGfnMF3xdvN3QQLDH+MVrlMJkvAB4uFRcA+hOTvV2wuwfBDd2yHJF5UUQYyhQIznHPgffRn98XOhR6BBsobCz/6/clPIIEfca/0+lMlW2lFC3tIIhjSeMYtxrD29y4RXCDxPI8jLsvMHF8Kq9GRfOgq80h5ioGe5n3joRqvICEIW1ubmpzczMDCYAanouBjCLMcX19PSejY2y4EhDnPcq8J2a7wuWyT1JWUlCI4O1Op5NzUdfT5LYAACAASURBVPAuOz9Swjwaya1WS8PhMIdmoHB5udeoBEeZW1FFR1HUNeIcRyH3ue5zrSzfwtchL68dgTL3FnAday2eC4AQ2uB3gL7l5eXMU+6V8BApfy5Cz73UsXtNPJ816kv+jLHK41F/Ps6z6Jo8G3yCEsVFnMHGwgOh5+bkJkTUK8bu0R6xk6B0sToUgxg9LljUMW53FnpHO3g2yBXhehAOPBq33367pEmuBshHRCQ8rr7sfrSPh4R9PHgOxu+lpKMmjxsB0bjwfRf49HCo2D6ud8Zgc3MzxxT7PajCwhjDLF6Fx3MOHFn0mtIwkaMwblygAPAbhDVP+3t7e5mxo4fAaW9vL4cN0c95LoouBKkWhXDa29vLG/lI08o7z0x/QEk4l7CIdrutVqulU6dOaWlpSa1WKysOtOnC0/ezoBqNNPbe8f663a42NjYyOup5ExgbnqtDvgtKh6NBGD8kqt9yyy3q9Xp69tlnNRwOtbe3p3q9nsNNHJXy+exeLDca/XzG+XooNYwBsnOWZ2MWKsm8ZKGKOQvRaHMPYdl3N5b9ODRL0ZtVsSbKxMhjcU+i6O1mzeE9xE0FI8UKh8gJ+kd7jNu8KFZ7iX/U2V9bW9Py8nJ+Tpcr29vb+uhHP6rNzU1tbW1pf39/akMv5CG5H41GQ6urq7nthYWF/J1cB/8Of4H2pjSdkzEajbS+vj4FlMCH5FzxnORqXblyRb1eL3sOn332WTUaDZ09e1arq6t64IEHsvzAY+loaafTySEsjUZD29vbWXbFfC7G1t9v3HxzHkTEAbzInInVgGYBSFczhiIPRA+Eg2z+fVa7sf3YXvQ8xHbjc/B7lDlX83ZGj0vcJyxWpYKHYy7uvCiuvT5+eBUXFhYO7VlFZbcoOz0c13WfpaWl7E10r0DMmeA7uhNlr4nCQM/E6AHsu+eeezJI4eOEDkVUARti4v1bWVnJ+gFeSfIrY/4N+goeWvrhHg0PtYYYN7wk1wLcVZ6Nm4i+FOF8NYs2xifOuk9Ey8rQKlf8HE2PqIJPakfx/ZgzT0Q2PE7TEXT3RMRwMH/eiIL4M7lb08dvXhTbAh1xpT+iKWVoZnymuJ8GHo7oUYjvg/GPJVJ907FabZzwivfD0VGENAKv3+/nMcbzAXoaPS4oL3hWUJBIkmcexWcv82wchZZW3o+KZlGUA5DLJUIQ8Oa5HNra2tLm5qYuXLigra2tvIM2sotwIjyFeDAIwcDYIJwKg9wNc2ki39wYLeuzI5z0XZpO/h0MBlpaWsoKDfxLOOVoNN5R3duJCmCUO61WS6PRKIfFehy7/1VU0bVQjADwY3Gex/UMKtNnfE66h8JDjFKa5EDBI6xRKPGs2+QuSZOy9uRJdjqdvNaxFkrKxpGkKWCCY6yzXAfvejQHsgAZ4SBgGfDsY+LrOuMR19MyOtLYANGngVhNCYuMwSLmMSL5eC7iC8VTAjFY0QNCTgQeBWL94i68MYk4VouIxO+g53gaeA6uiyhirBgzy83LggGaxfU8F+PJPhwx/vmlojIF6yjjwhVPD7Px9+sKORO5jGkpVeroR0rjuvL1ej27EH3TOFB6YvajUCiKya6b3meEwf7+vra2tnJVBvoPgg+6kVKaQgKcGfl0Dw4x1+zhUebhebHkaALxos1mU51ORzs7O9rb29Pu7m4WhNFr6MilNMlxWV1dVbvd1qlTp7LrFkHqYxOrefm+JsSXSsrz4uTJk7mv/X4/o5Ye77m/v589H41GI+8TAkqDkrW0tKTjx4/nfuFhSinpxIkTuTQoyG6j0dDx48ezxwXUiHFwz0ZEbeaNmkaifVDA6JmIngz6x3dkB+1E2RRlVJSpZTwjXd3jESl6YGgn5qQwD/ks22CqrF94r3lu2geFBu2L18Wdz2MVrnkRxm4szhDzN2Jo0fb2tra3t/Xwww/r8uXLeuKJJ7S/v6+VlZWsBICetlotveY1r1Gn09GpU6eykkLbyAPGBw+fGy3kpDG3kaPdbleDwWCqypU0kTPuSXOjv9lsan9/X+vr69k7s7e3p8uXL2tra0udTkerq6u64447VK/XM/izt7c3Jc+5z6lTp9TpdNTtdrW4uJjnN2ElbAbosn/eRERFnCPci7lbZoD5edFzUOY5999neTaihwKKXkd+px9RD4n38/xCJ3glthN5L7YbeS/qOciemOtCu+xfNk/yPpUBlf7JfPI/1jR/h1zD+gMAEL3/Kysr2t/f18WLF6dAU6pK+R5W0bNRr9d1+vRpNZtNHTt2LOtGRVFobW0t9wu9am9vT1tbW1mXkqbzHKXpinwYH+hC8OZgMMgAAc/iRRngfWQZ93ZvV+XZqOhIOsob8cVcwwLrE9SNjTIvA65Lj+X1Nn1iM9EjxY2G3AUavR5liiULWEzExaUKeZiPh93EPjEG0RMwb6LtVquV3dIRseGYU/TGeIUnDBQMhbKkUjfepMMVODzGEwFFTDnx6B7eRlhGSilvBgjCIykbDXyS4M8z0k+UPozJuIkgFL0a1yIkK6poFh2FApadJyl7NDY2NnTx4kVtbm7mXAX3FKKQdDodra2tqdPpqN1uT4EitOcyycEED+9wviMkFICCTUCRp7HUuHtn4CtJWS7jSaQ/Ozs7WVFaWFiYkrOEmPAMhHtJkxwO0Fkvy+nGUsWzFV0r+XrIdwdEo64QIxci4FJWCMd5AkWe38mhAMBzfvJ1lvPY1yrmi/peOtIkF8PvTf/c2IB3PAc2emjcQzFLlh0FQL9oYwPULVZNYt8JqkhhmUFYshHlmoVmxdi/WahbrLzCdQiqWC0hom3xPihRtBfjmWehczEWEooKM58RSXAXud8XFI/clOtNsZ8+wY6acJAzbayHP+v6GAuPVU4YkCvxKSVtb29nZZWwAY/F9EVTUo5ZptqUGz8ot1jvxDvSXxZJD/GhigroQnzGGDvMospxD2+aB7H4t9ttrays6N5771WtVtMjjzyiy5cv52fzEChXqt0owBggH8KrkvA7gssVFp6b98T7ozoNYylNUKA77rgjKw8u2DDiut2u9vb29MILL6jf72t7e1spJR07dmwq/p/FACWp2+1m17Ok7K164YUXtLS0pPX19Sx4vdqO/11NUF4PY5Fnwqsa48G5Z9xvA4rx0xE1jDIOirJ4Vhx3vH4WShpjgD0fx/sZPRrIPvLkWDNilSlkMzketMf79jhj/0RJJf6eNSzGk79YcvkiHQ4LjcBLSkndblcPP/ywLl26pKeeeirnGbnXmM3/7r//fq2srOjkyZOlc5j2ARwYXzfiXVZeuHBBu7u7eu6553KVPs+TwKPgm3cuLCzoxIkTWlpamkJneU73eB4/flzD4VCXL1/Wzs5OrorV7/dVr9d14sQJ1et1XbhwQfV6PYeFkc/CLundblfb29vZS+yKFqjqvA2O6B2M+kgMFY1zO3olYzWmOEej52BWrsasHI94fawuFfO3YoQH50G0w3H3AvtzRtnj6HfZczAezJk4Do899pjmSZEHy8ZtNBrluQWRvxn3pIIv8fKzLroX0b3mvV5v6l2Q6H327FmllHTx4sUpg4Yxw6tJuBXE3F9ZWck6jXsv3TCBHDRtNBoaDAbZ8OcZ3dtbr9dzIYqyUEdfR8pyN+LxMqo8GzcRlXkPIl2rpyOiA2W/H2W80BdHzlKa7AeBss49dnd3p5jdw0SI05d0CA2IFBVMVxZ80YXper1edjWWGRuuJLgS4JUj5k0wuMduzsoVQTkvU3zwHHmIxyz0whGPMmQDAcxzu6ITjRP6gFHDO3APCO9/dXU1n8P789A37u2fw+EwC07C9KBoWFSejYpeLJXNnzL5g5LMZpcbGxtTm116Nbh2u51DkQgldV72NqPS4nLZjY9+v6+NjQ3t7Ozo0qVL6vf72dOIxzaGs8KnKJpspsl93WvtYaMkpe7u7mbFxg0wN0odbGq1WhoMBmo2mxoMBod439Hkim8ruhaKaxfHPAJBmjaCfK10EIA57p6+OCeZ6/AsORiEOeG9i3mpzj9QXF/d88Lvvg47ucyQJsYeHo74bB7J4LwfjTfXJxw4vBodaWxEy5qa53ynFrrvXMgA+HfQKgQfKBa5H5zHYIAwxF1vYxwy32NuhtcYlg4jDjGOPSIBEYWDuP6pp56SdLgyTKzDzXfGh+O0GxGUmCNyvSgaAn7cJ5+kQ4sbBEOhhMcwJ2cAZ1TGxGP2YVBiiFNKucID51KxC+UYxJrEYnIHQCe4L8o/DEZ/YcLNzc2p8nGuDI9GI125ciWH6RRFkRdn2sQ1SoJjt9tVv9/XyspKNpLmuSgikFZXV9VqtXT+/PncPolobkCBZMTwLkruHT9+XM1mU61WK4+th0Lx3hgbBFj0kOBJicaqzzU8Pf6+XajVajWdPHlyygPl8wzkFQUF5UyauLZRhs6fP5+NHvfa+Di6wPQFqWwez1uxiaggNEvWxfNitaaIorrL32kW6hhR3OiZiAsZ/YooZ0S3OC/K1NjfiP7Sj+jBiDuIR290vG9ENa/H3jdlhoUrEfSffSc2Njb0zDPPaGdnJyeAs+HfiRMn1Gq1dObMGS0vL+vEiRNaXFzMldwgrxADn6DYFEWRS+mSS3HhwgXt7OzoC1/4gnZ3d/M+Gijw5GSAcnrVQHil3+9P7a7saG/cTZn5e+nSpSxrkJ+eFLu5uZnlJ3K90Whoc3NTrVZL3W5XvV4ve39mlRmdB0WvHu8O2RHzqDg/7iweeS/ycvSMxL1hYu5E1GNmfY+y10Pp/LxYzZO8J7xayBbaif2O7UXZEKt0xpySWK0T7+a8qMw763PWlWaex/mX82IYMWsU3jrf0wZ9AuAzpaSzZ89m4ICy1Hgt4U1p4rmQNMXTbgQ40MCnRyJ4eDjyaHl5OZ9HWLF7XQDmuBfn0x5rsBfb8fcZPWxXy6OqPBs3Cc1Slq7mwYjnRFSNxSqeHxnc0WeOu9IZXcU+cR0J43difgnDiUS/ypjQn4dwAzdg/fcYvxkV0RiSg4cBZH7eimpUyggjQhnnnjFMyPvsY+HH46LlFBGY2I6jIdL0ghTHPaVU+s6YJ/zvYYfeRhnaEwUzQtRDFGYhMNfq3aiQ1Ipm0SzvhisKJFJTqhIPKNWYlpaWculpymO6sYJMi7wL/2OAE9Y6GAy0tbWlXq+ny5cvq9vt5u/cP6KqcQOvmB/ingnndZebLksAcWiHfnKey02UOcIxGQOXqx7aeS1rV0U3L5WBRr4ezwJd4zrnSjVhTnEfKmkSXcA6RO5Fs9mcqg5HOCRtesI3gJ5HaEQjyZ+JNqL+5GspxgzyZmtrSylNikREQIZjboiU6YCcFwGnWXSksUEjoEh4JKgOhWcDTwS/R4saC44dutk3A4s6VmHCAma/ifvuu29qMCLiEHMhfNB8EKKiFj0d9CciGE8//bQk6ZFHHpm6L9UrYsUYiPswPrfeequkSSUVEJFY7WFW9awXQ3GSRgWwLEzGnyGOoS+Cw+EwLyLxzxdM2q3XJ5tS+aJEG5Rzg7m5xhF0PAYkVjrjp5TU7/fzpyviVI5IKeU4RjaEc4bm3pzPTr/MfYwckPa9vb2M/OH+9/GbF0UUjCoy3Ju56fWzXRH3MRoOx6UqEYiOkMRwB+ryRxQIgcP8wTNSpqDE/yH/HnkW5YkYbfe8oJgg4BHUnIPSRZgKfYzG4dUM6DKB+2KJezJPQPdiXLN7Av043k9yGnys/TvnR7R0Fio6q5/RAwLN8oxwXVlddz9/Vjx6rCYV1wiuizuHx+eP/aMf86Qy0IQ5QxjUzs6Odnd39eSTT2pnZyfnla2urqrZbOqWW27JFWhQNoqiyFX5HB12pYk57B7U/f19Xb58Wbu7u3r22WezJ2MwGOTzkJeU0OU7+xuBhHrlGr673EVWcD787efxLvb399XpdLKC5UbKlStXpuQL8eknTpzQzs5Olrd4XjCW5klxLsbjce5H3olzN17P+yG/EE9CzJ9i369Z+3lFfYf7ISuiHuNrrHR4DYneU/SVWREekYcjr8X+RI+G6wPS/HNUWQ/QPWYBhmXH43XSdF6i73WzvLyceYWKbpJyfhPl3judzlQRFubw7bfffsjg9vFzPdp5y6NB0D3gPx9vN3pYK/f397W2tqbt7W09+uij2t7ezvPQPVW+5rnM4b3NWjuPosqzcZNRmVJxrUpUVBp9wrnnwo2EWe34J33wCe1WtaN5jhQ6g5ZZ3/4X7xFRuIgY+nFfVJzxhsNhXvTc/ehhOfR3nsTiHhd7BE6ZMcBzlCWNuzHHNXHseP5I8Rj3cYPZ+8n3aNC6QsU5ZRTfYTRu4/st8+5c6xgf9dwVVeQU+YlPFBQQ/m63q93d3TxXKRnd6XTUbDZzOBO/+4J+lDz1TUcHg4E2NjaykeGJ4M7vsQpd9BJK0wq1/yHn+c0RUq5zPgeQAtiB+B1PNfcHPW42m1Px7YAPMTywooqgaMxI5bpG/B51AZ/fzD334HkhAPcUwtNe3dH/kAmUfyZMKRqdEN+9P+6VgRfdOGA9d4AREI6ckXa7naMwZulKsT+xD378RRkbPDyoG8oAFjf7UnAeHhAICx7PABYzsYIIHTwfsarDE088IWlSh/nBBx+UJL32ta+dum+sIhHR+PgSY64EccCch6X3e7/3e1P9v//++yVNPBNUSOE5IxLA/hn055lnnpk677bbbpv6PVaTmCc5A/piEhkzushiG47yOALGZmwew+soqnuhRqNR9lxEa9k9BSzGoAEwqzTZ9yKiEzyDuyndSIhoATvZOhInKZehA+VAWeD5aJOdd8k14d77+/s5p2Oexgb9uXDhQt6xtygKbW1tTRk+Pvb0lfcjTdApdjam2hcKC8ilV6XxUA7IPSbuKnZj09EYnwPSdElLYl790xEdv9aT8EBqUUKiV6IoiqlF5FrI56Y0f4MjeleRnYxLRAUjgUbiaYt7DkWvavQQxIUkJvCXAQJO0VNwNY8H94k7gM+qXOieLf+MHpGIIkdZHz0e3G9eVAZe0L+UktbX19Vut7W9va1+v6+LFy9qb28v729z7NgxLS8v68yZM2q1WnltRD56CW4UnjiP9/f3denSJe3u7urixYtT96G6HjxAjlk0LpAblKx176+vFzyX59KUgSzIqWazqdFoUj1OGs9V9ssBOAKNp0+MW6/XU6vV0sbGRpbftVotJ6DPk5gjcZ+ryBPMxcijsXoTegWf8bxYiY53iv5BBAj6R8wJpT32Noq8MSuXBJnA78ieeB5rYZRBPE9UiiOyPqv6KMRzzHsHcd8HgrWnVqvltazMoPY10tctjpNXVavVcugyOcysLYRC8unFV3xfjvX19Rw6CY/hvfdciTLZh47ieR2nT5/Oax/PBU/69VyHV+b222/Pu4pT/W0WyAugeZQ38WqgXuXZuEmozNqPCnr8vczKLbNmfTH0RSkqp3FRLvvfvRZ+P0f6pEmJ2WiweB+87Widc4+IcrhSKykjcfTLy1T64u/IhYfyePLXPMhRFRfiPt5xTCVNKRbelhtwZWh+2fjEeXGUIu/CqgxdYlyvFs4T+0FfaKPMwJnlOTuqn0c9/7wNjopePTRrkUU+sKGl7zdB4jSx3QAr0kQOxnnsygDfUVDY0LPX62l3dzfno/n1UZly5SH+VvZcURGJ8jLyi+dXoLDU63V1u12NRqOs+EaQy3NZKPKArIpVhCqqyKlsPT9qLXDjw6/14z7X4LlYpjsa8NwXPvZ1ysOZXR9xT7z3D3Kjgjbj+XyPgI/vFQZQR56YNAll92gVv2/ZOEf96yg60tjgIbFw6TQWORZ1jP3DU4FFDpr0/PPPS5rkbtA++3UgeOg08cggDR/60Iem2n/Tm94k6XBMZfRw0F6sAoVnIu6M/tGPflTSZII+9NBDU/3g/nh4QCpoBw/Gxz72MUkT9PEtb3nL1H1AhyA8Jl9syMe1UJkCHpV8fosUJ7Ifjx6QMjSadvF41OuT3SgdPaDqVLfb1ebmpobDYc6VoG12gvZ3DQoHch+TgR2po294ULDWiWWGSS9dupSV4Hq9nsMbiDWmegq1sYnPPHPmTK64EsOJ5kG+GKeU8k7bvrsu6EVKKefBeJKYpJzghguWqlvLy8v5Od2Lw/uOxkZM0nYFhveRUppy1SLw3NsR0a+otLBjeIwvdrQQBIt56iF2HuIHXc0Y8QXoqLDAL4W4N7IgehaibAGFjFWgOJ9xoJ3YHp+0FxfhaNxfrdpUVEodZPDrY1y3J1N6P+N9WXPi79EzR3usHcRNMx6sPdcr7KYMGJGm55yk7HFIKeVdhzudjk6ePKnl5eW8WV9UOLxtX9ypjre7u5sNGcKl3KjhPOQGPOc5MKCy5KZBoKiuUETwx4GKWq2W9zdyVJW4cYyira0tfeITn9DS0pLuvffenEBbq9Wmqhs2Gg2dOnVK7XZb586dU0opo74x7HKeNGvORqUq8hZ6Qdzni7wi1nfmamw35qCiN6BPoE8hG7gOvQpZwnV8umEqHdaPIPqDHgMxJ+A532Ol7PxZ1UFn5THOW9/BW8767WuTz19yGaLnShqPGZ4Hxs+9sPzuOgHRFxgi7IPBmhdltpf9dwMCnnXvNzzrHg9JOUpkOBxmbx+feHYYC/pDe9zrtttu02Aw0JUrV/KO5BgeblRFgJBjrPFX48fKs3GT0yxjw5WRWeh//O7nOiIQLXpPkuKeGBvE7lJvncnuSoiHBUQFsgyhjn3nPmWVsKK17sqTW/tR8fYqFQiEWSj6PMiNv6igzEJzygxOxuNq/fQx8evd1Rrnjp/H/1EgxWtjX48SYGV99vcWPTB+//j+rkbX811W9MqnMl4D8EARIDQUWUEJWC8n7cpBWVgCCpTz+qxQprKQkNhu9ES77IhKklPka9ryMEbnY+Q7f+y/QcnyWcm8bEjGvhvkv1xPY6OiVwfFNcVlftka5fqLA6KxPeZ1p9PJvOxAqnstpOkiOuRUeZ+guF6h25SBurFPLjNcr4nyw2WMg70ePo1HlLArSTlvCsDCjbTo5ZxFRxobWHIxrha3S6xCBRrFPhTsNP6BD3xA0sRT8bVf+7WSJtWZyM2gHeo8gwRguWNhf+pTn5I0sbTf8Y53SJJOnjw51f9ZblauI1aQmEU8ESAAd911l6RJzspb3/pWSRPEgH4wLlTP+uqv/uqp637hF35BkvTe97536vlB4RhP+nM9qlFFigaBu7ul6VhcaRrZ9EnlngxpsuB4EhTxijDmwsKCjh8/npVzZ6IzZ85ob29Px44d09bWlj7zmc9oa2sroyfEMmO0eIgThokLDY8fHY1GGfXb3t7OfXdEn2ddW1vLscQIitFoevMtf08IGZ6T/jjDzvO9OYF6OJLiyeIuJH2n4eFwvIP60tJSfn7eucfbukBzdKosfIn7RHeyJ8Exno7elF2HxwkBWpYs6wLcc1bi+dHzEg1NzikLrbsehAxC5sX9JKKXF28yngnQ0ogWxlyKGKYS0VqOxwozsQJN9AzM2v8jeibi9VFBjbKO5+B8PmM+G+3Tb9YMxtP3X5Em4xVzPK4HwWfsnD0ajXTp0iUNh0Otr6+r1WppbW1NnU4nn4csY6FH7tAeKKuHkzrg4veu1cZ7ASAP3UDZ2dmZ8o6CxkqaOp8/PMDevjThJxR/5DweEo8fl6bR5sFgoF6vp36/r2eeeUadTkf33Xef6vX61D467qG+9dZb1el0cm5gp9OZe85G9LbFXIrIE9Ebx3d4Dp6O+2xFpBz9gzUuegK4jnyeeF/0B3gg5ivxfrk+7hsS78d18Bh6CvoJOapcT7/5PeakxBwOfo+yaF7kHpwIvvkcZz3DQwEvrqysTO2fwXoAQIpRQQ4I/OTRGvTD15vLly/n3KVGo5H3EnOvhutkrqPRf4BSBwOoSOnRDYwrnofo4ZGmq5qRMN7v99VsNqfyvdArkEnINNri/ZVtfOxUeTZuMvLF/ihjw88vQ8r9e0TDZt03ogvSpAa0J7fhzsfLISkzWNmmXGXel/g9ogsIAjwQMd7SGQkGQ6CUxWY6AlhmtM2LorIWlWeOReWScxCGCAuMEj5x7ZahnEc9x9V+m+V9cYPIx82F9lEUEd14j7K2y+Zt/IyI8DzfYUWvHorzi/mCfHDvrIdlei39opgkglPGEg+upByqIZXn1nlforc3zlvkhAMiGAD0xcM/yzwb0eAgTJNwEg+thI9R/CAUlH6/r4WFhany48gf95AsLS1lJY0S39cbEKjolUlRx2E9KFszmW+xSpQbG5KyAe8GuocqY4z4/Rxg4zjAGNfF/kaKno/oyXR9JoZpodNhBLmxG9tBxyGfCh4cDodqNBr5GX2zXTcurkVHONLYiLkaWNBUocJy5Tws3CeffFKS9Nhjj0mSvuIrvkKS9N3f/d2SJrGNP/MzPyPpMGr1+OOPS5IeeOABSdLb3/52SdInPvEJSRPPAtUaPv/5z0uaVH+aVeca4nmw2B999NGpftx5551T533DN3yDpEms5C/+4i9OPQcvl+cmdvKbvumbpp77p37qpyRJn/70pyVJr3/96yVNPCB33313aX/nRY6MsRCgOMM4ZZ4NlHx35UVlkYnmtbyLosiLZnS7YUlTmYHY4wsXLuSJTeUGz+HwfvE/z4GnxF35INy4B0lMhBFBK1AAqNbinpiiKKZKRxL/KE0Wzc3Nzal9NhjD66GoulDifoRpODriCGiv18vfpUmuy9bWlhYWFrS7u5tjRdvtdq4hHgWcG4XuCXCPEuPm84hKGz5nmAuO3syqXALyy32iYHOB6XHmKEERKYrGqKM9jlZB1xJq9sUQ88fnihP9Aa1E1jLfQClBGWPVqegBiGGScR+LqMzOqsEPRbQynkc/Y/GBWKEmfo/9i1706Mnh+TjPK6/4+bFa17wohglxD+4LOsj4uOeXRdz5gz0l8EBglAyHwyn5FCvUcF/kGdeDQmKAeHJoSkk716kD6gAAIABJREFUOztT90fpX15ezvdzo8Flc61Wy/uEnD59emoMkMF4rh2NZadwEtq73a7OnTundrutM2fOqNFoZFnKWkG1OarmMU7zpJjTEOds9FxEDwHeM4i5hj7w2c9+VpL0uc99TtIkFwM9gvaYy1x/7733SppUr2Ruz/Jucn2ZF1g6XG3KQ2WkiYeCHBFybOP+HfBUfG5kG5EpyCruF2XEvHkS731ZWKH3nzVgbW1NrVZL6+vrWlpa0vr6+qHCKfV6Pe8YTl4S/WaecpxrMALIqWAeucx0XSwCv5B7Ytw4oNx15GnWSnQhjKbobfQqWBRhIPJjMBhoZWUl83pRFLmi3vPPPz+V8+qASuXZqOgQ6u9KXvx0NNfjjqXDCaORPOQFZE+aJEO5MVMWssK1oH7uSndjIyrxEWGMi2v0aLhLscw74NVQeF7P9aAdFFPCmBzBdOX2epAbXdd6vivV9NsRTt8AMSrjPl4RGaEP7hKOqEzsRzQ6+Cx7nuidiG1CvEOfBzHMJLY5a6zKxq2iimZRGUKJvCiTWY5+IlNIzsQIIKeDmvyz+MoVFWR1NMod3Y33j8YGMdpegANed+KenOfoqvOeeyp5LkkZGCB3A/IQD4ylWOnveoFyFb2yyRV+aZovmcMo5VSEa7VauWRtBCxZz/1vlpc8rvfwkudKcD7Ho9cPiutt9CJ4KBjeQzwYbmzw3Y2NOA6SslGBJ4QxYMNfaWwQb29vK6WUjWQHmr5kYwOLnNhCkH2Q/1i7ndhDPAZf9mVfJkn61m/9VkkTi/vd7363pIml/lVf9VWSpB/90R+VNPFc4OHgfnySE4JngH5wX6pbxZrqsUIKnojz589Lmnhs8DTgWSF349/9u3831V8s/R/+4R+WJL35zW+WNMlVmTUOv/RLvzTVb98TwY/Pm2ZZ0fGTc6XpzZliDKIvXo74s8iAZrE4sFDACDFfw2NxHfGlXeKZYZDIuN4nGNCrTbHwQ1jrrpTu7u5m4eMeH88dAHH0e5NXsr29raIotLq6eihGcl7vUJqgIiy6jIe/Izd2CHFwowKlhvNAKCXld0X/ywwzSfl9Y8BhKCJEHfn1cBKuZx6CXEqHQzfcNR0LAvBMCFLmCQKT/Qx8PvOu3Zgu89j59+jpeLHk48dzSBNPBrIX1BCZyLwDXUV2IevoI7IHdNKRdSfaJ98tVsyZtf9G3Mk7xs9zHM8L/WGBinl+9DvmptAOx1ljuH5WXHqsSoXXfd6yNS6wET2NAAlVYlCykZHIp8uXL2tvby+P0/LyclaC2u22VldXSw3uRqOR5ZKjoDFhG08t1yIffc+qhYXxzuLtdlt33HHHFOK5vb2dZYcDMFTgQ95Q6cf5TJoU0oBXpbH+cOHCBS0tLenUqVM5P4Xxc4/N8vJyVojmTa5cSoc9AXGfi3g+7wz9gX26PvnJT0qa7JQddwZnzsL7v+/3/T5JEw8I7cGjXMd9mduz9u2AokeB8YdXiBSJHhtkAv3mucnBRTYxXvAo1UqRJTFsaJYX+8WSrxHu1fPwJvJGV1ZWdPbs2VwdDq8j3jgPg+p2u6rX6zmk2vP+0HcIs/JIDgxxisfw3FSNcm+Ej28Zn7u88bUTXQAe9+eIHgfGwiu7sXanlHT8+PEcBSIph4ctLy9rf39fzWZTu7u7Wlpa0s7Ojp577rmcg3UUAFB5Nm4ycis//h9DTFxpx+XvlaGiZ2HWvfweMabXFViPacZIIdExKn+xXf6PyEM8DoPF/rkw8t+jMRXjPznHk8hntXE9KAqeiK7O8gT4uLuw83bKvEfxem83vkv3XpXdVzpcnz/+OSJadq+y7whQn2eRyjwl1/M9VXTzUJQZAADRWPdYa8KCut1uztmQJjltvV5P9Xo9J2T7fVAw3CMJ77i3IcolBye4hgpZ7XZb7XY756gh21wuuEfGk3/LEOBodACOEHqF4ouHA0XJwRI+Kc5xvZTVil7ZVLb2x4gMDOJms5nnFEp69Ma5ceR87MAecx1+wxPHPfkteiuivuAUj8Fr/LkXUZoAYvQL46MsVBegQZpOpIdHHXh2w0maGI/tdjvraRhNR9GRxgb7UHBTEH48ChyP8ba33HKLpAnSD3oGykWs4td93ddJkj74wQ9Kkt71rndJkr7ru75L0qS6E+dTDYrv7OD4hje8QdIEtcKyjsgD/aW6FaggiAC5HzzfPffcI2mCHHD+a17zGkmTKlO/9mu/Jkl63eteJ2mSO8LzgojgycFDw8sGMWBy8TJfSnJjwFFp+t5ut7W/v5+RN9/FUprE1MO0XtGBhO+FhYUccwxjE3pEKUQYd21tTSklnThxQpubm7p8+XLO2+B+vjsngiPWq4fJT548qb29PT311FM5VGBhYbwDuKPj6+vr2bPhCM/u7m5GzX3BBkF3r4t09coMXypFdMqVCgSFI/2cz1yMxhljefLkSbXbba2srKjVamVh6iFIrpRAzBnfk8MFHud6zgfKBd9Rchhr5gBJeG7IMc54qYgZ5Y97Mt+8Ms4sY8mpzNi5HgYjMgn0CFQTmYVng+/IQvqMjIu5H7FSDGgr/ee6ixcvTvWHcEHQyLNnz059gmLG2vn0j37QPjIW7/Rzzz0n6bBHA9kaZR/jwX3ZUwkZz30j8MFzIKsZ55gXOG9yhNOVbFfGa7XaVIlJ5N7e3p4uX76cjQxf8JGz586dm5KhyDkUmLW1NTUajYxO48kkNMLfF6EWo9FIGxsbeaO9Wq2mY8eOqd1u684778yGBv137zJhsRhHeB9Onz49tdcRY+EeJVf4FhYWpvj3ySef1PLycr4/ybrHjx/P8e+Li4u6cOHCoYpnL5a8Kpf3M+Y1xZwHCN4jYgJ9AhnKus8cjTzMPlzMdSItqHoJz3K9I8/ev1mejfhcXA+vxMiPWSg1ESrICvrjJV39ubkP7cX9RK5nmDHtO8KPlwHPBvOcsCP6HaMj8MwTcsgaN0sn8OpVi4uLufpUBEOhuN7EkHEqWZF3GsFQ+sU19IM2fQ2WdEjHI78WmRPXPtZg9LnTp0+r2+1qZ2dH29vb2VMziyrPxk1Es5TfsuNu3aKwM8lQMGE+2ojItiunMJLH3tIGE9zj7VmcORcjxhnFFzTadiMGwYL17YhctOBjTLAzvLtL3cByoeG/sbj7udeDXNBET0QURO4BcKHWaDRy6VvCNTzsyO9V5i2JffF+uOeEe0OODuHCjfdi3nn7nh/i5PPNF4vYv0iO5pSdd7Xnrqgi6eoeNwh5AuCBseEb8UnTaOJoNMqhG7u7u9nDAR9LyugicjEqTMxdPATkdriCRZgSQAvnRC8M/OJ8zXn0KXo3PAwvekR9DKlO5QCmjwXyCiOqoorKqGzOOVDnSr/rHRHF97UBgKvX65UaG15RDt5xb8be3l7WMXytprQs9/Zy4PCWF4FBt0E3Y00k1Jt+9Hq9qfa4H3wKIMEaB4+XAaX+nWsIfQb0vVoe1ZHGBjcAqb/99tslTVAnLGmQeSx0YhOxxLHgQalA6fAI4DH55m/+ZkkTzwHXY0nHHSyj5YznAPRs1k6WMU6aHBDaxTODwOM5ia183/veJ2niAfm2b/s2SRNLnwFn3ED5PvzhD0+1w3m0HxGK60FR2YyuQRArX5SIQ5TGXqBer6der5dzFFzR7na72Vjw3Y+5HgQBjwIMRBUqYo/xqq2trWk4HGp1dVWNRkNbW1sqiiK70hEaGCSdTkcpjeMri6LIaEJKaQr1dfdqSkkrKytqNBq5CgtMTBUqns/jHH0cPRZ7NBplgQQKMW+KaEU0hkAmJE0p7R4j2mg0cv36u+66S81mU51O51CdfP9jDjFuft+I2ng/+GOeuaHpigpjD+81m80ptMWvd4HNcRBf5jn38HhyV764zg1i6TCqN29DAxkV95kAHYyVAB1d8t+RYbEqFLwXczCQ1aCReFCQoXggaC/GdyPT4nHy1PgE3aW9uKcRFPcXgUdBW5HpjEPcSX0W+hx3Hp+l4L5YcmMUz1sZvzBfmceME/KTtlqtVpa/PPdoNMo5YfV6XZ1OJ3txqZwD362vr095A5nvhF8dO3Ysy0fPr2LfI0IjqDwT9/3gPVDhJno0AWs8byuOEe1RcWo0GuUdxnd3d6cUM+QEuWVU/FlfXz+UkznPd+oUwYvoLfMQMGmyfxiRG9H7GHk85j/98i//cun9GXt4isgL9CPmDLzhFcv8PnynvxxH36J9Pjmf+9HP97znPZImOamvfe1rJU30PNr3+e3tOWg3T4oJz+6Nk5T5h/nD/GNuufcTwJL8xl6vl/fM4Xd4lD3A8JQcP348hwMClC4uLuZ8Tq47d+5c5n0MIQfsAGLRaeAtZDr5XuRTeQU8lz0ACK7bLS4u6uTJk/kZh8Nhzjv1TQ09TIzcE/icfTmu5jWuPBs3CbmCNet3PqPy5X8sHiiBLGau/EoTpdbDJWDqohgnRNZqtRyrTPiAx/ByHscbjUa+3kNk3EMRFzdf5GBAL2XoYTvu5eB5IjJJm47aoRxjcHgs57w39St7ZwgSaVLNyRHS6PXgeL1ezyFj7XZbzWbzkPIQ50E8fpQnwL1IZX32PmG8uNu4TDH066OLWJpsrhg9bWXz37/H5znKe1NRRZHKvBllBg5GNXLRQztcBrHIe0gLqCrhS244R88sHg0Pt4TfkREoPxgwXO+hX9GjAd+5nIze4fhblBNl/M2zlHlS6CuKIB6ZiiqaRcx3BwijQRwBJw+99jUltut/0WPiPELbzF/3gBBSGfOd/B7wgodccp7nYhGCiG5DaHoMycSLKU30Ate1HET0cXGd0I/7RsLXouMcaWyAwJNzAarFccgrb/jvVGMA8edB2fH7x37sx/JgSZMqT1RfwHLjE7Q77jQOOjZrk5T4yUvleaheRT/wzDBhyOkgBwXkAcsf9O3cuXNT7dHfhx9+WNLh6lRUcZhVC3+e5Asf6NIslx19wAp2JlpdXVWn09H29ra2trbU7Xa1v7+fETm3xvv9vs6fP6+Uks6dO5fd9AsLCxkZAxE/derUlLeCTf0YQ9znKysrh1x6GCGgfaAEe3t72trayqjA1taWLly4oJ2dHe3v7+fwoaIoMqqBtc/zEI/Mb46gsrjCdDs7O3r++eenYlev1+ZTGHMIzrW1tRyG0ev1MkIMiuLCBOS01WppZWVFKysrWl1dnapUE0PUvKKYKxeeVOfzCcQy8iRUJpDdo4EShkAGZUXpoa8obyBXGEwIajxMLgzdI+WyIbrQjzKq5kXuSvfPWJM+5jgwzhiZeAyQ1cgaciaiJwBPA9+5Lu7fgVcWWRXvS39pD8Se6+KuyLGaFjKYfnB97B/E+6EfMU485nTEfTg8tG4e5Aqxgy703T1/krIigAeEeYcR4Ma3E/3e3NxUr9fLlanW19fz9YS5gko6+EK4w7Fjx7SwsJDRZ9Bf5Pjm5uYUOgq5bEBxcUOD8Ct2RY+VBiH3ROGh9rECsMGo8spVjUYj7+tx6tSpPPfmRdE4jP3nnURknrmIHhLX8VnetHi/WP0qejzQe6I3L3oGuA4ejtUuo8xB3/rIRz4iaZLn9ba3vU2S9MQTT0zd753vfKekyX5jv/IrvyJJeuihhyRJX/mVXylp4imBRzEQ6S88HvXJF0uscYBO8BzhgceOHctrH9EQrBmEImIMOFjlHkWnskp8eAH4Th+cJ5FNrFmswXgXt7a2co7GYDDIugzn8z69QpznMQLweg6VGwZ4NtmfjGpcy8vLGShFT5SmI4n6/b6efPJJ7ezs6MKFC+r1eoeqfUaqPBs3MZWhcfwPlSHZGAMod8QoeggKCycTG8VwZ2dnKpkZxdeVPV+0o/IbBThCxfNAOF6G8qOcw/ws/MREO+KIkeWoCO17qBHns1B6wuWshebFEsIMQcn76Ha7eQM/d6H6plqOvjgSGmOlHbnxhdc9QrxrhLSjQrO8QrThIV8Q/7tHJCpsfn+MWkIyYvxtRGPK+lLm0XD3sR+vqCKnKB8kTcktyHkJnhmNRlPGQCQ3yLkXRvTu7q4WFhZy+IIDANHjQDvILJddKFmOwEYZIE17SLnOvd3k1MXz6Hd8puhBjnyKPIXvPSQspTQV4lFRRbPIASvmHMAlZdE97M/XjKjzcJ57IBygdfDM/2L+ka9xzpteFQtDHK9nTFJ33aJsTQeAA5jzkEfnGzdA+XOvh4fAISfQMQAA6bs/WxkdaWxQXQk0KtaHpqPxdzwA0G/+5m9KmngCyNFgZ26qLWDpgrLhAYjVrKiQAtrmCp80Qc3izp9MEhAInoP+cDzShz70IUkTCxzPBRY/yIBbmtKkyhaT7e6775YkvfGNb5Q0QSo4H8s/7sg5T3IlUjpcipRJGxdGFgb6zK6buPeuXLmi0Wg0ladBDLM0QcZw7cFYnU5nKlfCk8BrtdrUXCvzVGGVr62tqd1uTzEorvd6va7jx49nb8Dm5qY+85nPTLkQMRAuXbqkZrOp22+//RCKUavVpmKq3dXJTrgg+qAE1ytfo9VqaXl5WW9605symlkUhR544AENBgO98MIL2t7e1mOPPaadnZ0czuXvHw8Q7wCBwXjwvhC0kvJuqFQPI74TgXPmzBk1m00tLy9PGaAIPu7tngueyY09n39evQfhDcJ74cIF7e7u6plnnsnGRr1e1+rqqur1euYp955E5Yf7M1c4n0WmTHl8seTGsn+P8dSzPBvOr9LEi4q3F08HXloq+OE5gE+JL48egpjzQO4HsimiqvSHfT3IBYHod0Rb8QZHxTTGw3Md/fAKKtLhPD3WAK/C5p/zIke3masgmcgJlJw431F8QDHxSBAH7UYB18PHFy9eVLfb1alTpzQYDHIOHOGtjC8eBvjCUVrGjZ28CcWgH/ST53BlLHo8qRoVvZn8OTrKdTF8FS93SinPF+QvCC1KIR7peVLMn2JOz9prxoEnnleaXc0pAhhQnPNcBw/BA+gnkUdn9ZvrIm/F47xrcmnR42JuiYdISxOvJNc/+uijU8eRPTGvLI7HvHnS88kA4Jg/tVpN7XZba2treQ8ZvGjkd0qTPCuPZkAXcR7gXMA/51XPZ4WIPHC9qtVqqdlsam1tbSoEeHV1Na/DjUYjVwLFA8PYec5WrVbTyZMnM8Do4zsLGEGXgf/Rnba3t6c8JAAArC0rKytaW1vT7bffrn6/r49//ONZhy+jyrNxk1D0TkRjo8zLEZFwV1JRQFGqSWKKVZ+iC5f/I7rgngD64Qwd0QaYC6+GL8zeXz8vpZRdhO12eyqUAUXUN/DBkncl2PtThsg5OnA1S/9LIVdu8GiwyU5RFGo2m1Plii9duqSFhYXsio0oiNetjyEmjsygzHlFDly7CHKS9PF2OQrkyrUbMxE5ZXw9XpX9B9xzhLGxsbGRDT2vG+7vPwpYf7ZI7pmJ77iiiiJFeTXLMEWZnmX0uCfQ5bCjn26A7+3tZTlFQjgy1I0bl4kR0YxrgoeeeK4HYVPOn26EOOpbxisuc1wOuFeD7xzzhF5vB49z5dWoaBYxB/HURXDJPXy+R4R7AqXJWl/GK84LtMt5AA8UuynzcriOEP88EsM9k84f7hnkPhynYqi3I02X4/e1F16PIVBRv4l5W4C7AItLS0tfeulbLFQUBnawnFUnGRQCDwA5GFjKoF4f//jHJU0sRe7DeVwfLXJQrVgRhUGKCANKF4oS/aR9PCOgX/QnxhdjrbGvB/elHzEemv6Qq0K9bDxFIBS0z3f681Lts8Hk908fUxBllHNHuHGfnT17VqPRSCdPnpyyuCOTeRIUDMF+HIThjEajrLy6u8+ZFGbznAkWOkd4YDzmLh4HckXuuOMODQaDjHZ49RRp7IVz5kKRPnHiREYfY8iWCx93x86TEIyUf0TR/tznPqfRaKRbb71VKysrev3rX5+f8/Lly3rf+96nK1eu6MqVKxkpabVaOnbsWPYKpZRyrLYr7nt7e/r85z8/5cFCKYkK/qc//ek8XlS6wKjkfaeUcqgXMecoEfDUpUuX1O129eSTT2avBp6Ufr+vixcvqt/va2NjY6pkqOcZ0Uf6J2kq7yMaHL6QOKLsgnue79H7FZVQZJbnsDgha2M1J2Qoso/jINmML7KGCjNUf8Kovv/++yVNKvMhk/geZTNzHy9vRGf5nX7gFQZlRSbHeGh4jH7FXZIZH2Q0v8+S/dcjZ8M9da6M+zv1uRiVDzcKUCDcY+Ny2Y31/f19Pf3009rZ2dGtt96aY7kp+rC/vz9VIcYNFt9duFYb76+BZ0HSoXKeoJ++5uNhpUKN53R5yU/IwyHhT1BUr0KHESVNqheS00JohyuF86IyYMspKmfRswHxzDEHY1Z/o0dkluci3of+wkNu9DrFXBNkNtejVz344INTzwWvUoWUHIz3vve9kibeVH6H96juiSyIXkjaR6Zcj3USL6CH93q4NesavAVvAJw5LzL34R10CSjqoWUJ30VRZA+BNAnpdqMhApbkXhCy5FWoFhcXs5JPZAMED8IzMY8yVnZkvBxchU+Hw+GhnBrG7NixY6rVxkV+arWaTp06dWT+TeXZuMkoKleOljnT+8LgCZCujEmT8CiPJ4xKHAwqKcfagqYjUD0ZqyznI1JczLlPGZLB8xTFJNxneXl5ajMeRxC8H858GFOOUsY/v6e3Ny+6mucHpZxnPXHiRC7DVxTFVBhLrF5DO47uU853e3tb3W53yiUtaUpgF0WRlSOEOyWQCZOjChnCE8PN8zeGw6GuXLmira0tnT9/fiqPhqS5zc3NnEjqAhyBHOdFHMNZFD0ZsxSPiiqCyjzALp+iEuhyN3pFpEkJXfjPFSLCmXw/oaiMoPR7mw6GeD/KPBq+JnBeVO7dixK9FGUykef0XKxo3LvcjxWp4j3mKVMrenWRz9+oPziP4g10T3iZ7HdejnpSXBdmAVhRNriHgPajsRHlA6GHcbfz2DfkgOdy0TfXayLo5nzl7TA+Dv4SsimNARE8G0fx5ZHGBrkRKBZ4NmJMIr+DSvHwWFRvf/vbJU0sW1ALLH8sZhRSLGQ+Y65FrBvteQLS4f03YtUK2uF5HBX1fjO4nAeqF2M3IdqhfRADvnM+8aj0l8ogUEQw5kVuVERLmjGNscXkPRCrj3LqVY5YDCRNLXI+Qd2DwnnSZEMbr2Tg7kSs8SgI/JkIt5EmC3nMEeC3drudN6/znXsdAa3Vavn52LeDxZ9dMnEhgtjHhZD/vYrLvAh0otVq6dSpUzm3pNfr6fLly+r3+3r00Ue1vLysN77xjTp9+nSuDPb+979f/X5fy8vLuXKMb+7T6XRy9a5er6dnnnlG3W5X58+f13A4zDkrGJYe3kQuB5+uJGFUUI2sXh+X23zooYfyjshFMU4+297e1m//9m/r4sWLeuaZZ/L+KxhLrnwQ205IHDvCYixhULpAh8qMNacyZWle5IiSdHhH7PjJefANuRPIXL6DRrKTNDLm3nvvnfoeKwzGHbfxePD8tItHBJkL3yHr2HuIalhcT14ctfjJg8PzwXMwDlyP7AQl5X4xPDN6p2OYTVyr5k3+Hl2xj153znXww+UG8s49Giz4jUZDx44dy0AN/NDtdrWxsaGiKPLvx44dyxXy8ER6WAby2Nt2sIbn8Th0+p7SZD+QaIS48RCv83h1CGXP48ZZD1JKunTpkvr9fq5wxXnk/l0Piut7vM8sT4MDV9LhPWF8TZUOe0qiDGLuz/LGxdyMWPnPkXV/Lo7HHcjjfiH0k7wucjp4TngRHmY8/u2//beSDudoRN5znWOehMzHUGVcMMh5Bt9vjbnqxrY0KSXPbz6fyyov+ZrvYCa/4YnwtYm57X12Y9+LISAz/LsnhLuRQBsxh8rLRvscd2/maDTKfWJ8+A7vAYRg2CwvLx+5R1zl2bjJaJaAjpauowL+nT83JjzOMJZ6je5/D2fw9jhHOqzkscA5xaQr3+ciPg/f4+JYr9fzRlmuzME8lJJ05M0rTc1C8jh/VgzziyHaB/VDWSDMzTf0YeFGsUcYkEDKnytm0ajzsAcEKUoBcZuMu6MmCCb6gGAmB4OcHeYQ403S/vb29pQXptPpZAFHqAWJqSmlvO8ACgl9j/N9FvI6y5itqKKr0Rczh8pQS/90BQGZ5bKLcE54hfMJuaAPeAsdRfW+4aGl3y6P/Vz3+NJHznfDKvJUvGcZAsxxrvd2kOmgpl79juvnHdpY0auDZvHYrOOsJ8x9n5vSYeOw7H6Qeyji+sN3N/iP8mxA6CfufXT9yEM3y/oVnyd+cj+XNfHTPSQxFJm2/JwyOtLYeMMb3iDpcOwdFKs0ROaPFnvcKZuXCLpGOzGm0FEWJzwC7kby7xENdCVEmqBgxB7OOh+EAUs+5lpwP47H542oGr/H8XwplB33bESlUpqEM8VKXl4lBwsYD4e70LCKfUMZfz73aFDPXZrsSMsEJiaZPvgihifDwwqIz4y7yjozFEWRz2fHWj758+pZHq/pxlVE8vj0hX5tbS3fa57vlTmzu7urxcXFHC8J8nDLLbfk8Zekz372s1kIeGKc70tCBRvenxsHq6uruWIVRgXvuNVq6aGHHtJwONTDDz+sra0tbW5uqtFo6MEHH9TCwoKef/55DYfDQ0h4u93WysqKbr/9dq2srOTEfapL0f7rXvc6ra2t6Q1veIOKotDHP/5xbW5u6sKFC0op5ed94oknVBSFjh8/rnq9nhFUN/6YDwhuadoYnve7Ooq4b0QjI3IfZRLH8TB4MQA/P+aEIOuizKRdZD0eDAj0k+P0m/tTTQpZSCVCBxX8fI7jNSePj4p83I81AQ9MzNVgDWFeIcNBZSOai3ybt4KKzEIRcGXYz3HZ5n1hjnryZlzIMerhQzx5rnj0er2pCmGgq16phrbZf8iTSfFsOHDE+LoyhGLmieHuyXAPnfPUUUm2MYxLmuQMnj9/Xp1OJ0cX7O7uTvHzPGmWUubvUTrs2Yj02shgAAAgAElEQVRgWVTyOG+WhyMCIlEmzGovRmJEPSiOUcw1ifoYa3HUx3weer+5Dp6Eh9ljJ/Ia94t5aEclFX8pBI+xjkWwEt2CUKStra1sbDhIiEHNXHbvIM/nHo6YX+bGiusWePljKFQ0oKP3kzAq/935Cpnh5d95LqcIPPj/jJXzfNSroRgdczWqPBsVHbJwo5eD774owgBH5S7Qpt8DYqFlUrOowSi+q6Xfn/Pjgh1D+1y59Od0IRL7wmLnz+dejBiG4GPmCD2C6XqFbEiTak2+r4ek7Llgsd7c3FRK480Yeb6y/IxZKAVhSjwzpYl5Z6urqxqNRlpdXZ0yotk8jGS0uKCxiRBKVKPRmEr0Xl5eVrPZ1MmTJ7W2tpYTWNfX17NyJSkbiysrK9n1y+/+PDHW1p+f/yuq6MVSVGycIrgTFXtHVR2tJLcJGcn1HsbB98jDEfWMoJIjl/Qfb6ikQ7KStmKuhnsluA99cpk9C2GObTAW7OEECOKK+7UoOBVVFPUSN/LLcjbKgEVXuqMnwu/B7+4ZiPcv00Oiwu46WNQ9XGbMUvRdZ4oyqUw+lbXjxsrV1skIpJTRkcYGO0J+5jOfyQ1Kk9g+lKloMUcEH0sPy9ZrjUuHLWBQrJjVPwudckTIv8e61Y5iShMLPu6CGy05zifTPlr88fm5jvtHVJDv3Bf07qj48XlTXNTiGHpoEn3ySiAYBCD5XgGB6/09OZL7/7f3Zj2SZVf59xORY2VlVtbUVdVdTfXkbrfBA4OxMVwgIRsZCySLSy65Q+KjwBdA3MMFV5aQGQw0FgbbGNw0NB7a3dVTDV2VmZWZkRmRY8R7Ue9vx3OePCeq2h2F9HedJaVORsQ5++yzz15rr/WsYYO8dzodnT9/vpTQHY1G2t7eLhu0oUj7opVzAcWX/SA8cVGqziHvK/3kWRytAeFnfFxQONI8Go2Kpwfl+8qVK+p0OlpcXFSnM871mBZ5mBjVmDqdjq5cuVLCqBAuR0dHun79ukajkZ599lkdHx9Xqrp4TClH3yF9OByWRPp79+6p0+nowoUL6na7FY/UaDTSJz/5yYKYdjodra6uqtPp6OrVq+p2u+U6cjwGg4EWFxd19uzZSkWl0WikU6dO6Td+4zfU7d6vp9/tdrW9va3Dw8NSapm5xLxgTx72FGHfAtql+hXGpC8o6f36v+A/noH3iRKVqKEbcNLJPYI4L73FV69erbSfCiPn067zg5+f1a7oD8Zeek6glLFcTy4Ga0RWu8pcE3JBcu+klM2071WM/HnSQzstchTV+5OoovcXlBM5g4cBvnSPBfsIPfHEE4UnKQUNqDAzM1PQWuK2oUz89kIc9A+QBBmGjHfQhWeoMy6Qqy4jWSt8vPFY+9pCmKVvOIYH+ujoSJubm6XE9uzsrJ5++mmNRiO98cYbJ7zYH5XSowA1haekXG/ybPB9gk/MZY6J8KehmPpG9jPLjEP0M3knPRzwOt7BjNxIPYp3iz6TuR5clzmwudY8qhxV+uDVqIiCwLtw6tQpXbx48YTiD++5UeH6pb9b51U+w2sOZDpA6f1xAMB5yMvDdzqdsncZ7fleVc6T9I35lOCie6jcqOEIj3I+PLq8vHzCaPO5Bj83UevZeEwoBd/DnM8EdMPBF1dXplHk+N4VOZ/g3hcYlD+/h1cjQaFkEcywhPR8JEIwCbXgvs6kaeU7g+X4+ALgYQVQMuQ0yJ8ZpT0FJIzvKIukktuBoPMFP5WJujAISly6UkXIE4oNSoo03mmdJDeEmdfL9/HiOQirI7yDfT0YY/rB3PBSpz43vN18b5NQoexPSy01Ucq6BCYmzUNHDetkFXlVnOMy0RUC70siqHXoK/f3eHXn+TpU1uUA7XGOH+t4K7/zvvmY+L34joIVi4uLRWb45q8tteQEP5K3lDyWHgMvkMI5vpb6dXXEnPTqTbQxHA4rXjjneeejpLp1yfUoX2MzhJHfk9dynfX/0+uY982+8mwPWkudJhobWCmJdkEMvu826g/C74lGZ8eJN8YSSzQu20vUj/OzznFa5DkY9Bv0LdHBppjC3AeDdrhfenzy+/T08DnjpadJ6QpjctQp4j65PD/Dn8UVVC+NBoKMsuhGgDR+pyifIAz0yUut7u/vl1Kr3BuF03fYdQ+Mt5VGhccvs3CurKxoNBqV+3icpIdzuRI+Go1rZnOfU6dOVbwFHtY1TWPDK10cHx/r9u3bGgwGunjxopaXl3Xu3Dl1Op2yMzBoyMrKSlFQ9vf3S2L81taW+v1+MZbcGOl0xvX1ES6DwUALCwslBp9dRnk/CEAQUpQjNvnjPAxW0EmvhgNy2+l0Cm+TgD4zc38fDXIPeL7t7e0KwuMGsc9nR4P83SBw64zxaRuLjIF0Ei3kc6KBmSeWNfxzbyD220hvaZPhlCgmHgjmMTIa2Zcy2QsASGNPR8pe+sV75zy8u3hkfH8caYya0i/uAzFeuWYlPQrDsdvtVt6LL97c06vxDYfDSpU6jGkveIARjYJNxZ/0fLApJ/IN4Ab55caJy0PkP6WonbcSlOEZUPJ5TveCp/Hg89o/+x5dtDccDkthC89BcTn8xhtv6Ny5c3rxxRd15swZvfDCC1P1GNMnJ+ZK8lDKg9RTmjwc2W7mvGZ7eT7HnNvpMeF6eCfR5pQ5qafA62nMwfu0n+2iR+GNxEuZOQ18To/KtAgeoojC+vp6ma/oCqwV/uyZf0ROhX8vVXOqfI10PRndgXLt8L2PH7kgCeDRLkaLe9zcC8k1vgeOeyZ8Dyz6zX25j8sk12dchiGDfE3y+zCuH8nYaOlnlxyxdcbJc+os5DoL2pUOR5YzDMkV/fRSwKCEUaFQJxLxIOTMv0+kHuaA4bK8mz+rH1MpdWTRnyvRumlTooh7e3tl505HJxFU5DRgPBDexXtKhdzDQfjzPA9C57hXCkB3yXp/PN7UF3UXwpALWC8o4ELeF0qft/kOck77O+ZeHH1sW2rpYSj5von/mxROn39+TqKQdSHB6YVw8Ijvub7OC5L98/P8+zoQKteOSe273G56ZjeI8trhcFh2St/Z2SlhsA/rpW/p8SLWJc+7TP2gaW1w3vGQbI6STrQBiOm6DjpGKu7J434/56nkteQZv9Y9G35eArzZB//fDYaMDvEIhNSD6iJfmmiisZFoURM1eSy4ccZCpuWflY+aLO4cvPRAJHqS921CLhIZAEXj/um5gBJ9bELT+N2TWuuoCQmZBvmYpUXvDMGEc5QGxMtzM/jNw3G8LZBsYpFR9GEUt5ZBsvr9ftnXwT0Hrnyy6zfXuyKcLlNQB9CJbvd+vCaeE8JxqM7k1Svos1eScCQPDwbIG+3C9A9a3D8KuUAD8QRFYcdUhEPGjXqoU/ZvOByWMCTaYfzOnDmj/f19ra2tlXk0Pz9fqlm5h2hvb083b97UcHg/cXxhYUGrq6sajUYl94LwEPrsQnM0qm4+OBqNKvt88D4llR3nEYjutnaDyIUz84n5Lo0RWDegcnweBV9mBRPPmfHv07uMjMIjkPltnJdVm5BlaVSDgvZ6vcpnn/N+HUdkWu6dRP/wgKXXGI9GVtKBOI8KN1SbYr6nR4XnBU1lXLICzrSJ+Yk8qEPbE8TwEKVcu6iyx7hQfckVGpfHHPFUeqgh4+IgQNbiT/Q24+2RY55TUofyErJJdSvmDzyEp3N7e7vIWUeZAZYYzxw7clRee+01Xbx4UZ/5zGemjoinXpH5Pk1AFMf0Tjad78/m16UekREaUOZ15X2ymmceXcGUTnpNqT6asijlX+pN9Aee//GPf1z53pVe6aSMmRa54t7pdEolxZ2dnTLnPOIiFfv8S2O6znOVaw3PRxSBewI93MqPPv9Go+qO3v6950y7DkS0SG7omdEVfPa8KpdhrnOlceJ5Hr7pJnw8qbJY69l4TCit0qS0duv+aKfumL+nFZ/tpEdjMBgURfXg4KBi7bvCU7eg1wnxRCeyX5TK9TApFxauEPh1TeiAe2ByTKdpcGT71Ncnod6Vl0mCg+ckFC2VEhLnOZ8wJ+YRBoLHfOf4dDr3w+lwE2cFHE/Oz+dKpQfi+0xcpa2cj/w/CdGto/8LQ6Olnw1qQi6lk560vM7/bzJW/BxHM9274f2oa8NBJFeiUnGq65srXvlsyMkEqerac0AL4wKFpW798XtzDzb9BBRpqaUk1oEEDnMeN4GBaWxAPkf9tzQ6JJU5nuG6CXylh5L+18mSpATb/R4JwvozuPHkxzodx/+y2IOfn7pPHU00Nnh4UBZQJh4CyzbRpYwVyxfa5HLh+rT4obqX69clipVxvxyJ9+Ylgc5lLGPdwuH3SyQEatp/JFHEFJaZs/IoKBnNLXxXylAM2TkcdCI3gSMcB1Q6LWpyOECdMSRgsoODA+3t7enNN9/UYDAoi8jly5crFXrIDaDakXSS6bHq5+bmSs5IUrfb1eHhYUHoR6P71Y+uXLlSCT/ysalDCbyuPsic73buYz1tckZnvO7cuaP9/f3y/I5SDodDbW1tFQ8Mnhl2GibZksXc4zDxUC0vL5e8FGKrR6P7ngreK+1L9/OPTp06pWeffVYzMzNaW1srfEZsOp4NwuY8dhSEl+pTIDrkZuAVY6d3hPvOzk6lqkddiFadUta0sKShOy1K2cO9U4bh4eHI74l+OpLs/c4dw9NbnQtVyiqI/mQFIN530+/cJ/Pi8j4p07mO50amkteWshVyFPBhvv+o5B6NVNCl8fNw3zSsvV+gowAAHseNsoThzjku7xYWFsr+RnhN4GPkuOe0IJdnZ2d1/vx5zc/Pl4oz3o57Qngm5k1uoMn7z/DM/f39igLW6/V0eHhYcjV4Po8Dl8abxILi4pW+fv265ufn9bnPfW5q77LJUGo670GgV+oJdUCIt5O5nkke+iuN9THeQcqClBXZ76aqoRl5wXUpcxyprzsmz6UHJ/XFaRE6gpe0BUDzvKa6AiUOVKYng7FwnQlC12Ftpew8c9uNCa716A50olT2E/Bz+b67u1sptnJ8fFxyoDyqhOtoJ6MdHITI5/TvnS8dAADo5Psmaj0bjwk9jADNyeZKWbof61CuOkqr2v+HQQeDQSlZipBwpZn7z8zc3/G7zmOQaEQdMlFnOBDu4eFHTaiCt8V5eBPqcg98HKdNLsjZHBGjwUOlfIGoQysQdnUCh/YZs5mZmWJwpJfJPRbd7ngTMk8IzSRV2kdQ+txKIcj+GQhTlDEvT8hcSYHI/RPB+TD0qN5jSz8bVIfku3JZN3ea0H9X8hPEqUNDXab5ZnrwgvMqih/3QAHykBZXihIJTnQ3ZbEDHA54pAejCVXOsfLndgMTj3RLLSVRip456gBiGhNOk+ZjHRDlPIDi7Qa1Awi+7tUZNNI4lLdJv/JrWEvd45AbudaBpdl+6keTKHmzLpphEk9ONDbSQkZxQEFLBJ8b5e63dTFuThmrCGX1grTMcydO+pMWPu0Qj4xng+fKHI2mmMuHRTwgR3K9fzkOHEEsqLhC5ZZHQU1eG39WrHSsaKqe8OdM5PH2zpypYKI0EsL09ttvq9/va319XZ1Op9ST53oSAZeXl4sF7VVLXGn27zJnJJ/13LlzOjg40ObmpnZ2drSysqJTp06VShwggZzvDNrt3q8gcXx8rLW1NW1tbZWwLM9pGQ6HlWpa0yLQSN7L3t6e3n//ffX7fb344oulb9L9OFxipfE+uOGRyjnhUt5/lJvZ2dlSacTr5fs7z/1Wdnd3K+FqbDroYRBu9FFdiufb2trSwsKClpeXiwGDgO10OuX5mBe+8zxtJ9Uh0Q9SCKft2UiZkt7Y9BQgE5ARuWsvvzN/szJMyp7c1RiZyf2a8t14T1wPGsr3yNimeHfazXjzpDwfSm96ViKEEnVtWmOmRa5411HT2pfKtxsH7tnz5yFnw3PR8Ob1+/3itTw+Pi5yyfvo/I4svnfvXmWvDd85HMMlDQz3ZPr7drlAVSk3bmZmZnThwoXSZ2LbHdjIazw+PZW0aVGTty0p9ZAmfabpWOdtlU7yTOo/XE9+FESeUua+Iis4pmzJoyu9df2Ectw5L2VByhqOqVc25bD+tIRng/lCdSqfS+718Ofyd08ekjRec/18D5nKNSjzPzF66FeGdgPkMfcJj3a+cw+Ie1C4D2v1aDSq5IPUAa/+vJPkVj5vXb4V/1PUp4laz8ZjRHUegUnnsnCwCHqYSypnvhDVWcsuQDEGBoOBtra2SgKVJ627oeqGxPz8fBECiYDX/SWT0a4nvXnYDsZQU1iECy0Wc0fZ/Hld8EyLEuVEGJLzgqfFNzIaDocVI8KfO99ljpUL1Ix7dYHJdyg/lPZEAOf4OyriXg8SxylbiCDzkBJC15xcuE8yGj6sgvIgYdxSS9LJeVXHU3UATxNymiCKKz11Xg3408OkmkAzB4dyR26MgyzNW2do+PrggI4rMRk2ls/hseUuW+rGzMfUZV9LLTnlxpbwhusDk+ZOHXqf89D/9zXM19QEbj3kPHULX0+9WEnTepb98Pvklg14O70/CX4k6JGGSF5T16ePlCCOB+D27duSTuYaZNxmonSJirmwq/ucEyARhPRoOPotndyhO6s5gPrxHKB/WP5vvfWWJOnixYu1/XS03j9njGLeNxci0Ln0ANGPmzdvSpJeeOEFPQqqWwjrlHJ3u8MouStlViXxd+bVplh0UCJv3bql7e1tvfnmmzo8PCz7Q4CE+94teCiOjo60srJSUe5hLhZLdyF6/2Fg3s3q6moFCd/Y2Ch9y/hkCEHgaN/Gxoa2trYqisFodD8B/fDwUFtbWxMZ8MNSomOUtiWX4e2339bFixf18ssvlxju4XCo1dVV7e3taX19veSXSONStm5c5n1Go/v1+DEcXLk5ffp0RZAx7v7MbvR4boajzBgUVJvC8+QhdKC6bgCzg3uv1yvIWsbS+nc535uEq/f9UVAi7cxb5n2ilHgMVlZWJI09Fngi7t69K2kss8l1yHw1T7D39iHOo11kFUdkbea5pSu/qZJOk7ca4jPt8E6zvcw5ydyRjG+HsoLORyVXHOCfumeqk7epPPtYwZfwjqPerpgwjyis4V4P2nWeAQxYWVkp8goPLZ7j+fl5Xb16tbJ/gIdkAM444u57LTHOniuSBpV00utNW4yrzy1y/Q4ODrS4uKj5+fmpI+KpeDX9noh/k4x4UERHPnPKqqzWBE+n54F20FtyXNL71OSRyepTWdUqPSzcNyNMqGbFeeldzHam7W30NQ0PAOs2YU11YVQJsPKMPn9dT4LXCS9HB+E6KrB5dAMhxu558L3GOp37RVWOjo6K4c/904MCb6XnxL378BH38metM47yOXk+/55xQrbMzc3p6Oio5Nw2UevZeIzoYVDaOovX/xINrxOk6VlwYuHo9XpFQXRE2l3rnU6nUquaxdIZwBVPDABnIhdsvrCz6HoRAY8FzjCttOZpG89GoiUg8hnW81Ep0YZEEbe3tzU/P18UD5QFhA6KC+fTz3xPaSgzzpzb5BFpUqbcCPF4U8jnlxtungvjixzJ5W4Uu0JEm96+/5YLx6Tx9mNLLTk1zY8EKZquTT5Io53fXFFsCiNyBUIaGxpSfRKuK2Qur7wdl5soGCkXvV/etvcz5ad7IpMn8RrnmNAvz0tpqaWkJsQ/vRsPaiM9Abm+NOlBbkQ6f8E/bpylzuCfaZdrHXT3NqUqvzlgg66ELpBjUqfb1I1FXUhdrrUfybPBDZ544glJ0nvvvVcGxX9PSxk0KX9PSz4t5CbFBcoHyYSYjCPOalSJNND+5cuXJUk/+tGPJEnr6+uV505EI/ubvzvq7PfNF+bJs3XjM03yPtQpWm7NS1XlzJFsr0SRrjnug1IKeo3Ce3x8rHv37ml7e1s//OEP1e/3NRqNKgbEzs6Out37+2HMz8+XHatBt0k6pma2u/MzF4D54G5JUD88AlStIvYdpByFOOcyi/dgMNDOzo62t7e1u7tbWfyGw2FJeMezME3iffhnqshcv35dd+/e1dLSks6dO6dr165pYWFBc3NzOnXqlC5duqRer1f2uwB18XhR3qO/VzdopLFg3N/fL/uUOLrCZomMJ7klVK/ifq6M0B/e9dzcXKnwgTB1r+Dh4WEJf6MdEFX6nvMyn8uVqRxj3uloNK4+Ni1KmUm1KPfEOSHriNfGgwE/EieNV5fPnO+eQunkfhrpQch+wEvU4s/+gbo2VcdqWjM45nUc6X96LlJGgqhlpZyMU39UCqor7rk41xnAucZl3iGGOagscgq5nGFU1NjnO/LrCDvFm+nlp0FBj4+PSxW3w8NDLSws6MqVK4UH4XtXaOgLCk9dUi5yFtTT5enm5mZtEi3eZc/hILxVUiWHbNpeqget51DTep7vOH9P0CSva9ofjL1l+EyEBnlaeDTweuYc43vuD++nvgGP0Q/ad+XVr0sZcefOncp16T3lmF7caYJxTswpV6pd/0kPh/OthxryrPBO6q3pHcj54ToUugHgH/lR6C5+LkBm7kBOm7lXDmuqV5wkUqMurNLHBEoZRvRHvkN0OvQ28q5az0ZLlYVgEnFOTsi0hOuuy8++qHFfqk8NBoOiqEpVY8fDY2ZmZoqyjDcihQff+X0ynMJd8qPRqHgy/M/Pp9/uOfFFlgWcBTMRCQ/tmqaSmqiJMz+CptvtamNjQ6PRSOfPny9995hR+uljXKeINSnjdf1KdAWByR8KSZ0bm3fJePqzMp5UKeN91Bm7dX3y//Nz/p/j7P8/iHdaevyoDtms43mXq3XzKJFRX/Bd9kBepUpSxfBwHuN7znevr6SimHjxCIAhFBUHGTwUQ6rP8XKZkYZWriX+fEn+PChL9MufoaWWnFK2u/FXp8ek/E/+dd2B3/0a90DSnrcLP/iazVrX6XRKWJT3B57EYEhDgfNYzz2vi/vWRXf4eGR7Tbqe63FpdDAuCZjX0URjAysF1O2pp56SNLaAsbCb0Jq08OoE8KTrQbf4HSUpa8NnPHJOjqwfTZUn2sHyv3TpkqSxB4fzPOlOGqNlqdhCTchHU5UHV7gl6WMf+5geBeVEq6O6d8QiRB+d2RIxANnOhQTF9s6dOwXdllSYiZ1w8Thsbm7q+PhYN27c0MzMjFZWVrSwsFD2hfBdMhNh517Zb2KTE/kGtZDuI2fsK+FKOa57kDe8GqCFXoLVlexUEj4qpbfG8xoYj729Pb322ms6deqU1tfXtby8rKtXr2p2drbEdtPP7e3t4kXyCmMpiJ2HXcjx3jIhvdfraTAY6O2339be3l5RYs6fP1+UGY/hpU9pAPIONjY2tLi4qKefflozMzMnYvg9FIS2M+zK34XHujYZz6nMTdNoTO8bO2ojy/AgJCKPTAadxJPh/a47j3HMnbzxSGQuFu3SHh4G0Et29KYSDu1yXVMFmiYZyO94ZJDZ3Be0NFHF9Ii45837D03bs+FrTpMi4+hhk4Lsi7mvX478+xymkIUr3r5+On/irYDP9vf31e/3i8dhNBqd2JtnMBhUvAe7u7vld47JIzzfcDisVDVzVHhnZ6fC56Cm8BeeTPck4un09RbP9qOg1EtSkWryUOT1+c6bPBpJzHnmNLyJNzN3/GYcyNv6x3/8x8rnz3zmM5Kkl156SdJYf4FXswpWU0RGVoKDaO/69euSTu6Nk/oN7WaOyLSobp8z10HITeJ54RVfG+oAL1e6nbydNFo6nU6lgiK6BGs3oGWdh2V3d1ej0aiAbKdPny7ru6Sy5w6V4zxUy9dsoky8j3XP4DyZICzj4l4cruE5mgBLqPVstFShJgE4CbFL1D8texiNxQ6E25OXUBaZyL5IwSD7+/vFtb+0tFQYTap6Mlx4olTC4F4dKpVbR/8TrfSjpBMCK5FNRzumiYg7OpoKjS/ag8FAx8fH2tjY0MHBQQk/y5J6viEQAtffaR1y6R6MOhQJAeSbU7oA83twHeEaCF7mA891eHhYKcSQ94XSne0KUdM1+X/T+2o9Gy0lNaGEH+Y6qT4Ep27Bd4DFDWGMM0cz3fj2nAw2Uj06Oiqb99G2gy++r5IbGa7I1PGE87b3Gz52+ZiyIxW6bMfla+vZaKmOMlSW/92jkJ6EBE/9f2+j6ZjrpN8TYND5pk6Jz3WM5yBMKo23uhK6PI8f4TX64dXfcmz8r05/qdP/HBD5qY0NLgQ9IocByzUHnfOxzLGQ8+HzuiZLkfsyyGnpv/LKK5LGFvXv//7vSxrHAmIB0g9Qvq997WuSxp6L3/7t35Y0trAR3Fltyyt81JFPDqk6Kfx7njcrpvDcoJzTpjpDoOkZOMcXAH9vMC1tMvH587roo9FIu7u76vf7JZeB/RF8EXIBkLHBOzs7Gg6HWltb08zM/Trti4uLunLlimZnZ0uSNjkbIO6g2hg4MMXp06eLpwODxlFXEqz93dBX6tK7h4Zz6oyNR0EeR8m74lnp4/7+vt577z3Nzs7q1q1b6na7FQ8DG+SlZ8GVGI8hZ1y63fvVbEjSlqo8d3BwoI2NDQ2HQ/3cz/2c9vf39frrr5fSvLyfubk5LSwsaGFhQU8++aS63W7ZdRWUc2NjoxggLBLMNa+E4YaSC1Z/L4wTR8/JaDKgHhU15UDkQpmylfGm4gv9zERgzvP9S6TxIocszapX5C4hOznixQYtpd23335b0kkPx/nz5yv9pp+0n2tC7vNB+6w1H3zwQaUf9J9xyMo9PE96bNwAngY1KbxpwE6SBchP6SSoQ1gk8svXEKr04Un1kNHNzc1KBb61tTUNBoNSdYpKNn6Ny1ze09LSUkUBox/u7c93LFUrNhJPfnh4WPib3/BEA474uPq64EDFo6LMQ2pSLps8FE0ekDpDsu77zJ9i7lN5LnkdXsg8K3g9ZQvn0S734TzPW/TnRFalt5Dv0avefPPNyv2a9gRKfehReDZA6F2uo7NQNcnfpyvKmYck1YcD+nt0Hk6Z6l5xz3rFzmYAACAASURBVD/iHLx5GPLoThcuXKjMNXiC6z000j+7TsZzcL17RR30c6MB3SLnW51BBJDxMF7G1rPxmNCHUXybjL9UpmHgnLh1aDjue9oHyfaF2PvoDOroNIyRi3BeU4dUdDrjRMqFhYUT4XhO/nzeHgKLMClvm+vSCJgm1Sk3TWg9gmY4HBYAAA/B2bNnK8ITweGx4P4+873XEUYO49rtdkviOF4VkuAODg40NzdXwuYIg3MkljlzdHRUFlYXyo7yJsJT562Y5NmoW1RaT0ZLD0tNgNqkOdRkhDjKCD9hUKPcJ/oIZRgUBkW/3y9AALlmfj1ypW6zM3+eDFVOuZ/tJXiV7UF1noo68KYO/GqpJadc+5NYQ+quk1T4y3UHX1Ny/tXxuF9Lm5zrwJLzOn2iiA1GQc7/Os9EHc/6c0hV3s21v463UqblMWUUIeVNNNHYwGPw6U9/ujTuRMOgWVhQaQk/CLVLcsEnnaySgGX+5JNPShqjbeyT8YlPfKLSHoNDLOOf/umfSpL+6I/+qNIeccE5YNw3a7Yn0pHPmTGOjE/GWmZ8M/HQGY89DXIFrel3JyYmyUx4AVBiqYjAX7oJSQy+e/duyQ3wnTgd5cKIcPc/aA876VK7mlyDXq+nhYWFgrSfPn1as7OzJcyKfILV1VUtLCzozJkzJWxob29Pa2trFTSU5/CkJ5B4PCe7u7u6d+9e8bYkqvGoYolpP9F5J59TidY7ypHKObkfXJMlhp1XvR12eAcZ2dra0nA41Llz50oOzdzcnJ588kltb2/rrbfeUr/fL14h0NmbN2+W3d55vwcHB1pbW9Pc3FzxxLABJGFhVJ/Cw5FuYEdn3Cjmf3ctpwGThsc0lRvaRaYwz5EZKUvz3sisppr6KXOyyhPzlXw1ZN5rr70mSfr+978vaZyjAbqKxwI089atW5LGHgniyl9++WVJ0sc//nFJY5kHapo5N5mjwfMhC9fW1ir94Dx+z0o5Ln+k8TjT7jQp59XDXuP8y3z149HRkXq9ng4PD3XmzJlKiCcopY9jpzOuRENuBMY61abwIDggII3ny6lTp3T69GmtrKyU/QBAT5EH+dxuKHgu2Wg0qhgvo9GoeJR3d3fLcTgcFqQ5dQTXJYbDYQm9fRSGRpPHoumYHp1Jhpb/nsecq7wLIhxyn4r01sFbzIWvfOUrlfZyrKi2CQ/DExzhZSI/0jMBcX/2BXv33XcljT0VmfPBERmB7Jl2VTGItdL5DA+5R2VA/O8V2KTqe/I10deQurXC9VjWVDdc0ClWVlYq1zrPAcTxm8+pBFJzXgFMsLajuzEOOW/536MC6tbMBJD7/b56vZ5u3bpV+lpHrWfjMaG6hfDDKlB1Ho4mAcq9iBMGTQOdcyXQ+8KC4oZLp9MpCx9ufWhhYaFS5nFpaamUeeUeJFhxDoxGXzA+8tmkqoHtSkDWlJ6EpEx7YZyk1LjwSTTfKRdsvAXZ1/z8IEQEpcT39hgOh2XseQ8YCQite/fuqd/vF0GPkpXoK8qUF2nwvzrUs8lrMWmc6sat9XS0lNQ0Z5p4/sPMIfgJGUrxBb8HIFDOf0dLPfcp++GyGHCA6n/wr/fZgRkHoVJhcaPJlSSX91yf4EATcpzP1hTC1tLjTQkYeYitVC1Vm3Ms5X6djuP38WuSv/LPw7m8b17sBmPD7+1Kfp2e9SD9Ip+9KZE8jem6cXWCF9Hv+GuiicbG17/+dUnSiy++KOlklQIeIKs10amspJLUpNhwbEJfsITJIfnSl75UuS/9IfQCFIxqDH/wB38gSfrkJz9Z+R2UDIub+6LY5EsBUaCfmZMBpaAFOeA+GU9Mu9Mmn2BNE4vFwX/Hek+PhCu9vijhCRiNRsWjgZIIEk7bHg84HA4r8Yij0ajkYJw5c0aLi4u6ePFiyUtwlOHo6Ejz8/N67rnnCmJA1RJJZZdO34Cv2+2WsQZhyzkLGsD35J14haW6+c0YTntBTN5KBYe/Oi+ivy9yJxhvPErU80eJ8FKzzq/Med9V3PkD7wf7efBun3/++WKM4Anp9/t6++23S0ULrvfcHEJBdnd3y/t3I9HHxRWYVMI8DtkXHeahv38X8tOm5D83zrm/dNI7ynnkKiCbqBhITf6UycicPP/s2bOV80Ed8fLiNUZG3r59W9IYBeV+mUvCmsHYcX8P8/Hr0vMC6pooLM+TO6SnbE6Zm1WqpkWOFErjeZXykXno8i6VB0dDaY/3gmfCDYFUUFBcXOYsLy8Xzx887nN8cXGxeIzn5uZ06dKlAuBI9997p9Mp4I1X+cODQQ5cp9M5EVrq/Ec/WR8Iq/R34gnkOb4AEzMzMyUcc5qUldJyLjXxbF7/IJDCDTbp5NzkCI9CdRWSpJMeDPcy+XXwDJ+5Dt7kemRO7gOWngra+bd/+7fK9VndKsvPE7HB52l7NgC6zpw5U+Yu1Ol0Ss7S3t6e9vf3debMmUp0hu/P5vzpSrob0vw5X3W73TLX0/ORlRHpl1eDdOObPascNPA1iXPpb64l6GKDwaAAdt3u/Z3Mvd/0K2WUUxpPo9FI29vbJ+RzHbWejceEmjwbk6jOkq3zavhvbnwQBuWbKnp8vVvxhAhkUhrXnTp1SgsLCyV8xktqEhIFeg5zOZpQhzhyf5i86XlciPf7/bLxHGFArsw2ISTTpDoj/UGohC8kCEH6jdBMBcivcyQl37GkimvZkU2P45yZmSk5HITD8T0eDTcUZ2dnixLkKC8Ck+dIwfthDDwXnFyfSBTnPQqjo6X/tynlqiumD/Ju1MmbbBce8/0v0hCuQ0/5342f+fn54pEFIKB8JuGO8Bv8SygKvIbcg995ZvrV9Kz+XHzvv9UBJn6+G2OeMN5SS0kovKnUOwDn8whyPqKd5OH09Llh7GsgSjtrmfMmek4dEJj6Q53Oles0nxOAdF5zSlnTpBfWAW11fUHHe1B440Rjg1i8v/qrv5Ik/eEf/qGkkztzJwLQ5JGA8uWnpZ+/c5/MmQCVw2LHYgbNBnXLqg6/93u/J2mMGoHeZR3r7F/mZPikrPvelTJ/vqb9SbDAv/nNb0qSvvjFL2qalJNs0jm5aPIOsZ4JR4KB+B1lcnNzs4THDIf3d4zOyiXOkB5T70YInpDl5eWKu5GcDBAvr6KT1SCkav6Q35t28bzwu8eAu+fprbfeUq/XK6gQYwCjscB7hY1pKql1qBXIvBtriayCbl66dEnz8/OV3UZBOTAOeF+8VxeG6eL1KmD0q9vtlgobXivc3zPtnz59uiSLS9KVK1c0Go108+bNSv1xlAsUJXI5zp07VzFEcvNAfy8pDP3d5mZldQbZNN9j5mo0vd9cCJh3eAbwJOS+GMg8ECdkX1aASTSTPX6+/OUvSxp7Fsjd+PGPfyzpZNw47bGbMfeD8nl4zqw+xfMhu7M6Vu5UnryesjvvN8nN/9OQ59aAYrpnED7JogZQhg/xLI6uk9twdHSk06dPF9DFS0F7hZ9Op1PkJl6Qa9euaTgc6t69e9rf39f6+noxEGZmZnT27NniHRmNRtra2qrkdvR6vYrx7cmmPDNyz+Wny9MMl3JvOXImc/joA38eTjZt4z/X7TSeHmTcpBH5sNenpwD9IPWPJn0qDTNkS+ZHwUtEqOCtTA9E5lwwt2iP67/97W9Lkt5//31JJ/OyMv+M58rqU02RLz8t4RFCWSYfyCMzMKR9l+004OElvgPYZC7j5fPcCo6j0ajilcb7R6gi+aWMLd5FX7MpmMKeMvCPexfd6E6PigOyzlsJrtLPBC64Pstnw4f7+/saDAba3NzUzs5O8Zw0UevZeEyozrMB5ULnDFmHOPmkdMHqi9FgMKjsjMn5vog6Y6fV765YPvsfCyNxxTCqI/WTFiN+8wovTf2EUQ8PD9Xr9SqJ4TmO2ddpKzdNaIX3IRXqNAYJKeO5XAlIhDKRRuJNHR1y9MbHr85TkiisNI4Vx6gcjUbFOM0+8IdgT9fyJMOgacz8t0SWfF601FKSy5qHBXL8/zreqLuOfDeOnrcAvzlvIdOYw3gHMfLZ5I9zOd8VJzc2/NxEV2dmZsrGZe7p9WeqQ2Xrzkn5VcfbdXKlpZYg5nqdziKNPfIYGzmPUulmDci1gPmIMk4xhn6/r9FoVIBX13HgXTcyfFNcD0Miz3WSsUEfCP8CVPM1v05v82erM2qd15vWPs/HpS+TeHKisUFOxHe+853SaUn66le/Wvmc8cBYvhmrB9Ut7N5eolCJioHicd+sQ537cWDp5k7i3Ae0jOoLPAftewUfP7oF6v1MDw/n8TmreNH/v/7rv5Yk/cmf/Imk/xvPhr+DpgUwr8Xqxa2fnob9/X3dvHmzlCx1V76jHu7VwNr277wPoGZnzpyp7CDucaAsyiyA3uc0hkDwHY2UqgmTXCfdR1q3trb07rvvam9vT0888UQFYaNtnovKSuzUOy0C7W+qnuSLcrc7zsmgOs358+e1sLBQjEHfi6bT6RQUhPAKns/RN0dPXPkHwXR3tefXME5SNVbfiwe4V5R9OEaj8T4r0n2+7fV6lbAs0CYUJJ7DUVdHexwRSgPDhbAvNNMM28hxyDLM6XnIfTM4sqCQ/4VM4b2m99lRLOlkBRy8u1/4whckjWUjlf7wJuP1ph2+f+aZZyRJL7zwQu19Ep0FXQV15TxyRvDMZO3/bBfKmOFpo6Z1hJHucgflIBdqNw6kkyh6epT9/Xc645wIn8t4UvjOZd/+/n4llhtFZ3l5uaL04BWkSpSjtv589Is2UXKGw2HJrUJ5oi/c11FhxqcOvPL1x8/hOv+bJjW9i6SmNdQNybp28p1nO+klzGpI/O77Jnj7KUv4HRmAZyFzS+vy3ryd1MvY8+a//uu/Kv2Ccp8xKPUk9Ljnn39e06SlpSUdHR1pY2OjzDd0iNFoVPJ97t69q9FopGvXrlX4FqqLYPG5Kt0fG9bSzc1N7e/va2NjQ51ORxcuXCiVMqX7XmIvLMOaRGj42bNnK0b/7u6uDg8PK15GX4vpG3rHhQsXCsAJr5C7Qri469KU2GVcfF8758csQgFP9/t97ezsqNfraTAYlPaaqPVsPKb0ILR20u+ujHEe6BjxeyiQkLsepXG4DUpx3tcNgzprPJPfsn/+v392hdXv1YRg08fd3V3t7OyUpDIYOY0TrpmdnS3u2WnTg4xCPvtzEX6wuLioU6dOFSXNN4FDyDkqkiiiHz080I0RV+Szr94nlCGvmOP9QhlCuXHDh3nj/eUv0dOHQUAT+Jh0TkstJfm85HOTDM3fEtzAaKnz3Hk4g1RNKk6QxT0fhBb6vZ0XPQwRxNLv77LcFZP5+fnCc24AJNiTffLn4vc85v9ucGQ1wJZagjzXgjnjBr7PdTzkTfONz5Iqegk6jwMKvo5JKnoKyjsRBYQYo8SjY/T7/YpeNBgMKpta0ocEGjE2AGy9eAN9pD/Zf4/qcOO4ySD2MSXJ3nPAJoE7EzUhFv5Lly5JGu9T8c4770iS/viP//h+I4YWSs25HFDmdKTnIuOBszZ8Vn9Ky57PmRuR53G9x7tKJxGFppwTKPvP76CMuU8IRB1t4q7/7M/+TNJJRGCalGEw6UFIBktjwBFfdwGy6Ny+fbvshcD1KOtubYP8+zn0ifYSSYBBiW+UVCY7iykVFtzD4cIAxNfdlj4OjiSCru/t7emNN97Q5uZmqZ6yvb2thYWFssunhxFJ9+NX3a05LaJ9PAk8R4ZTUFUKA3BtbU0rKyu6fPmylpeX1ev1dHR0pM3NTQ2Hw4oXAQTEDTCqRCH4EEgZM4/wJok+Y7G91jjvxI0V4j7TkwECTvUc0CA3Bnd3d4uAZrHw/A0QHPrrc4t54guLv8+6ZMGPQilbmvbV4Hsq5XmxBf+MZwBZ7QurH7NyDe0gm3LXYdBHZNXly5cr52c/eE+JsmY/8vmzKlXW/s81JdecRJHTy/yoyMOKJFXmeyKQ3l/mX8od5hxeUzw6EF4H9yA4TyH3kA88P+8dOeu7TnP+4eGhdnZ2tL+/X0I30iigj9z33LlzWlxcLHHovv8SnlLu4x6TOqPDv6efGSuOZ4cqd9OknCtN4EIi3ukR4JgeDY6J/jK26Avp2cj1I/uZ+5pldSm8hOhDWe0yIzfgxaYqV+zBQz4XyD1eSNrPPW7S+8o4EEEzLcIrj47B3k+MK/rCvXv3NDMzo93d3eJh8AgBqWpgsEawPgIootOQZ3bx4kXNzMzo0qVLWlxc1NmzZzU/P6/V1VV1Op2yR9ja2lpR1FnzfHwAK59++mnNzs5Wijk44Mp6jGfxzp07ZRNP+sZ+Vh7iTTvwmuvDLsPcyzEcDot8uH37tjY2NsrePXgym6j1bDyGNAl1azpfavYw+KQfDAZF2EljwcUiI1WFpaMIdW27Re/KnyuJMAFhQ76wO7P4s+T3fnQlgAWu1+up1+tVFmuMDH8eV05R0qdpbDwIffGwIYwBcmek+0rj6dOnK+Esx8fHJdGb8cPY4r00GTX5Hmmv3++fUBq4n4+7e5iksbGSSokbrzwfxgO/eyKbK3UuKLO/dcirI8Af1kPS0uNHLltynrjszO+aznEZ6IoP57pM9LDGRP+51uez8xO/wW/553zkgJobGwsLC0UJckSUvqZx2cR3+V2dd9K9J1TnaqmlJOYF4Fka+pIK8k+YoVdX83M4+rx2UBldgfnv3gxKSMMrDrB69AfrlgNq8NjMzIxWV1cre4Txu4PG9KfOA+IhjdlP9/ikPlgny1xWeIhmjk0dTTQ2UhEjnvfNN9+UJP3v//6vpHF8LyiMV8mo60Ba+k2UORY8VFryruz67+lRyF1rOT/rSWfMIcpzUyxnIh3005Vu7y8WPkgDVR1ABrLq1jSpTsgnIyXT+XvESqZEItbxwcGBBoOBNjY2tL+/X3bFZKdonp1qJdkH2qfdhYUFdbvdUknCN+7zSlheeWJ2drbscLy+vq6Dg4NK/WdHoDz22ZnVlfWjoyPduHFD9+7d0/vvv1/Zfdh3smYzQcbGn23aCyJIoVdxcQOHndRxp964cUMHBwean5/X2bNnde3aNS0tLekHP/hBJcRqfX29vBsQIEeHMT5cQHW73YJ841Gg2g2GmaNJtDk7O6vl5eVi1KQx46gn7wYjCISH6jkIPa+bzp8vBM7jfp/M7/Lx9DCTFKoflZAZKWOyUgvf5+69iU6mjEovalbe43dkkBtn0nieuSdQGueXIVuzX8g2+B1UFe9t5vOxZuT9WHgzby4pPRj0M73D6UGaFqWh4DIV2VJnaNDfVGiyvUzSzLAN0Eu8G+6pcLnt85jrnMdQfHq9XgXRxMO8srJSya1jDaC6XXqa+N/3w8D4QO6mAeRGBmG3tONgBh6Yab/LOi+TU51R6M+c63+Sv9O663KtSNmZ3r30QPA9vJMRFln5DW8p52fOLecjK27duiVJ+tGPflQ5L/UqKPuXObHoOdeuXdM0aXFxsSD1vi4wfszh7e3t4vlgLfNncA++A2Tu8YNPWBtHo/uJ4fPz81pZWSmb1x4cHKjf72t+fl7nzp1Tt9stXmLWI6/25Wsj48wa5GuXe1KpkrmxsaHhcKhz586VdR+dwAFi1vJJIY1p6KNTbW9vl7ByL84zKY+q9Ww8JlRntaeiD9VZp66ApQeCCetWOkzI51xgvAqEJ4W75d/pdCoJoKPRqLjsnGm4VhrnhiBcco8PiAWM/3lGFxxUgtje3i7ohxsTLNK5qHO/DFObBqEMe3hbepl4HySAHx8fa3V1VWfPntXS0lIJRfJ3iWBjvEj6dGXHz3chSx9QZtwtn8aXh364K9iTXXH9OvFc9MHr/vsO9Z3OeHMk+oWShHHsiqDfV6qGWLpCOAmxaenxpQehn37Og673duqAoDoEMcPUXMYmKokSj6xGRiDH3CuI8Q8iSrw5Rg2ggcts71vK1gwnrfPYpPFRZ2xwrgNYLbXkBA9gEGPUMI88BJg1Ej6Qqnzmc69pDXAlHoLXvB10F0IOXXfx+e3GhnQyB8W9e653+XN0Op3iWfHoE++v97VJRjkA4EYOho7rCO5lqaOJxgYdQHkAtcLi/e53vytJ+uxnPyvppAD0l+yUwimFhiu2db8nqkd1hLwP989YShQM2kl0k+fOHI5MSEtPR6JVqTBhoebzfe9736v0K9HNaZNPijQanPKzh69AtIOSyTV4JLzGtTQu54Yx4bWviZukBnW6Bf33paUlzc3NaXV1tVSpkqSf/OQnOjg40Pb2dnlvMN7c3JyeeOKJCoLlSY2uxNL3n/zkJ6WqhYcVoXD3er0K0p7GyLSTGEEKudfKyorm5+e1vLxcmH5vb09ra2saDAbq9Xo6ffq0Pv/5z+uZZ57RxYsXNRqNSvgDKBACa3Nzs7KDLLkf/m5RMNwliycIYXf16tUi6OkTKI3Hj1KmEKOIfiwsLOjw8FCDwaD0V1IxRFBWdnZ2tLOzUyp9nD17VouLi7p69apmZmYq4Vx4vnwPDQgDjWo8CFJygohhnxYxBxNNzTjvzMloitdGJhM3/NRTT5XxksaeBfbLoJpUehQ4H48g8xeZlFWyOB8PF+hlIt2gdqB56a3OtSaV6MzHa/Jq59oDPaw3/cOSyw5vP0McoHze/M4XdkklPApDGvnDvji7u7vl3qCpzG/64+AORgaK1tbWVqWYB7x84cKFEpueceLOr/BHXdKq8xf39zXSFRj4LXM7GGP3dLhncpqUnoam9TAjGaAmfSDby+tSFuTcp1/wJB52vIbwFr/Dy+g36alAlqTngT1yyM9yT4Akvf7665LGsoRcC/qR+WZ4TXle1uh8/mlHcvg4z87O6sqVK6ValJdpHQwG6nQ6unPnjobDoa5cuVLxCI9G40IN2a7rAkRhrK6uFt1jb29POzs7mpmZKVEP5IjcvXu3koPhYc0ArBgbzrPkQcIbdYDBaDQqFUDdo+F6CwYI8sTBYM5x+cSYEXbGjuEAn6zxdWWEnVrPxmNMdd4O6eGr7tQhU5C7Gf2Y6BxWuKPzWMswCedy/szM/fJuoM8w+2g0KgIFJZXrjo/v76DLztOOpPkiiMB35nI3cZ1r3Y0WXyjqEMaPSh7qhMBygQLqwOaGCwsLOn36tJ544gldvHixuHXZb8MNBqmaY8NzSCd3X/U//462cCW7ssQY+vgg2FyAuneMa1wQMkcwBrwqBjkyoLAoMIRVec6Ie9MwejwPBDTq6OiotNVSSw+ilHEP69lwjwbXet5Gfu9yFTDAjR+/d/ItfAavQZ6L4Z6LREOdj/0775uTo8r0J70ZTd4NN0imLU9b+tkkn4voB3j0JVXW6sFgUOS86yNSfU5R8hb3oaiDr2m+zpBnhBFDhUh+d12FPnqIYxobnm9J//A4zs3NVbwnOSaTni+/d/4jud4NC0Bhjxapo4nGBi8DC5UYPz7fuHFDkvTGG29Ikj7xiU9UrmvyaPiDOuX5npTmhAX/rW99S5L09ttvSxpb6E8++aSkMYq2trZW6S+W++c+9zlJYwud+2W8M9/nbr+JYPhErns+noP7X79+XZL07rvvShrXquf5HiXRN48BTmU6J6Qrlo60Ofni6AtEbvrCpPddNWEUrHtvj3dCAvHdu3dLvOXc3FwZU5C+1dVVzczMFIbc3d0t7wdPCGE4kioC6fDwUO+//35pm1Kxvthh9IDUE/fprtEmj9FHoeeee04zMzOl/ygM5Eq4sTU/P69PfepTunLlin75l3+57A0iqVS4eOONN3R4eKhTp05pOBwWBH9nZ6fi4VheXi7jhdDBO8W4D4fDUuubzcPoH8KccfTa/8xDF2y0zQ7vCHNkD9U2bt++rV6vp7t37+rw8FBra2vl3qdOndLp06d1fHxcfn/nnXfKfCJ35NSpU3r66ae1tLSka9eulfmBUQOSO00lJ1HARMYT3eT3rJnP76CW5GQgS1IWgYrSHqhi7tgNr4COUnkGL3IaoY5Ce/t4w8n3ox/IZp4DSiM3PRJ5TE8H7WfuRnpupkXO277Y+lxxJd0Vcfdg1HleXHa4EuPPBQBDrX/QXPLayHVyI8NzHghJXV5eLvzlnmXkXSbQslbs7e2p2x1vKOb9Runx691D6X9eGh2FyuWFl8h2xWyalLyYXsQmMK7O8HJqitxIXmHu5w7geCPZ6wb9wD3NTshteBneZk3liDeSfnA+n/GIwPs//vGPK/cj1wPC60m/M+eD9ng+9DS8odMi3pN75YbD+xUXySs8Pj4u3vX19fWy5hE94fOLdZ+2PW/RFW50GMKZUMp7vZ46nU7ZMXx7e1szMzMFjFtZWanMNQc8h8PxBpt4E30zPzdqzpw5U6IVAAv4S6+MAxhu0Pvcdi8oe/Cwr8bOzk7hWfcu/9RhVC09nuRKoH/XZN07UgwCBiP4ObTD9Y5WY4U3MQbfT0Lq/P7pcTg4OFC32y0LsisnjnBzTb/f12AwKKh3hoV4ojxKP9/7ePjzTINwgbrxhTDgfxcC58+f14ULF0oiO+O3srJSEtY87toFB0pAIqY+JzIEkbFyQe1/5L6QoAr5+dzbwQb3fEljY4P3hGCnHUpjMg9Qxjw0ACSK/vBZOhmHO23FpqWfDaoDzupQwjoQx39vui4VAG8DT4Gfh4LiXuL0jEjjYhjILy/M4SUsmzzTKQ+9D+mtqXvWOu9oE5rqYY08e3paWmqpjtxwZM30vWHQVTBsDw4OKiGeDpw26UQQXgsMbA/BQq/gf+4tNQO+viYCEnjYlPMLhpVXnso2/Y/7NgEkzoce2ujhjDyv63GT1smJxgaDgwXblOPA/htYqrlDZVanqkPD/fus+8z3WM54NF599dXK9/T3Bz/4gaSxxwA0DwEKMvC3f/u3kqSvfOUrku6jvdJJizwrm0A5KSC3dr1/tMNkJlcDj1He91FRKozSeMLXJfvxu1uu5E7wGQYDRQDpdqXciXAgEGt2BgcdQAFGcfWY4uFwWNAHysKxjDLfqAAAIABJREFUSK6srOj4+P5+Hyiix8f3dy4HySYHwRnF0aBer6f33ntP/X6/5HogkNgo54knnijXHB0d6e7duyVUazgclipceEymSUdHR1pbW9Ph4WFB7AkT8pKwp0+f1ssvv6xLly5pOBwWBX1ubk4f//jHdeHCBX33u98t70uSzp8/r9FoVDwcGGi8y/TYgGyivGO8eM6D52J4IqobqggueB+ElaOHcxwfH2tra0vb29v64IMP1O/3KzuGHx4e6vr16wXV8gWDfTxoC88GXrbd3V3Nz88XFAqvildRmwblTt4p6xKpTzQx0VK+R5bg4cBzAP/hxWWHb/g3d+LmPXCk/UQzWQuI2waFzfhyjqCk3JfqcaChicK6gunj5AueE8/Jc4Oi0p98zo9KHnbnCH6GDWV/08Co89jQXoIvyECeD9k0Go2KBxd+AlyhPX5jLef+nOMeSQ8/rDNW6jwvbpy7se4y1NFajKOMQ2et8RK89B9POOvotChzJ3Ldz5xOXzv9c0Y8pL6TekNW14THqPYJT5Hz8Gu/9muVz/DUs88+W7me/dDwct6+fVvSOMIDnkwvYMqM//zP/5Qkvf/++5LGPMv1qRfSDuNFLkh6Ifker+e0KNdbLxLDfhfMJYyAbrdb8Xg4mMrvTng2fG6yBp49e7aE3hL+5FEerpjPzMxoeXm55DeSwwFwcHBwoA8++KCAZc4jkkqUw8LCgi5evKiFhYXisZHGoKHzoqSy1jaFPiF70C0Gg4F2dnbKniSMI20D2E3KUW09G48JPayiVIdG+W8Pis3LRcgXHf+N/3ElEo6EsgrjeUlHPAWSyuZ9JETyR9iMb9DjXgdf/FGgfQHv9/tFeaX6FIvz8fFxRfmEyRKlY5FEoE3T2PB9KA4PDwuDp2cFQw2Xqj+zdF9JWVpa0vnz50vYFd4ER3Q8DO74+LgSY+rvGSWA711JTOEqVTeFc2MD8vmQ7fLcCGCPH/WQBPfGuPLGc83MzBTjkQWaCh4kvWL8NiHMLT3eVBdS44p5ytMm+eoeAT+37j7pXc75LalR/qTH1Y0dRyY9+bQOHHTjiHvXgW91odN18rLpLz0eLkfS0GyppSTnRV+rAOb8N0qou9eDNpI/kse9fXSWxcXFSv4FPOk8OKkkrYNi/J5hhL7GA4x6dIePQ46J6wO+5vszY9xgeHgECQUkIPfk1NFDGRuJckG8DFApYv5eeuklSeOqBY7W+EvJl5VKWYZ24KnAYudF/vu//7sk6Xd+53ckSb/wC78gaWzBuwCVxrkSKBh4SK5cuSLppGch0UUo0bbsd7qcseR/+MMfSpK+/vWvV35PhOBRUQru7LMjcolG4H3Z3d0tyccwqu/0vrCwUJi20xlXgyK3YmdnR7Ozs7p27VoxLBzpoRoKHgtyOzAuEmkGLe/3+6UqkS+IKKb37t2rIIEotSD4b7/9dtkjAmE0Ozurp556qggCPBl7e3va3Nws95bGtbCT2adFzF0oeYeqWQgfzvE5jJG0uLio3/qt39IHH3ygr33ta9ra2irnYVxlYjXX+2ZFXhXL91FgvBCaGGvSuLRsomAuLLmfo6fb29vq9/u6ceOGer2e+v1+JYG/zqjlWpQvCIHZ6XS0vr5e2qZa1tzcnM6cOVPJ8ZkWZc5GypimyneMRe4hlPtRgOTzPkArc5diZPR7771XaR+0ER7ifGQU37uR5kdkL79nbf/cX4N+49ngc+Za5PjQX+7LuBIHnrI0PTPToJSjgCN1Cnii4HVriHs6XC6jsHjYAvOa90oS6cbGRgmjdGMdkMQNFfcogsKS0+GykOvhaT96rporQx7S6gqM52rgefbvAT/S4KBd9lyaJmWltfSqJc/mO8rzM68p9YLUS9CfvvOd70ga5zV98YtflDTmYfQUroNX2a8C3sN7SWU6eASZgacCWUG78Di89z//8z+V5+O+WRU09ytjvOhf8i7P86j1ndTbcj6hNN+7d08HBwe6cOGCjo+PSxUp1lQHPqVqeG8+t+cXutfOvZ5SVb8i1JvfKfKCkYIOBZAKL6Ib5Q7oTeAY/ci8Sch5FBnQ7/e1u7ur/f39MhbkeXrFwklrZOvZaOlDkVu47mZ2N6E0RthZxGAgD6ljcUrjIRFtjBc3PnB1ukHjiD6MDnPPzMyUzeZAvX3BpYLTYDAopedy0aSvuWjCvL741yGR0yB3CfvRlRmMHVe8nRBG3W5X58+f13B4fwOg0WhUPCepzKDk8DyMn8dSuwDnPoRPECdLGwj43HDTDX33hKCUkKRGSFida9oXgEnkoAcuasLteDbC4h4kSFt6PCkXdpSFSd7huuPDkiP+8Cf94P54OIbDYUWhaJJJ6bHmOhQdV6rTI5J8l54IbzPzL9yQ8FytSZ4NH/fW09jSw1Cd1zB/l1TJfSSaoc6D8TDtO6/we4Zcwj8eISBV86S8LUrM8h1rqOdd+hrtPJIeGe9H3t/51XNZvN+0Sf+zjTqaaGxg8YJGo8CR2/Cxj32sdE4ao1KJOhHbRzsZj+wP6dfn9xsbG5LGaNzf/M3fSBp7OohJ/NKXviRpHPOIZU+MI/uDYMk/+//HOlJV6zOf+UzleXLXWxRmFL/cTySRCxATkIK//Mu/lDSOxaS9RAamST7Z0h3IPevchIlQ7ezs6ODgQKurqwWt7nQ6ZZdOr2ThC8z29nbZF8OTsbe3t3V0dKRnnnmm4uFA2ev3+4WhDg8P1e/3i6cEY8bdgTDm8vJyad9ReRgfpRJv0zvvvFP2bDg8PCzGD+Oxs7Nzggl5t91utyjpiepkFZePSomu8W55Lu6PYbayslLZGRWhwPu+dOmSVlZW9MUvflG3b9/WN77xDfX7/UrZWEddiAdFmen1eiW5mvhx3m232y2VNhCUnmuAQOc9gqSAbHY6nTKPNjc3tb29rddff129Xk+bm5uVUoX+bn0uJGKcqBTvB2NqOBwWBIv2Z2dnS3+mRciCROabkHtkELIUz0ZWruH7zHujndxXA5kOSknFG/LeIGQYY0s/uQ/tEj+OB4PzQDeRccSR04/MpaA97nfp0iVJJyv14cHAY8Eak3HqKaunSZ1Op8KXzLNUVJoUF1/40wOS748iCPAj89k9mlTWIadrbm5Oly5dqvAH403eHUUWyNWiPzyHHx2sIHeP95yJt+TuUb3GUV6XL/B9ekD8PKma8zJNucoY+j3qQJq6d/Kw63WuuR5iI415iXwrqkUxp8mLSo8OvIUMoD+5wzeEF5M58Nxzz1XagUdSr6I9+pWeGa5DBsCbmfvKfV588cVKPx4FuUHuHgU3kPkN8Or27dtaWVkpaxehQyj3aejWteVzMyNl6tYRX8M8LBGPhl+XHjJ/pjoQwNdGByI8Dzavh2d3d3e1trZWqca4v79f0WvwcMLfTdR6Nh4TSkGXFm+e2+SGy+pCtJlIVN4TVDsVf2LmXUn2HAoWUUe7XAD4wki7bsz4QpX9k1QsdhirbuFw5M374Dkl/vxcM+3FkDGtM9Lz/Xr/6GNex3jNz8/riSee0HA41JkzZ2orXjDejC/v8eDgoBg2GGmJ6IxG42Q8n1s5v1wZ8vseHx+r1+uVv93d3dI/D5d6EHrr7yYFa352l3e3W93UsKWWnFyWuiGf8wyqkwt1RsiD5G+dMZN8QAgG8xfeRM5lOynX4S343gtBEMrk+RNeEn04rCZ+8+de8PR2JOqacrWllh5ETboLlL+l8bezs1MAVL+GOevFPeC5bOtB61wCu6ytHgblhoPUHMWQOksaVP59gh5+/1z/2L/K8x+5b533ti6KwmmisYHF+8wzz0gaVzJJdBWE/rXXXpM0trzxfKQFjPWDhQwRs+hVfqQxKnXr1i1J0j//8z9LGsccgnpBoHL/8i//Uvmdgfi5n/s5SWOLHU/JT37yE0ljC53rEnnIqg1MygxtyUnwjW98Q9K4ugPXN+1LMk3yCesMklZw5m4k87CYUBYWZQxisnpiIdeiNLLj9nA41NraWqlGtbKyoieffFILCwu6fPmyut1uQdocPcGFCLLmz0Z9eN/czuPC8b6wA+8777yj/f193bp1q6DbMzMzlXfqTOVhX6AO7v730q4sro9ioUzFOd+bezaWl5fLJof0E+UAxP7555/X5cuXtbOzo9u3b+ub3/xmQco6nU5BMzY3NyvCiYS11dVVLS0t6fnnn9fS0pJWVlY0Go1K9QpHhFxYUrXLE9G9n3t7e9rd3dVrr72m7e3tslM4/SKWPKv85Fh5uJf3wRPByQ2SVHZIxWM1qcrGT0PIFGQAiD2yzxc+aYwG4p3l/Nzll/PT88H8SNnF83Kk0gyykCMyHr5gnLO/HEE1OYLOsjbkQsX40m+e7/nnn5c0lr3pJacfXA9aSj+Q7dDVq1c1bUrlxuf5g4wMz+1Knq5TijCy8SZyPwgFBDmJRwE5u7S0pLm5OS0tLRXQRlJJZvVqcbRNxRxy7ghj5ejVr8i9QA6Re+FV6VxJQk76JqzOr270u1H3IMXmp6EEyZpkyiSgro4SCErllncAD/76r/+6pHEu6T/8wz9IGs9t+pU7fcMbuV8HXkb0M3IwPvWpT1Xum4R+xfW5dw+8zPXcn711kFW+1kjj3BL0K76fJjlPNoFa+Ye+sL29rVOnTmlhYUGrq6u6evWqZmdnyz4cvmdQHdCXICvk3ko3/JnLi4uLlRwnQNgESeuMDNcBmgBT91amAe9A6mAwUL/f182bNwv/SmMgwgFNSYWnl5aWJnqOW8/GY0Rp0f601zpSlRauK5LptkSZrEveHQwGJa8CBYYwm3T7O2P6oo7RIqmUsSPu2A0jnoUNftivYTQalVKpdeOVyrKjEV6xaHZ2tiD/00blJqGi+b8jkolc+ELa6XRKwtfly5c1Go3KwrK7u1sEnwtPF9IYWTMzM8U7xE7l0thA89hO3ofvfupIJkbhYDDQ7u6utre3tbu7W+ZcJrk2oUc+LnXvMsfKlR0v1esVr1pqCapTNpp4Pr0d/n2inE33gj+QL1yfv2ebBwcHFe8y4Z+cmwUbIHiXfXowMrxaoFeKQ76ilOCZbMrJYB3hfJfPdZ7oujFrqaU6ajISIZRyn28YxsPhsBhZly9frqzl7r3L+9StQw8iNzhSl0owog4Aflg+qLs+x8M9GlnpEV6lX2m0scY30URj41d+5VckjS1W0C1yJzjyO59/6Zd+SdLYM+IPK41j/+gY8bZY5KBrtEfuBKgalvAv/uIvShqjVRm3++Uvf7ly/09+8pOSxkhCejbI4YC4z7lz5ySdRCiIc+Y8ngeLn8+0/8orr1Suc/TIPz8KD0dTHowzTSpv/rwZQ45yTvUeFguSoHZ2diqGCGNBLDH7apw7d07D4VA3b94sCiq/LS8v66mnnip7eIxGo5K8df78+WKc4K5fWFjQyy+/rKOjI33729/W7u6utra2dHh4WIyTs2fP6uDgQLdu3dLm5qZeffVVHRwclGpKHsbllJ4f3zCu0+kUJJEYaJDCaSPi/u7qKAWIh1FJVSQcxWM4HJZcmE9/+tN64YUXNDs7q9u3b+uVV17Rzs6OLl68WOaKI46ZkPruu+9qZmZG586d0+Lioi5dulTCLTxpzpFQqVrZ5PDwULdu3dLu7q5u3bqlwWBQjMfc9JHxYK75e6tTSDjPq1zx/mZmZio5QDzfo1BsEtVD9iE7MAYz/hrZSMUZ4rhB2zDK2ekb2ZVe5oxLxwPBXkkf//jHJY1lPvdDZqZspt/cj35npZnMneC56ffa2pqkMfpJRR7y73L3Y+6D7CYH680335Q0XjPIMyQfb5rkxoUr0VKz8pFKBfzQZKh4+5KKMY6HletQlBwdxWiH32ZnZ0uBD//rdDolDBKlAoT3woULpdqcK0e5QznFQ3JPAI54ODB+kN++T440RkvrEFj4ctqlb+s8SR+GmhStVEpdyfX78hleRl9h5+7XX39d0knPAce8P+2RA0IlQwxKvmccifhAj/rRj34kaayXZE4pPAzv443Ea5r9aKryOe19xRIYS57zuerGP3OXykvf//73tbq6qjNnzpSqhJ1OR9vb2wVUlFQJDffn4n2krHVdIhX31L2cB9Kjkbkb6elIHS4LMyTwR+nfGzdulBwu5MjR0ZG2trYKkCidrLL1oH1vWs/GY0YPUpqarF7/jcnnZdlYTJoUM1dUWWykakgeiw8byHQ6nYpHQxrXdffN6xydc+ucBTmRAcK5KHV6cHBQUX5coU1DzNvx31h86ROJmokUToPqkFH/3xEH758rBdkW44rygpKHUZeCmzYQ7C6oXenp9XolpMvzabhub2+v0n+qgeDJ6PV6JYRLOhnW1jQ+TeNS59Hw98p8TiVn2iEbLf1sUMqWOo+uU51C60rGw97LE7VTEamLp+Z35rfLRtpz5YM/cjQ4ej5cnUclczD86H9uhLhM8mMTkvuw49XS40keMdFkuKeMT0/acDgsOggejvPnz6vTOVn6dhKP1/0/ySvv/X0Yj0nqKXXPWsdT3n/XydC/AAGQDRlClTqGdLLEcNJD7SBO7B5H4oKzKgGWzbe//W1J45hAPA08dMbnYqlzPS8XRQ3Pyd/93d9V2kVhxWMAevXVr35VkvSv//qvlftxBCH4/Oc/X2nnn/7pnyr9zUonDCbPjccCRZX+J5r2F3/xF5LGSAIhKqCCiXw8CvJN1NxTIZ2MMXSrvalPIFHUc3ckD9e7T1C3+o+OjnTnzh3Nzc3p/Pnz5QjSvrOzo7feeksLCwtaX18v4T3kUrB75alTp0rOBkjFq6++qp2dHV2/fr3sbk0S9PHxsdbX13Xv3j29+uqrpYb8zMyMtre3SygUOSWeuM4YeQ6GI+S+CHa73ZKz4DGb0yD60fR+UAZAKVjMvUa4Cw1H0PFISdLLL7+sa9euqdvt6oMPPtC3vvUtDQaDEvN95syZSkIbXqn19XUNBgOtra1pe3u7IOPUAs/Sti54j4+Ptbm5qb29vZKbUYfC1An4FKCOHnG9kxtXMzMzZT5jJLlx5qjVtAgUkb2DsgZ+VnDhefEcQLm/Bv3E44FHAJlDe8i4zI/jd2QynhRq9iPzsuoV+XRUuqEiTe4Eznuif8heno9x+fmf/3lJ47h1ZCeeF9rBI4YnhrXDQyAk6bOf/WzluadJqWzXoY+QKwkuHwApPJzIz8/7kbOBAVCnRCQfuNLPPchVgofJwXBjw729HqftRCIp+3y40e4eDfiL8EQ8IP5sbhSlVwPjJb2Y0yD64eXapZPIdJ3M9fPz9/w+Ixe4nneNLIAH0SPwLuJVzCqYqYfBA+mBQIZkHhR6DjIFXobHci7Sb87PfkM8N3MdHkb2THufjaWlpUaj1fmU9QWCLxYWFnR0dKR79+5pf39f3/ve93T+/Hl9/vOf1+nTp0tbXqHRQxKd5/z50RW8gIpUnVcuQ3K8UyakIZMelcw1rgPshsNhAWZv376tvb29kucFIMBnv0/y4sPwY+vZeEwola8mSmu6ybp2a9iRbxbNOjeek5cclaobwrDo+uJ1+vTpEjZFicajo6PKfhsHBwdaX1/Xzs6OBoOBjo6OykKJ4ri1taXNzU31+/2SH9LpjEugHhwclHPrxirHI/9SiX9Ung3/3xdlf1ZHGLPSBc/nAsIRTjwclMy8dOlScZN7XLUbWCgwCF48TJznVWvou4cz4aolUc+V/ofN0Wj6PceNzz5eiW75uNYJ/5ZaSqAolX4/r8lwaGq3ab7VeTbyvvB1k7z3nAqpGsZXJ+PyHhkmRny3eyxcuXPlxRWVOk9M3iufvfVstDSJ6nJCpWbvI7/59fyNRiP1er2Sj8jayDVuHNQBBA5Y+XdpbPh3nhfp/fbrH0aHS09GyhTGAWByMBgUUMC9j/Cq62dNPDqJJydqQlwISpXVoiDijp999llJY8Q+44Y5LzPWs5pTom3EMH7uc5+TNN4xPKteUS2L64j7zepXnE8/eLmgf1npJC3UrMbgypI0tuxBFr71rW9JOpmbkjt1Por4fogx9Ylat6Cl0trkdYHRqNBA2yjvWPAwLDtR534IeEZ2d3e1uLioJ598UouLizp37lyJoaeKFNd1u11tbGyU8BxpXPuZBGWYA6RuNLpfGel73/te2fU7hcZwOCw7UrODdFZRaqqzzff08/Tp0yXvZNoInL+3ZHru5QKEnBQ3OIi3dGPNhRQu5JdeeknPP/+8nnvuOW1sbOiVV17R+vq6rl+/ruPj44LqkKvCDu6EaeDhunfvXonlHo1GlXK8jgRDLvCZS3Xj7WPCb/l9XYgJqDD9IKadOeGolytl0yK8pnh+eG/IHp4BFDPnEB4HZAyyFoIvkNlUwMMrTHvIvJRhicxnPDJHPB25/weGKTIvUWOuZ20g747n+u///m9JY3SV75HhzAc8F3y+efOmpDG6y/OCprI2TZOcHxOthOrm9yRlOo3sRP+Rd8iXNLTdaEB+prynHa+yeHBwUBLAqVoFKATvIEvd4IHfOeKxQBnjfA/NwAjx8XIDxBW4pEfp2ch32aRAPej3fOfJO24QSie9mMx18pH4Hv0jnz9zaZOn04NAP9LDgAcF3s19wuCtNDzRe5AVeX1GjCA7pr0TPGsRG7JmyF56NKRq3tTx8f0Ki2fPni1z9+joSD/84Q+1urqq559/vuSASvdlXafTKWssa0oq55JK9AYRF+5dB6j1ym30p06xr3sGabx2eXI3MoH7slYfHBxoY2OjRCPgcTw6OiqRH/Ah4+p7bgAs8v/E9/Kh32RL/09SKqj+3SQULZHz/N5dfiit3hYM7HG+aV2zmHW73eKaJBE8XYAwMWi9GxvD4bCSC4IAkVTi/3u9Xklqd2PKn4d+uHHkuQb+DPTHk8X4HuNrmiE4dQpDEu/B0UT3Kjja0oTCMr6Li4sajUZlp3FJxbMEkuljwhh5FTGMDPqV4SZubOR7awqbqjMqmsYrr/OcnzrU1s+dhCC11NKHnR8uc5vkbgIZdQT/wj8p27nWjfY6NJc/jAkUEYAieNV/x/h2Hibh26u3ZchK5ogk6spz1SG5k8aipZYeRE3gXN0a6OtYp9OplJuen5/X/v5+xcMBL6KjOKCa7Wafkke5XqrugZN9dspzco3NNTRzNAijAtjNvKsE2112fRj5N9HYyIZyl1sUK+JlqfTBLrTsyP2JT3yi0l5ayFDGSuIJodoC1af+/u//XtLYgv/N3/xNSWPUiuoN9BdPBpVKsKzpn8epS2MLHcs7PR6J/jHpUDh5rtyFl+/pp8cVSieTb6dJjC0LFFZzHSpHH5wB6pAm2kGp5j7OrFQrYRHLnY0hFqPr168XjwSlbL2eO6jB7u5uxZNC/1GqFxcXNT8/r6WlJfX7ff3Hf/yH1tfXC8KWDO1xt4eHh6XiBMbP6upqZcM6jA9ippkb7DTuzz/tUKpkeieMOtBMwsXm5+c1Pz9fjDGUBAQnz8W8SGTm8PBQg8FAN2/e1M2bN8u+JPDYwcGBTp8+rV/91V/VysqKlpaWdHR0pFu3bhXDDaVkf39f7733XmUnUjdEEeSJ6EJ1ylgqLc5LrpBRcezs2bPFE+VC3TfvexToKUTFFmQLOQyMZ+ZmZC16+sj56clAhjEeyESuA3UENeVztpeV8+gH53O/3Dm8ad8MxpO9m8jNQC5897vflTSuJoXsx7tN//F04NnAA0K/2ZMA7zb9ZLynTR8WBc98AAhlHl7Iec3/8BPzdWlpSZ1Op6COyGXaR046L9IPQABQV8/v2t3d1ezsrPb29oo847zhcFjq8NcZG66Q5GZ+/ruPRx1aWwcs0NY0KffRqgvDoU/+e8rgOtDCj00KWq6N8B6f0TPSi+j5mP4Z/SVzKhi3zJVlrlCFivvQD67PalP0i3ZzbnP89Kc/Xblf7ng+LXLvac6xVMKdF9yT7tcuLS1VUP2bN29qcXFRFy9eLOuqdF/GdbvdsicN60edjuKeDe6JkbK0tFQJc8TYoQ3OY11jDfM8KM4jX9XDwkgAv3v3rvb390voMuAke0zlePjeaTxLyqhJgEDr2XhMKBXUOoHXhHpBdRNpUm6DC2dchHznSUp+DUyWbnk3WJxxub/v4+D13w8PD7W3t6fNzU1tb29PRAH96Ising5JJxbRumQvjBcXJo+CElloWpDrBEcu8nWoC897cHCg27dvl3LBvV6veDVcEDHmi4uLeuqppySNQzM4j30z7ty5U0kA555ucPh4NqE6D0JX0pDB2HDj0udUnt9SSw9DD5orD1I4J3k68nw/B/7me1eUnOr4vw60QNZixKDoIJv9Hhg7KETu0ciE0szRcAUzebiJnxMMexSgXEs/G9Sk40iq8ABGNd/XHX0Nr9MJ0AHgEy+2kREF2T83djwqpI43nW+5zsPE4D0v8sNaSv8AGPnz/TTSo1HXX9cXJukfdfRQxgYNuptIGlu8eA6woEG3suII5zvK7u2mUoblnTtk4uEgF4KqUyg4mXPBdQwUnpd0W4FoYHGDnqVHhva5DvQtd/OlQgsIBJMayz4rsbh3YNqUFTPqGECqVgNxaztdg3WMkefBeITVdDqdSgUzPz/jG93o2NvbK3GRMD6lVMmNoA72yspKyZU4PDzUO++8o42NjbJvAwaAx706gs7ndPnv7OyUqkXENM/Pz1fykXgmduoFnZ8mJY+gjLty7kYGQsXHXDqJjjAenuh/fHysDz74QLdv39af//mf6+bNm3rvvfd0dHRUnj8RSlzNv/u7v6tz585pa2urGDtHR0e6e/eubty4oRs3blRivr3cMPPA+9uEYvo8yvFBONLW3NycLl26VKqaZQ6JVOU9F/TTVmzoJ/xAvhkyg+pPIPFU0Mv5xHOD/CPrkL3J52+//baksezBO0y1p6xcwzH3w0Bmc7/0fICqpqcDmcpawA7l7AFAxUPQU47ITD7jpUYmc//M0eA5eX+TdridJjWh4I6m+kLt8sd/zzCRv4h4AAACdklEQVRDj+/vdDolFwNeYk8i9xgOh8NSKIOcMsCfLPWM8UAJ8MPDwyJXu91uBY3G2EDOuJHhz44c9+fIsaozjurGk2clzHaalBEX6bngeT6sp5p2cr1Mj0meD8/l/hbwALwJr/FuaA/ZQnuZ64HeAuGdRNZkrivjQn8yJyNDienHSy+9JEn6whe+IGnsZXzYWP+PQglwAkyy5nhuAwo2POgV4sipQH8YDodaX1/X3NycLly4UPaSksY5HABwFLepM/bh4dS14EUfSw9jdiODHA/3dKGjEC48HN4vQ39wcFCqbN27d68U4Tk8PCz5rvTDK2/SB9eVMo+vDkB2aj0bjwlNskLrJogLdz8nEbE83z8nIp2KT1rf2ZYnc7mgxHshqRISxT3co0DYTuYXTELK68bD3ZoIIjdI8LA4Mv8o0PFUyP27OsMxPQdNz5gC0IGAwWCgO3fu6M6dO2WB8Z3WvU0E9tmzZ8tGYIwNwpPQDBdO/n/dPG0axweNb7bFgsFviSjXeUomeU5aaulBVDeHJ82phznPecV53BM2U/7WyfxJHg6OdbkWDlIgn93b0ISKfhgkdJKXZxJy3dLjTU3ePX6btA7WnZffQxj46CDuxeN798yn3lQ3j9N7l7+7R8N5j/+dmnI1mnIz+MtQsknj/DBjVn5vGballlpqqaWWWmqppZZaehTUbovbUksttdRSSy211FJLLT0Sao2NllpqqaWWWmqppZZaaumRUGtstNRSSy211FJLLbXUUkuPhFpjo6WWWmqppZZaaqmlllp6JNQaGy211FJLLbXUUksttdTSI6HW2GippZZaaqmlllpqqaWWHgn9f66uunYGOzxvAAAAAElFTkSuQmCC\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"image_size= 48 # image is a 48*48 grid of pixels\nbatch_size= 64 # number of images to process at a time\ntrain_datagen = ImageDataGenerator(rescale=1./255,\n                                   zoom_range=0.3,\n                                   horizontal_flip=True)\n\ntraining_set = train_datagen.flow_from_directory(train_dir,\n                                                batch_size=64,\n                                                target_size=(image_size, image_size),\n                                                shuffle=True,\n                                                color_mode='grayscale',\n                                                class_mode='categorical')\n\ntest_datagen = ImageDataGenerator(rescale=1./255)\ntest_set = test_datagen.flow_from_directory(test_dir,\n                                                batch_size=64,\n                                                target_size=(image_size, image_size),\n                                                shuffle=True,\n                                                color_mode='grayscale',\n                                                class_mode='categorical')","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:12:58.995377Z","iopub.execute_input":"2022-07-31T09:12:58.995718Z","iopub.status.idle":"2022-07-31T09:13:30.316667Z","shell.execute_reply.started":"2022-07-31T09:12:58.995687Z","shell.execute_reply":"2022-07-31T09:13:30.315715Z"},"trusted":true},"execution_count":7,"outputs":[{"name":"stdout","text":"Found 28709 images belonging to 7 classes.\nFound 7178 images belonging to 7 classes.\n","output_type":"stream"}]},{"cell_type":"code","source":"training_set.class_indices","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:13:30.318247Z","iopub.execute_input":"2022-07-31T09:13:30.318784Z","iopub.status.idle":"2022-07-31T09:13:30.325701Z","shell.execute_reply.started":"2022-07-31T09:13:30.318744Z","shell.execute_reply":"2022-07-31T09:13:30.324767Z"},"trusted":true},"execution_count":8,"outputs":[{"execution_count":8,"output_type":"execute_result","data":{"text/plain":"{'angry': 0,\n 'disgust': 1,\n 'fear': 2,\n 'happy': 3,\n 'neutral': 4,\n 'sad': 5,\n 'surprise': 6}"},"metadata":{}}]},{"cell_type":"markdown","source":"## Defining Model","metadata":{}},{"cell_type":"code","source":"def fer_model(input_size, num_classes=7):\n    fer_model = tf.keras.models.Sequential()\n    fer_model.add(Conv2D(32, kernel_size=(3, 3), padding='same', activation='relu', input_shape=(img_l, img_w,1)))\n    fer_model.add(Conv2D(64,(3,3), padding='same', activation='relu' ))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(MaxPooling2D(pool_size=(2, 2)))\n    fer_model.add(Dropout(0.25))\n\n\n    fer_model.add(Conv2D(128,(5,5), padding='same', activation='relu'))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(MaxPooling2D(pool_size=(2, 2)))\n    fer_model.add(Dropout(0.25))\n    \n\n    fer_model.add(Conv2D(512,(3,3), padding='same', activation='relu', kernel_regularizer=regularizers.l2(0.01)))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(MaxPooling2D(pool_size=(2, 2)))\n    fer_model.add(Dropout(0.25))\n\n\n    fer_model.add(Conv2D(512,(3,3), padding='same', activation='relu', kernel_regularizer=regularizers.l2(0.01)))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(MaxPooling2D(pool_size=(2, 2)))\n    fer_model.add(Dropout(0.25))\n\n    fer_model.add(Flatten())\n\n    fer_model.add(Dense(256))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(Dropout(0.25))\n    \n    \n    fer_model.add(Dense(512))\n    fer_model.add(BatchNormalization())\n    fer_model.add(Activation('relu'))\n    fer_model.add(Dropout(0.25))\n\n    fer_model.add(Dense(num_classes, activation = 'softmax'))\n    optimizer = Adam(lr = 0.0001 , decay = 1e-6)\n    fer_model.compile(optimizer=optimizer, loss = 'categorical_crossentropy', metrics = ['accuracy'])\n    fer_model.summary\n    return fer_model","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:13:52.061433Z","iopub.execute_input":"2022-07-31T09:13:52.061850Z","iopub.status.idle":"2022-07-31T09:13:52.093934Z","shell.execute_reply.started":"2022-07-31T09:13:52.061813Z","shell.execute_reply":"2022-07-31T09:13:52.093118Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"code","source":"fer_model = fer_model((img_l,img_w,1), num_classes)\nfer_model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:13:55.913274Z","iopub.execute_input":"2022-07-31T09:13:55.913634Z","iopub.status.idle":"2022-07-31T09:13:58.316515Z","shell.execute_reply.started":"2022-07-31T09:13:55.913602Z","shell.execute_reply":"2022-07-31T09:13:58.315696Z"},"trusted":true},"execution_count":10,"outputs":[{"name":"stdout","text":"Model: \"sequential\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d (Conv2D)              (None, 48, 48, 32)        320       \n_________________________________________________________________\nconv2d_1 (Conv2D)            (None, 48, 48, 64)        18496     \n_________________________________________________________________\nbatch_normalization (BatchNo (None, 48, 48, 64)        256       \n_________________________________________________________________\nactivation (Activation)      (None, 48, 48, 64)        0         \n_________________________________________________________________\nmax_pooling2d (MaxPooling2D) (None, 24, 24, 64)        0         \n_________________________________________________________________\ndropout (Dropout)            (None, 24, 24, 64)        0         \n_________________________________________________________________\nconv2d_2 (Conv2D)            (None, 24, 24, 128)       204928    \n_________________________________________________________________\nbatch_normalization_1 (Batch (None, 24, 24, 128)       512       \n_________________________________________________________________\nactivation_1 (Activation)    (None, 24, 24, 128)       0         \n_________________________________________________________________\nmax_pooling2d_1 (MaxPooling2 (None, 12, 12, 128)       0         \n_________________________________________________________________\ndropout_1 (Dropout)          (None, 12, 12, 128)       0         \n_________________________________________________________________\nconv2d_3 (Conv2D)            (None, 12, 12, 512)       590336    \n_________________________________________________________________\nbatch_normalization_2 (Batch (None, 12, 12, 512)       2048      \n_________________________________________________________________\nactivation_2 (Activation)    (None, 12, 12, 512)       0         \n_________________________________________________________________\nmax_pooling2d_2 (MaxPooling2 (None, 6, 6, 512)         0         \n_________________________________________________________________\ndropout_2 (Dropout)          (None, 6, 6, 512)         0         \n_________________________________________________________________\nconv2d_4 (Conv2D)            (None, 6, 6, 512)         2359808   \n_________________________________________________________________\nbatch_normalization_3 (Batch (None, 6, 6, 512)         2048      \n_________________________________________________________________\nactivation_3 (Activation)    (None, 6, 6, 512)         0         \n_________________________________________________________________\nmax_pooling2d_3 (MaxPooling2 (None, 3, 3, 512)         0         \n_________________________________________________________________\ndropout_3 (Dropout)          (None, 3, 3, 512)         0         \n_________________________________________________________________\nflatten (Flatten)            (None, 4608)              0         \n_________________________________________________________________\ndense (Dense)                (None, 256)               1179904   \n_________________________________________________________________\nbatch_normalization_4 (Batch (None, 256)               1024      \n_________________________________________________________________\nactivation_4 (Activation)    (None, 256)               0         \n_________________________________________________________________\ndropout_4 (Dropout)          (None, 256)               0         \n_________________________________________________________________\ndense_1 (Dense)              (None, 512)               131584    \n_________________________________________________________________\nbatch_normalization_5 (Batch (None, 512)               2048      \n_________________________________________________________________\nactivation_5 (Activation)    (None, 512)               0         \n_________________________________________________________________\ndropout_5 (Dropout)          (None, 512)               0         \n_________________________________________________________________\ndense_2 (Dense)              (None, 7)                 3591      \n=================================================================\nTotal params: 4,496,903\nTrainable params: 4,492,935\nNon-trainable params: 3,968\n_________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"plot_model(fer_model, to_file='fer_model.png', show_shapes=True, show_layer_names=True)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:14:03.806126Z","iopub.execute_input":"2022-07-31T09:14:03.806464Z","iopub.status.idle":"2022-07-31T09:14:04.435854Z","shell.execute_reply.started":"2022-07-31T09:14:03.806433Z","shell.execute_reply":"2022-07-31T09:14:04.434174Z"},"trusted":true},"execution_count":11,"outputs":[{"execution_count":11,"output_type":"execute_result","data":{"image/png":"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\n","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}]},{"cell_type":"markdown","source":"## Training the model","metadata":{}},{"cell_type":"code","source":"steps_per_epoch = training_set.n // training_set.batch_size\nvalidation_steps = test_set.n // test_set.batch_size","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:14:14.120123Z","iopub.execute_input":"2022-07-31T09:14:14.120496Z","iopub.status.idle":"2022-07-31T09:14:14.125841Z","shell.execute_reply.started":"2022-07-31T09:14:14.120460Z","shell.execute_reply":"2022-07-31T09:14:14.124069Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"markdown","source":"# Checkpointing","metadata":{}},{"cell_type":"code","source":"chkpt_path = 'fer_model.h5'\nlog_dir = \"checkpoint/logs/\" + datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n\ncheckpoint = ModelCheckpoint(filepath=chkpt_path,\n                             save_best_only=True,\n                             verbose=1,\n                             mode='min',\n                             moniter='val_loss')\n\nearlystop = EarlyStopping(monitor='val_loss', \n                          min_delta=0, \n                          patience=3, \n                          verbose=1, \n                          restore_best_weights=True)\n                        \nreduce_lr = ReduceLROnPlateau(monitor='val_loss', \n                              factor=0.2, \n                              patience=6, \n                              verbose=1, \n                              min_delta=0.0001)\n\n\ntensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1)\ncsv_logger = CSVLogger('training.log')\n\ncallback = [checkpoint, reduce_lr, csv_logger]","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:14:25.987900Z","iopub.execute_input":"2022-07-31T09:14:25.988334Z","iopub.status.idle":"2022-07-31T09:14:26.007225Z","shell.execute_reply.started":"2022-07-31T09:14:25.988290Z","shell.execute_reply":"2022-07-31T09:14:26.006336Z"},"trusted":true},"execution_count":13,"outputs":[]},{"cell_type":"code","source":"fer_model_hist = fer_model.fit(x=training_set,\n                 validation_data=test_set,\n                 epochs=100,\n                 callbacks=callback,\n                 steps_per_epoch=steps_per_epoch,\n                 validation_steps=validation_steps)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T09:14:28.930387Z","iopub.execute_input":"2022-07-31T09:14:28.930721Z","iopub.status.idle":"2022-07-31T10:19:28.513354Z","shell.execute_reply.started":"2022-07-31T09:14:28.930690Z","shell.execute_reply":"2022-07-31T10:19:28.512588Z"},"trusted":true},"execution_count":14,"outputs":[{"name":"stdout","text":"Epoch 1/100\n448/448 [==============================] - ETA: 0s - loss: 8.4183 - accuracy: 0.2231\nEpoch 00001: val_loss improved from inf to 7.48729, saving model to fer_model.h5\n448/448 [==============================] - 212s 473ms/step - loss: 8.4183 - accuracy: 0.2231 - val_loss: 7.4873 - val_accuracy: 0.2475\nEpoch 2/100\n448/448 [==============================] - ETA: 0s - loss: 6.7023 - accuracy: 0.2613\nEpoch 00002: val_loss improved from 7.48729 to 5.87606, saving model to fer_model.h5\n448/448 [==============================] - 38s 85ms/step - loss: 6.7023 - accuracy: 0.2613 - val_loss: 5.8761 - val_accuracy: 0.2699\nEpoch 3/100\n448/448 [==============================] - ETA: 0s - loss: 5.2504 - accuracy: 0.2977\nEpoch 00003: val_loss improved from 5.87606 to 4.59153, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 5.2504 - accuracy: 0.2977 - val_loss: 4.5915 - val_accuracy: 0.3153\nEpoch 4/100\n448/448 [==============================] - ETA: 0s - loss: 4.1533 - accuracy: 0.3263\nEpoch 00004: val_loss improved from 4.59153 to 3.62481, saving model to fer_model.h5\n448/448 [==============================] - 39s 86ms/step - loss: 4.1533 - accuracy: 0.3263 - val_loss: 3.6248 - val_accuracy: 0.3818\nEpoch 5/100\n448/448 [==============================] - ETA: 0s - loss: 3.3830 - accuracy: 0.3585\nEpoch 00005: val_loss improved from 3.62481 to 3.02373, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 3.3830 - accuracy: 0.3585 - val_loss: 3.0237 - val_accuracy: 0.3901\nEpoch 6/100\n448/448 [==============================] - ETA: 0s - loss: 2.8473 - accuracy: 0.3929\nEpoch 00006: val_loss improved from 3.02373 to 2.54038, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 2.8473 - accuracy: 0.3929 - val_loss: 2.5404 - val_accuracy: 0.4383\nEpoch 7/100\n448/448 [==============================] - ETA: 0s - loss: 2.4606 - accuracy: 0.4267\nEpoch 00007: val_loss improved from 2.54038 to 2.20632, saving model to fer_model.h5\n448/448 [==============================] - 37s 84ms/step - loss: 2.4606 - accuracy: 0.4267 - val_loss: 2.2063 - val_accuracy: 0.4699\nEpoch 8/100\n448/448 [==============================] - ETA: 0s - loss: 2.1913 - accuracy: 0.4509\nEpoch 00008: val_loss improved from 2.20632 to 1.97960, saving model to fer_model.h5\n448/448 [==============================] - 39s 87ms/step - loss: 2.1913 - accuracy: 0.4509 - val_loss: 1.9796 - val_accuracy: 0.4979\nEpoch 9/100\n448/448 [==============================] - ETA: 0s - loss: 1.9961 - accuracy: 0.4771\nEpoch 00009: val_loss improved from 1.97960 to 1.80974, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 1.9961 - accuracy: 0.4771 - val_loss: 1.8097 - val_accuracy: 0.5208\nEpoch 10/100\n448/448 [==============================] - ETA: 0s - loss: 1.8534 - accuracy: 0.4940 ETA: 2s -\nEpoch 00010: val_loss improved from 1.80974 to 1.72399, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 1.8534 - accuracy: 0.4940 - val_loss: 1.7240 - val_accuracy: 0.5197\nEpoch 11/100\n448/448 [==============================] - ETA: 0s - loss: 1.7451 - accuracy: 0.5126\nEpoch 00011: val_loss improved from 1.72399 to 1.58654, saving model to fer_model.h5\n448/448 [==============================] - 37s 84ms/step - loss: 1.7451 - accuracy: 0.5126 - val_loss: 1.5865 - val_accuracy: 0.5597\nEpoch 12/100\n448/448 [==============================] - ETA: 0s - loss: 1.6714 - accuracy: 0.5215\nEpoch 00012: val_loss improved from 1.58654 to 1.52753, saving model to fer_model.h5\n448/448 [==============================] - 37s 84ms/step - loss: 1.6714 - accuracy: 0.5215 - val_loss: 1.5275 - val_accuracy: 0.5665\nEpoch 13/100\n448/448 [==============================] - ETA: 0s - loss: 1.6235 - accuracy: 0.5290\nEpoch 00013: val_loss improved from 1.52753 to 1.51782, saving model to fer_model.h5\n448/448 [==============================] - 39s 88ms/step - loss: 1.6235 - accuracy: 0.5290 - val_loss: 1.5178 - val_accuracy: 0.5610\nEpoch 14/100\n448/448 [==============================] - ETA: 0s - loss: 1.5807 - accuracy: 0.5358\nEpoch 00014: val_loss improved from 1.51782 to 1.46619, saving model to fer_model.h5\n448/448 [==============================] - 39s 88ms/step - loss: 1.5807 - accuracy: 0.5358 - val_loss: 1.4662 - val_accuracy: 0.5784\nEpoch 15/100\n448/448 [==============================] - ETA: 0s - loss: 1.5536 - accuracy: 0.5446\nEpoch 00015: val_loss did not improve from 1.46619\n448/448 [==============================] - 37s 82ms/step - loss: 1.5536 - accuracy: 0.5446 - val_loss: 1.4759 - val_accuracy: 0.5725\nEpoch 16/100\n448/448 [==============================] - ETA: 0s - loss: 1.5189 - accuracy: 0.5518\nEpoch 00016: val_loss improved from 1.46619 to 1.42496, saving model to fer_model.h5\n448/448 [==============================] - 38s 85ms/step - loss: 1.5189 - accuracy: 0.5518 - val_loss: 1.4250 - val_accuracy: 0.5872\nEpoch 17/100\n448/448 [==============================] - ETA: 0s - loss: 1.4969 - accuracy: 0.5568\nEpoch 00017: val_loss improved from 1.42496 to 1.41496, saving model to fer_model.h5\n448/448 [==============================] - 40s 90ms/step - loss: 1.4969 - accuracy: 0.5568 - val_loss: 1.4150 - val_accuracy: 0.5925\nEpoch 18/100\n448/448 [==============================] - ETA: 0s - loss: 1.4762 - accuracy: 0.5657\nEpoch 00018: val_loss did not improve from 1.41496\n448/448 [==============================] - 39s 86ms/step - loss: 1.4762 - accuracy: 0.5657 - val_loss: 1.4626 - val_accuracy: 0.5667\nEpoch 19/100\n448/448 [==============================] - ETA: 0s - loss: 1.4639 - accuracy: 0.5703\nEpoch 00019: val_loss did not improve from 1.41496\n448/448 [==============================] - 39s 86ms/step - loss: 1.4639 - accuracy: 0.5703 - val_loss: 1.4427 - val_accuracy: 0.5770\nEpoch 20/100\n448/448 [==============================] - ETA: 0s - loss: 1.4460 - accuracy: 0.5773\nEpoch 00020: val_loss improved from 1.41496 to 1.36250, saving model to fer_model.h5\n448/448 [==============================] - 38s 85ms/step - loss: 1.4460 - accuracy: 0.5773 - val_loss: 1.3625 - val_accuracy: 0.6031\nEpoch 21/100\n448/448 [==============================] - ETA: 0s - loss: 1.4331 - accuracy: 0.5800\nEpoch 00021: val_loss did not improve from 1.36250\n448/448 [==============================] - 38s 84ms/step - loss: 1.4331 - accuracy: 0.5800 - val_loss: 1.3881 - val_accuracy: 0.6017\nEpoch 22/100\n448/448 [==============================] - ETA: 0s - loss: 1.4228 - accuracy: 0.5858\nEpoch 00022: val_loss improved from 1.36250 to 1.33888, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.4228 - accuracy: 0.5858 - val_loss: 1.3389 - val_accuracy: 0.6143\nEpoch 23/100\n448/448 [==============================] - ETA: 0s - loss: 1.4113 - accuracy: 0.5872\nEpoch 00023: val_loss improved from 1.33888 to 1.33788, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 1.4113 - accuracy: 0.5872 - val_loss: 1.3379 - val_accuracy: 0.6182\nEpoch 24/100\n448/448 [==============================] - ETA: 0s - loss: 1.4048 - accuracy: 0.5919\nEpoch 00024: val_loss did not improve from 1.33788\n448/448 [==============================] - 37s 82ms/step - loss: 1.4048 - accuracy: 0.5919 - val_loss: 1.3565 - val_accuracy: 0.6144\nEpoch 25/100\n448/448 [==============================] - ETA: 0s - loss: 1.3974 - accuracy: 0.5964\nEpoch 00025: val_loss improved from 1.33788 to 1.33742, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.3974 - accuracy: 0.5964 - val_loss: 1.3374 - val_accuracy: 0.6203\nEpoch 26/100\n448/448 [==============================] - ETA: 0s - loss: 1.3899 - accuracy: 0.5974\nEpoch 00026: val_loss did not improve from 1.33742\n448/448 [==============================] - 36s 81ms/step - loss: 1.3899 - accuracy: 0.5974 - val_loss: 1.3553 - val_accuracy: 0.6137\nEpoch 27/100\n448/448 [==============================] - ETA: 0s - loss: 1.3868 - accuracy: 0.6007\nEpoch 00027: val_loss improved from 1.33742 to 1.31978, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.3868 - accuracy: 0.6007 - val_loss: 1.3198 - val_accuracy: 0.6256\nEpoch 28/100\n448/448 [==============================] - ETA: 0s - loss: 1.3754 - accuracy: 0.6069\nEpoch 00028: val_loss did not improve from 1.31978\n448/448 [==============================] - 37s 82ms/step - loss: 1.3754 - accuracy: 0.6069 - val_loss: 1.3303 - val_accuracy: 0.6186\nEpoch 29/100\n448/448 [==============================] - ETA: 0s - loss: 1.3707 - accuracy: 0.6064\nEpoch 00029: val_loss improved from 1.31978 to 1.31355, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 1.3707 - accuracy: 0.6064 - val_loss: 1.3136 - val_accuracy: 0.6318\nEpoch 30/100\n448/448 [==============================] - ETA: 0s - loss: 1.3621 - accuracy: 0.6089\nEpoch 00030: val_loss did not improve from 1.31355\n448/448 [==============================] - 37s 82ms/step - loss: 1.3621 - accuracy: 0.6089 - val_loss: 1.3477 - val_accuracy: 0.6173\nEpoch 31/100\n448/448 [==============================] - ETA: 0s - loss: 1.3530 - accuracy: 0.6133\nEpoch 00031: val_loss improved from 1.31355 to 1.29844, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 1.3530 - accuracy: 0.6133 - val_loss: 1.2984 - val_accuracy: 0.6373\nEpoch 32/100\n448/448 [==============================] - ETA: 0s - loss: 1.3567 - accuracy: 0.6105\nEpoch 00032: val_loss did not improve from 1.29844\n448/448 [==============================] - 37s 83ms/step - loss: 1.3567 - accuracy: 0.6105 - val_loss: 1.3440 - val_accuracy: 0.6164\nEpoch 33/100\n448/448 [==============================] - ETA: 0s - loss: 1.3448 - accuracy: 0.6156\nEpoch 00033: val_loss improved from 1.29844 to 1.27805, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 1.3448 - accuracy: 0.6156 - val_loss: 1.2780 - val_accuracy: 0.6469\nEpoch 34/100\n448/448 [==============================] - ETA: 0s - loss: 1.3366 - accuracy: 0.6211\nEpoch 00034: val_loss did not improve from 1.27805\n448/448 [==============================] - 38s 84ms/step - loss: 1.3366 - accuracy: 0.6211 - val_loss: 1.2958 - val_accuracy: 0.6366\nEpoch 35/100\n448/448 [==============================] - ETA: 0s - loss: 1.3431 - accuracy: 0.6186\nEpoch 00035: val_loss improved from 1.27805 to 1.27604, saving model to fer_model.h5\n448/448 [==============================] - 38s 85ms/step - loss: 1.3431 - accuracy: 0.6186 - val_loss: 1.2760 - val_accuracy: 0.6444\nEpoch 36/100\n448/448 [==============================] - ETA: 0s - loss: 1.3301 - accuracy: 0.6207\nEpoch 00036: val_loss did not improve from 1.27604\n448/448 [==============================] - 37s 83ms/step - loss: 1.3301 - accuracy: 0.6207 - val_loss: 1.3102 - val_accuracy: 0.6369\nEpoch 37/100\n448/448 [==============================] - ETA: 0s - loss: 1.3288 - accuracy: 0.6249\nEpoch 00037: val_loss did not improve from 1.27604\n448/448 [==============================] - 37s 83ms/step - loss: 1.3288 - accuracy: 0.6249 - val_loss: 1.3136 - val_accuracy: 0.6390\nEpoch 38/100\n448/448 [==============================] - ETA: 0s - loss: 1.3357 - accuracy: 0.6222\nEpoch 00038: val_loss did not improve from 1.27604\n448/448 [==============================] - 37s 83ms/step - loss: 1.3357 - accuracy: 0.6222 - val_loss: 1.3190 - val_accuracy: 0.6324\nEpoch 39/100\n448/448 [==============================] - ETA: 0s - loss: 1.3269 - accuracy: 0.6279\nEpoch 00039: val_loss did not improve from 1.27604\n448/448 [==============================] - 37s 82ms/step - loss: 1.3269 - accuracy: 0.6279 - val_loss: 1.2861 - val_accuracy: 0.6449\nEpoch 40/100\n448/448 [==============================] - ETA: 0s - loss: 1.3210 - accuracy: 0.6305\nEpoch 00040: val_loss did not improve from 1.27604\n448/448 [==============================] - 37s 83ms/step - loss: 1.3210 - accuracy: 0.6305 - val_loss: 1.2776 - val_accuracy: 0.6468\nEpoch 41/100\n448/448 [==============================] - ETA: 0s - loss: 1.3160 - accuracy: 0.6318\nEpoch 00041: val_loss did not improve from 1.27604\n\nEpoch 00041: ReduceLROnPlateau reducing learning rate to 1.9999999494757503e-05.\n448/448 [==============================] - 37s 82ms/step - loss: 1.3160 - accuracy: 0.6318 - val_loss: 1.3342 - val_accuracy: 0.6286\nEpoch 42/100\n448/448 [==============================] - ETA: 0s - loss: 1.2723 - accuracy: 0.6454\nEpoch 00042: val_loss improved from 1.27604 to 1.24692, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 1.2723 - accuracy: 0.6454 - val_loss: 1.2469 - val_accuracy: 0.6590\nEpoch 43/100\n448/448 [==============================] - ETA: 0s - loss: 1.2291 - accuracy: 0.6514\nEpoch 00043: val_loss improved from 1.24692 to 1.21674, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.2291 - accuracy: 0.6514 - val_loss: 1.2167 - val_accuracy: 0.6655\nEpoch 44/100\n448/448 [==============================] - ETA: 0s - loss: 1.2013 - accuracy: 0.6594\nEpoch 00044: val_loss improved from 1.21674 to 1.19530, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 1.2013 - accuracy: 0.6594 - val_loss: 1.1953 - val_accuracy: 0.6674\nEpoch 45/100\n448/448 [==============================] - ETA: 0s - loss: 1.1936 - accuracy: 0.6571\nEpoch 00045: val_loss improved from 1.19530 to 1.17996, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 1.1936 - accuracy: 0.6571 - val_loss: 1.1800 - val_accuracy: 0.6723\nEpoch 46/100\n448/448 [==============================] - ETA: 0s - loss: 1.1622 - accuracy: 0.6666\nEpoch 00046: val_loss improved from 1.17996 to 1.17273, saving model to fer_model.h5\n448/448 [==============================] - 37s 84ms/step - loss: 1.1622 - accuracy: 0.6666 - val_loss: 1.1727 - val_accuracy: 0.6687\nEpoch 47/100\n448/448 [==============================] - ETA: 0s - loss: 1.1576 - accuracy: 0.6627\nEpoch 00047: val_loss improved from 1.17273 to 1.15146, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 1.1576 - accuracy: 0.6627 - val_loss: 1.1515 - val_accuracy: 0.6759\nEpoch 48/100\n448/448 [==============================] - ETA: 0s - loss: 1.1323 - accuracy: 0.6716\nEpoch 00048: val_loss improved from 1.15146 to 1.14351, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 1.1323 - accuracy: 0.6716 - val_loss: 1.1435 - val_accuracy: 0.6769\nEpoch 49/100\n448/448 [==============================] - ETA: 0s - loss: 1.1283 - accuracy: 0.6718\nEpoch 00049: val_loss did not improve from 1.14351\n448/448 [==============================] - 36s 81ms/step - loss: 1.1283 - accuracy: 0.6718 - val_loss: 1.1527 - val_accuracy: 0.6702\nEpoch 50/100\n448/448 [==============================] - ETA: 0s - loss: 1.1136 - accuracy: 0.6732\nEpoch 00050: val_loss improved from 1.14351 to 1.13480, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 1.1136 - accuracy: 0.6732 - val_loss: 1.1348 - val_accuracy: 0.6730\nEpoch 51/100\n448/448 [==============================] - ETA: 0s - loss: 1.1041 - accuracy: 0.6733\nEpoch 00051: val_loss improved from 1.13480 to 1.13155, saving model to fer_model.h5\n448/448 [==============================] - 36s 80ms/step - loss: 1.1041 - accuracy: 0.6733 - val_loss: 1.1315 - val_accuracy: 0.6763\nEpoch 52/100\n448/448 [==============================] - ETA: 0s - loss: 1.0956 - accuracy: 0.6752\nEpoch 00052: val_loss improved from 1.13155 to 1.11769, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.0956 - accuracy: 0.6752 - val_loss: 1.1177 - val_accuracy: 0.6770\nEpoch 53/100\n448/448 [==============================] - ETA: 0s - loss: 1.0798 - accuracy: 0.6810\nEpoch 00053: val_loss improved from 1.11769 to 1.11683, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.0798 - accuracy: 0.6810 - val_loss: 1.1168 - val_accuracy: 0.6779\nEpoch 54/100\n448/448 [==============================] - ETA: 0s - loss: 1.0764 - accuracy: 0.6801\nEpoch 00054: val_loss did not improve from 1.11683\n448/448 [==============================] - 36s 79ms/step - loss: 1.0764 - accuracy: 0.6801 - val_loss: 1.1179 - val_accuracy: 0.6776\nEpoch 55/100\n448/448 [==============================] - ETA: 0s - loss: 1.0625 - accuracy: 0.6819\nEpoch 00055: val_loss did not improve from 1.11683\n448/448 [==============================] - 36s 79ms/step - loss: 1.0625 - accuracy: 0.6819 - val_loss: 1.1195 - val_accuracy: 0.6755\nEpoch 56/100\n448/448 [==============================] - ETA: 0s - loss: 1.0593 - accuracy: 0.6859\nEpoch 00056: val_loss improved from 1.11683 to 1.11219, saving model to fer_model.h5\n448/448 [==============================] - 35s 79ms/step - loss: 1.0593 - accuracy: 0.6859 - val_loss: 1.1122 - val_accuracy: 0.6784\nEpoch 57/100\n448/448 [==============================] - ETA: 0s - loss: 1.0533 - accuracy: 0.6871\nEpoch 00057: val_loss improved from 1.11219 to 1.09759, saving model to fer_model.h5\n448/448 [==============================] - 36s 79ms/step - loss: 1.0533 - accuracy: 0.6871 - val_loss: 1.0976 - val_accuracy: 0.6837\nEpoch 58/100\n448/448 [==============================] - ETA: 0s - loss: 1.0484 - accuracy: 0.6896\nEpoch 00058: val_loss did not improve from 1.09759\n448/448 [==============================] - 37s 83ms/step - loss: 1.0484 - accuracy: 0.6896 - val_loss: 1.1001 - val_accuracy: 0.6814\nEpoch 59/100\n448/448 [==============================] - ETA: 0s - loss: 1.0378 - accuracy: 0.6914 ETA: 1s\nEpoch 00059: val_loss did not improve from 1.09759\n448/448 [==============================] - 37s 82ms/step - loss: 1.0378 - accuracy: 0.6914 - val_loss: 1.1113 - val_accuracy: 0.6744\nEpoch 60/100\n448/448 [==============================] - ETA: 0s - loss: 1.0395 - accuracy: 0.6903\nEpoch 00060: val_loss did not improve from 1.09759\n448/448 [==============================] - 37s 82ms/step - loss: 1.0395 - accuracy: 0.6903 - val_loss: 1.1003 - val_accuracy: 0.6809\nEpoch 61/100\n448/448 [==============================] - ETA: 0s - loss: 1.0257 - accuracy: 0.6947\nEpoch 00061: val_loss did not improve from 1.09759\n448/448 [==============================] - 36s 81ms/step - loss: 1.0257 - accuracy: 0.6947 - val_loss: 1.1016 - val_accuracy: 0.6833\nEpoch 62/100\n448/448 [==============================] - ETA: 0s - loss: 1.0185 - accuracy: 0.6958\nEpoch 00062: val_loss improved from 1.09759 to 1.09322, saving model to fer_model.h5\n448/448 [==============================] - 38s 86ms/step - loss: 1.0185 - accuracy: 0.6958 - val_loss: 1.0932 - val_accuracy: 0.6854\nEpoch 63/100\n448/448 [==============================] - ETA: 0s - loss: 1.0145 - accuracy: 0.7005\nEpoch 00063: val_loss improved from 1.09322 to 1.09131, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 1.0145 - accuracy: 0.7005 - val_loss: 1.0913 - val_accuracy: 0.6821\nEpoch 64/100\n448/448 [==============================] - ETA: 0s - loss: 1.0122 - accuracy: 0.7017\nEpoch 00064: val_loss improved from 1.09131 to 1.08632, saving model to fer_model.h5\n448/448 [==============================] - 37s 84ms/step - loss: 1.0122 - accuracy: 0.7017 - val_loss: 1.0863 - val_accuracy: 0.6814\nEpoch 65/100\n448/448 [==============================] - ETA: 0s - loss: 1.0087 - accuracy: 0.7015\nEpoch 00065: val_loss did not improve from 1.08632\n448/448 [==============================] - 37s 83ms/step - loss: 1.0087 - accuracy: 0.7015 - val_loss: 1.0874 - val_accuracy: 0.6855\nEpoch 66/100\n448/448 [==============================] - ETA: 0s - loss: 1.0010 - accuracy: 0.7027\nEpoch 00066: val_loss improved from 1.08632 to 1.08269, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 1.0010 - accuracy: 0.7027 - val_loss: 1.0827 - val_accuracy: 0.6882\nEpoch 67/100\n448/448 [==============================] - ETA: 0s - loss: 0.9994 - accuracy: 0.7039\nEpoch 00067: val_loss did not improve from 1.08269\n448/448 [==============================] - 37s 83ms/step - loss: 0.9994 - accuracy: 0.7039 - val_loss: 1.0838 - val_accuracy: 0.6886\nEpoch 68/100\n448/448 [==============================] - ETA: 0s - loss: 0.9976 - accuracy: 0.7032\nEpoch 00068: val_loss did not improve from 1.08269\n448/448 [==============================] - 37s 82ms/step - loss: 0.9976 - accuracy: 0.7032 - val_loss: 1.0955 - val_accuracy: 0.6855\nEpoch 69/100\n448/448 [==============================] - ETA: 0s - loss: 0.9955 - accuracy: 0.7038\nEpoch 00069: val_loss improved from 1.08269 to 1.08232, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 0.9955 - accuracy: 0.7038 - val_loss: 1.0823 - val_accuracy: 0.6855\nEpoch 70/100\n448/448 [==============================] - ETA: 0s - loss: 0.9898 - accuracy: 0.7056\nEpoch 00070: val_loss improved from 1.08232 to 1.08174, saving model to fer_model.h5\n448/448 [==============================] - 37s 83ms/step - loss: 0.9898 - accuracy: 0.7056 - val_loss: 1.0817 - val_accuracy: 0.6882\nEpoch 71/100\n448/448 [==============================] - ETA: 0s - loss: 0.9883 - accuracy: 0.7065\nEpoch 00071: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 82ms/step - loss: 0.9883 - accuracy: 0.7065 - val_loss: 1.0891 - val_accuracy: 0.6851\nEpoch 72/100\n448/448 [==============================] - ETA: 0s - loss: 0.9799 - accuracy: 0.7093\nEpoch 00072: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 82ms/step - loss: 0.9799 - accuracy: 0.7093 - val_loss: 1.0883 - val_accuracy: 0.6855\nEpoch 73/100\n448/448 [==============================] - ETA: 0s - loss: 0.9767 - accuracy: 0.7114\nEpoch 00073: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 82ms/step - loss: 0.9767 - accuracy: 0.7114 - val_loss: 1.0836 - val_accuracy: 0.6862\nEpoch 74/100\n448/448 [==============================] - ETA: 0s - loss: 0.9701 - accuracy: 0.7119\nEpoch 00074: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 83ms/step - loss: 0.9701 - accuracy: 0.7119 - val_loss: 1.0887 - val_accuracy: 0.6896\nEpoch 75/100\n448/448 [==============================] - ETA: 0s - loss: 0.9687 - accuracy: 0.7131\nEpoch 00075: val_loss did not improve from 1.08174\n448/448 [==============================] - 36s 81ms/step - loss: 0.9687 - accuracy: 0.7131 - val_loss: 1.0879 - val_accuracy: 0.6872\nEpoch 76/100\n448/448 [==============================] - ETA: 0s - loss: 0.9681 - accuracy: 0.7128\nEpoch 00076: val_loss did not improve from 1.08174\n\nEpoch 00076: ReduceLROnPlateau reducing learning rate to 3.999999898951501e-06.\n448/448 [==============================] - 37s 82ms/step - loss: 0.9681 - accuracy: 0.7128 - val_loss: 1.1038 - val_accuracy: 0.6857\nEpoch 77/100\n448/448 [==============================] - ETA: 0s - loss: 0.9486 - accuracy: 0.7211\nEpoch 00077: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 83ms/step - loss: 0.9486 - accuracy: 0.7211 - val_loss: 1.0837 - val_accuracy: 0.6913\nEpoch 78/100\n448/448 [==============================] - ETA: 0s - loss: 0.9425 - accuracy: 0.7208\nEpoch 00078: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 83ms/step - loss: 0.9425 - accuracy: 0.7208 - val_loss: 1.0827 - val_accuracy: 0.6908\nEpoch 79/100\n448/448 [==============================] - ETA: 0s - loss: 0.9402 - accuracy: 0.7235 ETA: 0s - loss: 0\nEpoch 00079: val_loss did not improve from 1.08174\n448/448 [==============================] - 37s 82ms/step - loss: 0.9402 - accuracy: 0.7235 - val_loss: 1.0864 - val_accuracy: 0.6896\nEpoch 80/100\n448/448 [==============================] - ETA: 0s - loss: 0.9368 - accuracy: 0.7265\nEpoch 00080: val_loss improved from 1.08174 to 1.07819, saving model to fer_model.h5\n448/448 [==============================] - 36s 81ms/step - loss: 0.9368 - accuracy: 0.7265 - val_loss: 1.0782 - val_accuracy: 0.6950\nEpoch 81/100\n448/448 [==============================] - ETA: 0s - loss: 0.9274 - accuracy: 0.7289\nEpoch 00081: val_loss did not improve from 1.07819\n448/448 [==============================] - 36s 81ms/step - loss: 0.9274 - accuracy: 0.7289 - val_loss: 1.0804 - val_accuracy: 0.6935\nEpoch 82/100\n448/448 [==============================] - ETA: 0s - loss: 0.9265 - accuracy: 0.7286\nEpoch 00082: val_loss did not improve from 1.07819\n448/448 [==============================] - 37s 82ms/step - loss: 0.9265 - accuracy: 0.7286 - val_loss: 1.0791 - val_accuracy: 0.6918\nEpoch 83/100\n448/448 [==============================] - ETA: 0s - loss: 0.9256 - accuracy: 0.7272\nEpoch 00083: val_loss did not improve from 1.07819\n448/448 [==============================] - 36s 81ms/step - loss: 0.9256 - accuracy: 0.7272 - val_loss: 1.0823 - val_accuracy: 0.6893\nEpoch 84/100\n448/448 [==============================] - ETA: 0s - loss: 0.9125 - accuracy: 0.7318\nEpoch 00084: val_loss did not improve from 1.07819\n448/448 [==============================] - 36s 81ms/step - loss: 0.9125 - accuracy: 0.7318 - val_loss: 1.0790 - val_accuracy: 0.6925\nEpoch 85/100\n448/448 [==============================] - ETA: 0s - loss: 0.9141 - accuracy: 0.7329\nEpoch 00085: val_loss improved from 1.07819 to 1.07793, saving model to fer_model.h5\n448/448 [==============================] - 37s 82ms/step - loss: 0.9141 - accuracy: 0.7329 - val_loss: 1.0779 - val_accuracy: 0.6927\nEpoch 86/100\n448/448 [==============================] - ETA: 0s - loss: 0.9130 - accuracy: 0.7333\nEpoch 00086: val_loss did not improve from 1.07793\n448/448 [==============================] - 37s 83ms/step - loss: 0.9130 - accuracy: 0.7333 - val_loss: 1.0807 - val_accuracy: 0.6900\nEpoch 87/100\n448/448 [==============================] - ETA: 0s - loss: 0.9148 - accuracy: 0.7334\nEpoch 00087: val_loss did not improve from 1.07793\n448/448 [==============================] - 37s 82ms/step - loss: 0.9148 - accuracy: 0.7334 - val_loss: 1.0781 - val_accuracy: 0.6927\nEpoch 88/100\n448/448 [==============================] - ETA: 0s - loss: 0.9048 - accuracy: 0.7394 ETA: \nEpoch 00088: val_loss did not improve from 1.07793\n448/448 [==============================] - 37s 82ms/step - loss: 0.9048 - accuracy: 0.7394 - val_loss: 1.0795 - val_accuracy: 0.6927\nEpoch 89/100\n448/448 [==============================] - ETA: 0s - loss: 0.9115 - accuracy: 0.7341\nEpoch 00089: val_loss did not improve from 1.07793\n448/448 [==============================] - 38s 84ms/step - loss: 0.9115 - accuracy: 0.7341 - val_loss: 1.0793 - val_accuracy: 0.6911\nEpoch 90/100\n448/448 [==============================] - ETA: 0s - loss: 0.9009 - accuracy: 0.7364\nEpoch 00090: val_loss improved from 1.07793 to 1.07430, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 0.9009 - accuracy: 0.7364 - val_loss: 1.0743 - val_accuracy: 0.6943\nEpoch 91/100\n448/448 [==============================] - ETA: 0s - loss: 0.9038 - accuracy: 0.7345\nEpoch 00091: val_loss did not improve from 1.07430\n448/448 [==============================] - 38s 84ms/step - loss: 0.9038 - accuracy: 0.7345 - val_loss: 1.0785 - val_accuracy: 0.6931\nEpoch 92/100\n448/448 [==============================] - ETA: 0s - loss: 0.8921 - accuracy: 0.7398\nEpoch 00092: val_loss did not improve from 1.07430\n448/448 [==============================] - 36s 81ms/step - loss: 0.8921 - accuracy: 0.7398 - val_loss: 1.0788 - val_accuracy: 0.6936\nEpoch 93/100\n448/448 [==============================] - ETA: 0s - loss: 0.8926 - accuracy: 0.7369\nEpoch 00093: val_loss did not improve from 1.07430\n448/448 [==============================] - 37s 82ms/step - loss: 0.8926 - accuracy: 0.7369 - val_loss: 1.0759 - val_accuracy: 0.6955\nEpoch 94/100\n448/448 [==============================] - ETA: 0s - loss: 0.8909 - accuracy: 0.7378\nEpoch 00094: val_loss did not improve from 1.07430\n448/448 [==============================] - 37s 82ms/step - loss: 0.8909 - accuracy: 0.7378 - val_loss: 1.0756 - val_accuracy: 0.6953\nEpoch 95/100\n448/448 [==============================] - ETA: 0s - loss: 0.8859 - accuracy: 0.7416\nEpoch 00095: val_loss did not improve from 1.07430\n448/448 [==============================] - 36s 80ms/step - loss: 0.8859 - accuracy: 0.7416 - val_loss: 1.0789 - val_accuracy: 0.6918\nEpoch 96/100\n448/448 [==============================] - ETA: 0s - loss: 0.8895 - accuracy: 0.7360\nEpoch 00096: val_loss did not improve from 1.07430\n\nEpoch 00096: ReduceLROnPlateau reducing learning rate to 7.999999979801942e-07.\n448/448 [==============================] - 37s 82ms/step - loss: 0.8895 - accuracy: 0.7360 - val_loss: 1.0774 - val_accuracy: 0.6957\nEpoch 97/100\n448/448 [==============================] - ETA: 0s - loss: 0.8874 - accuracy: 0.7396\nEpoch 00097: val_loss did not improve from 1.07430\n448/448 [==============================] - 37s 83ms/step - loss: 0.8874 - accuracy: 0.7396 - val_loss: 1.0754 - val_accuracy: 0.6961\nEpoch 98/100\n448/448 [==============================] - ETA: 0s - loss: 0.8826 - accuracy: 0.7423\nEpoch 00098: val_loss improved from 1.07430 to 1.07297, saving model to fer_model.h5\n448/448 [==============================] - 38s 84ms/step - loss: 0.8826 - accuracy: 0.7423 - val_loss: 1.0730 - val_accuracy: 0.6959\nEpoch 99/100\n448/448 [==============================] - ETA: 0s - loss: 0.8847 - accuracy: 0.7401\nEpoch 00099: val_loss improved from 1.07297 to 1.07274, saving model to fer_model.h5\n448/448 [==============================] - 38s 86ms/step - loss: 0.8847 - accuracy: 0.7401 - val_loss: 1.0727 - val_accuracy: 0.6967\nEpoch 100/100\n448/448 [==============================] - ETA: 0s - loss: 0.8855 - accuracy: 0.7396\nEpoch 00100: val_loss did not improve from 1.07274\n448/448 [==============================] - 38s 85ms/step - loss: 0.8855 - accuracy: 0.7396 - val_loss: 1.0739 - val_accuracy: 0.6968\n","output_type":"stream"}]},{"cell_type":"markdown","source":"## Loss and Accuracy plot","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(14,5))\nplt.subplot(1,2,2)\nplt.plot(fer_model_hist.history['accuracy'])\nplt.plot(fer_model_hist.history['val_accuracy'])\nplt.title('Model Accuracy')\nplt.xlabel('Epochs')\nplt.ylabel('Accuracy')\nplt.legend(['train', 'test'], loc='upper left')\n\nplt.subplot(1,2,1)\nplt.plot(fer_model_hist.history['loss'])\nplt.plot(fer_model_hist.history['val_loss'])\nplt.title('model Loss')\nplt.xlabel('Epochs')\nplt.ylabel('Loss')\nplt.legend(['train', 'test'], loc='upper left')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:26:19.527643Z","iopub.execute_input":"2022-07-31T10:26:19.528003Z","iopub.status.idle":"2022-07-31T10:26:19.796521Z","shell.execute_reply.started":"2022-07-31T10:26:19.527968Z","shell.execute_reply":"2022-07-31T10:26:19.795656Z"},"trusted":true},"execution_count":15,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1008x360 with 2 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"### Model evaluation","metadata":{}},{"cell_type":"code","source":"def model_data(model):\n    train_loss, train_accu = fer_model.evaluate(training_set)\n    test_loss, test_accu = fer_model.evaluate(test_set)\n    return print(\"Train accuracy = {:.2f} , Validation accuracy = {:.2f}\".format(train_accu*100, test_accu*100))","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:26:25.219259Z","iopub.execute_input":"2022-07-31T10:26:25.219616Z","iopub.status.idle":"2022-07-31T10:26:25.224873Z","shell.execute_reply.started":"2022-07-31T10:26:25.219585Z","shell.execute_reply":"2022-07-31T10:26:25.223889Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"model_data(fer_model)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:26:28.146167Z","iopub.execute_input":"2022-07-31T10:26:28.146512Z","iopub.status.idle":"2022-07-31T10:27:03.900835Z","shell.execute_reply.started":"2022-07-31T10:26:28.146475Z","shell.execute_reply":"2022-07-31T10:27:03.900076Z"},"trusted":true},"execution_count":17,"outputs":[{"name":"stdout","text":"449/449 [==============================] - 30s 67ms/step - loss: 0.6789 - accuracy: 0.8254\n113/113 [==============================] - 5s 46ms/step - loss: 1.0741 - accuracy: 0.6966\nTrain accuracy = 82.54 , Validation accuracy = 69.66\n","output_type":"stream"}]},{"cell_type":"code","source":"fer_model.save_weights('fer_model_bestweights.h5')","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:27:03.902487Z","iopub.execute_input":"2022-07-31T10:27:03.902735Z","iopub.status.idle":"2022-07-31T10:27:03.972057Z","shell.execute_reply.started":"2022-07-31T10:27:03.902711Z","shell.execute_reply":"2022-07-31T10:27:03.971258Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"markdown","source":"### Confusion Matrix and Classification on training set","metadata":{}},{"cell_type":"code","source":"y_pred = fer_model.predict(training_set)\ny_pred = np.argmax(y_pred, axis=1)\nclass_labels = test_set.class_indices\nclass_labels = {value:key for key,value in class_labels.items()}\n\nfrom sklearn.metrics import classification_report, confusion_matrix\ncm_train = confusion_matrix(training_set.classes, y_pred)\nprint('Confusion Matrix')\nprint(cm_train)\nprint('Classification Report')\ntarget_names = list(class_labels.values())\nprint(classification_report(training_set.classes, y_pred, target_names=target_names))\n\nplt.figure(figsize=(8,8))\nplt.imshow(cm_train, interpolation='nearest')\nplt.colorbar()\ntick_mark = np.arange(len(target_names))\n_ = plt.xticks(tick_mark, target_names, rotation=90)\n_ = plt.yticks(tick_mark, target_names)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:27:52.921852Z","iopub.execute_input":"2022-07-31T10:27:52.922175Z","iopub.status.idle":"2022-07-31T10:28:21.916647Z","shell.execute_reply.started":"2022-07-31T10:27:52.922146Z","shell.execute_reply":"2022-07-31T10:28:21.915664Z"},"trusted":true},"execution_count":20,"outputs":[{"name":"stdout","text":"Confusion Matrix\n[[ 554   63  463 1035  740  731  409]\n [  54    5   50  114   72   81   60]\n [ 519   57  449 1088  773  712  499]\n [ 953  101  796 1888 1422 1238  817]\n [ 625   77  551 1263  949  899  601]\n [ 621   82  574 1245  885  837  586]\n [ 441   57  320  799  634  559  361]]\nClassification Report\n              precision    recall  f1-score   support\n\n       angry       0.15      0.14      0.14      3995\n     disgust       0.01      0.01      0.01       436\n        fear       0.14      0.11      0.12      4097\n       happy       0.25      0.26      0.26      7215\n     neutral       0.17      0.19      0.18      4965\n         sad       0.17      0.17      0.17      4830\n    surprise       0.11      0.11      0.11      3171\n\n    accuracy                           0.18     28709\n   macro avg       0.14      0.14      0.14     28709\nweighted avg       0.17      0.18      0.17     28709\n\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"### Confusion Matrix and Classification on test set","metadata":{}},{"cell_type":"code","source":"y_pred = fer_model.predict(test_set)\ny_pred = np.argmax(y_pred, axis=1)\nclass_labels = test_set.class_indices\nclass_labels = {value:key for key,value in class_labels.items()}\n\n#from sklearn.metrics import classification_report, confusion_matrix\ncm_test = confusion_matrix(test_set.classes, y_pred)\nprint('Confusion Matrix')\nprint(cm_test)\nprint('Classification Report')\ntarget_names = list(class_labels.values())\nprint(classification_report(test_set.classes, y_pred, target_names=target_names))\n\nplt.figure(figsize=(8,8))\nplt.imshow(cm_test, interpolation='nearest')\nplt.colorbar()\ntick_mark = np.arange(len(target_names))\n_ = plt.xticks(tick_mark, target_names, rotation=90)\n_ = plt.yticks(tick_mark, target_names)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:28:25.720414Z","iopub.execute_input":"2022-07-31T10:28:25.720749Z","iopub.status.idle":"2022-07-31T10:28:31.184857Z","shell.execute_reply.started":"2022-07-31T10:28:25.720709Z","shell.execute_reply":"2022-07-31T10:28:31.183397Z"},"trusted":true},"execution_count":21,"outputs":[{"name":"stdout","text":"Confusion Matrix\n[[127  17  75 257 203 163 116]\n [ 13   2   7  36  23  15  15]\n [145  10 118 262 199 167 123]\n [257  16 207 426 373 275 220]\n [164  19 122 321 244 218 145]\n [163  16 141 308 266 200 153]\n [101  15  82 209 187 162  75]]\nClassification Report\n              precision    recall  f1-score   support\n\n       angry       0.13      0.13      0.13       958\n     disgust       0.02      0.02      0.02       111\n        fear       0.16      0.12      0.13      1024\n       happy       0.23      0.24      0.24      1774\n     neutral       0.16      0.20      0.18      1233\n         sad       0.17      0.16      0.16      1247\n    surprise       0.09      0.09      0.09       831\n\n    accuracy                           0.17      7178\n   macro avg       0.14      0.14      0.14      7178\nweighted avg       0.17      0.17      0.16      7178\n\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"emotions = {0:'angry',\n            1:'disgust',\n            2:'fear',\n            3:'happy',\n            4:'neutral',\n            5:'sad',\n            6:'surprise'}\n\npath = '../input/emotor-test-set'","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:28:31.186433Z","iopub.execute_input":"2022-07-31T10:28:31.186903Z","iopub.status.idle":"2022-07-31T10:28:31.201066Z","shell.execute_reply.started":"2022-07-31T10:28:31.186867Z","shell.execute_reply":"2022-07-31T10:28:31.198261Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"datagen = ImageDataGenerator(rescale=1/255.)\n\npred_generator = datagen.flow_from_directory(path,\n                                             batch_size=64,\n                                             class_mode=None,\n                                             shuffle=False,\n                                             target_size=(image_size, image_size),\n                                             color_mode='grayscale',)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:28:36.934200Z","iopub.execute_input":"2022-07-31T10:28:36.934555Z","iopub.status.idle":"2022-07-31T10:28:41.621447Z","shell.execute_reply.started":"2022-07-31T10:28:36.934523Z","shell.execute_reply":"2022-07-31T10:28:41.620495Z"},"trusted":true},"execution_count":23,"outputs":[{"name":"stdout","text":"Found 7178 images belonging to 1 classes.\n","output_type":"stream"}]},{"cell_type":"code","source":"preds = fer_model.predict(pred_generator)\npreds = tf.argmax(preds, axis=1).numpy()","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:28:59.354953Z","iopub.execute_input":"2022-07-31T10:28:59.355275Z","iopub.status.idle":"2022-07-31T10:29:23.078302Z","shell.execute_reply.started":"2022-07-31T10:28:59.355244Z","shell.execute_reply":"2022-07-31T10:29:23.077425Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"code","source":"images = os.listdir(os.path.join(path, 'EMOTOR_TEST'))\npred_dict = [{'img_name':img, 'label':emotions[pred]} for img, pred in zip(images, preds)]","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:29:23.080556Z","iopub.execute_input":"2022-07-31T10:29:23.080992Z","iopub.status.idle":"2022-07-31T10:29:23.097657Z","shell.execute_reply.started":"2022-07-31T10:29:23.080943Z","shell.execute_reply":"2022-07-31T10:29:23.096714Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"code","source":"import csv\nwith open('predictions.csv', 'w', newline='') as f:\n    fieldnames=['img_name', 'label']\n    writer = csv.DictWriter(f, fieldnames=fieldnames)\n    writer.writeheader()\n    writer.writerows(pred_dict)","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:29:23.100640Z","iopub.execute_input":"2022-07-31T10:29:23.100898Z","iopub.status.idle":"2022-07-31T10:29:23.126854Z","shell.execute_reply.started":"2022-07-31T10:29:23.100874Z","shell.execute_reply":"2022-07-31T10:29:23.126179Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"code","source":"from IPython.display import FileLink\nFileLink(r'./fer_model.h5')","metadata":{"execution":{"iopub.status.busy":"2022-07-31T10:29:32.737828Z","iopub.execute_input":"2022-07-31T10:29:32.738174Z","iopub.status.idle":"2022-07-31T10:29:32.746146Z","shell.execute_reply.started":"2022-07-31T10:29:32.738142Z","shell.execute_reply":"2022-07-31T10:29:32.745104Z"},"trusted":true},"execution_count":28,"outputs":[{"execution_count":28,"output_type":"execute_result","data":{"text/plain":"/kaggle/working/fer_model.h5","text/html":"<a href='./fer_model.h5' target='_blank'>./fer_model.h5</a><br>"},"metadata":{}}]}]}