Assignment-2-Three-Layer-Neural-Network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Building and Training a Neural Network!\n",
    "We have learned a bunch of modularized layers. It's time to put them all together, and train a NeuralNet to recognize the digits in MNIST. In this notebook we will walk you through training a simple three-layer NeuralNetwork and then set you free to build the best net that you can to perform well on MNIST dataset.\n",
    "\n",
    "## Goal:\n",
    "\n",
    "Open up the file `nnetwork.py`; you will see that the `three_layer_neuralnetwork` function computes the loss and gradients for a three-layer NeuralNetwork, your goal here is to complete the missing parts. Note that this function uses the \"modular\" layers defined in `layers.py`, you will have to complete that before proceeding with this assignment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "%pylab inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# import the layers\n",
    "from nnetwork import *\n",
    "# As usual, a bit of setup\n",
    "%matplotlib inline\n",
    "plt.rcParams['figure.figsize'] = (10.0, 8.0) # set default size of plots\n",
    "plt.rcParams['image.interpolation'] = 'nearest'\n",
    "plt.rcParams['image.cmap'] = 'gray'\n",
    "\n",
    "# for auto-reloading external modules\n",
    "# see http://stackoverflow.com/questions/1907993/autoreload-of-modules-in-ipython\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "def rel_error(x, y):\n",
    "  \"\"\" returns relative error \"\"\"\n",
    "  return np.max(np.abs(x - y) / (np.maximum(1e-8, np.abs(x) + np.abs(y))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Digit Recognition\n",
    "\n",
    "We are going to build a three-layer neural network and use it for recognizing the digits. For this purpose we will be using the famous MNIST dataset.\n",
    "\n",
    " Note that the neural network parameters will be stored in a dictionary (`model` below), where the keys are the parameter names and the values are numpy arrays. Below, we initialize toy data and a toy model that we will use to verify your implementations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train data shape:  (50000L, 784L)\n",
      "Train labels shape:  (50000L,)\n",
      "Validation data shape:  (10000L, 784L)\n",
      "Validation labels shape:  (10000L,)\n",
      "Test data shape:  (10000L, 784L)\n",
      "Test labels shape:  (10000L,)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdwAAAHpCAYAAAAoFI3rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJtoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa40a0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import mnist_loader as mnl\n",
    "# Invoke the above function to get our data.\n",
    "X_train, y_train, X_val, y_val, X_test, y_test = mnl.get_mnist_data()\n",
    "print 'Train data shape: ', X_train.shape\n",
    "print 'Train labels shape: ', y_train.shape\n",
    "print 'Validation data shape: ', X_val.shape\n",
    "print 'Validation labels shape: ', y_val.shape\n",
    "print 'Test data shape: ', X_test.shape\n",
    "print 'Test labels shape: ', y_test.shape\n",
    "\n",
    "# display the mean training image...\n",
    "mean_train=np.mean(X_train,axis=0)\n",
    "plt.imshow(mean_train.reshape(28,-1),cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.title('Mean Training Image')\n",
    "\n",
    "#\n",
    "# Centralized the train, test, and validation data...\n",
    "X_train=X_train-mean_train;\n",
    "X_val=X_val-mean_train;\n",
    "X_test=X_test-mean_train;\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Sanity check loss\n",
    "Now we have all the ingredients in place. Now we are going to build a complete network. After you build a new network, one of the first things you should do is sanity check the loss. When we use the softmax loss, we expect the loss for random weights (and no regularization) to be about `log(C)` for `C` classes. When we add regularization this should go up."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Layer 1 Variance = 0.000200034590056', 'Layer 2 Variance = 2.15525838512e-08', 'Layer 3 Variance = 7.35341953742e-09']\n",
      "Sanity check loss (no regularization):  2.30258513948\n",
      "Sanity check loss (with regularization):  2.30849286332\n"
     ]
    }
   ],
   "source": [
    "#%pdb\n",
    "#Initialize the weights\n",
    "model = init_three_layer_neuralnet(weight_scale=0.01,input_feat_dim=32*32*3)\n",
    "\n",
    "X = np.random.randn(1000, 3* 32* 32)\n",
    "y = np.random.randint(10, size=1000)\n",
    "\n",
    "\n",
    "# Compute the loss\n",
    "loss, _ = three_layer_neuralnetwork(X, model, y, reg=0,verbose=1)\n",
    "\n",
    "# Sanity check: Loss should be about log(10) = 2.3026, for a perfect random initalization of weights\n",
    "print 'Sanity check loss (no regularization): ', loss\n",
    "\n",
    "# Sanity check: Loss should go up when you add regularization\n",
    "loss, _ = three_layer_neuralnetwork(X, model, y, reg=1)\n",
    "print 'Sanity check loss (with regularization): ', loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gradient check\n",
    "After the loss looks reasonable, you should always use numeric gradient checking to make sure that your backward pass is correct. When you use numeric gradient checking you should use a small amount of artifical data and a small number of neurons at each layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Layer 1 Variance = 2.18852883633', 'Layer 2 Variance = 3.38955058375', 'Layer 3 Variance = 0.794160363111']\n",
      "W1 max relative error: 2.263985e-07\n",
      "W2 max relative error: 1.248766e-07\n",
      "W3 max relative error: 2.938098e-06\n",
      "b1 max relative error: 8.005930e-08\n",
      "b2 max relative error: 5.242194e-09\n",
      "b3 max relative error: 5.980053e-08\n"
     ]
    }
   ],
   "source": [
    "from gradient_check import *\n",
    "\n",
    "num_inputs = 3\n",
    "feat_dim = 5\n",
    "reg = 0.0\n",
    "num_classes = 10\n",
    "X = np.random.randn(num_inputs, feat_dim)\n",
    "y = np.random.randint(num_classes, size=num_inputs)\n",
    "\n",
    "model = init_three_layer_neuralnet(input_feat_dim=feat_dim)\n",
    "import copy\n",
    "\n",
    "mcopy=copy.deepcopy(model)\n",
    "\n",
    "loss, grads = three_layer_neuralnetwork(X, model, y,verbose=1)\n",
    "\n",
    "for param_name in sorted(grads):\n",
    "    f = lambda _: three_layer_neuralnetwork(X, model, y)[0]\n",
    "    param_grad_num = eval_numerical_gradient(f, model[param_name], verbose=False, h=1e-6)\n",
    "    e = rel_error(param_grad_num, grads[param_name])\n",
    "    #print e\n",
    "    print '%s max relative error: %e' % (param_name, e)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Implementing Gradient Descent and Its Variants\n",
    "To train the network we will use SGD with Momentum. Last assignment you implemented vanilla SGD. You will now implement the momentum update and the RMSProp update. Open the file `classifier_trainer.py` and familiarze yourself with the `ClassifierTrainer` class. It performs optimization given an arbitrary cost function data, and model. By default it uses vanilla SGD, which we have already implemented for you. First, run the optimization below using Vanilla SGD:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  5\n",
      "Final loss with vanilla SGD: 0.447344\n"
     ]
    }
   ],
   "source": [
    "input_size = 4\n",
    "num_classes = 3\n",
    "num_inputs = 5\n",
    "\n",
    "def init_toy_data():\n",
    "  X = np.linspace(-0.2, 0.5, num=num_inputs*input_size).reshape(num_inputs, input_size)\n",
    "  y = np.array([0, 1, 2, 2, 1])\n",
    "  return X, y\n",
    "\n",
    "from classifier_trainer import ClassifierTrainer\n",
    "\n",
    "\n",
    "X,y=init_toy_data()\n",
    "\n",
    "#print \"Input: \", X.shape\n",
    "#print \"labels: \",y.shape\n",
    "\n",
    "model = init_three_layer_neuralnet(num_classes=num_classes,input_feat_dim=input_size)#returning nueral net model\n",
    "\n",
    "trainer = ClassifierTrainer()\n",
    "# call the trainer to optimize the loss\n",
    "# Notice that we're using sample_batches=False, so we're performing Gradient Descent (no sampled batches of data)\n",
    "best_model, loss_history, _, _ = trainer.train(X, y, X, y,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             reg=0.001,\n",
    "                                             learning_rate=1e-1, momentum=0.0, learning_rate_decay=1,\n",
    "                                             update='sgd', sample_batches=False,\n",
    "                                             num_epochs=100,\n",
    "                                             verbose=False)\n",
    "print 'Final loss with vanilla SGD: %f' % (loss_history[-1], )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now fill in the **momentum update** in tmuhe first missing code block inside the `train` function, and run the same optimization as above but with the momentum update. You should see a ch better result in the final obtained loss:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  5\n",
      "Final loss with Momentum SGD: 0.476309. We get: 0.046405\n"
     ]
    }
   ],
   "source": [
    "model = init_three_layer_neuralnet(num_classes=num_classes,input_feat_dim=input_size)\n",
    "trainer = ClassifierTrainer()\n",
    "# call the trainer to optimize the loss\n",
    "# Notice that we're using sample_batches=False, so we're performing Gradient Descent (no sampled batches of data)\n",
    "best_model, loss_history, _, _ = trainer.train(X, y, X, y,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             reg=0.001,\n",
    "                                             learning_rate=1e-1, momentum=0.9, learning_rate_decay=1,\n",
    "                                             update='momentum', sample_batches=False,\n",
    "                                             num_epochs=100,\n",
    "                                             verbose=False)\n",
    "correct_loss = 0.046405\n",
    "print 'Final loss with Momentum SGD: %f. We get: %f' % (loss_history[-1], correct_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  5\n",
      "Final loss with Nestrov-Momentum SGD: 0.282566. We get: 0.046405\n"
     ]
    }
   ],
   "source": [
    "model = init_three_layer_neuralnet(num_classes=num_classes,input_feat_dim=input_size)\n",
    "trainer = ClassifierTrainer()\n",
    "# call the trainer to optimize the loss\n",
    "# Notice that we're using sample_batches=False, so we're performing Gradient Descent (no sampled batches of data)\n",
    "best_model, loss_history, _, _ = trainer.train(X, y, X, y,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             reg=0.001,\n",
    "                                             learning_rate=1e-1, momentum=0.9, learning_rate_decay=1,\n",
    "                                             update='nestrov-momentum', sample_batches=False,\n",
    "                                             num_epochs=100,\n",
    "                                             verbose=False)\n",
    "correct_loss = 0.046405\n",
    "print 'Final loss with Nestrov-Momentum SGD: %f. We get: %f' % (loss_history[-1], correct_loss)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now also implement the **RMSProp** update rule inside the `train` function and rerun the optimization:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  5\n",
      "Final loss with RMSProp: 0.144823. We get: 0.046405\n"
     ]
    }
   ],
   "source": [
    "model = init_three_layer_neuralnet(num_classes=num_classes,input_feat_dim=input_size)\n",
    "trainer = ClassifierTrainer()\n",
    "# call the trainer to optimize the loss\n",
    "# Notice that we're using sample_batches=False, so we're performing Gradient Descent (no sampled batches of data)\n",
    "best_model, loss_history, _, _ = trainer.train(X, y, X, y,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             reg=0.001,\n",
    "                                             learning_rate=1e-1, momentum=0.9, learning_rate_decay=1,\n",
    "                                             update='rmsprop', sample_batches=False,\n",
    "                                             num_epochs=100,\n",
    "                                             verbose=False)\n",
    "correct_loss = 0.046405\n",
    "print 'Final loss with RMSProp: %f. We get: %f' % (loss_history[-1], correct_loss)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now go and implement the (**`SMORMS3`**)[http://sifter.org/~simon/journal/20150420.html] update rule inside"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  5\n",
      "Final loss with SMORMS3: 1.251001. We get: 0.046405\n"
     ]
    }
   ],
   "source": [
    "model = init_three_layer_neuralnet(num_classes=num_classes,input_feat_dim=input_size)\n",
    "trainer = ClassifierTrainer()\n",
    "# call the trainer to optimize the loss\n",
    "# Notice that we're using sample_batches=False, so we're performing Gradient Descent (no sampled batches of data)\n",
    "best_model, loss_history, _, _ = trainer.train(X, y, X, y,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             reg=0.001,\n",
    "                                             learning_rate=1e-1, momentum=0.9, learning_rate_decay=1,\n",
    "                                             update='SMORMS3', sample_batches=False,\n",
    "                                             num_epochs=100,\n",
    "                                             verbose=False)\n",
    "correct_loss = 0.046405\n",
    "print 'Final loss with SMORMS3: %f. We get: %f' % (loss_history[-1], correct_loss)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Overfit small data\n",
    "A nice trick is to train your model with just a few training samples. You should be able to overfit small datasets, which will result in very high training accuracy and comparatively low validation accuracy.\n",
    "\n",
    "**Remember your model should start improving with in one to two epochs, if it is not improving then your initalization might be wrong. Print the variance across the layers to check whether your initalization is fine.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train distribution [4932 5678 4968 5101 4859 4506 4951 5175 4842 4988]\n",
      "Validation distribution [ 991 1064  990 1030  983  915  967 1090 1009  961]\n",
      "Test distribution [ 980 1135 1032 1010  982  892  958 1028  974 1009]\n"
     ]
    }
   ],
   "source": [
    "#Finding data distribution...\n",
    "print \"Train distribution\",np.bincount(y_train)\n",
    "print \"Validation distribution\",np.bincount(y_val)\n",
    "print \"Test distribution\",np.bincount(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Layer 1 Variance = 0.131559658903', 'Layer 2 Variance = 0.172737662145', 'Layer 3 Variance = 0.0740608756484']\n",
      "N:  10000\n",
      "Finished epoch 0 / 10: cost 2.761902, train: 0.115000, val 0.123900, lr 9.000000e-02\n",
      "Finished epoch 1 / 10: cost 0.450187, train: 0.892000, val 0.891300, lr 8.550000e-02\n",
      "Finished epoch 2 / 10: cost 0.762332, train: 0.908000, val 0.897900, lr 8.122500e-02\n",
      "Finished epoch 3 / 10: cost 0.398459, train: 0.927000, val 0.904500, lr 7.716375e-02\n",
      "Finished epoch 4 / 10: cost 0.549461, train: 0.926000, val 0.909300, lr 7.330556e-02\n",
      "Finished epoch 5 / 10: cost 0.379276, train: 0.926000, val 0.906600, lr 6.964028e-02\n",
      "Finished epoch 6 / 10: cost 0.749657, train: 0.910000, val 0.906500, lr 6.615827e-02\n",
      "Finished epoch 7 / 10: cost 0.465106, train: 0.926000, val 0.903600, lr 6.285036e-02\n",
      "Finished epoch 8 / 10: cost 0.445714, train: 0.932000, val 0.909300, lr 5.970784e-02\n",
      "Finished epoch 9 / 10: cost 0.756004, train: 0.937000, val 0.911500, lr 5.672245e-02\n",
      "Finished epoch 10 / 10: cost 0.519772, train: 0.933000, val 0.913700, lr 5.388632e-02\n",
      "finished optimization. best validation accuracy: 0.913700\n"
     ]
    }
   ],
   "source": [
    "from classifier_trainer import *\n",
    "trainer = ClassifierTrainer()\n",
    "model=init_three_layer_neuralnet(weight_scale=1,input_feat_dim=X_train.shape[1],num_neurons=(10,10))\n",
    "three_layer_neuralnetwork(X_train[:10000], model,y_train[:10000],0.001,verbose=1)\n",
    "best_model, loss_history, train_acc_history, val_acc_history = trainer.train(X_train[:10000], y_train[:10000], X_val, y_val,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             num_epochs=10, reg=0.01,batch_size=16,\n",
    "                                             momentum=0.01, learning_rate_decay = 0.95,\n",
    "                                             learning_rate=0.09,update='momentum', verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from vis_utils import visualize_grid\n",
    "def show_net_weights_mnist(model,weights='W1'):\n",
    "    plt.imshow(visualize_grid(model[weights].T.reshape(-1, 28, 28), padding=3,isgrayscale=True).astype('uint8'))\n",
    "    plt.gca().axis('off')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot the training loss evolution history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmoAAAHuCAYAAAAvGyAsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xee6cf28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAegAAAHpCAYAAABJDK+pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJztvXm0Zcd1n7cLIOYZaAzdjcY8z4QAURxgUpZMW4poccmy\nFVmKRVlaXHZiO4qTaLQsUYq4FCWKxGQpUsJFa0o00bZoJo4VhcvCQGgiQApTA2gA3egB3WjM80ze\n/HEvLnZ977067/a7771C4/vWwsKpV/ecU6eGU332r/auMhqNQkRERPrikPUugIiIiCzECVpERKRD\nnKBFREQ6xAlaRESkQ5ygRUREOsQJWkREpEOcoN9hlFKeL6Wcs97lkDallIdLKd+0yvc4a9IfyjJ+\ne04p5WullDV7Z6zHPdeCWepd3tkcVB1f3mLygn9p8iJ4vpTyXCnljNFodNxoNHr4AK73oVLK7oHf\n/EYp5WcPuNCrwNv4JT+a/FdRSnnvpC1L+tunl/jbrzZvMBrtmvSHFQdDKKX8dCnlt1d6nXcC86x3\nObh5u720ZPmMIuLbJi+C40aj0fGj0ejR1glzmMQWnVTWilLKu94B970txuP269LfboiI3Yv87aY1\nLJeIzBkn6HcYk6/J8ybHv1FK+dVSyv9TSnkhIj5USvnWUsrWyRfZnlLKPy+lHB0R/yEiNuWv8aVu\nscR9P1VK2VVKebaUclsp5QOTv59RSnmxlHJy+u21pZTHSimHTtL/cFKmp0opf1RKOQvP85+XUh6I\niPtnrIsTSimfKaXsnTzrz775j5RSyvmllP9YSnmilPJ4KeX/KKWckM59uJTyw6WUOyPi+cnvv1ZK\n+QellJ2Tc348/b6UUn60lPLg5Jq/X0o5KeX/Z5PznsjnkdFo9HpE/HlE/LXJeadFxGER8Qf420UR\ncXPrvrQulFLOLaXcPGnf/6+U8iuLfBV/L5+vlPK3IuLHIuK7Jv3jK5O/f6yU8tDkettLKX9/iXb4\n+kmfeLaU8mgp5ReH7jk574hSyi+XUh6Z/PdLpZTDJ3k3lVK+Y3L8/slzfusk/U0o4xdLKf/DpH9t\nnzzPoqS6fK6Uck8p5aMp74LJfZ+ZlPX3lrgG6/3GSd+7dVJ/ny+lbCil/J+TOvnLUsrZ6fxFx9Ik\n76hSym9OnmXrpI/uTvmbSin/pozH1/ZSyj9d6lmlA0ajkf8dhP9FxI6I+KZF/v61iDhvcvwbEfFM\nRLx3kj4yIvZFxPsn6RMi4t2T4w9GxO6Be/56RPzsEnnfExEnxfgfhf98cp/DJ3n/PiL+UfrtL0XE\npybH3x4RD0TExZNzfyIibsXz/L8RcWJEHLHIfc+Z/OaQRfL+MCJ+NSKOiohTI+IvIuLjk7zzI+Kb\nYjz5bYjx1+gvpXMfjogvR8TmiDgi3ed/m6SviohXIuLiye//y4j404jYNLnmr0XE70zyLouI5yPi\nAxFxeET8YkS8HhF/fYm6/JcR8bnJ8XdGxG9GxDfjbw8u475V3UTEn0XEL0TEuyLi/RHxbET8Fn67\n1PP91Ju/naSPmZx/4SR9ekRctsTz/FlEfM/k+OiIeM8y7/kzk2fbMPnv1oj4mUneJyLif54c/3hE\nPBgRP5/O+6XJ8cci4rWI+IEY/+PyH0XEI40+/p0Rccbk+O9FxAsRcfok/bsR8WOT48Mj4n1LXIP1\nfmNEbIuIcyPi+Ii4J8Z9/q9HxKGT9v1XyxxLPx8RfxLjsbs5Iu6MiF2TvEMi4vaI+BeTNj43Ih6K\niA+v9/vK/5bob+tdAP9bpYYdTyDPR8TTk//+7eTvnKB/A+ftjIiPR8Tx+PuHYgUT9CK/fSoirpwc\nf1dEfHFyfOjkhXPdJP0fIuIfpvMOiYgXI2JLep4PNe5TvQzT30+P8cv+yPS3746I/7jEdT4aEV9O\n6R0R8bFF7rMp/e0vIuLvTY7vjTThRsTGGE8Mh8Z4wv2dlHd0RLwaS0/QH4qIJybHn4rx5HJMRDya\n/vaZgfsekusmIs6K8T8Kcn38dkT89jKf76ff/O0kfcyk331HRBw10Bdumpy/YYm2W+qeD0bE30p5\nH46IHZPjb4qIO1If+oGI+LN0v49Ojj8WEQ+g7r8WEactsx9/JSI+Mjn+zRj/Y2LzwDnTep+k/yQm\nE/sk/T9GxL9P6W+LiK8scyw9FBF/I+X9QEzGbUS8JyJ24twfizT5+19f/2niPngZRcS3j0ajkyb/\nfccSv+HCr78TEd8aEQ9PTG/fMI/ClFL+m4nJ7ZlSytMx/hf+hkn2v4uIy8p4dfnfiIhnR6PRbZO8\nsyPiU6WUpyfnPTn5++Z0+ebitSU4O8ZflPvStX8txl/SUUo5vZTye2Vs+n42xpPVKbjGYvfNOv9L\nEXFsut8fpnttjYg3YvwPhY0RsefNk0aj0Uvx1nMuxp9HxLGllCtirDXfMhqNXoyI3ZO//bWIuHkZ\n981sioinRqPRKwf4fBWT8nxXjL9I95ZS/u9SysVLPM8PxNgkf+/EnPufLPOem2L8D8o32TX5W8T4\nq/yiibn/moj4rYjYUko5JSKuj7fqp7r+pO5jqecqYwnjK6k+r4i3+vEPx/gr/C9LKXeXUr5/iedd\njP3p+JWIeAzpaXkGxtKmqNttTzo+O8Yy1dOp/D8WEafNUE5ZQ9ZlUY30y2Ri/GgZ67//NMba5lmx\n/MVfi608viEi/tsYf8ndM/nbUzHRq0ej0SullM9GxPdGxCUxfpm+ya4Yf5X/7iz3XAa7Y/yVespo\nNPraIvmfjIivRsQVo9HomYnW+L+s4L67IuL7R6PRnzGjlLIvIi5N6aNj4T8G3rrpuL6+FBF/OyI2\njkajbZOsWyZ/uzLemoBa9z0nJfdFxMmllKNGo9HLk7+tqN1Ho9EfR8Qfl1KOiIifi4hPx0Qnx+8e\njIi/PynT34mIf13SmoQGe2P8NXpvKu/eyTVfKqXcHhE/FBF3jUaj10spfxoR/3WMzf9PLfO5pkx0\n4P89xqbnPxuNRqOJlv1mP94fY+tTlFLeHxFfKKXcNBqNts94qyXrfGgsxbgdt0TEfZP0lnT67hhb\nGC6asTyyTvgF/c6mWtBVSjmslPI9pZQTRqPRV2NsIv/qJHt/RJxSSjl+4HrvKqUcmf47PCKOi/FX\n2xOllMNLKf8yxlpb5rci4vtjPMHkhUm/FhE/Xkq5bFLGE0opf/cAnjWX6cjJ8/xxRPxPpZTjSimH\nlPFCrzcnkGNjbEp/rpSyOcYvxZXwaxHxyTJZ4FZKObWU8rcnef86Ir6tjBczHR5jjXRobN4cY335\n1vS3L07+tnc0Gu1Yxn2njEajnTFeIf7Tk37w3hibVpc7QT8aEeeUMnb1KqWcVkr59lLKMTE2nb8Y\nb/WlilLK95ZSTp0kn53cc7F/NJHfjYh/MVlQtSHGUkHuOzdFxH8Rb61mvzEi/kkc+Or2YyZleyIi\nDpl8IV+RnuPvllLOnCSfmeE5Iuqx2PKPHhpLfxARP1ZKOXHSb/9JvNWGfxnjBY0/PFlMdmgp5YpS\nynXLLKOsMU7Q7zxGOOYL+HsjYsfErPvxGC9IidFodF+MX4jbJytEF1vFPYqIH42xGfLN/74QEX80\n+W9bjLXxl2P8ZffWiaPRrTF+md0+Go12p79/LiL++4j4vUmZ7oqIv7nE87R4AeX6UET8gxgv5tka\nYx3vsxHx5nN9IiKujfGE8X9FxL9Zxr1a+Z+KiM/H+IvyuRibYL8+ImI0Gm2N8UTyOzH+Anwqhs32\nN8XYrPnF9LdbY2yiv2U5912kzN8TEe+NsXn9ZyPi92OsVy/n+T47+f+TpZQ3XcH+q4h4ZHK9GyLi\nHy9x7t+MiLtLKc/HeIHgfzoajV5dxj3/uxj/o+LOyX+3Tf72JjfF+B9ab1oTbo7xJJvN24uNgUXv\nOWmnX4xxHT4a48k51/91EfHnk+f4dxHxz0ZLxxxo3bNVpqGx9DMxNmvviPE/QD8bkzac/KP722Js\n8t8eEY/H2CLQ+ke3rCNlNDoQ66DI/CmlfCHGi6X+1XqXRSJKKb8fEVtHo9En1rsscmCUUv5xjBfV\nfeN6l0Vmxy9o6YJSyvUx/mL9/fUuyzuVUsp1EzP/IaWUb4mx3PC59S6XLJ8yjivw/kkbXhxjN6w/\nXO9yyYHhIjFZd0opvxljf+d/Nln9K+vDGRHxb2O8QG13jH3T71jfIsmMHB7jdQfnxlgH/92I+F/X\ntURywGjiFhER6ZA1/4L++Mc/vux/Eczyj4cysDHMvP4hwvu0rruS35K1us+nP/3pZn7mB3/wB5d1\nz8Xue6Dt8bWv1YtiDzlkfipNLuPb/R+un/nMZ2b6/c///M9Pj+fVVu80WE9DY225/OiP/uiyf/vr\nv/7rc7mnrA4f+9jHZuoUatAiIiId4gQtIiLSIeu+SCybhWgSGjK1zcskyXNpRs0MmVTzuUMmr3mZ\nZ4fu06qbedbbvH7b4tBDD132fWY1h7fO/epX6xgbhx122PT4YDMJ91L+1RrfS91jpazkWgdbH1ot\nOP6POOKI6fHLL79c5a3kvTdLW65m2/kFLSIi0iFO0CIiIh2y7ibulilhJaZbkk2UPO/VV1+t0rOY\nxF5//fUlf5vNoEP5vOe73lU3Tcs824N5bJ5lWImJnqbozEsvvbRk3qzk9jjyyCOrvFafiJif6Xae\nK9p7WME+S1syL5/7xhtvVHlMH3744dPjIdmJ47DFkESX6dGEPcsq9KHy5zo/9th6YzBe9+ijj170\nvIi23Mhy8LevvfZalc75LANN57nd18pDaDH8ghYREekQJ2gREZEOcYIWERHpkHXXoFvM0wUi6wTU\nr6gZ5nzqli+88EKVpn79yiuvTI/POKPekZF61jHHHDM9pvaVdbKIWuekXrISd6d51fFqas5ZO6LO\nRNeKrDu16nAxchuwTLxPTp900klVHvtEyzWE7U4NLvdN1ss8o6gdaNS+WdaJ8LdMcxzmsZSPI9ru\ndlnTXCz94otvhXtnW7HPcMyeeOKJ02PWP589t+U819OsFrO4tXIs8be5njZs2NC8zwknnDA9Puqo\no6o8vm+fe+65Jc9lW952221VOo/ZIQ06P9/Qe6PVr1f6fvULWkREpEOcoEVERDrECVpERKRD1lyD\nXokGyt+2fOSoO+X7tnTjiFpbos5EPYJlevrpp6fH1LpPPvnkKp3zsx692HWz9nL88cdXeUO+mlkL\nm6duOS9YT2zXXBd81tZvWaf0i6TulPUvXve8886r0rnPsE+cdtppVfqxxx6r0lnnZF/kmodWX2z5\nCc/KLH7QrXyWKT9fHhsRw/7iuX2OO+64Ko/9ILcH83huHg/sI0NrAvLvs/4ZEfHss89W6Zbf/Tzb\nbl4MadB57UTWmCMWasf5HcXrMj5EHi+8LjVnjstdu3ZNj/kev+6666r0tm3bpsf33HNPlcf2yH1o\n48aNVR7bfTXp720tIiIiTtAiIiI94gQtIiLSIWuuQbd0jiE9mvpQ1jKoFdFvNesTvA9jxV5wwQVL\n3vPBBx+s0i3N5PHHH48Wp5xyyvSY2gr1lKyD8NlYxqGtElvnzoshPSu3F3XY1rlPPvlklUe9Meu7\n1OqpQXNNwDPPPLPkdR999NEl78v6prad/W4jai2P2h11y9zW/O16+dLm+1BHZt/MddrylY1YqAfn\ndNYaIxbqvZs2bVqyDFl7jIi4/PLLp8dnnnlmlcexsnfv3iqd9VL2Eabz+Gbfo38v02uhUQ/FIed7\nMT/7kG9zboPTTz+9ynvkkUeqdH526rtbt26t0vv376/S+V3NNQ4sf37nUFe+9957q3Qe0ywT15y0\n1gDpBy0iInIQ4gQtIiLSIevuZjULDMOXzWutLeUiahMTzQ5nnXVWlaZ5OXPZZZdV6e3bt1fp559/\nfnrMUJ80W+U0zaQ0i2bzE69LkwtdCLLJlfVCd7RZaLnm0Mw7i4scTW3ZnMnyMsxmTp966qnNMvE+\n2YxKc9nu3burdDat0Y2HZtMtW7ZU6Ww2veiii5Ysf0TEU089NT2mOZlmUZpRZ2GWcZnbbqjd829p\nCqTJnn0ztwHr9JprrqnSWXridWmizH3x4YcfrvJojuV9s+mWJnnKE/m9snnz5irvgQceqNIsB9t2\nXuT2obsTxyj7de6bWaZZjHwtPgvf1fn9RVMzpQyWKctjbLv777+/Sufx86EPfajKO/vss6v0fffd\nNz3mfJDf8RELpbT87Jq4RUREDkKcoEVERDrECVpERKRDunKzItQIW9vRUZukHpT1Fupk1O6y+we1\nyKxNRCx0u8p6IzU16jZZe6EmSzeGJ554Ysky5byIheHxct0MbRM3C/ncWVy7mM/fsn1y+bkGgC41\nuX8xb0gLy/1pz549VR7103379k2PsztNxMLnoYaV+1dLQ4+o25o6eCuU5GqSx0srTG5E3d+YR83w\nwgsvrNLZTZE6Psl9hm48rNNcDrrasU/Qve7cc8+dHnMscd1IDg9Ltx6GfyUHql0ObUmaxynLz37M\nPp+1V9YT34vZ7Y1rAugqmbV8jlFqw3S7ym58+Z4RC9+/uR/w2bjWIJd5586dVR77fGtt1ErxC1pE\nRKRDnKBFREQ6xAlaRESkQ9Zcg24xpIfS9p/T9I+j9pJ1AmoV1COy7yn1XmokzM/6L7cd5H2zBsQ8\n6hhZQ2e4O2ro1IeyVklteyW0dLKWz3dEXWY+K/1Wc5mpGfLcrBmyTllP1Orz81AzpD7Xus9Q6NLs\nn821E/SpzOsWqFvSj3W1Qn2y/Pn5qIOz7VrbZXJNBsdSvjZ1Pl4rrzdorVWJqPsB24rjg77Ouc6z\nL3zEQr/7Wfr40Jary2VojU9O873HZ+e7Lr8X2Rdb7yT6gHMcZn9ltjPXkXA9Qa5XtlVeLxBRvxep\nv7M/td6ZfFaem+t1pWPSL2gREZEOcYIWERHpkK5CfQ6ZRenGkPNbblXMH9q1Jd+HJpehXadymmEc\nuZQ/m4lomqJrQjbtsAwtM3tEbUZtmbxmJZ87tKMWTXytcHitXY5ovqdZK9fTkPsZzcnZ7MUy0XUn\n9wu6YfC6JEsfNJfR7JjNwKxDjg/WzVrA+meZ8thi2EaWl2Fzs4sNJQa2R643hs1k++Qy0vxKF7+W\n2xV3q2P/ymP4kksuaf52aEe0eZGvy7bje4Tvtvy+2rFjR5XHvnnHHXcseZ1WiNqrrrqqyqOb23d+\n53dW6fe9733TY77L6M6VXSfZn7LbZEQ9/jnv8N28mrvI+QUtIiLSIU7QIiIiHeIELSIi0iFduVlR\nxxgKmZZ1QP6W7hMtjYrbN2btgqE8uZVj3jowotYXqfFQq8huGdRwqU1efPHF02PqZjncHa8bUesp\n1DhXC2p31Cbz89JFjr/NWjLLz/vka3FbuFYoyYiI6667bnp84403VnnU57I2xrajPkrNKuuPbCvq\nZrlvUhdnn18tLYzPl13d6AbTWuOQ+/Bi1z3nnHOqdB5LdK+jjpnHGtuVOmZe40A3N7YzQ77mdQrM\nY3/LfZFjlG4+a9WW+XlZXtYT33VZJ+c7NGvOEXU/Zh/h+iCOgQzbneMjh/7keiGuLcrjhyFDuU4h\n684sP9eYqEGLiIi8w3CCFhER6RAnaBERkQ7pSoMeorU9Gv0IqW9lPzdqztQ1sxZDf8SWf3JExLd8\ny7dMj+n7eMUVV1Tpu+66a3pMP+53v/vdVTqH2aO2zbCBrItcb9R3V0s/YZ0Ohb/M0O8w63PUvlgX\nWVdjO1Pn++AHP1ils7Z/9913V3kMbdjajo6hDJnOuhrDRVInz1or246aG/1N1wKWgWsCcj2xD5x/\n/vlVmjpt1v7Yx7n243Of+9z0+POf/3yVxzGa78vrPPTQQ1Wavtu5jtm/SO7HDGvK8JFch7Fa5D7E\nfklfYNZF1ufZ57ndZK5jhs1lu+c+zzF67bXXVmm+23KYULYzx1bui9u2bavy+Ky5X7OPE84J8/Rh\n9wtaRESkQ5ygRUREOmTNTdyt3Va4/J6mgtNPP71KZ3cJhs6jy0A2s9BUTtNgXvZPlw2aUWjGzq4i\nDPVJc3I25wy5e2QzC0NY8lyaErN5jc+6WuEhh0KK5jZge9CMlc3ldD9jWNBcp2wrmqrorpbrkRID\ny5RdLbL8ELHQVMj2yu4fNFvTzJsZqtN5yRVDOyLlPkOTJN2fsjn/0ksvrfLYF1nH+X2Q5aDF7pM5\n66yzlrxORF3HNDUzBCTLmPsU+xP7cZatWqFuI1ZPauJ18/Pw/UpzPvsizcCZa665pkrndx/HGWWc\n3C/YB77whS9Uab7rcvuxj9DdMb8Had7n+zb3mdYOcxEL6zHXeUvKWw5+QYuIiHSIE7SIiEiHOEGL\niIh0yLq7WWUbfUunjFi4nD2HfaPGw/BxWTegfkWtMpeDZdiyZUuVpgaal/bTBYj6StZPqGvQrafl\nQsNnZwi/XMcsw0o1kqVgW/L58vNQ/2F7ZJ2W+iJD9mU9iPVCtwvqvzm0JPXS7M4REXHvvfdOj6mp\nsa2oVWYdjWscqGdlrYzX4fOxbQ8Uth37SK43utNxjGb9jmsjqANyjUkeP9QBWaY83qkrcyxlbfL6\n66+v8ti/+Hy5zEOhiLMbFt2o2JZDdT4vcv9qaf4RC99t+XmYR505a9Icz7xPLgfftzfccEOVvvXW\nW6t01s3ZVgz1mTV0rjVoafVDa1mGwg+vBL+gRUREOsQJWkREpEOcoEVERDpk3TXoDG35DNdJzSTr\nBNSDqBVnX1SGkmSYzaxF8jrUhqjF5DJRg6YOm/UT+hxSN8vl53Wo4dInPNcrNZ7Vgm3Jesv1xC09\nqfHkdmd957aKqPUirjVgeEL2r6yxMXQhfZ2zXkrNn77a7Aet0KvUZXPfpIY7FDdgXlCfy+3D/kSt\nOMPYBNTq6OOa250+uHwX5PHP+2zatKlKZ92Z4Tp5XeZnLZnrB/heyXVBDXq12oq0tijls3L9AEN0\n5rHGemEf//qv//rpMcPmsh/nscX4A3y38T75WvztN3/zN1fpPF4Ym4BrD3IfYh1y3PE9Mk/8ghYR\nEekQJ2gREZEOcYIWERHpkK40aDKkz2U9hbol48pmPYj+cSRrvPSnZpmoe2RthjoNy/hHf/RH02PG\njeV2aFkHpF7CeqEuSD14PWjFGqa+m/3bI2qtsrWFXETbr/7KK6+s0txuMuvbO3bsqPLYPrt3754e\nUzulPy816tyHqFvSPzPrXex7XHexWloY6zH3p6HYBfm31F2pRbZ05rz1ZERd/xH1eGEZON7zGg32\ntbPPPrtKUzs+77zzpsfUIodi5Geovw/pnPMi32fIb5jb4+Y1HRx3XMeT24daNp8ttwfHCvs4tx3N\n71SWiWtxcmyD7du3V3mt2AUck0M+6vlc9vFZ8QtaRESkQ5ygRUREOmT9bZ8NaNKjWSibVWhqY3i2\n7CKwZ8+eKo+mkFnKQLNidumguezOO++s0tn8QRMqXSDys9Isx9/SDSs/A82M8wopyOsMmVuz6Ypl\nam29SXM4zWXZTEeJgWY4mp++9KUvTY+/8pWvVHk0sWazF91t6ObDcuQy0xzLushtN+RmtRJaW+S1\ntrVkndLMuHfv3ukxn419hLJClhxYxxwD2TRKWYpm0VxvrbCyEQtNn7luaLrlOMx9phUyNGLt3K5a\nLpesN6bz2HvggQeqPMoT+dlpIqZL6Xve857pMV0UWU+s4+x2NeT6lfsm37dZuoio+whN/+wjHB95\nTLjdpIiIyEGIE7SIiEiHOEGLiIh0yJpr0K1t1Ya0L2oXWXeiVkStsuXuwXNPPvnk6TFDfVLfpX6d\ntZc/+IM/qPJ27dq1ZBmpu/K62YWD7lrUeFjGtYBtxTpuhcej6xSvlZ+doT6pSWW9kRoadafPfvaz\nVTqvTWCfoNZKTTpDbY/rFnJfbLlVkSE9ayXbTc6ilWUNjuexDLmdWWdMc41GPpdaJNsya9JXX311\nlcdQn3krRJZ/aJ1CDsFLdzq2c3btpG7J8cFzW26J84L99Pzzz6/S1O5zXTCkK9ce3HHHHdNjhtyl\nO1pe70G3Q173+77v+6p0doNjW/FdkfP57AwxmrVw9jW2Ddt2nvgFLSIi0iFO0CIiIh3iBC0iItIh\na65Bt7Qu6jCEumzWiqk30G8y69XUcKmFZR3qxhtvrPLoN0mtOz8f9RRud5jD4X3d131dlUe/vKzt\nUWOmXkr/0uwPOKTzz0IrrCZ1mVb4ReZRl3344Yenxwy9SJ05tzv1xFtuuaVK33DDDVU6h4ukFsw1\nAvfff/+S5ee6BfbNXGbWG9s232do+9LVCvXZ6jP076Vu2QotyTUlJLcfQ/fyXZFjDlx00UVVHrXV\nfC1qqdy+lHppZmjsZF//1tqbiIVtd6D+s0P3ydo3n51aPd+huf/x/cQ+nt99fD9RV851zvfGOeec\nU6UZYyCPNb432EfyWGI7s/7ztbj1L98NDE+6krUgxC9oERGRDnGCFhER6ZCuQ33SXEPTQTaR0cxI\nt4xsGqG5jKbobAaiGY4mVZo3aJLJ0JSey0hzLM1POc16GHJxapmiV0LrWiwDy5zbizIBzU3ZDYth\n92jCy2mGg2Rb0n0itw/LT/NsDivIXbKGzKTZXE5zH/tmvm/rWVeTlmvbkIyTzct0g6EJkmbT3C+y\nnLXYubk/0czecnGiWZS/pTkzm63ZVqynPL55XdbFaoXcJbnM7HscH2yPBx98cHrM9xNdzvL7lrvv\ncXxfdtll02O+JyhpcSxlmXAolGw2VVP2YH/KblYs09Cucvm9uNL3rV/QIiIiHeIELSIi0iFO0CIi\nIh3SlQZNDYE6Mt2j8vJ9LuWnLrB169Ylr0PtJesR1HSoP2zYsKFK53CRQ9sBZu2S2io19Xwfhs5j\nPVFbam1lt1Y6ZmubTur61HxyGD7qStTysm5JzT9rXRERmzdvrtK5z3A7PWqVue247oDrEqiF52eg\ntsq2zRrceoRwjVg4llrQFfLee++dHtONh65rF154YZXOejbLwDrO44P1xPGetWKW4bbbbqvSfJ6c\npuZ85plnVulcDmqn83R3nIU8tvheoFsVny+/k7g1KLd6zG25ffv2Ko/rFrKOzLYbckPM17r77rur\nPLZ71tRQqEmGAAAgAElEQVQvvvjiKo/PnvVqvkf4Xuf7N+N2kyIiIgchTtAiIiId4gQtIiLSIV1p\n0IRaJLWArD/u27evyqPtP2tU9LulT1++L3VKQv0ha03UlXM4wohaA6LWQv+/XCbes+V/yWuvleZM\nrb6lk1Mzp6abdSdq2ffcc8+S96FOSf2a6xZy2EDWKdcpXHvttdNj+rBTY2M/boVeZV3k37JOVyu0\nJ+Hz5P5EnY9rMvLz5XUgEQv9bDnW8rjkWoMc/jWi9sPleOYagFYIUa79YDo/78aNG6s8arp8J2XY\ndvMMD5lh/8p1Q99+vhc5trJOy+tmH+mI+v3Fsc/75nyG8mSf4XsxrwXJ638iIi699NIqnfvIkC6e\n3x1c/9Dye543fkGLiIh0iBO0iIhIh6y5ibu12wrzuJy9FcKSJiIux8/mS5qe6WqRzWn8LU2sNN/k\nEHE01XInrGxi5bPx2bPpk+Wl2YeuOtkkM08Td+taQ7tbZVhempOzy8PTTz9d5dH0mdM0odIVhCa9\nXI80QbZkBLYHTeu8bzantUxrEWsnScxCHms07w+FRcx8+ctfrtJsr3wuQ0lecsklVTrXMc2idGXL\nY4l9jWFaW89HUy2ftRXyca3cqlrhkvlsfJex7+XnpbTBcLdZjqSUxPfV3r17p8ccKzQnX3XVVVU6\nu/HRFTJfl+mhkM15XLbmncXS88QvaBERkQ5xghYREekQJ2gREZEOWXMNumWvH7LlU/PJmi71BupO\n+Vy6a1EDze5QDE9IzTDroxG1VkmXGeri2eWBW+RRg6YGmhnafrKl869EP5lFR2vpdWyPu+66q0pn\nHYrbxNE1J+td1ASHQjNmLZnPxlCA2Z2OejVdjViOHC4yhzlc7Fo9atAZ9j2WP9cj+wC3IeQYyHCt\nAcdh3jqU+nRr20HqoXTV4X3z+4DrIeiyxfUqs7BaGnXL5ZJ1wTrOa2o4Zu+7774qndfusF3pBpfr\ncShM8bZt26p0fjfQ1Yt1mF3mHnrooSqP74Y8f6ykHVeKX9AiIiId4gQtIiLSIU7QIiIiHbLuftCZ\nWfW2rOky7F4rZCL9LenTl/0DqREObb2X/evoQ03dJt+XWhfvkzVplpe/pS64Hjom25laUi4z64Wa\nYX5e5lE3y76c9L9shaxkORgekj6Wed0CtVW2B8uR23ooXGcuY2vb0LUkl5l9sbUlLLVH+qLy3NyP\n2ae5RiP7tHM9Cts5a4rUR9mfWOf53cAx29Lj2UeGNObVGrOt+7bWuUTU9cY1PgyrmdcEMN4Ax3t+\nd3NNCdsyXzei1v1Zx0znvsj5oTWGqUGzn64mfkGLiIh0iBO0iIhIhzhBi4iIdMi6+0G34tVSF8g6\nMq/FPJL1B2qErS3lhmILU5vM+grLxHTWcajLUCPJ+UPbGc7CWsUEZp1nrZg6LNsj64DUhumLmvsM\nfTOpX7Ets77IMlA/zX2P/ZR9nH72+XmpcbJMPfpB5/Kz7ajX5efhs3FrwZYfNOuYMbRze7HOGAM/\ntwe3hKReSr0xa558do7ZnL9W42wWGH96aN1Ihu1B8rPnfRAiFr4Hc7/gWKEGzVgS2S+daz1acTPY\nF3ndVh9fyzHpF7SIiEiHOEGLiIh0yJqbuMks5gKakLKZYhbzDN0JeC5dCDI0qfLc7KZBszXNptmc\nRrM1n7W1tSa3iZvFlW2tzDUt16mhUKX5+VinNKlmsxa3DmQ90ezecoth++T+1QpvGbGwjnOZmdeL\nK1WL/HxD8kp2h6JrFJ+d2x/m8U2XpparC6/DdC4/25VtSdNnfoa3Y9tl2E9zCNqh38/yvmVbcQy3\n+hOv22p3urW2tolkO/PdsBLZcJ74BS0iItIhTtAiIiId4gQtIiLSIese6nMWDZSaYWsbRWpWWXNo\nuXpFLHThaNEqP3UMuhBknXNIj8v5/O1QaM+WdrRebjwtDZfkMtLdieS64LMyTT0r34d6Kd0yWsyi\nRQ49+9uNVv8a6mustzw+uH6jdV+uH+D4yPnsA0Nubq0x28vYOlCGynugz0N9l+3caruVMMuYbbGS\ndl5pHzi43g4iIiIHCU7QIiIiHeIELSIi0iFlrUPQlVL6i3knIiKyyoxGo5lEab+gRUREOsQJWkRE\npEOcoEVERDrECVpERKRDnKBFREQ6xAlaRESkQ5ygRUREOsQJWkREpEOcoEVERDrECVpERKRD1ny7\nyV/4hV+o0msdavSdztDWaT/yIz+y7Gv93M/93JJ53OJvteB2jat139a2kEP3XMm5reu0zv3Jn/zJ\nZV83IuKTn/zk9HhoTLa2eZ2Foa34DvTaQ318tVit+/7ET/zEsn/7K7/yK8383Gdm3ep0JecudZ1Z\nrzXLuWyPvP3kPO/ZGocr3U7WL2gREZEOcYIWERHpkDU3ca+XSXvInLbUb3nekEky//61116r8t71\nrqWre5byDdGqY+at5L7rZcZebhlmMdUOsRKz3LxYzfpu1VXLdLuSepnlXTCLjJBNmYudm9Mck4ce\neuiyyzhPk/Y8x3+L1ZJq5lUGwnrhubn92O5MH2j5h85bTZnNL2gREZEOcYIWERHpECdoERGRDllz\nDboFNRza9mfRePjbrFUceeSRVR61ipYeRM3qqKOOqtLUsDKHHXZYlX7++eenx3zWl156qUpnXeON\nN96o8lj+WXg7uLm1NB3mzVIXrfUFQ5pgy2WjpZMNsRIXjrVaE9Aqw0pg+V9//fXpMfspf5vHHdd+\ntLTIofIffvjhS95nqI+slh6/Wsyiv8/C0Hs81yPbdUiDbtVx6z6zPNt6ue1F+AUtIiLSJU7QIiIi\nHeIELSIi0iFdadCz2vpbGhDPzfovNehjjjmmSmfd6cUXX6zyqGfx3JZGxd+eccYZ0+Mjjjiiynv0\n0Uer9LPPPjs9fuqpp6q8V199tUpT614LH8vV1ENbGm5Lg2a9sB6oueW25rm8Ty4Ty0fdspWe1c++\n9dt56cFD4/BAtTyWl+dSO859fkjXz2OC1+F6jqOPPnp6zLbh2GGZN2zYsOh1Fvtta00Dy9+DBj2L\nzj90bv4tnzWvLYio19S01iEsdp/8Lmce1+rktuX4Zv/Kv511XM1zXYZf0CIiIh3iBC0iItIhTtAi\nIiId0pUGPQT1iaw/UOdo+UIO6X7ZP/m5556r8l555ZUqvXPnziqdy3HyySdXedS3sjZ20kknVXkn\nnnhilX755ZeXvA51tKeffrpKH3fccYuWby1p6TLUikh+Pup+1J1eeOGFJa/Ddub6gv3790+Pn3nm\nmeZvc/uccMIJVR7XE3DNQ/adX4lffUvzXAlDeugsemkeh1y/wX7L/Pzs1CKpIeb24XWoQef06aef\nXuUN6Ye5rTdt2lTl5TEasXAcZob03rWI9T7rPXOZ+Q5lOrc766EV72Kob3F85D7E8vO9ksvE37bi\nJwz5XrfG6ErxC1pERKRDnKBFREQ6pGsTdytcZ8RCc0eGps58Ls3WNF9mExjNYzRx0xR9yimnTI8v\nuOCCKo+mkIceemh6/OSTT1Z5NNHnNE2DNLlkk3ZEbRaeZ5jQzKzm1fz7lgtTRP28NO/TfJZdc9gH\nmH7ssceqdG4PXnffvn1VOrf7OeecU+WdffbZS5Y/om4D9p/sesdzZwlzuBJmccsbKkN+Vv6WZmvm\nZ5mBboeUFbLJm2Pp4YcfrtKXX3759Jguixwf7IvHH3/89JjSBftBfhfwHcPnmdc4nAW+j4bKkPsF\nZRy23RNPPDE9Zh1mCZFpXnfjxo1VumXipizFPpLP5fjmuccee+yi91iMIde8leAXtIiISIc4QYuI\niHSIE7SIiEiHdKVBDy2xb4UKpMZDnTlrG9Q56LKRdSjqV9S3slYRUWvU1Dmo+bzvfe+bHm/dujVa\nXHbZZdNjuoZQc26FvKMetFba1yzbJlKXPe2006bHrENqhln7otsLtSRqhuedd970mGsAuJ4gr024\n++67qzz2PZLLxTUO1GWzjsZ2pgY6L3cP9h/2kZxPtzeWIdcj+x7bmWMtrxmgLs41AbmMXCfC9njg\ngQemx+wT73nPe6p0Szdn/bPtWuEih7ZgXK1xOcs45Lut5dbKMMa5X/DdzHTuIxx37CNs2/wuZzvz\n3ZDLlEO2RrRDx/KeQ66qeQysdJ2IX9AiIiId4gQtIiLSIU7QIiIiHdKVBt0KFxexUDvOGsOQzpG1\ngJY/ckStvdBXNmucERG33HJLlc7PQB2JekouI7VI1sW55567ZPnpV9iCmtu8fPaGtpuk/2LWcVj+\nlu9pDscZsbB9csjUU089tco788wzqzR1M4ZuzPB5Hnnkkenx9u3bqzz6X7IfZK2Mz85+m/viUHjF\n1aJ1X7YVf5t1WfY9as7Us/OzU+/lu2Hv3r3TY9Yh/d9z/X/kIx+p8j74wQ9Wafq4btu2bXpMHZPj\nktplZq3akuMy34flzeMsYmH44VwX9PWnP3lrLHHcXXjhhdPjPK4iFo536sx5rOW1KsyLqP3SOe5Y\nptyv6SPNvjfLtpyz4he0iIhIhzhBi4iIdEhXJu4hNyuagbLJm6YcmsP37NkzPaabEk072aTBMm3e\nvLlZxmzuoEmP5pnbbrttenzRRRdVedwJK5vWs6kmIuLxxx9vlimHj6S5dZ5h6TI087A9ssl7yNyX\n3dVovmyZrflbmrzoRpLbi/XypS99qUpnExnNr3SHonk/uwjRrMt+nMtEt5F57oCUr8W2oxk7jwk+\nO8uYXRj5W46t1q5ThC50uS1pfs3ujBG16fb666+v8rKUFLHQBTP3Y75HOA5zms/Kdw7N4as1LvPY\n4vuJJvvWbnz8LcdSHmtsd0p9eWxt2bKlyqN7I9sj9xGauEkuI+UvyiC5PYbCPZPc1ittR7+gRURE\nOsQJWkREpEOcoEVERDqkKw2aMHQe9a2sBWQ3i8XIOgjdO+gSkO9DjZPaFzWTrOvs2rWrymPozxxq\nkiEG6c6V3Q9YhiFdNuvOQ1t4HijUjaljttwY+FteK+tDQ9sB5vwhzZk6Wta77rrrruZvs65MNwyW\nn+2Rz6U+RzcTjoHMXN05Gvo1nyf/lm5I7Is5n+WlGw9Dpma9l/VAbTK7+dxwww1V3vnnn1+l8zoS\nPhs1Q4b2zWOP4UbzVqc8l+sS2I/XYytRjh2uE+F4yfrv7t27q7yLL754yfuwDql9v/vd754e020y\nu7VFLNT9cxlb5eV9+Q6l61R+d7C8Q+/M/OyG+hQRETkIcYIWERHpECdoERGRDulKg6bORH2otSUY\n/XvpK5zzqZtRH8raBXUZaqkME5o1h40bN1Z51CIfffTRRe8ZsdBPL/tU0mePWmtL76LmOS+obVO3\nYX7WfIZ8s7P/OK9LzSpfizolteIPfOADVTr7ylM3Y2jAXKfZzzxi4bOyv2Xtlc9OrTu3O599nuEh\nW9diXu5DQ22X+yI1QrYd/WPzmg2OZ+rK2feZ7U4/6Nx23OaV60bYdvl5qWNSg86aO99t9I0n82rb\nVhhK6uBDIYPz+gKe29rCl+82kst01llnVXlcL8TnaYWDpb9yLjPfkVxPkN/N9IPmfVrrB/SDFhER\nOQhxghYREekQJ2gREZEO6UqDbsX8jWj7lNHWTz0lb2lGf0tqFfk+l156aZXHrQUZfzfrhN/93d9d\n5d13331VOvu8Mo4sdY6swVG7ox83dfKsvVA7oq55oLD+W/67LAe1PJYxa+zsI4ydnPUiavWMe33H\nHXcseS79IpnO/u+MWfwnf/InVbql4VKfY5lzXdBveJ7xmvNYY3k5DluxhqlJ57HEZ73yyiurNHXA\nvGaAbUef6tw+HDt8F+R2p985NemWhsu1H3w/5TUN1LKpta7WdpOzbGvJvQL4nslptgc19Xwtrv1g\nPeVxx9/u2LGjSl9zzTVVOq9b4DhkvPOsK3N7TL7Hcz9mn+azt/aQ0A9aRETkIMQJWkREpEPW3cTd\nCotGcybdGFpmCLp0ZJNLDrEZsdB0mM01NHGzDNl0HlGHvOO2dzSlX3755dNjmr/pZpVNZDTlEIYf\nZVi+tYDtQZN3TtPETZPR/v37p8fc7vPBBx+s0tlNiWZp/vbLX/5ylc6mUZr+2Za5jDfffHOVR5cT\npnMYV5rz+ey5HofCQa7WFoUk34flz20VUY8l9ku6VdG1Ldcbny2bKyPqdqcp/aabbqrSt99++/SY\nJm26c1F6yu45bKtWH6eJm1IG33XzCsFLcpnZdjRp0+Usl5kmbUp/+X1LyY2yQjZjswwcOxyXV111\n1fSY71ea7HPfZNvx3NxevCffbSxj7quauEVERA5CnKBFREQ6xAlaRESkQ9Zdg85QZ6JLQMv1hdok\ntyXLMBQgNYasO9Nlg1px3iototazqfG0trWkHsff5mel+wDTLHN2KWAdrhZ0g6HmljUsht2jRpV1\nf24rypCPWcOivkg9iPWWtSRel7/NGhXdLvg8dOHI/Zp9nGsccv6QxrxWGnRLW+Xaj6zpttaFRCxc\nW5HPpasU9dHsosVxd+edd1bpPD7e+973VnnUQOkOlcO6Mswv9ffcL/iOWavtJVsucqwnvhvYttkd\nivXEMdvqi/xtbjuOJb6rWU95/Qq1YJaxdR8+e85nPczi/rtS/IIWERHpECdoERGRDnGCFhER6ZB1\n16CzPZ+6JXUM+tNlzYr6FXWBHJaPfpLc4i9vE8mtxhiKjnrj1VdfPT2mbx2f5/77758eU2PLW7tF\n1PoK9Sxel9pS1tyGQnDOi6H75Lam7so6zrC+85aEEbXORP9vtjN9XHO9Us+ij27uF9Sz6FfPsIK5\nz7P8vG/LH5N1vFpaGPtX9tFleanZ5jFL7Y59ntux5nqjX/B5551XpfMaB5aJ181jlDoltw7leyWX\nieXn2o+8zoIaJ99lQ1vtzovcliwT/XvZXq0wukNbA2e4Pij3Ga714NocjuF8X75HOL5zmnMA/d/z\n+2mo7bi2gnWzEvyCFhER6RAnaBERkQ7pysRNMwlN3jRdZXMnTSM09+VQjTRR0NyRTc80cdNlhrur\nZDMkTR0sYzbXcFcmmojys7IeGCqztQMPzVZrYUpbjFyvrFOSTUhs14cffnjJ87JUEbHQFM1zs9sP\nzXA0deZQrAw7yfZgWMRszhzq47k/8dnp/rFaLnSt3XpackpE3ceHpBlKG9mljnnsM3lsfcM3fEOV\nR5e5v/iLv5geD40luvVlszbrn/0rh3Tlu4D1xDrmtedFrvOhcJd8L+bno0m+1Q8oRzCMcZaP2M4c\nhzm0Z0QtEbHtaLLP/Y/P1grnyWdl2/Ba85Sa/IIWERHpECdoERGRDnGCFhER6ZB116CzBkrNgDA/\nL8GnxkNNIWu4XLq/Z8+eKp31rC996UtVHkOK0i0j6ynUBB944IEqnfUgatB81qxj0o2Hug01kawP\nUR9p6YuzMKvukrU/al+ttuOz8rc5RCpdmKgFs57yfVq6cUQdPpLrBegiR+iylaE2me/LZ2X55xXq\nc6hP5PsOhT3MY4BjhZrttm3bqnQe09SCSXah4baWXEfSqieGH2XfzP12qEy5r7KP0EWL62LmRWuN\nCduK6yFabqJsO75Ts+sUXaNa7lB0jaKLHF2ach1zfLDOc9tld8yIhXWRxyFdsPgu4DqGPCZWOib9\nghYREekQJ2gREZEOcYIWERHpkHXXoDO011MXaG3NR82Q+kPWoagjU0/J+u/FF19c5VE/pN6YfWsv\nuuiiKo963c6dO6fH9NXkdpM5xGAr7GREe8vC1fKDHtJaqKnn37Ot6M+Yn4+6Pts962ZDZbrwwgur\ndA4HS92JWwvmehvyWW35OlN7bNUTn51tt1ahPnOaawJYptyWvA61vLvvvrtKZ02a9cRx+dGPfnR6\nzLHzxS9+sUpnn/abbrqpyuPYyr7MEbU2SR2Zz5P1U/ZT6parFY+gtZ5gyK+eZc5tyZCuLf2XfYT3\nyW3LdzNhzIF8H+rgrNP8vuV7m/fNWjfHZGtdQkQ9Tlcam8AvaBERkQ5xghYREemQrk3cTHOJfTbR\n0PTR2lFox44dVR7dcbLpiqHm9u/fv2jZ3ySbWVgGmsCyuYwml+3bt1fpbHpjyFCaZ2iWoztOZl5u\nVoRma7ZldqGhSZ4uc9lNjOZiusVkc9Ms4UZ5H5pbaXbPJj66bNBVhGXOpkOW4ZFHHqnSuf/xeWg+\nm5eb1VDI3XxfthX7Uz6XY5QmVu4slccSzZc0Y+c+zv7Ocff5z39+esyxT3MmQ4rmctDEzWfP/Ylt\nQzMv06zz1WBIMqHZOvdrvhfZllu3bp0eMwxolpIiIq677roly9RySYyoxz+lSpqec/9j/fLZc1uy\nbXhdmrznuWOgX9AiIiId4gQtIiLSIU7QIiIiHdKVBj2kw3CbsqzlDW1dd/PNN0+Pr7322iqPmkFe\n9s/Qi9RIWOasr3A7Q+pbWct48MEHqzzq17lMrBeGHKSumWlpLbOSdachFzmms27DemF7ZK2P5aWu\nmd2jqKEx1CrXNOR1ANSCqYHmOqZ2mkOGRix8nryOgW3V2oaQzz4vzZkMueLltqQGTT0+Pw/HDuuF\n/Tq7Q+U1JBEL1whknZMucdSK85oH1j/XNHBtSB6zrbUdhM/G8cC2XC23q1xPHHfUcDl+Mhw7XEeS\nNWo+K7duzflcP8P3IOtp9+7d02P2L4Zwzv2N722uRcjX4pofjg/WY06v1PXRL2gREZEOcYIWERHp\nECdoERGRDulKg6bvbMuvMKLWv4b0rKwjUOegjpbDblJDYBm4HVr2z6SuTF/aDH0DGWoy6xrUxZhm\nGYdCUR4oLQ10KMRd1ruoZ7EfZP2RWhG1uqx5sg9QK6K2lLVj+nlSG866P/0ieZ+WhsvftrYApOY5\npGPOQiscbEv7Zr20dH7+lto99d/LL798ekwtkn0mb/9J7ZTjI/cnhqyk1s1zc5/hNoqspzzumDf0\nrlut+AS5zwxtz8jxk99Rra1/CddkcNvRXMfM4/uV6wsyfO9RF89tx1gFrbCm7Lesl9Z7faX4BS0i\nItIhTtAiIiId4gQtIiLSIV1p0EPaHTWqrIVRK6I+l6+dfeciFuonWZuh1nXNNddU6VtvvbVK3377\n7dNjbpFHLSNrl9RDSdbrqEXy2enPmLWleWpbLT9o0tIxqZHTtznr2fSFp2aVr0utiJohtePcD1r+\nyBF132Qs3l27dlVprifI6yWoX9FHlGMiM08/6Px8HHetccm1E1u2bKnSeczS55i/5bXyfdlWjFWf\n63TI5zjrjYz1zGfnuyFrokNrD5a652LMMz5Bi/xuo+bM8jM/j0vWC9/N1PYzfLY8fjhW+K5mfIvL\nLrtsesz3IN+/OZ/6O99BuYz0o2/tCRFRt2WrTywHv6BFREQ6xAlaRESkQ7oycbfcUSIWLqNvhVuk\nyTj/lmZFmjNvueWW6fFNN93ULCPDyV144YXTY7oE0SSWzbM0J3Fby2yCoXmGLk28VouVmNJmMbGy\n3nJb0hRFt4xs9sr1G7Hw2dmWmSH3m2xqo9ma5td8Ls2kNNO15JahrR0zQ+NjXiZv9olW2Fb2abrF\ncLvGDM2+bPedO3dOj/lsdJXcuHHj9JguNLxudqHjOKMbD82ZeRwOyWq5L1J2GgqnOq9Qn60QvEOh\nY9k+uV6ZR5N3lqJaYTQjarM1TcKU82g6z7+nPETXvOxOy3cB75PnC5q4OQ+15AlDfYqIiByEOEGL\niIh0iBO0iIhIh6y7Bt3SQBm+s7WcnZozNemsG1A/yWECI2othsv8qSUxnTU3at3UYrJrArXTCy64\noEpnvYv66FDIvpYLzWq5cwyRdWe6OLBOs1ZPrYjbG2bdiXXKkK5sj6w1sY7ZlrkeqTnzvrxWft7N\nmzdXeWzLlhY5Tzer1n2GwpFm+Ox5zQDbdZYyUAds6e98b+TQvTyX7wK68TA/p1vbDEbU6wmG9N6V\napXLvW7OZ5m4noBa6xVXXDE9ZlvxPZlD4bI9uH4gvwep63PtB8dh/n1rrUdEPQ4Z0pVlZDrDPr5a\n275G+AUtIiLSJU7QIiIiHeIELSIi0iHrrkG3dCdqIC2on9Anjvpjhv5yWffgtoPUpKj35t9TB6dP\naPbBpO8mdY2sF1HTbIUy5LWG/C/XiqxBs/5JXl9AXYnPk/1wGX7wvvvuq9LUs7NGSr9O1lPWiqmL\nUb9iW+bnZVvRDzqfu1p+zxHttQgtfXRIn851sWPHjmYZ6K+c24MhXlnHuS2pV7Pd8zaE1MWpQbfq\nmOXl+6oVYndo3B3ouFyJls26YBmy3svYBXwn5Xff0Fabucz0c2Ydc3xkPZtjiVp31p2H4g9kX+1W\nbIXVxi9oERGRDnGCFhER6ZCuTNw0hdDlhOamlgm8ZSLmeXTzyeaabA6LWGgCo/kjm0ZoYuW5OU0T\nPc2B2dRDl7KhemmZy1biZjXLblY0veXnZfloAsvtQ5MX0/laeaebxdKsp2zmYvjOltsF65AuHLxP\n7iMMe0hyva6mO0fLlE5yHQ+FsMwhE9lvW6ZOlokyyKZNm5YsE3c4Y5lyW1KGyu5BEQtlkFzGIffA\nWULuknm5P84S6pPPSrNvdmnks5PcBjR/07Uwtx3rjNIky5jbkuOM0lN+9pZLXEQtcTGP95mXi9xi\n+AUtIiLSIU7QIiIiHeIELSIi0iHrrkFnjWQoDCX1lKxXUBPhtbJWyetSB8z3ocbJc6lz5Hy66jB9\n3nnnTY+pt1Pbbum91G1mcdFYiZtVSxMdcgnK9dbSdyNqrZ7uHay33A8YwnWojLnOqXVRd2qFlmSd\ntp5vaNu7WXT+lTCLjpb7OPs/nyeneQ+OWY6tXOfcBpLXyvovw7+y/vN96ZrD7UzZ3zJDYVlnGR/r\nAcs3tL4ml3koDHN+J7Gedu3aVaWzxsvtPQn17Px+5hila16Ga4tamnRrC9jVZv17iYiIiCzACVpE\nRKRDnKBFREQ6ZN016JYGOuQf22KW8GytbfCGQvRRx5nF5zhrJuulI8+LWcNQ5vwhX+Csd7U0fzIU\nzvsam6oAABQ5SURBVK9Vb0O6ck4P+UmyjC1fVGphS50Xsbqa9HJhmVqxCdjHW7EK+HtuO8j7trZU\npdadr0t/6qEYAvm+zJtlW06yWtvAtsrAvKHtGlu+8i1f81a8AcItYNlH+K7IbTs0PnKauvgs45Cs\n1tahEX5Bi4iIdIkTtIiISIeUeZlSln3DUtb2hiIiIh0wGo1m0iP9ghYREekQJ2gREZEOcYIWERHp\nECdoERGRDnGCFhER6RAnaBERkQ5xghYREekQJ2gREZEOcYIWERHpECdoERGRDnGCFhER6ZA1327y\nh37oh5b926Et2HL+0LaQszDLFoYr+W2r/K1rDT3rSsr0y7/8y81yZH7qp35qyesO0doecLXqqfXb\ntWRez9469xOf+MSKyyki64tf0CIiIh3iBC0iItIha27insUcO2SinMVUOAst0+csZtN3vauu3jfe\neGPJa83TTN36/VpvL7oUs5RpljK36nSWejv00EOrNNvytddemx5/7Wtfa96nxSzSzCzmcBF5++MX\ntIiISIc4QYuIiHSIE7SIiEiHrLkGTVai1y03j/nUDKkv5jS1SJ771a9+dcn8rFMudm5mSC9dCaul\nO8+rjEPXaenK5Igjjlgy7/DDD1/yuhERL7/88vSY6wWYzu3OPkDYvw45ZOl/F6/ExayX9QUiMh/8\nghYREekQJ2gREZEOcYIWERHpkDXXoOfpnzyLr3BOH3bYYVUedctXX311evzUU081f0tyPjVn6phZ\noz7qqKOqvBdeeKFK52sdffTRzTKtxJd2Fmap/5VADTdDXTnXDc9jPb3yyitV+vXXX1/yty+99FKV\nfvHFF6fH1KBPPPHEJcsbsXBtQoZ9M9ejftAi7yz8ghYREekQJ2gREZEO6crNasjcSpNxy12FedlE\nSZNky1Xq+OOPr/KOO+64Kv3MM89U6WyipLkyu/Hw3L1791Z5dO/KJu9TTjmlytu8eXOVbtUj62Ul\nZtGVhBDNdcxnbbVrNkNHRJxwwglV+rTTTpseZzN0RMSpp55apffs2VOlc3sdc8wxVR7liZNPPnl6\nTHmCZWQ634d5bI+WCxfrSTcrkYMLv6BFREQ6xAlaRESkQ5ygRUREOmTdNehZtlyk5tYKt0htL+e3\nNMGIWl9sufhERHzjN35jlc6/37dvX5V31llnVelnn312ekw3HmrbWadlHtPURPPztfTdlTBLuE6W\niXVM16n87GwrrgnIHHvssVWa7c77Zi2fblZ0jXruueemx3SJo37NtQf5vlzj8PTTT8eBslptKyLr\ngyNaRESkQ5ygRUREOsQJWkREpEPWXYPODPlx0hc167bUF6ljZo2Ueeeee26VPv3006fH1A+pRdKX\nNt9n//79VR59dvO1h/yrN27cOD3OGnlExGOPPbbkdSNqn2NqxUOhS5cL224ozGn+/dD6gazpUoPm\ndfPzsQ537dq1aNnfZMuWLYveM2Khn3rWnc8444wl8yIWat3Zj/3JJ5+s8tifsq7MEKIM+SoiBxd+\nQYuIiHSIE7SIiEiHdGXiHnLVoekwu9/QJElzYDYz0hxLt6SdO3cueR2ak5netGnT9Jg7YT3//PNV\nOpuqaULlb3OaJnq66tBlK5vlaRZthZKcBbYdXX5o5s0mcP6W7dzKY/mzPMHr5jCgi5Up1xtNzQ8/\n/HCVzpIEQ4hSBtmxY0eVzn2KblW8b+7jRx55ZJXHNEOmisjbG7+gRUREOsQJWkREpEOcoEVERDpk\n3TXolu5MfZH6Yw7lSA2arjo5NOOQS1b+LV2WWmEbIyIeffTR6TG1yewqxTIPuQ/lMmedO2Kh5pzL\nHxHxxBNPTI9ZpysJD9kK08o09dH8fNTFuU1kfh5q89Tfs64/tAZgw4YNS5aZ5aWLXNaRWd9cT7B7\n9+4qnduS4WBZRvaLzCzbdIrI2w9HtIiISIc4QYuIiHSIE7SIiEiHrLsGnXXMoVCf3D6QfqAZhlDM\nmmcOtRix0G/1nHPOmR6/+uqrVR61yIceeqhKP/jgg9NjatCvvPJKlc566ZVXXlnlUdvO4UhZBuqW\n9LvNGjt1yqE6P1B4XWrfOc06pqae07wO+8DWrVunx0Pbl37lK1+p0llLps903ho0ou4zLD/7BEO+\n5r5IH3yGDc06MzXnoecTkbc3fkGLiIh0iBO0iIhIhzhBi4iIdMiaa9At3WwoFjc16LxVInU++kyf\nf/7502P6I1PHzNc96aSTqrzs5xyx0C8651NfzNsZRtSaNH979dVXV+lcDmrZQ77ZWS8dquNZaF2L\nefQ9z/69XC/AdNZ/6TNNH/aWzzd9yxlH/aqrrpoen3nmmVUeY7JnX+ccuz1ioX/15s2bq3R+PtYT\n1xfk5+WWpEO+5yLy9sYvaBERkQ5xghYREemQNTdx0xUkm+WYl7ckXCyd3YvogkKzdTYr0mWGoTOz\niZKmTYaWvP3226t0drGh6ZPuQ9dee+30OLtcRSw0w2czNcM/0nWKz8MtDVeDVrtGLCxjy02J6Wzm\npZma20BmEzfrmxIJTca5z7A92GeyaZ1tRbmCfTGb+1kvPDeXmXmUDUTk4MIvaBERkQ5xghYREekQ\nJ2gREZEO6SrU5xDc1i9re9T5qGNmvXpoe8MMdb577723StOlKbtADYXgzFoxtW3qxtnth8/KMlD/\nzVrrarnmzHqdXDd0q6IbWe4jzzzzTJVHV6nt27cvuwwf+chHqnQOt8o6pntdLmMODcvyRizsQ7n/\ntZ6VDG3haahPkYMLv6BFREQ6xAlaRESkQ5ygRUREOqTrUJ/U1Bi+M/srD23Fl32HW2EnI2p/a2qR\njz/+eJW+7bbbqnT2n6UPNX1psw8vw3VSQ7/mmmumx+9+97urPGrzfJ6cZp3OS7cc8oMm2Z+Z/u1s\ny7xGgHn0Zc71z+tyK8cPf/jDVfr000+fHt91111V3llnnVWld+3aNT3et29flcfwsOxvuVx8HoYu\nzflDbWWoT5GDC7+gRUREOsQJWkREpEPWPdRnC4Z1pPkvmyxpKjzmmGOqdHaTYQjO66+/vkpnN5in\nnnqqyqP5mC422V1q27ZtVd4FF1xQpXfv3r1k+bkTU/4tzboMF0nytVbLNWdW82p2DTvllFOqPLq9\n5XCXNP3zvuedd970OJusIyLOPffcKn3JJZcsmWYZvvjFL1bpXKesf4YYZZlzP26ZvyNqiYXjQZO2\nyMGNX9AiIiId4gQtIiLSIU7QIiIiHdKVmxX1UIawpNvSCSecMD2m5kxdM+uC73vf+6o86spZr2Zo\nSbpKURfM2uTZZ59d5VH/zWXkfagz520U9+/fX+XRrYc6Zn4+6qWrtWUh27K1HWhux4iF4S9zn2EI\nVLZ7fr7smhaxMPQqt5/Maxq43STd7fJ96SLHMjKUad6ecmjr0HyfVr0sli8ib2/8ghYREekQJ2gR\nEZEOcYIWERHpkK78oOn7O6SPHnHEEdNj+rweddRRVTpv7UjNk9tPZu37nnvuqfIeeuihKs37Zn2b\nWyE+8sgjVTproNQeqbdn31nqudQxqUXmumCdzsuXlu1KHZz3ydpq1mQjIl577bUqncOr8jpbtmyp\n0lnXP/XUU6u87CMdsXCLz9zuQ3pu7qt8dt6X/Trr2QwhynNz29Gfmj76+kWLHFz4BS0iItIhTtAi\nIiId4gQtIiLSIWuuQbegTzH9S7PmzHzqb/R5zVre1VdfXeVlfToi4o477pge33///VUe9Wv6Omcf\nWOrVF154YZXOvtnUoKmPZq2bfs/UJltxvYfidh8orH+2Jcn6O7fLbPn7nnnmmVVe3nKU+dTmqe/y\n3LxmYM+ePVUetxnNaWrm1N/Z7ll/py5O8rV4HzVnkYMbv6BFREQ6xAlaRESkQ7oycROaPlthHXMo\nzIj2Fob33Xdflbd9+/YqnU3nNFcyPCRDdGYzPE2qLRchblFIM2/OZz0w/ChdtHI98T4rIbsXDZlb\naZ7NkkMOrRqx0Jy8efPm6TFd4tjO+dl5T4ZPZdvluqG8Qve0LDnQ3Yl9ke2VoTsgQ4zmfsvxwOdb\nLflCRNYHR7SIiEiHOEGLiIh0iBO0iIhIh3SlQdMthm5VzM8aKPU3npv1YP62FdaRmmBrO8CIWoul\nqxR1zKx5Mlwkw0PmMJRD2iO3O6SOPi/ysw6Vv7UFJt3cSN6uMYc8jVio4ebr0v1s586dVZp6fHbN\n47kXXHDBkufmbSojInbv3l2lqX1nVz26A7b6JuuUtMLoisjbD7+gRUREOsQJWkREpEOcoEVERDpk\nzTVo+su2fGmpNzI/64TUDBlmM2vS1Kfp2/zCCy9Mjxmy8pxzzqnS9OHNPru7du2q8qhXZ201nxex\nMFxnTu/fv7/K4/OwLlra5Lx0S7YNy8/7ZP9eauoMn5rrkfe58847q/RFF100Pc66fcRCf/F9+/ZV\n6ex/nbeEjFgY4jXDdQoM38l2z37T7D/0dc6aNPVqogYtcnDhF7SIiEiHOEGLiIh0yJqbuGmGa7nq\nMKwjTbfZREl3G5pCs1mR16UZO5skc0jHiIXhIU877bQqnc3lNN3S/Sm7zWSTb8TC58nhI2m6pbsQ\nzcstN6t57YjEtqOJmO5D+dkpZWSJIaI2eT/yyCNVHuWJbA5n/dOczDrPJvBzzz23WaYM3elaUkZE\nbaqm6xR/m+UL9lNN2iIHN35Bi4iIdIgTtIiISIc4QYuIiHTIuof6bIXrpA5L7bWlC9IVKeuP1Ay5\nZWHeavCSSy6p8riFJLeuzBoptyxkOuvX1I35bFk3p2ZLLZWac2sbwpXomC0XuSG3q1wmhnClfp23\n5eR1WKd5rQFd7RhqlSE6cz9gvTBMaHalGtqaklp4hroy11m01miIyMGNX9AiIiId4gQtIiLSIU7Q\nIiIiHbLuoT4z1Ni4DSS11qxJ/+mf/mmVR20vw5CJ3O7wmmuumR5zm0Hq4Fu3bq3SWSM96aSTqrzz\nzz+/Smedefv27VUey581W5ZhKKxmi5X4QbfOpQ5OnTbnU3dle/zVX/3V9Hjv3r1VHvXe7MvMe3Lb\nR/avrFnTDzrr4BG1fzL90Bl6lSFG83oI+kyz7WZpSzVqkYMLv6BFREQ6xAlaRESkQ9Y91GcLhkGk\nu0o2LQ7t4JTNsXTrufHGG6t0NnXSdE7XL7ow5d2UaJKn61c2Y9PVi+bjbBKmaXbI5N0DLZM9n4fm\n/mwiprsTySZjusQR3je737HtmD7xxBOXvC7Dj9Lcn9uH5nH2zdwPWiFb+VsRefvjF7SIiEiHOEGL\niIh0iBO0iIhIh6x7qM8MNTS6oLS2WRwKf5ldbqhtU0fO4TrpkkXt8bnnnqvSWQsf2j4zu+MwvCU1\n9VxGlndIe2yF5Fwr1xy6HmWo2ba04UsvvbTKY1vmeqObG5+d7k9Z/+V1ufYg9yf2NW4ZyXLkfsx2\nZx/PurMas8g7C7+gRUREOsQJWkREpEOcoEVERDpk3UN9Zg10SA+ljpk1w1l0WPqT0ke3tQUm/Vap\nGeZy0B+ZPq6Zlq8s4bPOokmvVzhI3jfXOeuFem8+l23V2uaSoT5ZT5s3b17y3KH2yCFGN2zYUOVR\nv2755LNMs2wNqiYtcnDjF7SIiEiHOEGLiIh0iBO0iIhIh6x7LO5ZdDTqgi29jrR0P2qgK9ni70B1\nQV6npW0Pla9VpvXygyYtPb5VJtZL61z60Q/dp7V+gGS/e9bpUBkz7NOt9hlqOzVpkYMLv6BFREQ6\nxAlaRESkQ8pamzhLKetjUxUREVlHRqPRTDqUX9AiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIi\nIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQt\nIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO\n0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId\n4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi\n0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAi\nIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIE\nLSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIh\nTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIi\nHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0i\nItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7Q\nIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3i\nBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLS\nIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhZTQarXcZREREBPgFLSIi0iFO0CIiIh3iBC0iItIh\nTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIi\nHeIELSIi0iFO0CIiIh3iBC0iItIhTtAiIiId4gQtIiLSIU7QIiIiHeIELSIi0iFO0CIiIh3y/wMF\nx/E/dMhPnQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa90cda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(2, 1, 1)\n",
    "plt.plot(loss_history)\n",
    "plt.xlabel('iteration')\n",
    "plt.ylabel('loss')\n",
    "\n",
    "plt.subplot(2, 1, 2)\n",
    "#print train_acc_history\n",
    "plt.plot(train_acc_history)\n",
    "plt.plot(val_acc_history)\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('accuracy')\n",
    "plt.show()\n",
    "\n",
    "#Show the learned weights...\n",
    "plt.title('First Layer Learned Weights shown as image')# they must look meaningful\n",
    "show_net_weights_mnist(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train the net\n",
    "Once the above works, training the net is the next thing to try. You can set the `acc_frequency` parameter to change the frequency at which the training and validation set accuracies are tested. If your parameters are set properly, you should see the training and validation accuracy start to improve within a hundred iterations, and you should be able to train a reasonable model with just one to five epochs.\n",
    "\n",
    "Using the parameters below you should be able to get around 94% accuracy on the validation set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Layer 1 Variance = 0.135429766531', 'Layer 2 Variance = 0.147470745082', 'Layer 3 Variance = 0.050059276536']\n"
     ]
    }
   ],
   "source": [
    "#Initalizaing the Network\n",
    "model=init_three_layer_neuralnet(weight_scale=1,input_feat_dim=X_train.shape[1],num_neurons=(60,40),num_classes=10);\n",
    "#testing variance across the network they must be reasonable values, if the weights are initalized properly...\n",
    "three_layer_neuralnetwork(X_train, model,y_train,0.001,verbose=1); "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N:  50000\n",
      "Finished epoch 0 / 17: cost 0.098245, train: 0.995000, val 0.980200, lr 1.000000e-01\n",
      "Finished epoch 1 / 17: cost 0.107425, train: 0.995000, val 0.979200, lr 9.500000e-02\n",
      "Finished epoch 2 / 17: cost 0.095621, train: 0.996000, val 0.979100, lr 9.025000e-02\n",
      "Finished epoch 3 / 17: cost 0.126695, train: 0.998000, val 0.979200, lr 8.573750e-02\n",
      "Finished epoch 4 / 17: cost 0.089477, train: 0.994000, val 0.979200, lr 8.145062e-02\n",
      "Finished epoch 5 / 17: cost 0.100690, train: 0.996000, val 0.979700, lr 7.737809e-02\n",
      "Finished epoch 6 / 17: cost 0.091755, train: 0.997000, val 0.979900, lr 7.350919e-02\n",
      "Finished epoch 7 / 17: cost 0.100238, train: 0.997000, val 0.979100, lr 6.983373e-02\n",
      "Finished epoch 8 / 17: cost 0.118427, train: 0.995000, val 0.979100, lr 6.634204e-02\n",
      "Finished epoch 9 / 17: cost 0.102727, train: 0.997000, val 0.979600, lr 6.302494e-02\n",
      "Finished epoch 10 / 17: cost 0.090373, train: 0.994000, val 0.979900, lr 5.987369e-02\n",
      "Finished epoch 11 / 17: cost 0.129021, train: 0.992000, val 0.978700, lr 5.688001e-02\n",
      "Finished epoch 12 / 17: cost 0.102144, train: 0.994000, val 0.979000, lr 5.403601e-02\n",
      "Finished epoch 13 / 17: cost 0.136621, train: 0.993000, val 0.980600, lr 5.133421e-02\n",
      "Finished epoch 14 / 17: cost 0.114305, train: 0.996000, val 0.979400, lr 4.876750e-02\n",
      "Finished epoch 15 / 17: cost 0.126840, train: 0.997000, val 0.979900, lr 4.632912e-02\n",
      "Finished epoch 16 / 17: cost 0.108627, train: 0.994000, val 0.980800, lr 4.401267e-02\n",
      "Finished epoch 17 / 17: cost 0.102031, train: 0.996000, val 0.980500, lr 4.181203e-02\n",
      "finished optimization. best validation accuracy: 0.980800\n"
     ]
    }
   ],
   "source": [
    "#Training the Network\n",
    "#rmsprop, nestrov-momentum,momentum\n",
    "trainer = ClassifierTrainer()\n",
    "best_model, loss_history, train_acc_history, val_acc_history = trainer.train(X_train, y_train, X_val, y_val,\n",
    "                                             model, three_layer_neuralnetwork,\n",
    "                                             num_epochs=18, reg=0.001,batch_size=64,\n",
    "                                             momentum=0.5, learning_rate_decay = 0.95,\n",
    "                                             learning_rate=1e-1,update='nestrov-momentum', verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnkAAAHuCAYAAADjghrHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xf0feb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeIAAAHpCAYAAABeLj9gAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkSGaiIUQQogM+T8MgrsTsrCtRAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21adbd30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(2, 1, 1)\n",
    "plt.plot(loss_history)\n",
    "plt.xlabel('iteration')\n",
    "plt.ylabel('loss')\n",
    "\n",
    "plt.subplot(2, 1, 2)\n",
    "#print train_acc_history\n",
    "plt.plot(train_acc_history)\n",
    "plt.plot(val_acc_history)\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('accuracy')\n",
    "plt.show()\n",
    "\n",
    "#Show the learned weights...\n",
    "plt.title('First Layer Learned Weights shown as image')# they must look meaningful\n",
    "show_net_weights_mnist(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Testing the Network...\n",
    "When you are done experimenting, you should evaluate your final trained network on the test set. \n",
    "\n",
    "**We will give you extra bonus point for every 1% of accuracy above 95%.**\n",
    "\n",
    "**Remember You should only use test set once you have completed your experimentation. Try improving the performance on Validation Set not on the test set.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy= 0.98\n"
     ]
    }
   ],
   "source": [
    "scores_test = three_layer_neuralnetwork(X_test, model)\n",
    "y_pred_test = np.argmax(scores_test, axis=1)\n",
    "test_acc = np.mean(y_test == y_pred_test)\n",
    "print ('Test Accuracy= {}'.format(test_acc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted    0     1     2     3    4    5    6     7    8     9    All\n",
      "Actual                                                                 \n",
      "0          971     0     1     0    0    1    2     1    4     0    980\n",
      "1            0  1126     2     1    0    1    3     0    2     0   1135\n",
      "2            6     1  1010     2    1    0    3     6    3     0   1032\n",
      "3            0     0     2   991    0    6    0     6    3     2   1010\n",
      "4            1     0     3     1  959    0    3     2    0    13    982\n",
      "5            2     0     0    10    0  868    3     1    5     3    892\n",
      "6            5     3     1     1    7    3  934     0    4     0    958\n",
      "7            1     7    10     3    0    0    0   999    1     7   1028\n",
      "8            4     0     1     3    1    1    2     2  958     2    974\n",
      "9            4     2     0     5    8    2    0     4    2   982   1009\n",
      "All        994  1139  1030  1017  976  882  950  1021  982  1009  10000\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2f407940>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdsAAAHaCAYAAABIJuORAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0dHPRR5wmAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xf807a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Confusion Matrix Across classes...\n",
    "from tools import print_confusion_matrix\n",
    "df=print_confusion_matrix(y_pred_test,y_test)\n",
    "print df\n",
    "plt.imshow(df.as_matrix())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.save('model-10-epochs-94',model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Your Turn\n",
    "\n",
    "**Tuning**. Tuning the hyperparameters and developing intuition for how they affect the final performance is a large part of using Neural Networks, so we want you to get a lot of practice. Below, you should experiment with different values of the various hyperparameters, including hidden layer size, learning rate, number of training epochs, and regularization strength. You might also consider tuning the momentum and learning rate decay parameters, but you should be able to get good performance using the default values.\n",
    "\n",
    "**Approximate results**. You should be aim to achieve a classification accuracy of greater than 95% on the validation set. Our best network gets over 95% on the validation set.\n",
    "\n",
    "**Experiment**: You goal in this exercise is to get as good of a result on this dataset as you can, with a fully-connected Neural Network. For every 1% above 95% on the Test set we will award you with one extra bonus point. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}