PyTorch Tutorial-loading-saving-model-lenet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "cab4753a",
   "metadata": {},
   "source": [
    "# PyTorch Tutorial, Loading and saving model (LeNet):\n",
    "#### We define and train LeNet with the MNIST data, and then save the trained model\n",
    "- Download the **MNIST** dataset\n",
    "- Define a **LeNet** (model)\n",
    "- Define loss function **MSE**  and optimizer as **SGD**\n",
    "- Train the LeNet and evaluate it\n",
    "- Save the LeNet parameters with **torch.save()** and load them with **torch.load()**\n",
    "- Save the entire LeNet model with *torch.save()* and load with *torch.load()*\n",
    "\n",
    "**Hint:** In previous post, we learned how to train a model with the MNIST dataset. This time, we use the LeNet model for the classification of handwritten digits.\n",
    "\n",
    "https://github.com/ostad-ai/PyTorch-Tutorial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "75354f9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# importing the necessary modules\n",
    "import torch\n",
    "import torch.nn as tnn\n",
    "from torchvision import datasets\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision.transforms import ToTensor, Lambda,Resize,Compose\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c5e5ce0",
   "metadata": {},
   "source": [
    "As we did before we download the MINST data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5ae0fc7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "training_set=datasets.MNIST(root='./MNIST',train=True, download=True, transform=ToTensor())\n",
    "test_set=datasets.MNIST(root='./MNIST',train=False,download=True,transform=ToTensor())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47462aff",
   "metadata": {},
   "source": [
    "Resize the input images of size 28-by-28 to size **32-by-32**, because the original **LeNet** gets images of size 32-by-32."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "d2515d44",
   "metadata": {},
   "outputs": [],
   "source": [
    "image_resize_transform=Compose([Resize(size=(32,32)),ToTensor()])\n",
    "training_set.transform=image_resize_transform\n",
    "test_set.transform=image_resize_transform"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "478c0f7a",
   "metadata": {},
   "source": [
    "Converting the class labels to one-hot vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "9ba42428",
   "metadata": {},
   "outputs": [],
   "source": [
    "one_hot_transform=Lambda(lambda y:torch.zeros(Nclasses,dtype=torch.float).index_put([torch.tensor(y)],\n",
    "                                                                         values=torch.tensor(1.)))\n",
    "training_set.target_transform=one_hot_transform\n",
    "test_set.target_transform=one_hot_transform"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2626bd63",
   "metadata": {},
   "source": [
    "Using **DataLoader** to make dataset iterable and in a form of batches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "dad868ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "batchSize=16\n",
    "train_dataloader=DataLoader(training_set,batch_size=batchSize,shuffle=True)\n",
    "test_dataloader=DataLoader(test_set,batch_size=batchSize,shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "fded2a91",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The device chosen is: cuda\n"
     ]
    }
   ],
   "source": [
    "device=torch.device('cuda' if torch.cuda.is_available()\n",
    "         else 'mps' if torch.backends.mps.is_available()\n",
    "         else 'cpu')\n",
    "print(f'The device chosen is: {device}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c9bede5",
   "metadata": {},
   "source": [
    "Defining the **Lenet** model, and set the device of model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "354a4b66",
   "metadata": {},
   "outputs": [],
   "source": [
    "cin,Nclasses=1,10\n",
    "model=tnn.Sequential(\n",
    "      tnn.Conv2d(cin,6,kernel_size=5),\n",
    "      tnn.ReLU(),\n",
    "      tnn.MaxPool2d(kernel_size=2,stride=2),\n",
    "      tnn.Conv2d(6,16,kernel_size=5),\n",
    "      tnn.ReLU(),\n",
    "      tnn.MaxPool2d(kernel_size=2,stride=2),\n",
    "      tnn.Flatten(),\n",
    "      tnn.Linear(16*5*5,120), #16*5*5=400\n",
    "      tnn.ReLU(),\n",
    "      tnn.Linear(120,84),\n",
    "      tnn.ReLU(),\n",
    "      tnn.Linear(84,Nclasses)      \n",
    ")\n",
    "model=model.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c9c4fc5",
   "metadata": {},
   "source": [
    "Choose the loss function and optimizer:\n",
    " - Loss function is **Mean Squared Error** (MSE)\n",
    " - The optimzier is **Stochastic Gradient Descent** (SGD)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e19b3c81",
   "metadata": {},
   "outputs": [],
   "source": [
    "learningRate=.001 # this is the learning rate in the SGD algorithm\n",
    "optimizer=torch.optim.SGD(model.parameters(),lr=learningRate)\n",
    "lossFn=tnn.MSELoss().to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8e05b45",
   "metadata": {},
   "source": [
    "We train the model with the training set.\n",
    "<br>Each run over the whole training set is called an **epoch**. We train the model over several epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "367566fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch0, loss:349.38851753622293; Epoch1, loss:337.70612182468176; Epoch2, loss:335.6845654696226; Epoch3, loss:334.7944259047508; Epoch4, loss:333.91106417030096; Epoch5, loss:332.830480478704; Epoch6, loss:331.4548827111721; Epoch7, loss:329.6796717569232; Epoch8, loss:327.3902518302202; Epoch9, loss:324.37757478654385; Epoch10, loss:320.28624837100506; Epoch11, loss:314.5762178003788; Epoch12, loss:306.73349337279797; Epoch13, loss:296.69125117361546; Epoch14, loss:285.1521411202848; Epoch15, loss:272.75507967174053; Epoch16, loss:259.2937738597393; Epoch17, loss:244.70307748019695; Epoch18, loss:229.1961415298283; Epoch19, loss:213.2877288274467; Epoch20, loss:198.28081243112683; Epoch21, loss:185.34943034686148; Epoch22, loss:174.62257538549602; Epoch23, loss:165.69931664131582; Epoch24, loss:158.00047400780022; Epoch25, loss:151.15065685845912; Epoch26, loss:144.8699934706092; Epoch27, loss:139.07992906030267; Epoch28, loss:133.69492955878377; Epoch29, loss:128.66235998366028; "
     ]
    }
   ],
   "source": [
    "Nepochs=30\n",
    "loss_epochs=[]\n",
    "model.train()\n",
    "for i in range(Nepochs):\n",
    "    losses=0.\n",
    "    for x,y in train_dataloader:\n",
    "        x,y=x.to(device),y.to(device)\n",
    "        output=model(x)\n",
    "        loss=lossFn(output,y)\n",
    "        losses+=loss.item()\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "    print(f'Epoch{i}, loss:{losses}', end='; ')\n",
    "    loss_epochs.append(losses)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4ad51a9",
   "metadata": {},
   "source": [
    "At this step, we  show the curve of training loss of each epoch as epoch advances:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "f13860c6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj0AAAHHCAYAAABUcOnjAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/P9b71AAAACXBIWXMAAA9hAAAPYQGoP6dpAABh2klEQVR4nO3dd1gT9wMG8DeMhL1kKyCCggho68Rtse5tq9ZdtdZVq1Vr6dYOtbbWWuvocrVat1Zbt4KKuHChIgqiqCwB2Tu53x/U/BrBSiLhAnk/z5NHuFySN2favN5973sSQRAEEBEREdVyBmIHICIiIqoOLD1ERESkF1h6iIiISC+w9BAREZFeYOkhIiIivcDSQ0RERHqBpYeIiIj0AksPERER6QWWHiIiItILLD1EOmjs2LGoX7++Ro/99NNPIZFIqjZQJT1PbtIfEokE06ZNEzsG6SGWHiI1SCSSSt1CQ0PFjkpERE8wEjsAUU2yYcMGld/Xr1+PQ4cOlVveuHHj53qdn376CQqFQqPHfvjhh3jvvfee6/WJiGojlh4iNYwcOVLl99OnT+PQoUPllj8pPz8fZmZmlX4dY2NjjfIBgJGREYyM+J92dSotLYVCoYBUKhU7ChH9Bx7eIqpinTt3hr+/PyIjI9GxY0eYmZnh/fffBwDs3r0bvXv3hqurK2QyGby8vPDZZ59BLperPMeTY2Pu3LkDiUSCr7/+Gj/++CO8vLwgk8nQsmVLnDt3TuWxFY3peTyGYteuXfD394dMJkOTJk2wf//+cvlDQ0PRokULmJiYwMvLC6tXr36ucUJ5eXmYNWsW3NzcIJPJ4OPjg6+//hqCIKisd+jQIbRv3x42NjawsLCAj4+Pcrs99v3336NJkyYwMzODra0tWrRogY0bNz71tVNSUmBkZIR58+aVuy8mJgYSiQTLly9XLsvMzMSMGTOUWb29vbFo0SKVvW7//rtYunSp8u/i+vXrlcr4tHFPFW3jymyTp/ntt9/QvHlzmJqaws7ODsOGDcO9e/dU1vn3Z7Vt27YwNTWFp6cnVq1aVe75UlNTMX78eDg5OcHExARNmzbFunXryq2nUCjw3XffISAgACYmJnBwcECPHj1w/vz5cutW5vNIVJX4z0EiLUhPT0fPnj0xbNgwjBw5Ek5OTgCAtWvXwsLCAu+88w4sLCxw9OhRfPzxx8jOzsbixYuf+bwbN25ETk4O3nzzTUgkEnz11VcYNGgQbt++/cy9QydPnsSOHTswZcoUWFpaYtmyZRg8eDASEhJQp04dAMDFixfRo0cPuLi4YN68eZDL5Zg/fz4cHBw02g6CIKBfv344duwYxo8fj2bNmuHAgQOYM2cOHjx4gG+//RYAcO3aNfTp0weBgYGYP38+ZDIZYmNjER4ernyun376CdOnT8crr7yCt99+G4WFhbhy5QrOnDmD4cOHV/j6Tk5O6NSpE7Zs2YJPPvlE5b7NmzfD0NAQr776KoCyvXGdOnXCgwcP8Oabb8Ld3R2nTp1CSEgIkpKSsHTpUpXHr1mzBoWFhZg4cSJkMhns7Ow0yvg0ldkmT/PFF1/go48+wpAhQzBhwgQ8fPgQ33//PTp27IiLFy/CxsZGue6jR4/Qq1cvDBkyBK+99hq2bNmCyZMnQyqVYty4cQCAgoICdO7cGbGxsZg2bRo8PT2xdetWjB07FpmZmXj77beVzzd+/HisXbsWPXv2xIQJE1BaWooTJ07g9OnTaNGihXK9ynweiaqcQEQamzp1qvDkf0adOnUSAAirVq0qt35+fn65ZW+++aZgZmYmFBYWKpeNGTNG8PDwUP4eHx8vABDq1KkjZGRkKJfv3r1bACDs2bNHueyTTz4plwmAIJVKhdjYWOWyy5cvCwCE77//Xrmsb9++gpmZmfDgwQPlslu3bglGRkblnrMiT+betWuXAED4/PPPVdZ75ZVXBIlEoszz7bffCgCEhw8fPvW5+/fvLzRp0uSZGZ60evVqAYAQFRWlstzPz0946aWXlL9/9tlngrm5uXDz5k2V9d577z3B0NBQSEhIEATh/38XVlZWQmpqqtoZn9xGjz3591aZbVKRO3fuCIaGhsIXX3yhsjwqKkowMjJSWf74s/rNN98olxUVFQnNmjUTHB0dheLiYkEQBGHp0qUCAOG3335TrldcXCwEBQUJFhYWQnZ2tiAIgnD06FEBgDB9+vRyuRQKhfLnyn4eiaoaD28RaYFMJsPrr79ebrmpqany55ycHKSlpaFDhw7Iz8/HjRs3nvm8Q4cOha2trfL3Dh06AABu3779zMd27doVXl5eyt8DAwNhZWWlfKxcLsfhw4cxYMAAuLq6Ktfz9vZGz549n/n8Ffn7779haGiI6dOnqyyfNWsWBEHAvn37AEC552H37t1PHcBtY2OD+/fvlzuc9yyDBg2CkZERNm/erFx29epVXL9+HUOHDlUu27p1Kzp06ABbW1ukpaUpb127doVcLsfx48dVnnfw4MHl9oBpmrEildkmFdmxYwcUCgWGDBmi8j6cnZ3RsGFDHDt2TGV9IyMjvPnmm8rfpVIp3nzzTaSmpiIyMhJA2d+js7MzXnvtNeV6xsbGmD59OnJzcxEWFgYA2L59OyQSSbm9agDKHbp71ueRSBtYeoi0oG7duhUOar127RoGDhwIa2trWFlZwcHBQTkIOisr65nP6+7urvL74wL06NEjtR/7+PGPH5uamoqCggJ4e3uXW6+iZZVx9+5duLq6wtLSUmX547Pb7t69C6CszLVr1w4TJkyAk5MThg0bhi1btqh82c+dOxcWFhZo1aoVGjZsiKlTp1bqUI+9vT2Cg4OxZcsW5bLNmzfDyMgIgwYNUi67desW9u/fDwcHB5Vb165dAZRtn3/z9PQs91qaZqxIZbZJRW7dugVBENCwYcNy7yU6Orrc+3B1dYW5ubnKskaNGgEoG78ElP09NWzYEAYGql8ZT/49xsXFwdXVFXZ2ds98f8/6PBJpA8f0EGnBv/foPJaZmYlOnTrBysoK8+fPh5eXF0xMTHDhwgXMnTu3Uv+aNzQ0rHC58MSg4Kp+rLaZmpri+PHjOHbsGP766y/s378fmzdvxksvvYSDBw/C0NAQjRs3RkxMDPbu3Yv9+/dj+/btWLFiBT7++OMKByr/27Bhw/D666/j0qVLaNasGbZs2YLg4GDY29sr11EoFHj55Zfx7rvvVvgcj4vAvzM/qTIZnzYg/MnB7JXZJhVRKBSQSCTYt29fhetYWFhU+LjqpsufR6q9WHqIqkloaCjS09OxY8cOdOzYUbk8Pj5exFT/5+joCBMTE8TGxpa7r6JlleHh4YHDhw8jJydHZW/P40N5Hh4eymUGBgYIDg5GcHAwlixZgi+//BIffPABjh07ptzbYm5ujqFDh2Lo0KEoLi7GoEGD8MUXXyAkJAQmJiZPzTFgwAC8+eabykNcN2/eREhIiMo6Xl5eyM3NVb6Wpp6V0dbWFpmZmeUe93hvyb9VZps8ycvLC4IgwNPTs1xRq0hiYiLy8vJU9vbcvHkTAJRnmXl4eODKlStQKBQqe3ue/Hv08vLCgQMHkJGRUam9PUTVjYe3iKrJ43/Z/vtfssXFxVixYoVYkVQYGhqia9eu2LVrFxITE5XLY2NjlWNv1NWrVy/I5XKV08IB4Ntvv4VEIlGOFcrIyCj32GbNmgEAioqKAJSdEfdvUqkUfn5+EAQBJSUl/5nDxsYG3bt3x5YtW/DHH39AKpViwIABKusMGTIEEREROHDgQLnHZ2ZmorS09D9fo7IZvby8kJWVhStXrijXS0pKws6dO1UeW5ltUpFBgwbB0NAQ8+bNK7fXRBCEchlLS0uxevVq5e/FxcVYvXo1HBwc0Lx5cwBlf4/Jyckq46JKS0vx/fffw8LCAp06dQJQNs5JEIQK97xxDw7pAu7pIaombdu2ha2tLcaMGYPp06dDIpFgw4YNOvVl8Omnn+LgwYNo164dJk+erCws/v7+uHTpktrP17dvX3Tp0gUffPAB7ty5g6ZNm+LgwYPYvXs3ZsyYoRzIOn/+fBw/fhy9e/eGh4cHUlNTsWLFCtSrVw/t27cHAHTr1g3Ozs5o164dnJycEB0djeXLl6N3797lxgxVZOjQoRg5ciRWrFiB7t27q5y2DQBz5szBn3/+iT59+mDs2LFo3rw58vLyEBUVhW3btuHOnTsqh8MqUpmMw4YNw9y5czFw4EBMnz4d+fn5WLlyJRo1aoQLFy4on6sy26QiXl5e+PzzzxESEoI7d+5gwIABsLS0RHx8PHbu3ImJEydi9uzZyvVdXV2xaNEi3LlzB40aNcLmzZtx6dIl/Pjjj8ppECZOnIjVq1dj7NixiIyMRP369bFt2zaEh4dj6dKlyvfWpUsXjBo1CsuWLcOtW7fQo0cPKBQKnDhxAl26dOH1tkh8YpwyRlRbPO2U9aedthweHi60adNGMDU1FVxdXYV3331XOHDggABAOHbsmHK9p52yvnjx4nLPCUD45JNPlL8/7ZT1qVOnlnush4eHMGbMGJVlR44cEV544QVBKpUKXl5ews8//yzMmjVLMDExecpW+L+KTsfOyckRZs6cKbi6ugrGxsZCw4YNhcWLF6ucwnzkyBGhf//+gqurqyCVSgVXV1fhtddeUzl9fPXq1ULHjh2FOnXqCDKZTPDy8hLmzJkjZGVlPTOXIAhCdna2YGpqWu7U6yezhoSECN7e3oJUKhXs7e2Ftm3bCl9//bXy9O3/+ruobMaDBw8K/v7+glQqFXx8fITffvut3N9bZbbJf9m+fbvQvn17wdzcXDA3Nxd8fX2FqVOnCjExMcp1Hn9Wz58/LwQFBQkmJiaCh4eHsHz58nLPl5KSIrz++uuCvb29IJVKhYCAAGHNmjXl1istLRUWL14s+Pr6ClKpVHBwcBB69uwpREZGKtdR5/NIVJUkgqBD/8wkIp00YMAAXLt2Dbdu3RI7ClWhzp07Iy0tDVevXhU7ClG14JgeIlJRUFCg8vutW7fw999/o3PnzuIEIiKqIhzTQ0QqGjRogLFjx6JBgwa4e/cuVq5cCalU+tRTuYmIagqWHiJS0aNHD2zatAnJycmQyWQICgrCl19+iYYNG4odjYjouXBMDxEREekFjukhIiIivcDSQ0RERHqBY3pQdq2axMREWFpaPvW6OERERKRbBEFATk4OXF1dy10QtyIsPSi79oybm5vYMYiIiEgD9+7dQ7169Z65HksPoJxC/d69e7CyshI5DREREVVGdnY23NzcKnUpGoClBwCUh7SsrKxYeoiIiGqYyg5N4UBmIiIi0gssPURERKQXWHqIiIhIL7D0EBERkV5g6SEiIiK9wNJDREREeoGlh4iIiPQCSw8RERHpBZYeIiIi0guilp6VK1ciMDBQORNyUFAQ9u3bp7y/c+fOkEgkKrdJkyapPEdCQgJ69+4NMzMzODo6Ys6cOSgtLa3ut0JEREQ6TtTLUNSrVw8LFy5Ew4YNIQgC1q1bh/79++PixYto0qQJAOCNN97A/PnzlY8xMzNT/iyXy9G7d284Ozvj1KlTSEpKwujRo2FsbIwvv/yy2t8PERER6S6JIAiC2CH+zc7ODosXL8b48ePRuXNnNGvWDEuXLq1w3X379qFPnz5ITEyEk5MTAGDVqlWYO3cuHj58CKlUWqnXzM7OhrW1NbKysnjtLSIiohpC3e9vnRnTI5fL8ccffyAvLw9BQUHK5b///jvs7e3h7++PkJAQ5OfnK++LiIhAQECAsvAAQPfu3ZGdnY1r16499bWKioqQnZ2tctOG3KJSnIpN08pzExERkXpEv8p6VFQUgoKCUFhYCAsLC+zcuRN+fn4AgOHDh8PDwwOurq64cuUK5s6di5iYGOzYsQMAkJycrFJ4ACh/T05OfuprLliwAPPmzdPSOyqTmV+M4T+dQWxqLjaMb4XWDepo9fWIiIjov4leenx8fHDp0iVkZWVh27ZtGDNmDMLCwuDn54eJEycq1wsICICLiwuCg4MRFxcHLy8vjV8zJCQE77zzjvL37OxsuLm5Pdf7eJKliTHc7cxwPSkbb6w/jx1T2sLb0bJKX4OIiIgqT/TDW1KpFN7e3mjevDkWLFiApk2b4rvvvqtw3datWwMAYmNjAQDOzs5ISUlRWefx787Ozk99TZlMpjxj7PGtqhkaSLB0WDO84G6D7MJSjF1zDg9ziqr8dYiIiKhyRC89T1IoFCgqqrgcXLp0CQDg4uICAAgKCkJUVBRSU1OV6xw6dAhWVlbKQ2RiMjE2xM+jW8CjjhnuPyrA+HXnkF/M0+mJiIjEIGrpCQkJwfHjx3Hnzh1ERUUhJCQEoaGhGDFiBOLi4vDZZ58hMjISd+7cwZ9//onRo0ejY8eOCAwMBAB069YNfn5+GDVqFC5fvowDBw7gww8/xNSpUyGTycR8a0p1LGRY+3or2JoZ48r9LLy18SJK5QqxYxEREekdUUtPamoqRo8eDR8fHwQHB+PcuXM4cOAAXn75ZUilUhw+fBjdunWDr68vZs2ahcGDB2PPnj3KxxsaGmLv3r0wNDREUFAQRo4cidGjR6vM66MLPO3N8fOYlpAZGeDIjVR8uucadGymACIiolpP5+bpEUN1zdOzLyoJUzZegCAAIT198WYnzQdjExER6bsaO0+PPugZ4IIPe5eNNVqw7wb2XE4UOREREZH+YOmpZuPbe+L1dvUBALO2XMbZ+AxxAxEREekJlh4RfNjbD92bOKFYrsAb688j7mGu2JGIiIhqPZYeERgaSLB06Ato5maDrIISjF1zlnP4EBERaRlLj0hMpYb4ZUzZHD73MgowgXP4EBERaRVLj4jqWMiwZmxL2JoZ4/L9LEzfdBFyhd6fTEdERKQVLD0ia+BggZ/HtIDUyACHo1Mxj3P4EBERaQVLjw5o7mGH74Y2g0QCrI+4i59PxIsdiYiIqNZh6dERPQNc8EGvxgCAL/6Oxl9XkkROREREVLuw9OiQ8e09MbZtfQDAzC2XcO4O5/AhIiKqKiw9OkQikeCjPn542c8JxaVlc/hExKWjsEQudjQiIqIaj9feQvVde6uyCorlGPbTaVy+lwkAMDKQoKGTJfxdrRBQzxr+da3R2NkKplJDcYMSERGJSN3vb5Ye6F7pAYC03CJ8svsaIm6nIyOvuNz9hgYSeDtYwL+uNfzrWiGgrjX8XK1gJjUSIS0REVH1Y+nRgC6WnscEQUBSViGiHmTh6j+3qAfZSMstP4OzRAJ4OVggoO7jvUGWcLSSoY65DNamxjAwkIjwDoiIiLSDpUcDulx6KiIIAlKyi/4pQFm4llj2Z0r20y9lYWQggZ25FHUsZLC3kMLeQoY65lLYW/7zp4WsbJmFFHUspJAZ8dAZERHpNnW/v3kspAaSSCRwtjaBs7UJuvo5KZen5hTi2oNsRP1ThuIe5iItpwjZhaUoVQhIzSlCaiWv8WVpYqQsRmVFSAZ7ZUkqK0f2FlLuRSIiohqDe3pQ8/b0qKu4VIGMvGKk5RYhLbcI6bllP6f/s0z5e24x0vOKUCJX7yNh+HgvknKPUVlJcrKSwcnKBC7WpnC2MoGjlQwmxtyDREREVYN7eqgcqZGBcs/QswiCgOyCUjzMLUJGXjHSc4uQlleMtJwipOf9U4xyi5H2z89ZBSWQKwQ8zCn650rxOf/5/LZmxnCyKsvibGWi8vPjP23MjCGRcM8RERFVLZYeUiGRSGBtZgxrM+NKrf/vvUjp/5Sk9NxiPMwtQnJWIZKzC5GSXYjkrEIUlSrwKL8Ej/JLcCP56eVIZmQAVxtTuNuZwd3ODB51zOD2+E9bM5jL+LElIiL18duDnktl9yIJgoCsghIk/1OAyopQkUopSskuRHpeMYpKFYhPy0N8Wl6Fz2VvIVUWIvc65spi5G5nBkdLGfcSERFRhVh6qFpIJBLYmElhYyaFr/PTj7sWlcqRml2E+48KkJCRh4SMfNxNz8e9jHzczchHZn4J0nKLkZZbjAsJmeUeb2JsAHc7MzR0tISPsyUaOVnC19kSbnZmMORgayIivcaBzKj9A5lrk6yCEtzLyFeWoYSMfGU5SswshFxR8cfZxNgAjZz+X4Ie/+nAPUNERDUW5+nRAEtP7VAiVyAxswDxaXm4mZKDmORcxKRk41ZKLopKFRU+xsbMGD5OZXuFfJwt4eNkCV8XK1hw3BARkc5j6dEAS0/tJlcIuJteVoRuJOco/7yTloeKdgxJJICPkyVe9LDFi+62eNHdBp725twjRESkY1h6NMDSo58KS+SITc39Z69QDmL++TMpq7DcunbmUrzgZqMsQk3drHmdMyIikbH0aIClh/4tNacQF+5m4kLCI1y4+whXHmSh+InDY4YGEjR2sfxnT5AtmnvYop6tKfcGERFVI5YeDbD00H8pLlXgWmIWLiRk4sLdR4i8+wjJ2eX3BtlbyNDcwwbtGzrgJV9H1LUxFSEtEZH+YOnRAEsPqSsxs+CfPUGZiEx4hOuJWeUu3+HrbIkuvo4I9nXEC+62PGWeiKiKsfRogKWHnldhiRxRD7JwNj4Dx26k4kLCI5VB0jZmxujcyAFdfB3RuZFjpWe8JiKip2Pp0QBLD1W1R3nFCLv5EEdvpCI0JhXZhaXK+wwNJGjubouXGpftBfJ2tOBYICIiDbD0aIClh7SpVK7AhYRMHLmRgmM3UnEzJVfl/nq2pgj2dUQXX0e0aVCHV6InIqoklh4NsPRQdbqXkY9jMak4Ep2KiNvpKmeGWcqM0P8FVwxv5QE/V34WiYj+C0uPBlh6SCz5xaUIj03H0RupOHojBSnZRcr7mrnZYHgrd/Rp6sI5gYiIKsDSowGWHtIFCoWAiNvp2HgmAQeuJaP0n5HQljIjDHihLoa3dkdjF34+iYgeY+nRAEsP6ZqHOUXYFnkfm84mICEjX7n8BXcbvNbKHX0DXWEq5dgfItJvLD0aYOkhXaVQCDgVl45NZ5/Y+2NihEEv1MVrrd3h68zPLBHpJ5YeDbD0UE3wMKcIWyPv4Y+z91T2/rz4z96fPtz7Q0R6hqVHAyw9VJMoFALC49Kw6WwCDl5LUdn7MyaoPiZ19oKFjAOfiaj2Y+nRAEsP1VSpOYXYFnlfZe+PvYUM77zcCENa1IORoYHICYmItIelRwMsPVTTKRQCDl5PwaL9NxCflgcAaORkgfd7NUZnH0eR0xERaQdLjwZYeqi2KC5V4LfTd7Hs6C1k5pcAADo0tMcHvRtzwDMR1TosPRpg6aHaJiu/BN8fvYV1EXdQIhdgIAGGtHDDO90awdHSROx4RERVgqVHAyw9VFvdTc/Dov038HdUMgDATGqIyZ28MKFDA57pRUQ1HkuPBlh6qLY7fycDn/8VjUv3MgEALtYmmN3NBwNfqAsDA17hnYhqJpYeDbD0kD4QBAF7riRh0b4beJBZAADwr2uFD3r5IcirjsjpiIjUx9KjAZYe0ieFJXKsCb+DFcdikVNUCgB42c8JIT190cDBQuR0RESVp+73NyfxINIzJsaGmNzZC6FzOmNUGw8YGkhw6HoKui89jrXh8eC/g4iotmLpIdJTdSxk+GyAPw7M6IDOPg4okQv4dM91vLXpInL/2QNERFSbsPQQ6TlvR0usGdsSH/fxg5GBBHuvJKH/8pO4mZIjdjQioirF0kNEkEgkGNfeE5vfbANnKxPEPcxD/+Xh2HXxgdjRiIiqDEsPESk197DDX9Pbo723PQpK5Jix+RI+3BWFolK52NGIiJ4bSw8RqahjIcO6ca0wPbghJBLgt9MJeHVVBO79c0FTIqKaiqWHiMoxNJDgnZcbYc3YlrAxM8aV+1no8/1JHLuRKnY0IiKNsfQQ0VN19nHEX9M7oKmbDbIKSvD62nP4+kAM5Aqe1k5ENQ9LDxH9p7o2ptjyZhuMDvIAACw/FotRv5xBWm6RyMmIiNTD0kNEzyQzMsT8/v74blgzmEkNcSouHb2XncD5OxliRyMiqjSWHiKqtP7N6uLPae3g7WiBlOwiDP3xNH4+cZuzOBNRjcDSQ0Rq8Xa0xO6p7dCvqSvkCgGf/xWNKb9fQE5hidjRiIj+E0sPEanNXGaE74Y1w2f9m8DYUIJ9V5PxysoIjvMhIp3G0kNEGpFIJBgVVB9bJ7WFo6UMMSk5eO3H00jNKRQ7GhFRhVh6iOi5NHOzweY3g+BsZYJbqbkY9uNppGSz+BCR7mHpIaLn5mlvjs1vtkFdG1PcfpiHYT+eRlJWgdixiIhUsPQQUZXwqGOOPya2QT1bU8Sn5WHo6tN4kMniQ0S6g6WHiKqMm50Z/pjYBu52ZkjIyMfQ1bxmFxHpDpYeIqpS9WzNsPnNNqhfxwz3HxVg2I+ncTc9T+xYRETilp6VK1ciMDAQVlZWsLKyQlBQEPbt26e8v7CwEFOnTkWdOnVgYWGBwYMHIyUlReU5EhIS0Lt3b5iZmcHR0RFz5sxBaWlpdb8VIvoXF2tTbH4zCA3szfEgs6z4xKex+BCRuEQtPfXq1cPChQsRGRmJ8+fP46WXXkL//v1x7do1AMDMmTOxZ88ebN26FWFhYUhMTMSgQYOUj5fL5ejduzeKi4tx6tQprFu3DmvXrsXHH38s1lsion84WZngj4lt4O1ogaSsQgz7MQJxD3PFjkVEekwi6Nj88XZ2dli8eDFeeeUVODg4YOPGjXjllVcAADdu3EDjxo0RERGBNm3aYN++fejTpw8SExPh5OQEAFi1ahXmzp2Lhw8fQiqVVuo1s7OzYW1tjaysLFhZWWntvRHpo4c5RRj58xnEpOTAwVKGjRNao6GTpdixiKgWUPf7W2fG9Mjlcvzxxx/Iy8tDUFAQIiMjUVJSgq5duyrX8fX1hbu7OyIiIgAAERERCAgIUBYeAOjevTuys7OVe4sqUlRUhOzsbJUbEWmHg6UMG99oDV9nSzzMKcJrP51GTHKO2LGISA+JXnqioqJgYWEBmUyGSZMmYefOnfDz80NycjKkUilsbGxU1ndyckJycjIAIDk5WaXwPL7/8X1Ps2DBAlhbWytvbm5uVfumiEhFHQsZNr3RBn4uVkjLLcZrP53G9UT+Y4OIqpfopcfHxweXLl3CmTNnMHnyZIwZMwbXr1/X6muGhIQgKytLebt3755WX4+IAFtzKTa+0RoBda2RkVeM4T+fxtUHWWLHIiI9InrpkUql8Pb2RvPmzbFgwQI0bdoU3333HZydnVFcXIzMzEyV9VNSUuDs7AwAcHZ2Lnc21+PfH69TEZlMpjxj7PGNiLTPxkyK3ya0RlM3G2Tml2D4T6dx5X6m2LGISE+IXnqepFAoUFRUhObNm8PY2BhHjhxR3hcTE4OEhAQEBQUBAIKCghAVFYXU1FTlOocOHYKVlRX8/PyqPTsRPZu1qTE2jG+FF91tkF1YihE/n8HFhEdixyIiPSBq6QkJCcHx48dx584dREVFISQkBKGhoRgxYgSsra0xfvx4vPPOOzh27BgiIyPx+uuvIygoCG3atAEAdOvWDX5+fhg1ahQuX76MAwcO4MMPP8TUqVMhk8nEfGtE9B+sTIyxfnxrtKxvi5zCUoz65Sz3+BCR1olaelJTUzF69Gj4+PggODgY586dw4EDB/Dyyy8DAL799lv06dMHgwcPRseOHeHs7IwdO3YoH29oaIi9e/fC0NAQQUFBGDlyJEaPHo358+eL9ZaIqJIsZEZY+3ortPa0Q25RKcatPYeEdF6ygoi0R+fm6RED5+khEk9OYQmGrj6N60nZ8LQ3x/bJbWFnXrk5tohIv9XYeXqISD9ZmhhjzestUdem7Ors49edQ0GxXOxYRFQLaVx6iouLcf/+fSQkJKjciIjU5WRlgnXjWsLKxAgXEzLx9h8XIVfo/U5oIqpiapeeW7duoUOHDjA1NYWHhwc8PT3h6emJ+vXrw9PTUxsZiUgPeDta4ucxLSE1MsDB6yn49M9r4NF3IqpKRuo+YOzYsTAyMsLevXvh4uICiUSijVxEpIdaedph6dBmmLrxAjacvgtXG1NM7uwldiwiqiXULj2XLl1CZGQkfH19tZGHiPRcrwAXfNjbD5/tvY5F+2/AxdoEA16oK3YsIqoF1D685efnh7S0NG1kISICAIxv74kJ7csOl8/Zdhnhsfx/DhE9v0qVnn9fjXzRokV49913ERoaivT0dF6tnIi04v1ejdE70AUlcgGTNkQiOon/fyGi51OpeXoMDAxUxu4IglBuLM/jZXJ5zTvVlPP0EOmmwhI5Rv96FmfjM+BkJcPOKe3gamMqdiwi0hHqfn9XakzPsWPHnjsYEZG6TIwN8dOoFnhl1SncSs3F2DVnsXVSW1ibGosdjYhqIM7IDO7pIdJ1DzILMGhFOFKyi9Da0w7rx7eCzMhQ7FhEJDKtz8i8Zs0abN26tdzyrVu3Yt26deo+HRHRM9W1McWasa1gITPCmfgMzNpyGQpOXkhEalK79CxYsAD29vblljs6OuLLL7+sklBERE/yc7XC6lHNYWQgwd4rSViwL1rsSERUw6hdehISEiqcednDw4OXoSAirWrnbY/FrwYCAH46EY9fT8aLnIiIahK1S4+joyOuXLlSbvnly5dRp06dKglFRPQ0A1+oh3d7+AAAPvvrOvZFJYmciIhqCrVLz2uvvYbp06fj2LFjkMvlkMvlOHr0KN5++20MGzZMGxmJiFRM7uSFUW08IAjA25sv4dydDLEjEVENoPbZW8XFxRg1ahS2bt0KI6OyM94VCgVGjx6NVatWQSqVaiWoNvHsLaKaR64QMOm3SBy6ngIbM2PsntoOHnXMxY5FRNVI3e9vjU9Zv3nzJi5fvgxTU1MEBATAw8NDk6fRCSw9RDVTQbEcw36MwOX7WWjoaIEdU9rC0oRz+BDpi2orPUDZLMwAavyV1ll6iGqulOxC9Ft+EinZRQj2dcSPo1vA0KBm/z+JiCpH6/P0AMD69esREBAAU1NTmJqaIjAwEBs2bNDkqYiInouTlQl+HNUCMiMDHLmRisUHYsSOREQ6Su3Ss2TJEkyePBm9evXCli1bsGXLFvTo0QOTJk3Ct99+q42MRET/qambDb56pexU9lVhcdh58b7IiYhIF6l9eMvT0xPz5s3D6NGjVZavW7cOn376KeLja968GTy8RVQ7LNp/AytD4yA1MsDmiW3wgrut2JGISIu0fngrKSkJbdu2Lbe8bdu2SErifBlEJJ453XzQtbEjiksVeHNDJJKzCsWOREQ6RO3S4+3tjS1btpRbvnnzZjRs2LBKQhERacLAQIKlw15AIycLpOYUYeKG8ygskYsdi4h0hJG6D5g3bx6GDh2K48ePo127dgCA8PBwHDlypMIyRERUnSxkRvh5dEv0/+EkrtzPwpxtV7BsWLMaf5YpET0/tff0DB48GGfOnIG9vT127dqFXbt2wd7eHmfPnsXAgQO1kZGISC3udcywYkTZxUn3XE7EitA4sSMRkQ54rnl6agsOZCaqnX47fRcf7roKAPhxVHN0a+IsciIiqkrqfn+rfXgLAORyOXbu3Ino6GgAgJ+fH/r376+8LAURkS4Y2cYDMck52HD6LmZsvoQdU9rC15n/sCHSV2rv6bl27Rr69euH5ORk+PiUXen45s2bcHBwwJ49e+Dv76+VoNrEPT1EtVeJXIHRv5xFxO101LM1xe6p7VDHQiZ2LCKqAlo/ZX3ChAlo0qQJ7t+/jwsXLuDChQu4d+8eAgMDMXHiRI1CExFpi7GhAVaMeBEedcxw/1EBJv9+AcWlCrFjEZEI1N7TY2pqivPnz6NJkyYqy69evYqWLVuioKCgSgNWB+7pIar9bqXkYOCKU8gtKsXw1u74YoA/z+giquG0vqenUaNGSElJKbc8NTUV3t7e6j4dEVG1aOhkiWWvNYNEAmw8k4ANp++KHYmIqpnapWfBggWYPn06tm3bhvv37+P+/fvYtm0bZsyYgUWLFiE7O1t5IyLSJS/5OmFuD18AwLw91xEemyZyIiKqTmof3jIw+H9Perxr+PFT/Pt3iUQCubxmzITKw1tE+kMQBMzachk7Lj6Atakxdk9th/r25mLHIiINaP2U9WPHjmkUjIhIF0gkEnw5KAC30/Jw6V4mJqw/jx1T2sLKxFjsaESkZZycENzTQ6SPUrML0W95OJKzC9GjiTNWjnyRA5uJahitD2QGgBMnTmDkyJFo27YtHjx4AADYsGEDTp48qcnTERFVO0crE6wa1RzGhhLsv5aMX8PviB2JiLRM7dKzfft2dO/eHaamprhw4QKKiooAAFlZWfjyyy+rPCARkbY0c7PBB70aAwAW/B2NCwmPRE5ERNqkdun5/PPPsWrVKvz0008wNv7/MfB27drhwoULVRqOiEjbxrStj94BLihVCJj2+wU8yisWOxIRaYnapScmJgYdO3Yst9za2hqZmZlVkYmIqNpIJBIsHByA+nXMkJhViHe2XIJCofdDHYlqJbVLj7OzM2JjY8stP3nyJBo0aFAloYiIqpOliTFWjGgOqZEBjsU8xKrjcWJHIiItULv0vPHGG3j77bdx5swZSCQSJCYm4vfff8fs2bMxefJkbWQkItI6P1crzO9Xdnmdrw/E4PTtdJETEVFVU3uenvfeew8KhQLBwcHIz89Hx44dIZPJMHv2bLz11lvayEhEVC2GtnTD2TsZ2HHhAd7adBF/T+8AB0tekZ2ottB4np7i4mLExsYiNzcXfn5+sLCwqOps1Ybz9BDRY/nFpei/PBy3UnPR1qsONoxvDUMDzt9DpIuqZZ4eAJBKpfDz80OrVq1qdOEhIvo3M6kRVo58EabGhjgVl47vjtwSOxIRVRGNSw8RUW3l7WiJLwf5AwC+P3oLx28+FDkREVUFlh4iogoMfKEeXmvlDkEAZmy+hKSsArEjEdFzYukhInqKT/r6wc/FChl5xXhr40WUyBViRyKi58DSQ0T0FCbGhlgx4kVYyoxw/u4jfH0gRuxIRPQc1D5lHQBu3bqFY8eOITU1FQqF6r98Pv744yoJRkSkC+rbm+OrVwIx+fcLWH38NlrUt8PLfk5ixyIiDah9yvpPP/2EyZMnw97eHs7OzpBI/n8qp0QiqZHX3+Ip60T0LPP2XMOa8DuwMjHCX9M7wM3OTOxIRHpP3e9vtUuPh4cHpkyZgrlz52ocUtew9BDRsxSXKjBkdQQu3ctEYD1rbJ0UBJmRodixiPSa1ufpefToEV599VWNwhER1VRSIwP8MOJF2JgZ48r9LHz5V7TYkYhITWqXnldffRUHDx7URhYiIp1W18YUS4Y0BQCsi7iLvVcSRU5EROpQeyCzt7c3PvroI5w+fRoBAQEwNjZWuX/69OlVFo6ISNe85OuEyZ29sDI0Du9tj4KfixUaOHBWeqKaQO0xPZ6enk9/MokEt2/ffu5Q1Y1jeohIHaVyBYb/fAZn4zPg62yJXVPbwcSY43uIqpvWBzLXRiw9RKSulOxC9PruBNLzijGhvSc+7OMndiQivVNtFxwlItJnTlYmWDQ4EADw88l4nIpNEzkRET2LRpMT3r9/H3/++ScSEhJQXFysct+SJUuqJBgRka7r6ueE11q5Y9PZBMzaehn73+4IazPjZz+QiEShduk5cuQI+vXrhwYNGuDGjRvw9/fHnTt3IAgCXnzxRW1kJCLSWR/2boyIuDTcSc/HR7uvYtlrL4gdiYieQu3DWyEhIZg9ezaioqJgYmKC7du34969e+jUqRPn7yEivWMuM8K3Q5vB0ECCPy8nYvelB2JHIqKnULv0REdHY/To0QAAIyMjFBQUwMLCAvPnz8eiRYuqPCARka57wd0Wb73kDQD4cNdVPMgsEDkREVVE7dJjbm6uHMfj4uKCuLg45X1paRzIR0T6aVoXbzRzs0FOYSlmb7kMhULvT4wl0jlql542bdrg5MmTAIBevXph1qxZ+OKLLzBu3Di0adOmygMSEdUERoYG+HZoM5gaGyLidjp+ORkvdiQieoLapWfJkiVo3bo1AGDevHkIDg7G5s2bUb9+ffzyyy9VHpCIqKbwtDfHR//M17P4QAyik7JFTkRE/8bJCcHJCYmo6giCgDfWn8fh6FTO1kykZTVqcsIFCxagZcuWsLS0hKOjIwYMGICYmBiVdTp37gyJRKJymzRpkso6CQkJ6N27N8zMzODo6Ig5c+agtLS0Ot8KERGAssvxLBwcCHsLKW4k5+DrAzHPfhARVQtRS09YWBimTp2K06dP49ChQygpKUG3bt2Ql5enst4bb7yBpKQk5e2rr75S3ieXy9G7d28UFxfj1KlTWLduHdauXYuPP/64ut8OEREAwN5CxtmaiXSQTh3eevjwIRwdHREWFoaOHTsCKNvT06xZMyxdurTCx+zbtw99+vRBYmIinJycAACrVq3C3Llz8fDhQ0il0me+Lg9vEZE2hOyIwqazCXCxNuFszURaoJXDW9nZ1TMYLysrCwBgZ2ensvz333+Hvb09/P39ERISgvz8fOV9ERERCAgIUBYeAOjevTuys7Nx7dq1Cl+nqKgI2dnZKjcioqr2UZ/G8LQ3R1JWIT7afVXsOER6r1Klx9bWFqmpqQCAl156CZmZmVUeRKFQYMaMGWjXrh38/f2Vy4cPH47ffvsNx44dQ0hICDZs2ICRI0cq709OTlYpPACUvycnJ1f4WgsWLIC1tbXy5ubmVuXvh4jITMrZmol0SaWuvWVhYYH09HQ4OjoiNDQUJSUlVR5k6tSpuHr1qnIOoMcmTpyo/DkgIAAuLi4IDg5GXFwcvLy8NHqtkJAQvPPOO8rfs7OzWXyISCuaudngrZe8sfTwLXy46ypa1LdDXRtTsWMR6aVKlZ6uXbuiS5cuaNy4MQBg4MCBTx0rc/ToUbVDTJs2DXv37sXx48dRr169/1z38RxBsbGx8PLygrOzM86ePauyTkpKCgDA2dm5wueQyWSQyWRq5yQi0sS0Lt4IjXmIS/cyMXvLZfw+oTUMDCRixyLSO5UqPb/99hvWrVuHuLg4hIWFoUmTJjAzM3vuFxcEAW+99RZ27tyJ0NBQeHp6PvMxly5dAlB2CQwACAoKwhdffIHU1FQ4OjoCAA4dOgQrKyv4+fk9d0Yiouf1eLbm3stOKGdrfqNjA7FjEekdtc/e6tKlC3bu3AkbG5vnfvEpU6Zg48aN2L17N3x8fJTLra2tYWpqiri4OGzcuBG9evVCnTp1cOXKFcycORP16tVDWFgYgLJT1ps1awZXV1d89dVXSE5OxqhRozBhwgR8+eWXlcrBs7eIqDpsOpuAkB1RkBoaYPe0dmjswv/fED0Pdb+/n+uU9ccPlUg02037tMetWbMGY8eOxb179zBy5EhcvXoVeXl5cHNzw8CBA/Hhhx+qvLm7d+9i8uTJCA0Nhbm5OcaMGYOFCxfCyKhSO7JYeoioWpTN1hyJw9EpnK2ZqApUS+lZv349Fi9ejFu3bgEAGjVqhDlz5mDUqFHqJ9YBLD1EVF3ScovQY+lxpOUWY0J7T3zYh4fhiTSl9ctQLFmyBJMnT0avXr2wZcsWbNmyBT169MCkSZPw7bffahSaiEhfcLZmIvGovafH09MT8+bNw+jRo1WWr1u3Dp9++ini4+OrNGB14J4eIqpu7++MwsYzCahna4oDMzrCXFa5w/FE9H9a39OTlJSEtm3bllvetm1bJCUlqft0RER66YNejVHXxhT3HxXgq/03xI5DpBfULj3e3t7YsmVLueWbN29Gw4YNqyQUEVFtZy4zUh7mWhdxF2dup4uciKj2U3t/6rx58zB06FAcP34c7dq1AwCEh4fjyJEjFZYhIiKqWPuG9nitlRs2nb2HuduvYN/bHWEq5dlcRNqi9p6ewYMH48yZM7C3t8euXbuwa9cu2Nvb4+zZsxg4cKA2MhIR1VohvRrDxdoEd9Lz8c3BGLHjENVqzzVPT23BgcxEJKZjN1Lx+tpzkEiAbZPaormHrdiRiGoErQ9kJiKiqtXF1xGDXqwLQQDe3XYZhSVysSMR1UosPUREOuDjPn5wsJQh7mEevjtyS+w4RLUSSw8RkQ6wMZPiiwH+AIAfj9/GlfuZ4gYiqoVYeoiIdES3Js7o29QVcoWAd7ddQXGpQuxIRLXKc5ee7Oxs7Nq1C9HR0VWRh4hIr83r1wR1zKW4kZyD5cdixY5DVKuoXXqGDBmC5cuXAwAKCgrQokULDBkyBIGBgdi+fXuVByQi0id25lLM698EALDiWCyuJ2aLnIio9lC79Bw/fhwdOnQAAOzcuROCICAzMxPLli3D559/XuUBiYj0Te8AF/Ro4oxShYA52y6jRM7DXERVQe3Sk5WVBTs7OwDA/v37MXjwYJiZmaF37964dYtnHBARPS+JRIL5A5rAxswY1xKzsTosTuxIRLWC2qXHzc0NERERyMvLw/79+9GtWzcAwKNHj2BiYlLlAYmI9JGjpQk+6esHAFh2JBY3U3JETkRU86ldembMmIERI0agXr16cHV1RefOnQGUHfYKCAio6nxERHprQLO6CPZ1RLFcgTnbrqCUh7mInovapWfKlCmIiIjAr7/+ipMnT8LAoOwpGjRowDE9RERVSCKR4IuBAbA0McLle5n4NTxe7EhENdpzX3tLLpcjKioKHh4esLWtmdeL4bW3iEiXbTl3D+9uvwKZkQH2vd0BDRwsxI5EpBO0fu2tGTNm4JdffgFQVng6deqEF198EW5ubggNDVU7MBER/bdXW9RDh4b2KCpV4N1tVyBX6P11ook0onbp2bZtG5o2bQoA2LNnD+Lj43Hjxg3MnDkTH3zwQZUHJCLSdxKJBAsHB8Jcaojzdx9hfcQdsSMR1Uhql560tDQ4OzsDAP7++2+8+uqraNSoEcaNG4eoqKgqD0hEREBdG1OE9GoMAPhqfwzupueJnIio5lG79Dg5OeH69euQy+XYv38/Xn75ZQBAfn4+DA0NqzwgERGVGd7KHW0a2KGgRI65269AwcNcRGpRu/S8/vrrGDJkCPz9/SGRSNC1a1cAwJkzZ+Dr61vlAYmIqIyBgQRfDW4KU2NDnL6dgY1nE8SORFSjqF16Pv30U/z888+YOHEiwsPDIZPJAACGhoZ47733qjwgERH9n3sdM7zbwwcAsODvaDzILBA5EVHN8dynrNcGPGWdiGoShULAq6sjEHn3Ebr4OODXsS0hkUjEjkVU7bR+yjoAhIWFoW/fvvD29oa3tzf69euHEydOaPJURESkJgMDCRYNDoDU0ADHYh7iz8uJYkciqhHULj2//fYbunbtCjMzM0yfPh3Tp0+HqakpgoODsXHjRm1kJCKiJ3g7WmLaS94AgHl7riMjr1jkRES6T+3DW40bN8bEiRMxc+ZMleVLlizBTz/9hOjo6CoNWB14eIuIaqLiUgX6fn8SMSk5GPRiXSwZ0kzsSETVSuuHt27fvo2+ffuWW96vXz/Ex/O6MERE1UVqZICFgwMgkQA7LjxA2M2HYkci0mlqlx43NzccOXKk3PLDhw/Dzc2tSkIREVHlvOBui7Ft6wMA3t8RhbyiUnEDEekwI3UfMGvWLEyfPh2XLl1C27ZtAQDh4eFYu3YtvvvuuyoPSERE/212Nx8cvJaCB5kF+ObgTXzc10/sSEQ6SaNT1nfu3IlvvvlGOX6ncePGmDNnDvr371/lAasDx/QQUU0XGpOKsWvOlR3qmtwWL7jbih2JSOvU/f7mPD1g6SGi2mHm5kvYefEBfJ0t8ee09pAaaTQrCVGNUS3z9BARke75qI8f7MyluJGcg9VhcWLHIdI5lRrTY2trW+nZPjMyMp4rEBERacbOXIpP+vrh7T8u4fujsegZ4AJvRwuxYxHpjEqVnqVLl2o5BhERVYV+TV2x8+IDhMY8RMiOK9g8MQgGBrxEBRHAMT0AOKaHiGqXB5kFeHlJGPKL5fhsgD9GtfEQOxKRVnBMDxGRnqtrY4p3u5ddiX3RvhtIyuKV2IkAlh4iolppVFB9vOBug9yiUny06yq4U5+IpYeIqFYyNJBg0eBAGBtKcDg6FX9FJYkdiUh0LD1ERLVUIydLTOlcdiX2T/+8hsx8Xomd9BtLDxFRLTalixcaOlogLbcYn/8VLXYcIlGpfe2tgQMHVjhnj0QigYmJCby9vTF8+HD4+PhUSUAiItKczMgQCwcH4pVVp7At8j4GNKuL9g3txY5FJAq19/RYW1vj6NGjuHDhAiQSCSQSCS5evIijR4+itLQUmzdvRtOmTREeHq6NvEREpKbmHrYY/c9p6yE7ryC/mFdiJ/2kdulxdnbG8OHDcfv2bWzfvh3bt29HXFwcRo4cCS8vL0RHR2PMmDGYO3euNvISEZEG5vTwhau1Ce5lFODbQzfFjkMkCrUnJ3RwcEB4eDgaNWqksvzmzZto27Yt0tLSEBUVhQ4dOiAzM7Mqs2oNJyckIn1w7EYqXl97DgYSYNfUdgisZyN2JKLnovXJCUtLS3Hjxo1yy2/cuAG5XA4AMDExqfS1uoiIqHp08XVE/2auUAjAu9uuoESuEDsSUbVSu/SMGjUK48ePx7fffouTJ0/i5MmT+PbbbzF+/HiMHj0aABAWFoYmTZpUeVgiIno+H/fxg62ZMa/ETnpJ7cNbcrkcCxcuxPLly5GSkgIAcHJywltvvYW5c+fC0NAQCQkJMDAwQL169bQSuqrx8BYR6ZOdF+9j5ubLMDaU4M9p7dHYhf/fo5pJ3e/v57rgaHZ2NgDU+KLA0kNE+kQQBEzcEIlD11PQ2MUKu6e2g9SI07ZRzVOtFxy1srJiSSAiqmEkEgm+HBgAWzNjRCdlY9mRW2JHIqoWapeelJQUjBo1Cq6urjAyMoKhoaHKjYiIdJ+DpQxfDAwAAKwIjcXFhEciJyLSPrVnZB47diwSEhLw0UcfwcXFhWdpERHVUL0CXNC/mSt2X0rErK2X8ff0DjAx5j9eqfZSu/ScPHkSJ06cQLNmzbQQh4iIqtO8fk0QEZeO2w/zsPhADD7q4yd2JCKtUfvwlpubG55j7DMREekQGzMpFg0OBAD8Gh6P07fTRU5EpD1ql56lS5fivffew507d7QQh4iIqlsXX0cMa+kGQQBmb72M3CJem4tqJ7VPWbe1tUV+fj5KS0thZmYGY2NjlfszMjKqNGB14CnrRKTvcgpL0GPpCTzILMBrrdyxYFCA2JGInknd72+1x/QsXbpUk1xERKTDLE2M8fWrTfHaT6ex6WwCujdxQmcfR7FjEVWp55qcsLbgnh4iojLz9lzDmvA7cLKS4eCMTrA2M372g4hEopXJCR/PvPz45/+6ERFRzfVud180sDdHSnYRPvnzqthxiKpUpUqPra0tUlNTAQA2NjawtbUtd3u8nIiIai5TqSG+HtIUBhJg16VE7ItKEjsSUZWp1Jieo0ePws7ODgBw7NgxrQYiIiJxvehui0mdvLAiNA4f7LqKlp52sLeQiR2L6LlxTA84poeI6ElFpXL0Xx6OG8k56ObnhNWjmnMGftI5Wj97CwAyMzNx9uxZpKamQqFQqNw3evRoTZ6SiIh0iMzIEEuGNEP/H07i4PUU7Lz4AINerCd2LKLnovbkhHv27IG7uzt69OiBadOm4e2331beZsyYodZzLViwAC1btoSlpSUcHR0xYMAAxMTEqKxTWFiIqVOnok6dOrCwsMDgwYORkpKisk5CQgJ69+4NMzMzODo6Ys6cOSgt5eRaRETPw8/VCm8HNwQAfPLnNSRlFYiciOj5qF16Zs2ahXHjxiE3NxeZmZl49OiR8qbuxIRhYWGYOnUqTp8+jUOHDqGkpATdunVDXl6ecp2ZM2diz5492Lp1K8LCwpCYmIhBgwYp75fL5ejduzeKi4tx6tQprFu3DmvXrsXHH3+s7lsjIqInTOrkhab1rJFTWIp3t13hZYioRlN7TI+5uTmioqLQoEGDKg/z8OFDODo6IiwsDB07dkRWVhYcHBywceNGvPLKKwCAGzduoHHjxoiIiECbNm2wb98+9OnTB4mJiXBycgIArFq1CnPnzsXDhw8hlUqf+boc00NE9HSxqbnovewEikoV+HyAP0a28RA7EhEALc3T82/du3fH+fPnNQr3LFlZWQCgPFMsMjISJSUl6Nq1q3IdX19fuLu7IyIiAgAQERGBgIAAZeF5nDE7OxvXrl2r8HWKioo4vxARUSV5O1rg3R6+AIAv/47G3fS8ZzyCSDepPZC5d+/emDNnDq5fv46AgIBy197q16+fRkEUCgVmzJiBdu3awd/fHwCQnJwMqVQKGxsblXWdnJyQnJysXOffhefx/Y/vq8iCBQswb948jXISEemj19vWx8FryTgTn4E5W69g08Q2MDTg2VxUs6hdet544w0AwPz588vdJ5FIIJfLNQoydepUXL16FSdPntTo8eoICQnBO++8o/w9Ozsbbm5uWn9dIqKaysBAgq9fbYruS4/j7J0MrAmPx4QOVT/MgUib1D68pVAonnrTtPBMmzYNe/fuxbFjx1Cv3v9PiXR2dkZxcTEyMzNV1k9JSYGzs7NynSfP5nr8++N1niSTyWBlZaVyIyKi/+ZmZ4YPe/sBAL46EINriVkiJyJSj9qlpyoJgoBp06Zh586dOHr0KDw9PVXub968OYyNjXHkyBHlspiYGCQkJCAoKAgAEBQUhKioKOVlMgDg0KFDsLKygp+fX/W8ESIiPfFaKze85OuI4lIFJv0Wicz8YrEjEVVapc7eWrZsGSZOnAgTExMsW7bsP9edPn16pV98ypQp2LhxI3bv3g0fHx/lcmtra5iamgIAJk+ejL///htr166FlZUV3nrrLQDAqVOnAJSdst6sWTO4urriq6++QnJyMkaNGoUJEybgyy+/rFQOnr1FRFR5mfnF6Lv8JO5lFKBjIwesGduS43tIFOp+f1eq9Hh6euL8+fOoU6dOub0xKk8mkeD27duVDvu0Kc3XrFmDsWPHAiibnHDWrFnYtGkTioqK0L17d6xYsULl0NXdu3cxefJkhIaGwtzcHGPGjMHChQthZFS5IUssPURE6rmemI1BK8NRWKLAtC7emN3d59kPIqpiWik9tR1LDxGR+nZdfIAZmy8BAFaPao7uTSoeR0mkLVqfp4eIiAgABrxQF6+3qw8AmLXlMuIe5oobiOgZNLrg6P379/Hnn38iISEBxcWqg9iWLFlSJcGIiEj3vd+rMa49yMbZOxl4c0Mkdk1tBwuZRl8tRFqn9ifzyJEj6NevHxo0aIAbN27A398fd+7cgSAIePHFF7WRkYiIdJSxoQGWj3gBfb8/idjUXMzechkrR7741DGbRGJS+/BWSEgIZs+ejaioKJiYmGD79u24d+8eOnXqhFdffVUbGYmISIc5WppgxYjmMDaUYP+1ZKwKq/wJLUTVSe3SEx0djdGjRwMAjIyMUFBQAAsLC8yfPx+LFi2q8oBERKT7mnvY4tN+TQAAiw/cwIlbD0VORFSe2qXH3NxcOY7HxcUFcXFxyvvS0tKqLhkREdUow1u5Y0iLelAIwFubLuJeRr7YkYhUqF162rRpo7w+Vq9evTBr1ix88cUXGDduHNq0aVPlAYmIqGaQSCSY398fgfWskZlfgkm/RaKwRLPLExFpg9qlZ8mSJWjdujUAYN68eQgODsbmzZtRv359/PLLL1UekIiIag4TY0OsHNkcduZSXEvMxvs7o8Dp4EhXqDU5oVwuR3h4OAIDA2FjY6PFWNWLkxMSEVWtU7FpGPnLGSgE4LP+TTAqqL7YkagW0urkhIaGhujWrRsePXqkcUAiIqr92nrb472evgCAeXuuI/JuhsiJiDQ4vOXv76/W9bWIiEg/vdGhAXoHuKBUIWDSbxeQml0odiTSc2qXns8//xyzZ8/G3r17kZSUhOzsbJUbERERUDaw+atXAtHQ0QIPc4ow5fcLKC5ViB2L9Filx/TMnz8fs2bNgqWl5f8f/K8ZNwVBgEQigVxe80bqc0wPEZH23H6Yi/7Lw5FTVIoxQR6Y199f7EhUS2jtKuuGhoZISkpCdHT0f67XqVOnyiXVISw9RETadfh6CiasPw8A+ObVphjcvJ7Iiag2UPf7u9LX3nrcjWpiqSEiInF19XPC9OCGWHbkFt7fGQUfZ0v417UWOxbpGbXG9PACckREpKkZwQ3RxccBRaUKjFt7DnfS8sSORHqm0oe3DAwMYG1t/czik5FR805L5OEtIqLqkZVfgqE/RuBGcg7q2phi66QguNqYih2LaiitHd4CymZgtrbm7kgiItKMtZkxNoxvjSGrIxCflocRP5/B5jfbwNHSROxopAfU2tOTnJwMR0dHbWeqdtzTQ0RUvRIzC/Dqqgg8yCyAj5Ml/pjYBrbmUrFjUQ2jtRmZOZ6HiIiqiquNKX6f0BqOljLEpORgzJqzyCksETsW1XKVLj28YBwREVWl+vbm+H1Ca9iaGePK/SyMX3seBcU1b643qjkqXXoUCkWtPLRFRETiaehkiQ3jW8NSZoSzdzIwccN5FJWy+JB2qH0ZCiIioqrkX9caa8e1hKmxIU7cSsNbGy+iRM7LVVDVY+khIiLRNfeww89jWkBqZICD11Mwe+tlyBUcVkFVi6WHiIh0Qjtve6wc8SKMDCTYfSkRH+6K4nhSqlIsPUREpDOCGzth6bBmMJAAm87ew2d7o1l8qMqw9BARkU7pE+iKhYMDAQC/hsfj20M3RU5EtQVLDxER6ZwhLdwwr18TAMCyo7FYFRYnciKqDVh6iIhIJ41pWx9ze/gCABbuu4ENEXfEDUQ1HksPERHprMmdvTCtizcA4KPd17At8r7IiagmY+khIiKdNqtbI7zerj4A4N1tl/F3VJK4gajGYukhIiKdJpFI8HEfPwxr6QaFAEzfdBG7Lz0QOxbVQCw9RESk8yQSCb4YGIABzVxRqhDw9h+X8MOxWJ7OTmph6SEiohrB0ECCJUOaYWLHBgCAxQdi8P7OKJTykhVUSSw9RERUYxgYSPB+r8aY16+JcgLDCevPI7eoVOxoVAOw9BARUY0zpm19rB7VAibGBgiNeYihqyOQkl0odizScSw9RERUI73s54Q/JgbB3kKKa4nZGPhDOG6m5Igdi3QYSw8REdVYzdxssGNyOzRwMEdiViEGrzyFU7FpYsciHcXSQ0RENZp7HTPsmNwWLevbIqewFGPWnMXOi5zEkMpj6SEiohrPxkyKDeNbo0+gC0rkAmZuvozvj9ziKe2kgqWHiIhqBRNjQywb9gLe7FR2Svs3h27ive1RKOEp7fQPlh4iIqo1DAwkCOnZGJ8N8IeBBNh8/h7GrzuPnMISsaORDmDpISKiWmdUGw/8NLoFTI0NcfzmQwxZfRrJWTylXd+x9BARUa0U3NgJm99sA3sLGaKTsjFwRThuJGeLHYtExNJDRES1VmA9G+yc0hZeDuZIyirEqysjcPzmQ7FjkUhYeoiIqFZzszPDjsnt0MrTDjlFZae0f3f4FuQKntmlb1h6iIio1rM2M8aG8a0wtIUbBAH49vBNjPn1LNJyi8SORtWIpYeIiPSCzMgQi14JxJIhTWFqbIiTsWno9d0JnL6dLnY0qiYsPUREpFcGvVgPf05rh4aOFkjNKcLwn07jh2OxUPBwV63H0kNERHqnoZMldk9rh8Ev1oNCABYfiMHYteeQzsNdtRpLDxER6SUzqRG+GdIUX70SCBNjAxy/+RC9l53EuTsZYkcjLWHpISIivTakhRt2T20PLwdzJGcXYtiPp7EyNI6Hu2ohlh4iItJ7Ps6W+HNaewxo5gq5QsCi/TcwYf15PMorFjsaVSGWHiIiIgDmMiN8O7QZFgwKgNTIAEdvpKL3shOIvPtI7GhURVh6iIiI/iGRSPBaK3fsmtIOnvbmSMwqxNDVEfjp+G0IAg931XQsPURERE/wc7XCn9PaoU+gC0oVAr74OxpvrI9EVj6v1l6TsfQQERFVwNLEGN+/9gI+G+APqaEBDkenoNeyEzgbz7O7aiqWHiIioqeQSCQY1cYDO6a0hbudGR5kFmDojxGYv+c6CorlYscjNbH0EBERPYN/XWv8Nb09hrUsu3bXr+Hx6LXsBM5zTp8ahaWHiIioEixNjLFwcCDWvN4SzlYmiE/Lw6urI/D53usoLOFen5qApYeIiEgNXXwccWBmR7zSvB4EAfj5ZDx6fXcCFxJ4aruuY+khIiJSk7WpMb5+tSl+HdsCjpYy3E7LwysrT2HBvmju9dFhLD1EREQaesnXCYdmdsKgF+pCIQCrw26jz/cnceleptjRqAIsPURERM/B2swYS4Y2w0+jW8DBUobY1FwMWhGORftvoKiUe310CUsPERFRFXjZzwkHZ3RE/2auUAjAytA49P3+JK7czxQ7Gv1D1NJz/Phx9O3bF66urpBIJNi1a5fK/WPHjoVEIlG59ejRQ2WdjIwMjBgxAlZWVrCxscH48eORm5tbje+CiIiojK25FN8NewGrRjaHvYUUN1NyMXDFKXx9IIZ7fXSAqKUnLy8PTZs2xQ8//PDUdXr06IGkpCTlbdOmTSr3jxgxAteuXcOhQ4ewd+9eHD9+HBMnTtR2dCIioqfq4e+MgzM7oW/Tsqu2Lz8Wi/7Lw3H1QZbY0fSaRNCRK6hJJBLs3LkTAwYMUC4bO3YsMjMzy+0Beiw6Ohp+fn44d+4cWrRoAQDYv38/evXqhfv378PV1bVSr52dnQ1ra2tkZWXBysrqed8KERGR0t9RSfhw11Vk5BXDQAKMbeuJd7o1goXMSOxoNZ663986P6YnNDQUjo6O8PHxweTJk5Genq68LyIiAjY2NsrCAwBdu3aFgYEBzpw589TnLCoqQnZ2tsqNiIhIG3oFuODQzI7oE+gCxT+zOQd/E4q/o5J45fZqptOlp0ePHli/fj2OHDmCRYsWISwsDD179oRcXnZcNDk5GY6OjiqPMTIygp2dHZKTk5/6vAsWLIC1tbXy5ubmptX3QURE+q2OhQzLh7+I9eNawaOOGVKyizDl9wsYu+Yc7qbniR1Pb+h06Rk2bBj69euHgIAADBgwAHv37sW5c+cQGhr6XM8bEhKCrKws5e3evXtVE5iIiOg/dGzkgAMzOmJ6cENIDQ0QdvMhun17HN8fucWBztVAp0vPkxo0aAB7e3vExsYCAJydnZGamqqyTmlpKTIyMuDs7PzU55HJZLCyslK5ERERVQcTY0O883Ij7J/RAe297VFUqsA3h26i53cncCo2Tex4tVqNKj33799Heno6XFxcAABBQUHIzMxEZGSkcp2jR49CoVCgdevWYsUkIiJ6pgYOFtgwvhW+G9YM9hYy3H6Yh+E/n8GMPy7iYU6R2PFqJVHP3srNzVXutXnhhRewZMkSdOnSBXZ2drCzs8O8efMwePBgODs7Iy4uDu+++y5ycnIQFRUFmUwGAOjZsydSUlKwatUqlJSU4PXXX0eLFi2wcePGSufg2VtERCSmrIISfHMwBhtO34UgAJYmRni3hy+Gt3KHoYFE7Hg6S93vb1FLT2hoKLp06VJu+ZgxY7By5UoMGDAAFy9eRGZmJlxdXdGtWzd89tlncHJyUq6bkZGBadOmYc+ePTAwMMDgwYOxbNkyWFhYVDoHSw8REemCK/cz8cHOq4j6Zz6fpvWs8cXAAPjXtRY5mW6qUaVHV7D0EBGRrpArBPx2+i6+PhCDnKJSGEiA0UH1MatbI1iaGIsdT6fUunl6iIiI9ImhgQRj2tbHkVllMzorBGDtqTsI/iYMOy7ch0Kh9/sqNMY9PeCeHiIi0l0nbj3ER7uu4k56PoCyQ14f9fFDi/p2IicTHw9vaYClh4iIdFlhiRy/hsfjh6OxyCsum8+nd4AL3uvpCzc7M5HTiYelRwMsPUREVBOk5hRiycGb2Hz+HgQBkBoZYHx7T0zp7KWX431YejTA0kNERDXJ9cRsfP7XdZyKK7sepb2FFLO6+WBICze9OsWdpUcDLD1ERFTTCIKAw9Gp+PLvaMSnlV2/y9fZEh/18UM7b3uR01UPlh4NsPQQEVFNVVyqwIbTd/Hd4ZvILiwFAHRt7Ij3ezVGA4fKz1lXE7H0aIClh4iIarpHecX47sgtbDh9F3KFACMDCUYFeeDt4IawMZOKHU8rWHo0wNJDRES1RWxqLr78OxpHb5RdkNvGzBgzghtiRBsPGBvWrun5WHo0wNJDRES1zYlbD/H53mjEpOQAABrYm2NWNx/09HeGQS0Z7MzSowGWHiIiqo1K5QpsPn8PSw7eRHpeMQDAv64V5nT3RceG9pBIanb5YenRAEsPERHVZjmFJfj5RDx+PnFbOblha087vNvDF809bEVOpzmWHg2w9BARkT5Izy3CitA4bDh9F8WlCgBlZ3rN7u4DX+ea9/3H0qMBlh4iItIniZkF+O7wLWyNvAeFAEgkQP+mrpj5ciN41DEXO16lsfRogKWHiIj0UdzDXCw5eBN/RSUBAIwMJBjWyg3TX2oIRysTkdM9G0uPBlh6iIhIn0Xdz8LigzE4fvMhAMDE2ABj23piUqcGOj3HD0uPBlh6iIiIgNO30/HV/hu4kJAJALA0McKkTl54vV19mEmNxA1XAZYeDbD0EBERlREEAUeiU/H1wRjcSC6b48feQobJnb0wvJU7TKWGIif8P5YeDbD0EBERqZIrBOy5nIglh24iISMfQNnV3Cd2bICRbTx0Ys8PS48GWHqIiIgqVlyqwLbI+1gRGov7jwoAAHbmUkzo4InRQfVhIROv/LD0aIClh4iI6L+VyBXYefEBfjgWi7vpZXt+bMyMMaG9J0a3rQ8rE+Nqz8TSowGWHiIiosoplSvw5+VELD8ai9tpeQAAKxMjjGvvidfbesLarPrKD0uPBlh6iIiI1CNXCNh7JRHfH41FbGouAMBSZoSx7epjXDtP2Jpr/1R3lh4NsPQQERFpRq4QsO9qEr4/Equ8oru51BBj2tbHhA4NYKfF8sPSowGWHiIiouejUAg4eD0Z3x2JRXRSNgDATGqIUW08MKFDAzhYyqr8NVl6NMDSQ0REVDUEQcDh6FQsO3ILUQ+yAJTN8Pzz6JZo39C+Sl9L3e9v8U+yJyIiolpDIpHgZT8ndG3siNCYh1h65BbupOWhqZu12NFYeoiIiKjqSSQSdPF1RGcfB9x/VABLEU5pf5KB2AGIiIio9pJIJHCzMxM7BgCWHiIiItITLD1ERESkF1h6iIiISC+w9BAREZFeYOkhIiIivcDSQ0RERHqBpYeIiIj0AksPERER6QWWHiIiItILLD1ERESkF1h6iIiISC+w9BAREZFeYOkhIiIivWAkdgBdIAgCACA7O1vkJERERFRZj7+3H3+PPwtLD4CcnBwAgJubm8hJiIiISF05OTmwtrZ+5noSobL1qBZTKBRITEyEpaUlJBJJlT1vdnY23NzccO/ePVhZWVXZ89Z23G6a4XZTH7eZZrjdNMPtppn/2m6CICAnJweurq4wMHj2iB3u6QFgYGCAevXqae35rays+AHXALebZrjd1MdtphluN81wu2nmadutMnt4HuNAZiIiItILLD1ERESkF1h6tEgmk+GTTz6BTCYTO0qNwu2mGW439XGbaYbbTTPcbpqpyu3GgcxERESkF7inh4iIiPQCSw8RERHpBZYeIiIi0gssPURERKQXWHq06IcffkD9+vVhYmKC1q1b4+zZs2JH0mmffvopJBKJys3X11fsWDrl+PHj6Nu3L1xdXSGRSLBr1y6V+wVBwMcffwwXFxeYmpqia9euuHXrljhhdcizttvYsWPLffZ69OghTlgdsWDBArRs2RKWlpZwdHTEgAEDEBMTo7JOYWEhpk6dijp16sDCwgKDBw9GSkqKSIl1Q2W2W+fOnct93iZNmiRSYt2wcuVKBAYGKicgDAoKwr59+5T3V9VnjaVHSzZv3ox33nkHn3zyCS5cuICmTZuie/fuSE1NFTuaTmvSpAmSkpKUt5MnT4odSafk5eWhadOm+OGHHyq8/6uvvsKyZcuwatUqnDlzBubm5ujevTsKCwurOaluedZ2A4AePXqofPY2bdpUjQl1T1hYGKZOnYrTp0/j0KFDKCkpQbdu3ZCXl6dcZ+bMmdizZw+2bt2KsLAwJCYmYtCgQSKmFl9lthsAvPHGGyqft6+++kqkxLqhXr16WLhwISIjI3H+/Hm89NJL6N+/P65duwagCj9rAmlFq1athKlTpyp/l8vlgqurq7BgwQIRU+m2Tz75RGjatKnYMWoMAMLOnTuVvysUCsHZ2VlYvHixcllmZqYgk8mETZs2iZBQNz253QRBEMaMGSP0799flDw1RWpqqgBACAsLEwSh7LNlbGwsbN26VblOdHS0AECIiIgQK6bOeXK7CYIgdOrUSXj77bfFC1VD2NraCj///HOVfta4p0cLiouLERkZia5duyqXGRgYoGvXroiIiBAxme67desWXF1d0aBBA4wYMQIJCQliR6ox4uPjkZycrPK5s7a2RuvWrfm5q4TQ0FA4OjrCx8cHkydPRnp6utiRdEpWVhYAwM7ODgAQGRmJkpISlc+br68v3N3d+Xn7lye322O///477O3t4e/vj5CQEOTn54sRTyfJ5XL88ccfyMvLQ1BQUJV+1njBUS1IS0uDXC6Hk5OTynInJyfcuHFDpFS6r3Xr1li7di18fHyQlJSEefPmoUOHDrh69SosLS3FjqfzkpOTAaDCz93j+6hiPXr0wKBBg+Dp6Ym4uDi8//776NmzJyIiImBoaCh2PNEpFArMmDED7dq1g7+/P4Cyz5tUKoWNjY3Kuvy8/V9F2w0Ahg8fDg8PD7i6uuLKlSuYO3cuYmJisGPHDhHTii8qKgpBQUEoLCyEhYUFdu7cCT8/P1y6dKnKPmssPaQzevbsqfw5MDAQrVu3hoeHB7Zs2YLx48eLmIxqu2HDhil/DggIQGBgILy8vBAaGorg4GARk+mGqVOn4urVqxxjp6anbbeJEycqfw4ICICLiwuCg4MRFxcHLy+v6o6pM3x8fHDp0iVkZWVh27ZtGDNmDMLCwqr0NXh4Swvs7e1haGhYbmR5SkoKnJ2dRUpV89jY2KBRo0aIjY0VO0qN8Pizxc/d82vQoAHs7e352QMwbdo07N27F8eOHUO9evWUy52dnVFcXIzMzEyV9fl5K/O07VaR1q1bA4Def96kUim8vb3RvHlzLFiwAE2bNsV3331XpZ81lh4tkEqlaN68OY4cOaJcplAocOTIEQQFBYmYrGbJzc1FXFwcXFxcxI5SI3h6esLZ2Vnlc5ednY0zZ87wc6em+/fvIz09Xa8/e4IgYNq0adi5cyeOHj0KT09PlfubN28OY2Njlc9bTEwMEhIS9Prz9qztVpFLly4BgF5/3iqiUChQVFRUtZ+1qh1rTY/98ccfgkwmE9auXStcv35dmDhxomBjYyMkJyeLHU1nzZo1SwgNDRXi4+OF8PBwoWvXroK9vb2QmpoqdjSdkZOTI1y8eFG4ePGiAEBYsmSJcPHiReHu3buCIAjCwoULBRsbG2H37t3ClStXhP79+wuenp5CQUGByMnF9V/bLScnR5g9e7YQEREhxMfHC4cPHxZefPFFoWHDhkJhYaHY0UUzefJkwdraWggNDRWSkpKUt/z8fOU6kyZNEtzd3YWjR48K58+fF4KCgoSgoCARU4vvWdstNjZWmD9/vnD+/HkhPj5e2L17t9CgQQOhY8eOIicX13vvvSeEhYUJ8fHxwpUrV4T33ntPkEgkwsGDBwVBqLrPGkuPFn3//feCu7u7IJVKhVatWgmnT58WO5JOGzp0qODi4iJIpVKhbt26wtChQ4XY2FixY+mUY8eOCQDK3caMGSMIQtlp6x999JHg5OQkyGQyITg4WIiJiRE3tA74r+2Wn58vdOvWTXBwcBCMjY0FDw8P4Y033tD7f6BUtL0ACGvWrFGuU1BQIEyZMkWwtbUVzMzMhIEDBwpJSUnihdYBz9puCQkJQseOHQU7OztBJpMJ3t7ewpw5c4SsrCxxg4ts3LhxgoeHhyCVSgUHBwchODhYWXgEoeo+axJBEAQN9zwRERER1Rgc00NERER6gaWHiIiI9AJLDxEREekFlh4iIiLSCyw9REREpBdYeoiIiEgvsPQQERGRXmDpISICEBoaColEUu76PkRUe7D0EBERkV5g6SEiIiK9wNJDRDpBoVBgwYIF8PT0hKmpKZo2bYpt27YB+P+hp7/++guBgYEwMTFBmzZtcPXqVZXn2L59O5o0aQKZTIb69evjm2++Ubm/qKgIc+fOhZubG2QyGby9vfHLL7+orBMZGYkWLVrAzMwMbdu2RUxMjPK+y5cvo0uXLrC0tISVlRWaN2+O8+fPa2mLEFFVY+khIp2wYMECrF+/HqtWrcK1a9cwc+ZMjBw5EmFhYcp15syZg2+++Qbnzp2Dg4MD+vbti5KSEgBlZWXIkCEYNmwYoqKi8Omnn+Kjjz7C2rVrlY8fPXo0Nm3ahGXLliE6OhqrV6+GhYWFSo4PPvgA33zzDc6fPw8jIyOMGzdOed+IESNQr149nDt3DpGRkXjvvfdgbGys3Q1DRFWn6q6RSkSkmcLCQsHMzEw4deqUyvLx48cLr732mvIq6X/88YfyvvT0dMHU1FTYvHmzIAiCMHz4cOHll19WefycOXMEPz8/QRAEISYmRgAgHDp0qMIMj1/j8OHDymV//fWXAEAoKCgQBEEQLC0thbVr1z7/GyYiUXBPDxGJLjY2Fvn5+Xj55ZdhYWGhvK1fvx5xcXHK9YKCgpQ/29nZwcfHB9HR0QCA6OhotGvXTuV527Vrh1u3bkEul+PSpUswNDREp06d/jNLYGCg8mcXFxcAQGpqKgDgnXfewYQJE9C1a1csXLhQJRsR6T6WHiISXW5uLgDgr7/+wqVLl5S369evK8f1PC9TU9NKrffvw1USiQRA2XgjAPj0009x7do19O7dG0ePHoWfnx927txZJfmISPtYeohIdH5+fpDJZEhISIC3t7fKzc3NTbne6dOnlT8/evQIN2/eROPGjQEAjRs3Rnh4uMrzhoeHo1GjRjA0NERAQAAUCoXKGCFNNGrUCDNnzsTBgwcxaNAgrFmz5rmej4iqj5HYAYiILC0tMXv2bMycORMKhQLt27dHVlYWwsPDYWVlBQ8PDwDA/PnzUadOHTg5OeGDDz6Avb09BgwYAACYNWsWWrZsic8++wxDhw5FREQEli9fjhUrVgAA6tevjzFjxmDcuHFYtmwZmjZtirt37yI1NRVDhgx5ZsaCggLMmTMHr7zyCjw9PXH//n2cO3cOgwcP1tp2IaIqJvagIiIiQRAEhUIhLF26VPDx8RGMjY0FBwcHoXv37kJYWJhykPGePXuEJk2aCFKpVGjVqpVw+fJllefYtm2b4OfnJxgbGwvu7u7C4sWLVe4vKCgQZs6cKbi4uAhSqVTw9vYWfv31V0EQ/j+Q+dGjR8r1L168KAAQ4uPjhaKiImHYsGGCm5ubIJVKBVdXV2HatGnKQc5EpPskgiAIIvcuIqL/FBoaii5duuDRo0ewsbEROw4R1VAc00NERER6gaWHiIiI9AIPbxEREZFe4J4eIiIi0gssPURERKQXWHqIiIhIL7D0EBERkV5g6SEiIiK9wNJDREREeoGlh4iIiPQCSw8RERHpBZYeIiIi0gv/A89t8eZOBIF4AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(loss_epochs)\n",
    "plt.title('Training loss versus epoch')\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('Training loss of an epoch')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "358be4ab",
   "metadata": {},
   "source": [
    "As we see above, during training phase, the loss value is reduced after each epoch.\n",
    "<br>Let's evaluate the trained model with unseen dataset, called **test set**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "e4fb8ae3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total test loss : 20.583334581926465\n",
      "Accuracy(%): 88.52\n"
     ]
    }
   ],
   "source": [
    "model.eval()\n",
    "test_loss=0.\n",
    "accuracy=0.\n",
    "#Nbatches=len(test_dataloader)\n",
    "NtestSamples=len(test_dataloader.dataset)\n",
    "with torch.no_grad():\n",
    "    for x,y in test_dataloader:\n",
    "        x,y=x.to(device),y.to(device)\n",
    "        output=model(x)\n",
    "        predicted=torch.argmax(output,axis=1)\n",
    "        ydesired=torch.argmax(y,axis=1)\n",
    "        accuracy+=(predicted==ydesired).type(torch.float).sum().item()\n",
    "        loss=lossFn(output,y)\n",
    "        test_loss+=loss.item()\n",
    "print(f'Total test loss : {test_loss}')\n",
    "print(f'Accuracy(%): {100*accuracy/NtestSamples}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dacb43cf",
   "metadata": {},
   "source": [
    "Now, we actually give a batch of test images to the model, and observe the predicted labels. \n",
    "<br>We compare the predicated labels with the true labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "7f7bbd9e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy of a test batch (%): 93.75\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1600 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracy_batch=0.\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    x,y=next(iter(test_dataloader))\n",
    "    x,y=x.to(device),y.to(device)\n",
    "    output=model(x)\n",
    "    y=y.argmax(1)\n",
    "    predicted=output.argmax(1)\n",
    "    accuracy_batch+=(y==predicted).type(torch.float).sum().item()\n",
    "    accuracy_batch/=batchSize\n",
    "    print(f'The accuracy of a test batch (%): {100*accuracy_batch}')\n",
    "    #plt.title(f'correct number is: {true_label.item()}, predicted number: {predicted_label}')\n",
    "plt.figure(figsize=(15,16))\n",
    "for i in range(batchSize):\n",
    "    plt.subplot(4,4,i+1)\n",
    "    plt.title(f'true: {y[i].item()}, predicted:{predicted[i].item()}')\n",
    "    plt.imshow(x[i].cpu().numpy()[0],cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f5be9ec",
   "metadata": {},
   "source": [
    "Let's save the model weights or load them. The model parameters are in internal state dictionary. This state dictionary is part of the model object, named **state_dict**.\n",
    "<br> For saving a model, we use **torch.save()**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "6fce6c51",
   "metadata": {},
   "outputs": [],
   "source": [
    "# saving the trained model parameters\n",
    "modelPath='./MNIST/mylenet.pth'\n",
    "torch.save(model.state_dict(),modelPath)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "edbb1b4a",
   "metadata": {},
   "source": [
    "We can load the model parameters from the saved dictionary state with **torch.load()** as shown below:\n",
    " - Hint: For this purpose, we need an instance of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "243b2f6d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_to_load=tnn.Sequential(\n",
    "      tnn.Conv2d(cin,6,kernel_size=5),\n",
    "      tnn.ReLU(),\n",
    "      tnn.MaxPool2d(kernel_size=2,stride=2),\n",
    "      tnn.Conv2d(6,16,kernel_size=5),\n",
    "      tnn.ReLU(),\n",
    "      tnn.MaxPool2d(kernel_size=2,stride=2),\n",
    "      tnn.Flatten(),\n",
    "      tnn.Linear(16*5*5,120), #16*5*5=400\n",
    "      tnn.ReLU(),\n",
    "      tnn.Linear(120,84),\n",
    "      tnn.ReLU(),\n",
    "      tnn.Linear(84,Nclasses)      \n",
    ")\n",
    "model_to_load.load_state_dict(torch.load(modelPath))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f52637cd",
   "metadata": {},
   "source": [
    "Another way to load and save models is to save the entire model, and then load the entire model as shown below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "469b10fc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sequential(\n",
      "  (0): Conv2d(1, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (1): ReLU()\n",
      "  (2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (3): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (4): ReLU()\n",
      "  (5): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (6): Flatten(start_dim=1, end_dim=-1)\n",
      "  (7): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (8): ReLU()\n",
      "  (9): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (10): ReLU()\n",
      "  (11): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "modelPathEntire='./MNIST/mylenetentire.pth'\n",
    "# saving the entire model\n",
    "torch.save(model,modelPathEntire)\n",
    "# loading the entire model\n",
    "model_entire=torch.load(modelPathEntire)\n",
    "# let's print the loaded model\n",
    "print(model_entire)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7ad81cba",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}