{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "47848e98",
   "metadata": {},
   "outputs": [],
   "source": [
    "import mindspore\n",
    "# 载入mindspore的默认数据集\n",
    "import mindspore.dataset as ds\n",
    "# 常用转化用算子\n",
    "import mindspore.dataset.transforms.c_transforms as C\n",
    "# 图像转化用算子\n",
    "####____####\n",
    "import mindspore.dataset.vision.c_transforms as CV\n",
    "from mindspore.common import dtype as mstype\n",
    "# mindspore的tensor\n",
    "from mindspore import Tensor\n",
    "\n",
    "\n",
    "# 各类网络层都在nn里面\n",
    "import mindspore.nn as nn\n",
    "# 参数初始化的方式\n",
    "\n",
    "from mindspore.common.initializer import TruncatedNormal\n",
    "# 设置mindspore运行的环境\n",
    "from mindspore import context\n",
    "# 引入训练时候会使用到回调函数，如checkpoint, lossMoniter\n",
    "from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor\n",
    "# 引入模型\n",
    "from mindspore.train import Model\n",
    "# 引入评估模型的包\n",
    "from mindspore.nn.metrics import Accuracy\n",
    "\n",
    "# numpy\n",
    "import numpy as np\n",
    "# 画图用\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "####____####\n",
    "# 下载数据相关的包\n",
    "import os\n",
    "import requests \n",
    "import zipfile\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "985e397f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2022-03-04 10:42:54--  https://ascend-professional-construction-dataset.obs.cn-north-4.myhuaweicloud.com/ComputerVision/cifar10_mindspore.zip\n",
      "Resolving proxy-notebook.modelarts-dev-proxy.com (proxy-notebook.modelarts-dev-proxy.com)... 192.168.0.172\n",
      "Connecting to proxy-notebook.modelarts-dev-proxy.com (proxy-notebook.modelarts-dev-proxy.com)|192.168.0.172|:8083... connected.\n",
      "Proxy request sent, awaiting response... 200 OK\n",
      "Length: 170441801 (163M) [application/zip]\n",
      "Saving to: ‘cifar10_mindspore.zip’\n",
      "\n",
      "cifar10_mindspore.z 100%[===================>] 162.55M   247MB/s    in 0.7s    \n",
      "\n",
      "2022-03-04 10:42:55 (247 MB/s) - ‘cifar10_mindspore.zip’ saved [170441801/170441801]\n",
      "\n",
      "Archive:  cifar10_mindspore.zip\n",
      "   creating: data/\n",
      "   creating: data/10-batches-bin/\n",
      "  inflating: data/10-batches-bin/batches.meta.txt  \n",
      "  inflating: data/10-batches-bin/data_batch_1.bin  \n",
      "  inflating: data/10-batches-bin/data_batch_2.bin  \n",
      "  inflating: data/10-batches-bin/data_batch_3.bin  \n",
      "  inflating: data/10-batches-bin/data_batch_4.bin  \n",
      "  inflating: data/10-batches-bin/data_batch_5.bin  \n",
      "   creating: data/10-verify-bin/\n",
      "  inflating: data/10-verify-bin/test_batch.bin  \n",
      "   creating: images/\n",
      "  inflating: images/01.png           \n",
      "  inflating: images/lenet.jpg        \n",
      "   creating: results/\n"
     ]
    }
   ],
   "source": [
    "!wget https://ascend-professional-construction-dataset.obs.cn-north-4.myhuaweicloud.com/ComputerVision/cifar10_mindspore.zip\n",
    "!unzip cifar10_mindspore.zip\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0b97a255",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#创建图像标签列表\n",
    "category_dict = {0:'airplane',1:'automobile',2:'bird',3:'cat',4:'deer',5:'dog',\n",
    "                 6:'frog',7:'horse',8:'ship',9:'truck'}\n",
    "\n",
    "####____####\n",
    "current_path = os.getcwd()\n",
    "data_path = os.path.join(current_path, 'data/10-verify-bin')\n",
    "cifar_ds = ds.Cifar10Dataset(data_path)\n",
    "\n",
    "# 设置图像大小\n",
    "plt.figure(figsize=(8,8))\n",
    "i = 1\n",
    "# 打印9张子图\n",
    "for dic in cifar_ds.create_dict_iterator():\n",
    "    plt.subplot(3,3,i)\n",
    "    ####____####\n",
    "    plt.imshow(dic['image'].asnumpy())\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.axis('off')\n",
    "    plt.title(category_dict[dic['label'].asnumpy().sum()])\n",
    "    i +=1\n",
    "    if i > 9 :\n",
    "        break\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4d520645",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_data(datapath):\n",
    "    cifar_ds = ds.Cifar10Dataset(datapath)\n",
    "    return cifar_ds\n",
    "\n",
    "def process_dataset(cifar_ds,batch_size =32,status=\"train\"):\n",
    "    '''\n",
    "    ---- 定义算子 ----\n",
    "    '''\n",
    "    # 归一化\n",
    "    rescale = 1.0 / 255.0\n",
    "    # 平移\n",
    "    shift = 0.0\n",
    "\n",
    "    resize_op = CV.Resize((32, 32))\n",
    "    rescale_op = CV.Rescale(rescale, shift)\n",
    "    # 对于RGB三通道分别设定mean和std\n",
    "    normalize_op = CV.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010))\n",
    "    if status == \"train\":\n",
    "        # 随机裁剪\n",
    "        random_crop_op = CV.RandomCrop([32, 32], [4, 4, 4, 4])\n",
    "        # 随机翻转\n",
    "        random_horizontal_op = CV.RandomHorizontalFlip()\n",
    "    # 通道变化\n",
    "    channel_swap_op = CV.HWC2CHW()\n",
    "    # 类型变化\n",
    "    typecast_op = C.TypeCast(mstype.int32)\n",
    "\n",
    "    '''\n",
    "    ---- 算子运算 ----\n",
    "    '''\n",
    "    cifar_ds = cifar_ds.map(input_columns=\"label\", operations=typecast_op)\n",
    "    if status == \"train\":\n",
    "        cifar_ds = cifar_ds.map(input_columns=\"image\", operations=random_crop_op)\n",
    "        cifar_ds = cifar_ds.map(input_columns=\"image\", operations=random_horizontal_op)\n",
    "    cifar_ds = cifar_ds.map(input_columns=\"image\", operations=resize_op)\n",
    "    cifar_ds = cifar_ds.map(input_columns=\"image\", operations=rescale_op)\n",
    "    cifar_ds = cifar_ds.map(input_columns=\"image\", operations=normalize_op)\n",
    "    cifar_ds = cifar_ds.map(input_columns=\"image\", operations=channel_swap_op)\n",
    "    \n",
    "    # shuffle\n",
    "    cifar_ds = cifar_ds.shuffle(buffer_size=1000)\n",
    "    # 切分数据集到batch_size\n",
    "    cifar_ds = cifar_ds.batch(batch_size, drop_remainder=True)\n",
    "    \n",
    "    return cifar_ds\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "550ceac1",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = os.path.join(current_path, 'data/10-batches-bin')\n",
    "batch_size=32\n",
    "status=\"train\"\n",
    "\n",
    "# 生成训练数据集\n",
    "cifar_ds = get_data(data_path)\n",
    "ds_train = process_dataset(cifar_ds,batch_size =batch_size, status=status)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "45c0fcde",
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"LeNet.\"\"\"\n",
    "\n",
    "\n",
    "def conv(in_channels, out_channels, kernel_size, stride=1, padding=0):\n",
    "    \"\"\"weight initial for conv layer\"\"\"\n",
    "    weight = weight_variable()\n",
    "    return nn.Conv2d(in_channels, out_channels,\n",
    "                     kernel_size=kernel_size, stride=stride, padding=padding,\n",
    "                     weight_init=weight, has_bias=False, pad_mode=\"same\")\n",
    "\n",
    "\n",
    "def fc_with_initialize(input_channels, out_channels):\n",
    "    \"\"\"weight initial for fc layer\"\"\"\n",
    "    weight = weight_variable()\n",
    "    bias = weight_variable()\n",
    "    return nn.Dense(input_channels, out_channels, weight, bias)\n",
    "\n",
    "\n",
    "def weight_variable():\n",
    "    \"\"\"weight initial\"\"\"\n",
    "    return TruncatedNormal(0.02)\n",
    "\n",
    "\n",
    "class LeNet5(nn.Cell):\n",
    "    \"\"\"\n",
    "    Lenet network\n",
    "\n",
    "    Args:\n",
    "        num_class (int): Num classes. Default: 10.\n",
    "\n",
    "    Returns:\n",
    "        Tensor, output tensor\n",
    "    Examples:\n",
    "        >>> LeNet(num_class=10)\n",
    "\n",
    "    \"\"\"\n",
    "    def __init__(self, num_class=10, channel=3):\n",
    "        super(LeNet5, self).__init__()\n",
    "        self.num_class = num_class\n",
    "        self.conv1 = conv(channel, 6, 5)\n",
    "        self.conv2 = conv(6, 16, 5)\n",
    "        self.fc1 = fc_with_initialize(16 * 8 * 8, 120)\n",
    "        self.fc2 = fc_with_initialize(120, 84)\n",
    "        self.fc3 = fc_with_initialize(84, self.num_class)\n",
    "        self.relu = nn.ReLU()\n",
    "        self.max_pool2d = nn.MaxPool2d(kernel_size=2, stride=2)\n",
    "        self.flatten = nn.Flatten()\n",
    "\n",
    "    def construct(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.max_pool2d(x)\n",
    "        x = self.conv2(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.max_pool2d(x)\n",
    "        x = self.flatten(x)\n",
    "        x = self.fc1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.fc2(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.fc3(x)\n",
    "        return x\n",
    "# 构建网络\n",
    "network = LeNet5(10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "e4ba3696",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 返回当前设备\n",
    "device_target = mindspore.context.get_context('device_target')\n",
    "# 确定图模型是否下沉到芯片上\n",
    "dataset_sink_mode = True if device_target in ['Ascend','GPU'] else False\n",
    "# 设置模型的设备与图的模式\n",
    "context.set_context(mode=context.GRAPH_MODE, device_target=device_target)\n",
    "# 使用交叉熵函数作为损失函数\n",
    "net_loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction=\"mean\")\n",
    "# 优化器为Adam\n",
    "net_opt = nn.Adam(params=network.trainable_params(), learning_rate=0.001)\n",
    "# 监控每个epoch训练的时间\n",
    "time_cb = TimeMonitor(data_size=ds_train.get_dataset_size())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e7962957",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mindspore.train.callback import Callback\n",
    "\n",
    "class EvalCallBack(Callback):\n",
    "    def __init__(self, model, eval_dataset, eval_per_epoch, epoch_per_eval):\n",
    "        self.model = model\n",
    "        self.eval_dataset = eval_dataset\n",
    "        self.eval_per_epoch = eval_per_epoch\n",
    "        self.epoch_per_eval = epoch_per_eval\n",
    "\n",
    "    def epoch_end(self, run_context):\n",
    "        cb_param = run_context.original_args()\n",
    "        cur_epoch = cb_param.cur_epoch_num\n",
    "        if cur_epoch % self.eval_per_epoch == 0:\n",
    "            acc = self.model.eval(self.eval_dataset, dataset_sink_mode=False)\n",
    "            self.epoch_per_eval[\"epoch\"].append(cur_epoch)\n",
    "            self.epoch_per_eval[\"acc\"].append(acc[\"Accuracy\"])\n",
    "            print(acc)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bb4fb1c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "============== Starting Training ==============\n",
      "epoch: 1 step: 1562, loss is 1.6773098\n",
      "{'Accuracy': 0.43607954545454547}\n",
      "epoch: 2 step: 1562, loss is 1.1719999\n",
      "{'Accuracy': 0.4886963828425096}\n",
      "epoch: 3 step: 1562, loss is 1.2762606\n",
      "{'Accuracy': 0.531770166453265}\n",
      "epoch: 4 step: 1562, loss is 1.3210876\n",
      "{'Accuracy': 0.5662011843790012}\n",
      "epoch: 5 step: 1562, loss is 1.1372198\n",
      "{'Accuracy': 0.5700424135723432}\n",
      "epoch: 6 step: 1562, loss is 0.9634448\n",
      "{'Accuracy': 0.5782050256081946}\n",
      "epoch: 7 step: 1562, loss is 0.94368756\n",
      "{'Accuracy': 0.6145366517285531}\n",
      "epoch: 8 step: 1562, loss is 1.049093\n",
      "{'Accuracy': 0.6105153649167734}\n",
      "epoch: 9 step: 1562, loss is 1.0961374\n",
      "{'Accuracy': 0.6070942701664532}\n",
      "epoch: 10 step: 1562, loss is 0.9355309\n",
      "{'Accuracy': 0.6252600832266325}\n",
      "epoch: 11 step: 1562, loss is 1.0560404\n",
      "{'Accuracy': 0.6420254481434059}\n",
      "epoch: 12 step: 1562, loss is 1.1965395\n",
      "{'Accuracy': 0.6423455505761844}\n",
      "epoch: 13 step: 1562, loss is 1.0756748\n",
      "{'Accuracy': 0.6483274647887324}\n",
      "epoch: 14 step: 1562, loss is 0.66533405\n",
      "{'Accuracy': 0.6499879961587708}\n",
      "epoch: 15 step: 1562, loss is 0.9417937\n",
      "{'Accuracy': 0.650268085787452}\n",
      "epoch: 16 step: 1562, loss is 1.3492303\n",
      "{'Accuracy': 0.6543293854033291}\n",
      "epoch: 17 step: 1562, loss is 1.4647925\n",
      "{'Accuracy': 0.6426456466069143}\n",
      "epoch: 18 step: 1562, loss is 1.1954968\n",
      "{'Accuracy': 0.6606113956466069}\n",
      "epoch: 19 step: 1562, loss is 1.0046768\n",
      "{'Accuracy': 0.6628521126760564}\n",
      "epoch: 20 step: 1562, loss is 0.9836742\n",
      "{'Accuracy': 0.6568101792573624}\n",
      "epoch: 21 step: 1562, loss is 1.093524\n",
      "{'Accuracy': 0.6728553137003841}\n",
      "epoch: 22 step: 1562, loss is 1.0372887\n",
      "{'Accuracy': 0.679657490396927}\n",
      "epoch: 23 step: 1562, loss is 0.8163093\n",
      "{'Accuracy': 0.6691941421254801}\n",
      "epoch: 24 step: 1562, loss is 0.78369534\n",
      "{'Accuracy': 0.676056338028169}\n",
      "epoch: 25 step: 1562, loss is 0.98369205\n",
      "{'Accuracy': 0.6844590268886044}\n",
      "epoch: 26 step: 1562, loss is 0.91965413\n",
      "{'Accuracy': 0.6779769526248399}\n",
      "epoch: 27 step: 1562, loss is 1.267575\n",
      "{'Accuracy': 0.6793373879641486}\n",
      "epoch: 28 step: 1562, loss is 0.9518202\n",
      "{'Accuracy': 0.6708546734955185}\n",
      "epoch: 29 step: 1562, loss is 0.86306566\n",
      "{'Accuracy': 0.6802576824583867}\n",
      "epoch: 30 step: 1562, loss is 0.81422985\n",
      "{'Accuracy': 0.6839788732394366}\n",
      "epoch: 31 step: 1562, loss is 1.1574671\n",
      "{'Accuracy': 0.6856794174135723}\n",
      "epoch: 32 step: 1562, loss is 0.80724\n",
      "{'Accuracy': 0.6888204225352113}\n",
      "epoch: 33 step: 1562, loss is 0.5779889\n",
      "{'Accuracy': 0.6859795134443022}\n",
      "epoch: 34 step: 1562, loss is 0.8925457\n",
      "{'Accuracy': 0.6881602112676056}\n",
      "epoch: 35 step: 1562, loss is 0.5016556\n",
      "{'Accuracy': 0.6842189500640204}\n",
      "epoch: 36 step: 1562, loss is 0.55656344\n",
      "{'Accuracy': 0.6837788092189501}\n",
      "epoch: 37 step: 1562, loss is 0.7363762\n",
      "{'Accuracy': 0.682558418693982}\n",
      "epoch: 38 step: 1562, loss is 0.88136625\n",
      "{'Accuracy': 0.7002040653008963}\n",
      "epoch: 39 step: 1562, loss is 0.8513111\n",
      "{'Accuracy': 0.6965829065300896}\n",
      "epoch: 40 step: 1562, loss is 1.1958897\n",
      "{'Accuracy': 0.6828585147247119}\n",
      "epoch: 41 step: 1562, loss is 0.7181257\n",
      "{'Accuracy': 0.6829585467349552}\n",
      "epoch: 42 step: 1562, loss is 0.7690495\n",
      "{'Accuracy': 0.6906810179257362}\n",
      "epoch: 43 step: 1562, loss is 0.8978268\n",
      "{'Accuracy': 0.690400928297055}\n",
      "epoch: 44 step: 1562, loss is 0.7274618\n",
      "{'Accuracy': 0.685179257362356}\n",
      "epoch: 45 step: 1562, loss is 0.77508724\n",
      "{'Accuracy': 0.6862395966709347}\n",
      "epoch: 46 step: 1562, loss is 0.9851495\n",
      "{'Accuracy': 0.7002440781049936}\n",
      "epoch: 47 step: 1562, loss is 1.2256715\n",
      "{'Accuracy': 0.6980033610755442}\n",
      "epoch: 48 step: 1562, loss is 0.8616962\n",
      "{'Accuracy': 0.694242157490397}\n",
      "epoch: 49 step: 1562, loss is 0.9762864\n",
      "{'Accuracy': 0.7004241357234315}\n",
      "epoch: 50 step: 1562, loss is 0.5436758\n",
      "{'Accuracy': 0.7025048015364916}\n",
      "epoch: 51 step: 1562, loss is 0.6838038\n",
      "{'Accuracy': 0.6987035851472471}\n",
      "epoch: 52 step: 1562, loss is 1.1647059\n",
      "{'Accuracy': 0.6915212868117798}\n",
      "epoch: 53 step: 1562, loss is 0.4193041\n",
      "{'Accuracy': 0.6977832906530089}\n",
      "epoch: 54 step: 1562, loss is 1.1223296\n",
      "{'Accuracy': 0.7077664852752881}\n",
      "epoch: 55 step: 1562, loss is 0.8014877\n",
      "{'Accuracy': 0.7016845390524968}\n",
      "epoch: 56 step: 1562, loss is 0.90530443\n",
      "{'Accuracy': 0.6964828745198464}\n",
      "epoch: 57 step: 1562, loss is 1.0512853\n",
      "{'Accuracy': 0.7057258322663252}\n",
      "epoch: 58 step: 1562, loss is 0.9477448\n",
      "{'Accuracy': 0.6972831306017926}\n",
      "epoch: 59 step: 1562, loss is 0.8948573\n",
      "{'Accuracy': 0.707246318822023}\n",
      "epoch: 60 step: 1562, loss is 0.43573475\n",
      "{'Accuracy': 0.7053457106274008}\n",
      "epoch: 61 step: 1562, loss is 0.54602766\n",
      "{'Accuracy': 0.7082466389244558}\n",
      "epoch: 62 step: 1562, loss is 0.8343785\n",
      "{'Accuracy': 0.7040653008962868}\n",
      "epoch: 63 step: 1562, loss is 0.7558348\n",
      "{'Accuracy': 0.7057858514724712}\n",
      "epoch: 64 step: 1562, loss is 0.71530366\n",
      "{'Accuracy': 0.6876400448143406}\n",
      "epoch: 65 step: 1562, loss is 0.81608075\n",
      "{'Accuracy': 0.7029449423815621}\n",
      "epoch: 66 step: 1562, loss is 1.0041242\n",
      "{'Accuracy': 0.6997639244558259}\n",
      "epoch: 67 step: 1562, loss is 1.143812\n",
      "{'Accuracy': 0.7050656209987196}\n",
      "epoch: 68 step: 1562, loss is 1.1727843\n",
      "{'Accuracy': 0.6988636363636364}\n",
      "epoch: 69 step: 1562, loss is 1.0248824\n",
      "{'Accuracy': 0.7052856914212549}\n",
      "epoch: 70 step: 1562, loss is 0.66688704\n",
      "{'Accuracy': 0.7118677976952625}\n",
      "epoch: 71 step: 1562, loss is 0.8246734\n",
      "{'Accuracy': 0.7065460947503202}\n",
      "epoch: 72 step: 1562, loss is 0.86741304\n",
      "{'Accuracy': 0.7088468309859155}\n",
      "epoch: 73 step: 1562, loss is 0.6880111\n",
      "{'Accuracy': 0.6971230793854033}\n",
      "epoch: 74 step: 1562, loss is 0.5727052\n",
      "{'Accuracy': 0.7172695262483995}\n",
      "epoch: 75 step: 1562, loss is 0.66865665\n",
      "{'Accuracy': 0.7123679577464789}\n",
      "epoch: 76 step: 1562, loss is 0.8614229\n",
      "{'Accuracy': 0.7164492637644047}\n",
      "epoch: 77 step: 1562, loss is 1.0337029\n",
      "{'Accuracy': 0.7030049615877081}\n",
      "epoch: 78 step: 1562, loss is 0.588214\n",
      "{'Accuracy': 0.709807138284251}\n",
      "epoch: 79 step: 1562, loss is 0.75220346\n",
      "{'Accuracy': 0.698243437900128}\n",
      "epoch: 80 step: 1562, loss is 0.73503876\n",
      "{'Accuracy': 0.7185099231754162}\n",
      "epoch: 81 step: 1562, loss is 0.98983157\n",
      "{'Accuracy': 0.7129481434058899}\n",
      "epoch: 82 step: 1562, loss is 0.96853757\n",
      "{'Accuracy': 0.7178697183098591}\n",
      "epoch: 83 step: 1562, loss is 0.5855131\n",
      "{'Accuracy': 0.7105073623559539}\n",
      "epoch: 84 step: 1562, loss is 0.87325096\n",
      "{'Accuracy': 0.7111875800256082}\n",
      "epoch: 85 step: 1562, loss is 0.9137762\n",
      "{'Accuracy': 0.7164692701664532}\n",
      "epoch: 86 step: 1562, loss is 0.8291276\n",
      "{'Accuracy': 0.7147487195902689}\n",
      "epoch: 87 step: 1562, loss is 1.04989\n",
      "{'Accuracy': 0.7071062740076824}\n",
      "epoch: 88 step: 1562, loss is 0.6412367\n",
      "{'Accuracy': 0.7150288092189501}\n",
      "epoch: 89 step: 1562, loss is 0.6106802\n",
      "{'Accuracy': 0.7177696862996159}\n",
      "epoch: 90 step: 1562, loss is 0.56246006\n",
      "{'Accuracy': 0.7141285211267606}\n",
      "epoch: 91 step: 1562, loss is 0.6070729\n",
      "{'Accuracy': 0.7214908770806658}\n",
      "epoch: 92 step: 1562, loss is 0.6557622\n",
      "{'Accuracy': 0.7238316261203586}\n",
      "epoch: 93 step: 1562, loss is 0.9780799\n",
      "{'Accuracy': 0.7162892125480154}\n",
      "epoch: 94 step: 1562, loss is 0.8116702\n",
      "{'Accuracy': 0.7229513444302177}\n",
      "epoch: 95 step: 1562, loss is 0.46478817\n",
      "{'Accuracy': 0.7254721510883483}\n",
      "epoch: 96 step: 1562, loss is 0.6958112\n",
      "{'Accuracy': 0.7189300576184379}\n",
      "epoch: 97 step: 1562, loss is 0.7830975\n",
      "{'Accuracy': 0.7067661651728553}\n",
      "epoch: 98 step: 1562, loss is 0.7695555\n",
      "{'Accuracy': 0.7114476632522407}\n",
      "epoch: 99 step: 1562, loss is 0.71506894\n",
      "{'Accuracy': 0.7197703265044815}\n",
      "epoch: 100 step: 1562, loss is 0.6726586\n",
      "{'Accuracy': 0.7181298015364916}\n"
     ]
    }
   ],
   "source": [
    "# 设置CheckpointConfig，callback函数。save_checkpoint_steps=训练总数/batch_size\n",
    "config_ck = CheckpointConfig(save_checkpoint_steps=1562,\n",
    "                             keep_checkpoint_max=10)\n",
    "ckpoint_cb = ModelCheckpoint(prefix=\"checkpoint_lenet_original\", directory='./results',config=config_ck)\n",
    "# 建立可训练模型\n",
    "model = Model(network = network, loss_fn=net_loss,optimizer=net_opt, metrics={\"Accuracy\": Accuracy()})\n",
    "eval_per_epoch = 1\n",
    "epoch_per_eval = {\"epoch\": [], \"acc\": []}\n",
    "eval_cb = EvalCallBack(model, ds_train, eval_per_epoch, epoch_per_eval)\n",
    "print(\"============== Starting Training ==============\")\n",
    "model.train(100, ds_train,callbacks=[ckpoint_cb, LossMonitor(per_print_times=1),eval_cb],dataset_sink_mode=dataset_sink_mode)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "02788df2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test results: {'Accuracy': 0.7216546474358975}\n"
     ]
    }
   ],
   "source": [
    "data_path = os.path.join(current_path, 'data/10-verify-bin')\n",
    "batch_size=32\n",
    "status=\"test\"\n",
    "# 生成测试数据集\n",
    "cifar_ds = ds.Cifar10Dataset(data_path)\n",
    "ds_eval = process_dataset(cifar_ds,batch_size=batch_size,status=status)\n",
    "\n",
    "res = model.eval(ds_eval, dataset_sink_mode=dataset_sink_mode)\n",
    "# 评估测试集\n",
    "print('test results:',res)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ac1311a3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAJNCAYAAAA8mg+EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOy9d5QsV3nu/bydJ/SkMydHHWUJJYKEAAn5Eg3G/sBcDAaDDNjGcdnGwIVLEAaDfM0HNh++xp/BFhcMCDBghIkSUSggBEooSyfnmTOxp3Pv+0fVEc08z/TpozQzdd7fWrPW9Fu7au/atfeu3dXPfspCCHAcx3Ecx0kSqcUugOM4juM4zqONT3Acx3Ecx0kcPsFxHMdxHCdx+ATHcRzHcZzE4RMcx3Ecx3ESh09wHMdxHMdJHD7B6QIz+5SZXdZl2mvN7NKHmc8x72tmGTMLZrZlge2vMbOvP5zyOM58lmJfMLOTzMz9LpzHlaXYFx5tzOy9ZnbFYuT9aLAsJjhmNtv21zKzctvnVy52+ZYyIYRPhBB+dbHL4Tw6eF9wnAjvC4CZvd7MvrfY5ViqZBa7AN0QQug/8r+ZbQfw+hDC1QulN7NMCKHxeJTNcR5PvC88fnjdLW28L3SHmaVDCM3FLsdisCye4ByN+DHalWb2GTObAfCq+Y8PzezZcSc48nmDmX3JzA6Z2TYz++Mu81phZl+L95sws6vMbP28ZCeb2U/MbCrOY7ht/6eb2Q1mNmlmt5jZxV3me4qZ/SA+5piZfXpekueZ2f1xmT7ctt9DM/y2n7P+ND7nMTO73MwS0Q6c46YvpM3sQ2Y2bmYPAHj+vO1DZvZvZrbPzHab2V+3t/G4T9wdl/nrZrYxjh/pH39kZvcDuLub8jhLk6T3BTM7C8BHAFxk0VOrsTj+KTP7RzP7hpmV4u2/9DOXzXvyY2ZnmdnVZnbYzPab2ZtFfjkz+1z8l+2mXhabJN3YXgzg0wAGAVzZKaGZpQF8FcBNANYDeA6AN5nZs7rIJwXgXwBsArAZQB3AP8xL8+r4bx0AA/ChON+NAL4C4F0ARgD8DwBfNLMVoownxI19XRz6GwD/BWAYwAYA/zhvlxcAeBKA8xB15Gd3OIffAPBEAE8G8NK4rE5ySHpf+EMAzwVwDoDzAbxs3i6fAlAGcCKiNv5CAL8bH+ulAN6EqA+sBHAjorpq59cBPAXAWV3UgbO0SWxfCCHcDuBPAPwwhNAfQhhtS/rbAN4NoAjg+qOc9yCAqwFcBWAtgFMAfG9eml4A/wlgBsDLQwj1TsdcKiRpgnNtCOGqEEIrhFA+StqnAhgIIbwvhFALIdwP4OMAXn60TEIIh0IIXwohlEMI0wDeB+CZ85J9IoRwZwihBOCdAF5uZoaocX8lhPDNuJzfAHAr5n0DjfPZFkIYCiHsjUN1AFsArA0hVEIIP5q3y/tDCFMhhO2IGue5HU7j8hDCRJz2wwBecbTzdpYVSe8LLwPwoRDC7hDCOIDLj6SNvzU/C8BfhBDmQgj7Afx92/n8AYD3hRDuiX+ueC+A8+d9235f3D+OVnfO0ifpfWEhvhRCuD4+XvUoaX8dwK4Qwj+EEKohhOkQwo/btg8C+CaiJ5qvDyG0jnK8JcOy0OB0ya5jSLsZwCYzm2yLpTFv1qowsz5EM/PnAhiKw8UOZdkBII9oZr4ZwCvM7MVt27MAvtFFmd8I4D0AfhI/ivxACOETbdv3t/0/B6AfCzO/fOsWSugsS5LeF9aJ4x5hc5zHgejeASD6Ire9bfs/mln7t+sWoqeiB0SZneVN0vvCQhzLeW8EcH+H7U9HNFf472GZvZ07SROc+RVfAtDb9nlN2/+7ANwXQjj9YeTzZgAnADg/hLDfzJ6M6JFmOxvb/t8EoArgcJzvv4UQ/vBYMw0h7APwegCIf5/9tpn9AA9vMN4I4J628h3t24CzvEh0XwCwTxz3CLsQTfBHFvimuQvAO0II9HOFmR0ZD5fVIO50JOl9YaG2eqzn/WIszNcA3AvgGjO7JIRw6JhLuUgk6Seq+dwC4IVmNmxmawH8Wdu26wHUzOyNZlaIRYtnmdmTujhuEdEAOhH/RvpOkebVZnZaPKt/N4DPxTPfTwJ4sZk9J86zYGa/0qYtWBAze1nbY/RJRA344Srj32yREHMTonq5Ms7jpFhkueFhHtdZmiSqLwD4HIA/N7P1cb5vObIhhLALwPcBfMDMBswsFbfrI6LNjwL4n2Z2OvCQIPmlnTKLBZpv76JcztInaX3hAIANXYh+bwHwm2bWY2anAHht27avIHpy9ScWCYkHzOz89p1DCO8D8AUAVx/RBlkk0F7Sq9KSPMG5AsBdiB4FfgPAZ49siH97fwEigeJ2AGMA/hnAQBfH/SCi3yTHAVwHQJnofRKR0HEfokecfx7nux3RTPkdAA4B2Inopye6Dma21SJl/JFGfgGAmyxSxX8RwB+HEHZ2UV7FVYga/M8AfAlRXQHRN4wH8cs/dznLnyuQrL7wTwCuAXA7om/JX5i3y6sA9AG4E8AEgM8j/sYaQvh8XO7Pm9k0gNsAPO8o57kBwHzNm7M8uQLJ6gvfBnAfop9kO43bH0D0pfgggH+Ny4E4/ylEgurfjLffC9YPIYTwrvi8vm1mQ4juF0u6X9gy+0nNeQTEj+DrAE6IO9X87ZchEpt9/HEumuMsSSxyCP9kCOGiRS6K4ywpLHI4/mQI4ZrFLstC+ATnOOJoExzHcRzHSQpJ/onKcRzHcZzjFH+C4ziO4zhO4vAnOI7jOI7jJA6f4CyAmV1iZrvbPv/czC55HPK9wsze+zD2224LvJ7BzC4ys3vUNsfphuXSH2Kbg5MeyzI5xzfLpS88ynlfambXLkbej4RlPcGJB7NSvGxuj5l90KL3iTzqhBDODCF8r8syLakBNoTwwxDCqYtdDuexxfuD40Qcb31h/qTLiVjWE5yYc0II/YjeP/PbAH5vfgL7hUOp4yQd7w+PMl5fyxbvC20cT+d6hCRMcAAAIYS7AfwQwBOAh36yeYuZ3QagZGYZM1tnZv9h0Svtt5nZQy6WscPjFRa96v5ORG8TRtv2h34Cit0m32ZmD5jZjJndbGYbLXp1AgDcGn9z+K04/a+Z2S0WvQX2OjM7u+2455nZT+PjXAmgsNA5mtmomX01Ps5hM/uhmbVfw3PN7DYzmzKzK82sEO83/5HqdjN7q5ndGZ/vvx1J6ySD46E/xOnfZGb7zGyvmb123ra8mX3AzHaa2QEz+6iZ9bRt71QOqq9jqH5nCZH0vmCRM/LXAayLjz0bn89lZvYFM/uURaaWl9q8n7nEvWGjmX0xrodxM/vIAnn+nUUO34NdX4jFIISwbP8QOTOeFP9/BiIH3tfFn7cjcuvdCKAH0WTuZkQW2jkAWxG59j4vTn85ok4wEu9zB4DdbXltB/Ds+P83IXJRPRXRa+/PAbBifpniz09E5A55ASL3ytfEx8rH5dgB4C8QvVztpYh8at7btv8kgGfE/78fkdV8Nv67CL9YCbcdwI8RvYhwBJFb5xvibZeIc7kjPs8RRG6U732k18P/vD88zv3h+Yis6p+AyLn40/Pq4O8R2dCPILLSvwrA+49WDlVfi31t/c/7wlH6wiXtZYpjl8X7/D/xOfYgcnJuP8ZD+8VluBXAh+L+VGg7/qUAro2P8y+I3i7eu9jX+ajtYLEL8Cg04mlEduwPAHgvgFRbo3ttW9oLAOyct/9bEb3kDHGDfn7btt/v0IjvAfAbR+tY8ed/AvCeeWnuQWSFfTGiF11a27brsMBkA8BfA/jP9uPPK9+r2j7/LwAfnd+I29K+oe3zCwA8sNjX0/8e2d9x2B/+FcDlbZ9POZIfoptLCcCJbdsvBLDtaOVQ9eV/y+vvOOwLvzTGx7HLAPxgXuwKLDzBuRDRqyIy4viXArgR0XsL/wNAbrGvcTd/SXjs+sQQwkKvem9/0/ZmRI/wJttiaUQzcyB68jH/dfYLsRFRp+mGzQBeY2Z/2hbLxfkFAHtC3IK6yPfvEDXab5kZAPz/IYTL27a3v4tkLs5jIeafazcvdnOWPsdTf1iH6Ju3SrsS0duTb477ChBNeo4ITTuV4wjt5+8sP46nvrAQx9KGNwLYEaJ3cilOQvRE6vwQQu1hlOVxJzEanAVobxy7EH17G2r7K4YQXhBv3wd+nf1C7AJwYpdl2AXgb+bl2xtC+Eyc53prG4E75RtCmAkhvDGEsBXAiwD8pZk9q8tyzGf+ue59mMdxlg+J6g9HKeMYgDKAM9vyGQyR6PRo5ThCe305ySJpfWGhtjo/XkI08T/Cmnnl2dRBb3YXgN8F8HUzWxarcpM+wWnnxwCmY3FZTywGe4KZHRGMfQ7AW81s2Mw2APjThQ+FjwF4j5mdbBFnW/wKeUSagK1taf8FwBvM7II4bZ+ZvdDMigCuB9AA8Gex0O0liN5kK4kFaSfFjX4aQDP+ezj8sZltMLMRAG9D9OjROX5Y9v0hLuOlZnaGmfUCeNeRDSGEVpzXh8xsFQCY2XozO/Lm8E7lcI4vktAXDgBY0YXo9xYALzCzETNbg/iN5m31sA/A5XFZCmb29Pad48nX2wBcbWbdTuQWjeNmghNCaCJ66nEugG2IvuF9DNEr7gHg3YgeAW4D8C1Er7ZfiA8iavTfQjTR+DgiARcQ/YT0CYtU8S8LIfwE0fLEjyD6Pfh+RL9nIn7M95L48wSA3wLwxfaMLFLEH3mT8ckArgYwi6gD/O/Qhf/CAnw6Lv+D8d+iGEg5i0MS+kMI4euIhMTfiY/znXnlekscv8GiVSRXIxJ/olM5nOOLhPSFuwF8BsCD8fEXkhx8EpGQeHtcxoe+2LbVw0kAdgLYHef7S4QQPoFID/odM9vSoS4WHX8X1XGImW0H8PoQwtWLXRbHcRzHeSw4bp7gOI7jOI5z/OATHMdxHMdxEof/ROU4juM4TuLwJziO4ziO4yQOn+A4juM4jpM4OjoZ/95b/oJ+v3rKOc+gdNUqmxoWcjmK7d9/gGK33fcgxdaPrKDYAwfYh66nv5diA31sY1FttCiWyxinq1Qols/yHLA3l6fYzFyZYrUm59sSPwm2mnWKpbPi0rT4eJUqx1YPDVAsNDnfap3z3bOb67ki6qVHnEc2y7HpA4coNtzL123n1CzFbvnJ9/giLTI/2z5HJ5lJcXtIGxe9NFelmKW4zgoFPl4QX0Vaga+9OBwsiGrsLoQAzgMiVquxHdPMHBui9hezFMs9ZC7cnm93qDKrEqtf4k1cIxNHVGUJIqrqSh6vJvpiRYwfNW4vFz95zZLrE9v38QmlRV1k06LoKW4j4rIgneHxUF4/EUunuX2lUo/9d/tKla/fN7/9bYp99+rvUWzXtt0UUxd+9aoRiq0cHaVYobeHYjlxj964aT3FTjrlFIqtW8er0FetWkUxSz/29VwsyKoB4E9wHMdxHMdJID7BcRzHcRwncfgEx3Ecx3GcxOETHMdxHMdxEkdHkfFsicXD23Zsp9jMDAtEVwyx+AkNFiI2GpxHyLEQcdXqtRTL5wsUy4lY0Xge12iwAKxXCJSVPhNCsFvMcr5KyGZiTplN8/malEqKogghc0OIlscmJjhf8Z7OdGWOyzLL+zYyfL6r1qyhWK3G4slGg+svk2ch4FLkgQdY/LdqJbfN3j6un4mZaYrVKyyyXLt+NcXMuM4sLWIpITymCBAC13ejyWVRbSmXZRH0obEpit173x6KPfFJLFjM9vAw1AhcFoUSlSpVcGiJulIi1ZQSGat+LMoiXntbK5co1hALEgq9XAcrV/eJXJYiajEFX79mS10rrjR1SRWmxtcuhcfd+r8FcaVVviVxD3xw+3aKTUzNUOwnN/+UYtsf2EaxfJbvE3294n5X7KfYhk0bKXbSifyuzJkZ7sd33XU3xdas4THv3CeeQ7HTzzydYr39ql0L0b64+WZSHacshD/BcRzHcRwncfgEx3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRwdFTvW4vlPPsfixGqOD1NvscgsK5x4WzUWGZerLMLbc5jFWbkCuzNahvfNCJFxuczpUkLopARq9QafW0sINKUjp3H9NYUjrXL9rAuRNoTgTaQChIvu6iI7WTaUg3KexdcHJlmMtu2WB7gsDRapjmZUmZfHS19/dP0tFFu7doxivf0sxC3kuS4mhDi3OMDtul7n61cTQvk+IVYtTbNwvDg4TDHlOloW7WblSi5fJsexmVnuYzMzoiw9XFdKfCrdiIUoWEmAVTrVP5tN4cg8I8aeDPfjdIuvb1+Br8fgCnYa7+tnsWg2u+RMiyVp5VAsBMVKKK8ughiuu0a6cYuGI4XHYt+WOOKdd9xJsa/919codvvtt1PszHPPo9jAMPfF2bnbKFYVjsxzZRY3T4nFDIcnJym2dw+71g8UWQCcz3P/zGbvoNitt7JY+rwnsfD4qU99KsWUW7JaRHSs+BMcx3Ecx3ESh09wHMdxHMdJHD7BcRzHcRwncfgEx3Ecx3GcxNFRZNxf7KVYoy6cc4XodnaOHTxHe1lcB+FMKIxucXB8nGIjK/i18FMzhyk2KM6jUmFxc6HAYqqaOLdqncWEqAsXWOEKnE3xyc0Kt9O1KwYppsTXlmYhVm8fCz5zA0KwJcSBs00hbk5zveSEsHxSCMaVYrAiHXOXh8h43w52GM0Ix81Sk8XDEzPsgpxPsevoU55yFsWq1QrFvvXtqynW38fC8dI0X5eBoRUUO+e8sym2cwe7EZ/MekCUa0JAPcFix6kprpcNq4YolhNi/JYQ45uo+yAExUp4b0JW2hLOzfv2H6TYUJH72OknrKLYyKBYCCG+VqoxtCVXCxybk+vjg7guynm7Oz24dqJWQbVoRbkbC5GxFB6LfaviPvH5Kz9Psau+chWXRZzcppNOpdiJJ59MsUMH91Nsdpr7Tl30u5oYh6eEUH5yioXHWXGRenq4DaeF4DkvFqjcdQsvyrjlhh9T7KKLL6bYs57/fIqNrllHsU74ExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdx+ATHcRzHcZzE0VGx1ptmwWI2y+Ksvh7hzmssas2khPtnjl8BnxJir75hFiK2hEg2k+fj5XtZyNnKsYKvMceiq2qJnVxTeRZdBePjCTNPNMS5ZXMs4h1YyefRL4S4uRRfo1UjLL5rGotUIVyG1zQ43x3bWKBWbimRHseaDSH4FC6wtboQKC9BQuUQxQ5uYzfRwhBfl70P3EWxNSMsmtt5P7uENoWzr4HzLZc4XbGPBeuzMwcotmc/C6j37NlJsUaT80WBz7eW4v5UqXFsaob7U2WC21xGCBsLPTzO1IX4NC3SWUY4sNe4/pot7jvFAXafHRjkcSYId+OG6DsAn1tYkoJigahvJeBWptNBOQory2qBKYWyUHC3TLQHIU5Pi7F5/34W+956660UKwu38KEN6ym2euMJFNtz/z0Ue8LZ51Is38dtuNkQbwIo8WKUsQPstj5+iGOlaV4YUK/xPbAhhMwlsRBiYpKP15hjQf26tRspdtGvcLpamcsCsTjoCP4Ex3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3F0FhkLh8oH77mbYv1D7FBcB4tVB0bXUqzVYiFRRbzuHRUWMAmNMYZ6WOyYEe6kWSFkywonxvwAi/8aQlQHIVrr7WMH5XROlC/Lgq1qjc93sJ/noysGuf4mZvh8Jyc5XVq4jdaEU/VAL+fbmOF0J24aodi2B1mUq9TXphSIS5Byil1tZ2dZ+LayZyXFBop87cuzfLwffve/KJbv4WvQP8Ci1oGhIsUGe4UQscbX/uD+2ynWqLOg+OBBdhUvDPMY0Fdkwe6ePbdxbCe7nY49uItiKSFILfSy8LI8yyLLniEWWq/dyoLPRpX7TlWMPZkWO6afvJ6Fx9ksX3MTgnElrE2J/qmtfxeZbk3IxRipMHGdpQuyqgslHhbJKtMsYv/if32dYgcOcVsvFrmtD/Vz+5qZ43YzOcNtM59jkb3SoWf7uL8PFoSjsFgctHnziRRTLsj1Bo9lVeG0XxVvKqjUed+WEO2fumULxS551iUUu+vHP6XY/h17KfaGt/8RxY7gT3Acx3Ecx0kcPsFxHMdxHCdx+ATHcRzHcZzE4RMcx3Ecx3ESR0eR8TnnnUSx+77MDqhzMywkHRxmsePIpq0U2zrKYsypOgux6odZ1NcssDhL+Bxi7VoWgDUqLJKtNbg6anXhACxEy0EIB2sNjmWFA2RvhtMVhkRMCKgPTXC6cpVjBw4KoViNhccWuF6UaG1ACN6KfSx4Kw7zvvkUC9AbU+wEuhQ58WksTA1lru8e4Xa9HpspliuxmrAo3KkzKb4uJSH+Kwj9+5oit/9dU9yfDMKxd5TPo54RbrEtXhhQEE7BjRb37akZPo9DB3dTbPKwcFkVbb0lFgH0rOijWH6QzyOf5VhKiFmrJXZPbRmfW0u446qxp9nkvliQDq3cNhablqgf5TKs03WnUFbmxiZE2BkRy2Z4vPnpT39Gsb9//99R7KLnP5dimzZvotidP/s5xU485TSKBbFwp7ef+1heOPJv2ryBYsUii+wnx6coNn2YY5k898++Ye4nfb3str5CLCoo9vOimjWrVlFsaBWL8Wd27qHYNz/zWYpNTvB9wkXGjuM4juMcV/gEx3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRwdRcaf/cxXxQ68y/gUiwT3j/Gr2NcNr6FYscjCpEZZWE+mhTNyPwvKrJfL1xLS40ZDiL0KXJa1K1nod+DQJOfR5Lnipo0sxC0d4H1hLM6t1bh8M5Msbp7Yw8LQuhLasbYZ/Q2u09kW11V/P1+PnHAjnhxnAdhIP4vWClmODQ+xC/JSpJFlMWhLXKuUqkdjQexID4vx1+S43YyMsDBvtsl5KDfu1Su4bptNbiPDaeGomuPj7d2/j2LVWXY8Xrt+PcV68iySrQ5xexhZy3W6P/BihvExFk+mRR5DK4STsXCkXTnC1yOf5jGgp8DpmtMsMp5LcXup1kUbanIHbdU434EBdoNfbJpiLGhJVTCH0sqyV6RT7saKmVkWok/tZLfwb3z5yxSbm+a2VBzgNjI1wy7IpTKPfVnRDhtigUpR9LEVo9xnTzqF3Yi3nMDC44mxCYrdcO2NFCuX1MIO7ndlsZihXOAyF3J8D5yd4ftdfoDbdV6IpRtCTz8nXPU74U9wHMdxHMdJHD7BcRzHcRwncfgEx3Ecx3GcxOETHMdxHMdxEkdHkfHPH9xFsZUjQxQzMU3aM86Cu9t23EuxM09md+OhNAuxzjuBRX1DfaxCumeMBWBVYUY8IFyQT9goRGEtFk8Ot7jaKhUu8+gQi6l2HWRB8f4pLnN/r3CQLXMezRILTeeEk+tgjsvSaHLFNIUzbH2G862wThIzoixp8PHG6yyCO/vsk/mAS5CWEEWmhLC32uLzTrGODlNCAD9dYqHk4BSL9YrCoVhoBLFvx36KzZWEG7cQS2dFH5u58z7ed47bUnoVu5T35Lkd9me4ree2skBZuaOXSjzOZMB9rE8IPtekeVFBsSXcg+vCMVc4nO+78y6KmVgc0TThIK7cl3OizBuWnshYORQjcCwtxMhmfN6mBMrCEVp9PZ8p83h9zfe+Q7G5Eo+5z3/B8yi25SR2Lr/pppspNjvHQtxKnYXjysk4k+H2Wihwu2m1uN1MzPBY2jvAov2sOF61Iq6HcNRWVQ9xbtUKj2V54YQ+LBaZVFNcBy963e9Q7JrvXisKszD+BMdxHMdxnMThExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdxdBQZl4XWa1q4cJbmWFxUq7OA6ad37aRYr3AOXTfI4sneQXaU7O0ZpdhgnV0rlUNxAAv4TljHr3a/6bbbKVY7yALDilDd5sD5HpjmfSfLLEhtBq7TIByKZ+o8R82xXguVBpdvThyvwkVBbxCiauHIOVvj2Moiu/Ju3MDX955t3DaWIqUyixOlGWuar8tMiQWGYw0WCaaE8DI1y/0pTPOFVk6pWaESVKJWCHFnwVicOFrgNlzayy7D941v4+MVVlCsLsSJEG0zPczn2yxygxX6bow3uf4OzAonV1HPWSGAzGS4TzRFHzMhwM2JfbNCGN0jHJTPwjMptuhIjTFfvyaE87xoc8q0WBwOEG34wQcfoNjNN7GL7wuf/SyKnXfh0yl2j2jXt93yc4r1FIUDtuh3ocJtLj/I97FbbuXzmBTCaKFhxwlb2PF4bB8vNFDjQn+fcJ4XAv1iL4/rg+K+vXKUz23tSn6jwYGpcYqNruc+cdJWPrdO+BMcx3Ecx3ESh09wHMdxHMdJHD7BcRzHcRwncfgEx3Ecx3GcxNFRZJzP8PxnsJ93SWVZAXbCShbI9RRZwDSwkkWRG07l4ylXz7yx6GpT4NfHW0qI/7IcW1FkddualZspdtMt7PDcADs7nnUqCyrXTbAqbHw/C882n8AusJPTfL7XX7+by9IS4lMh8BswVmMWMlwvI6PsXr3K+Hw3pllktnUD19+gUEGfl15NsSVJittrRlp9sgAyCPFrENcgpIVAMy1ExsIlul7j69cS5cvnhZhWCDlbou+cfBa7j/90DzvI3vhtdnx9yUv/G8UO19i5eaaPCxOEEDeV5jqoVbhttoTIeDbwNaqLayT04nLkbAnBbCrwzhlRllaD21UqxWPF60VRFpuUcNhtibYZRP0orXtKqIxTor63b2MR+92330Gx8889h2I9A3x/OjS+j2JN4VB/5hlPoNjOnbzvkFhgMTfOYtrh09jFvUcsjNnz4A6K5YWL+t77eMHG5CQ7oSuB90A/36PXrVtHsfUb+D47Pn6YYlMi30NjLNweFIuIZmfYHfrCp11IsU74ExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdx+ATHcRzHcZzE0VFkfOnrnkOxESEGUtaTa4dZYHvw0AGKTU6xIHBD3wjnIV4pv2KQ01XKBYql8ix4q7RY2Fuqs1Cyt4/FXk+/mEWWU9MsiCrkWSR48gquq/Ub2QVTieqGVrJorVFlMXJPP6erCCFgrcqxDLjM9SpfIxMiwlyOC36TmBgAACAASURBVJ3vFeLJFjvXNhpC3bkEEVpfpISAOwgBq9CbotHi+lGOwqmW2FmIgnOqfKJ/CrNroMntoV4XQtyBHordcpAFhtfdcBvFNk9PUWz1iOizwuE83ct90YQ4NyfaUqaXyzx46lqKTQ5yGzYhGE8J4Xa9KRSz4jukpUWdCmG5bBtLkFaD+7MJN26TjU4IrkW7TjU5WKtxvmvW8GKF2cPs4rvtwQcpds8D2yk2vOYEim1cv55iJ27ldN+++hsU27CO933GM55GsRe+4AUUazR4HC7khJWxqL9qlevqgQfYLfneu+6k2P79XH+q/c/N8T3whuvHuDCClSv5Pnbq6WdQ7GkX8L2tE/4Ex3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3F0FBl/9P9cTbFslsV6SPNhshl2isxl2I145RC75I7exYLFFX0sbh4aYKFwf5HddAuhTLFrf/Zj3ncll6Vc4jlgqofFf9PTShDI9SIMmVGpsCiyWmFRWCYjBGVKGNpk98iMUJ+2lJpPWIvWq8K9NytElhUuS1McTwlwc0Is964/4eItNjVxXep14ZQtxL55ITo3E67FDT5eKsv7BpEvjNtco85lVorPlFV4XyHYnWrwvrOB8w1K7bhnD4Wy+znduDheK89i5BUDPC70imx7imJxRD8vUpjrFUNijxAeN7gNK7fkIATKCOK6CYFyapl8/2wqUby49vUajyMtIRQu9rGb7uTUBMVKMyxYV+NIsciC9aEBFquWG+I8hI11o8b95MwzTqXYj374XYr1FrgNZ8Rgkc9wm0uLes6IftwjFsaMDA9TbO2aNRQ7/ylPolhNCJSbTR4XmuJeVKlwXc3N8f04K8a30VUsPD50iF2QO7E8epDjOI7jOM4x4BMcx3Ecx3ESh09wHMdxHMdJHD7BcRzHcRwncXQUGfcNsMgntKSVK4fSLH7K9rHwuCTEhDMHWCS7Xwh7IdLl8ny826+9gWIT+9hVeWgFC5TPvuS/UWxuRjilZlh8bcYVc/gwC6wsI5xmUyyCawo320JgEVdNuH7OVThWFQ7FWSEe7u0dpZgwskRIcVlyonEEITyu1ZeHa2tDius41lLphONrRogJlbC3VeNYWrjAiiYnxapp0T9NCJSVC/LuSV4EUA/clopZPt6mVSzkX5MRTtBljh2ucBvp72dBakq061aKx57JXTx+tFZxmcuqsUvTYq7TlmjreaVAD7xvTq1IWII8uIudbsW6E8zNsih4Yu8Oiq1eyePNwQM8XpeF+DXfw3XWKwSsWWFJbmINhxqV6kJkXOzj8f93Xv1KihUywrVb9WPRvtTTiLRoS0GMPTURE9nKWKGHz00tIBBrR1AcYHGzyrYlBNRqMUqtxXXfCX+C4ziO4zhO4vAJjuM4juM4icMnOI7jOI7jJA6f4DiO4ziOkzg6ioxf98bfp5hyVK0Kp9RslsVeo8MsHhseZBFSPs/7KufcVJrLkheCwLe99q8odu1edlQtDLLI+K3v+UuKWYoFUQ3hbKqcm0sVdl+uNVjKNtDL4sm62Pf2a35EsROefC7FCsPs5DpXFoJnIR7ra7IA8mfXX0exzAgLo0c3beR8Z0sUO3CIRYRLkZYQcDfqQhQs3Gpbwq22JlyiU0Lsq0R9ohnChDoxLY6XF4bkpRKXRbnPzjV2UmzVydx3age4b69euYJiQ8JluznNbbNXKD4Hhah0MvD12F7mvrPnwRmKrTudRdCVgnCbFqJISwnHYyECVaL0Wo2PlxaGx0uRbdu2U0w5dCvl7MwYO9OOH9pHsb4+Hlv6+jmmnNhbwvG7IlyL68LZvSH6XWmO21KPKN/IMLelNPiepRYkKPLCBbkgnJtVOtUOgxL2SvW8HGhESCwoCaL9q2skxsa0cJGem+G674Q/wXEcx3EcJ3H4BMdxHMdxnMThExzHcRzHcRKHT3Acx3Ecx0kcHUXGG1ayQDSVYcGpsiZMC1fPlLC3rFbZmVA5O86VWChWyIlXz2fYsfTiFz6XYsNr11Ps9M18viizgLpfiKUtK0SHQuyYzQgRVw8L1HJCsFVqscqyTzgoj/SwuK3YWxT5sgg0U+Y8bvzBDyh2549uodjUFDuVnnbO2RTbeM6pXJRsL5dvCVKucHtQYr1ahd10c1nRJ0Q/qQpRZEO48+ZEf8qIRQAQLsNBtE3pdir0hbk+bptrTl9FsfKOgxSri+9Ug+vWUay4jseA8TlumyVRL2OHxil21ziPM7aKy6zEky2xCCAt+mdKqb6FoLIuXLurdd43oxyPlyCTB3ZTTDlgF1exwLy3TywoEZbCaSGmrbW4n6RF3YYmp6unub4LYmHM2GFuSz+95ecUe+rTL6FYPs9jWlbcP1viPFot5YROIdRSXM+phkioXkCghMKiwyuXYaiYWODTFBmnRDrLKLd1vuaHJ6Y53w74ExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdx+ATHcRzHcZzE0VFkbELAlMnwLi352nUOZoXIMmcsxMqJPAZ6hWtlo7tXwK8cm6DY5IN3U2y8yi6JmeavchZKBCocIHuEaK3ZFKKrgnC3FGKvWSFwbQlxVl247RaEYKsqXCGv+wE7I+/btYti60/aQrFX/CqLuW+/9VaK3fKt71Ps1GeeT7GlSFmIzlOinzSUK6oQHqeFULgh1IRKEFgR4sS8Eh0KUd9sg699yrh/NoUYudXgdEGsPShuZrF7Y5xFoNmNLO5fuW4zxfbdfg/Fbrr1ZxQrlVhQfGiWr9sTnsji5sE+XqQw22RXZeUsnRYiYzWG5nNq2OVrKQyolyQT4ywmLwg33XQP1+2gGPtyYkyDGL8g7hNlIahviYUsPRVu1wcP8ThXLvB9Rzn21sU9IcfDv7o9yUUKTeHsqzTnTXHzVeOHykOtIDCRrqWch1VM7avOWIqbGTVuHZ4YEykXxp/gOI7jOI6TOHyC4ziO4zhO4vAJjuM4juM4icMnOI7jOI7jJI6OMra1oyspJr01hWhICYSUGDMrhLjy1e5iLpYqcLrevj6KFU86kWLDu3fyvsMjFOvv5+MFcR4DxQGKtYQIuincYnvyLL5rCjfKASG0GxlhV+XBDKcrgFWgN36fBcUTE+zcuWbLBopt2LqVYkFctyde8gyKbTr7dIrNjR2m2FKkWhNCbyEoVv0kneZr0BDu1Orap4TiNCWuaSojchZC4aZom/kc75sWQk4lYm8qwfMK7hNDIyw+HU9zvj/dze64N+/hPntAOIhfcuETKbb9uz+h2OjIIMVmRb3UlEGxEHJWRMKMcK7NinaQyfIYYKLulyZKcMr1ODPJiz0y/UKd3sttpKkstY37ztQsO93+/IYbKLalwfeTB+5/gGKDF1xIsZ4edoWfm2PRfr7ALvMV0XdSov2HIETGKW4PqszX/vCHvG+XrtjVMp9HpcIibSVaFsMgajUWfadEu1aLlwYHefxYMcLu+53wJziO4ziO4yQOn+A4juM4jpM4fILjOI7jOE7i8AmO4ziO4ziJo6OKbWquRLG0kE+me1gonBJzp96UsHYUjoiWFcUSzo5Cx4Ybv/oNiv3X1zn2tS9+hWIrh9l59fm/zk7G++/fRrE1p51EsY1btlAsLZRYOeWMKYSIdeFkPDzAQsnx7Sw8+9l1+yg2JkR/uSwLlMuHpyh279RtFEsJcfPmJz2BYg0htG4Jl+ulCLdCoNHi88kJG9O0qB+l/csI4X1DOKUqZ/CWECcG0VFSQvzaEu3QhLtxMCEyFqLDhurGQywcnBBfs763/T6KTRW4/a/cxC7IQ/0svN984lqKpVNcL3smWWRZywsRubhuWdV3qkK0nOFrlBMibauLAW4JIpoNynMzFJs7xAsYds4Jl3kxLpWEC3hLCHZXioUxN9/Abte7Ds9SLDfIC0oO3b+DYg0h9lWX6tzzWIwM0XeyYrGAdAoW42a1yn1i1052ZJ6d5fNVwt6Zqcmu9j3xRF64c+5551HsxzfeRLH+IrtDD43wuLBl8wkUK5V4TtIJf4LjOI7jOE7i8AmO4ziO4ziJwyc4juM4juMkDp/gOI7jOI6TODqKjKeFq2E6x7tkSyx+6u3tpdh4lYVnrRK7JE5NsRvlivWrKZYS7p83fP/7FLv9uh9TTLkkzlbLFLv8L9/G6YTo6tkvfhHFfvMPXkOxZoHr777b7qTY97/wRYr193GZn/K851Bs+918vFkh0oYQt03uO0ixH3yH67Qp3Dc3ncRC63VPOpNiLdGGckUhyFuCZISQVDp5C5G4ugLKoFW5gJuo7yZYmNpssrtrEDnn0tw/m0LI3BJOpE3RlpQ7aRBlwSoWAA8V2J00N72fYqtzLKjPpbkCZ5osvDzrKWdRbHyGx7dqmc835IXburhw1bqoF+E+rnyulYg2lV4e3z+//93vUkyJsFeOrqFYs87XYM+evRQrlfk+MTHGgtiB0VUUmxV9LCdE8RDttbn3AMX6B3is2rOHnbdPO/00Pp7xvdJS7NycSvH4oVzwTxIu/W980xsppu5ZSrQMsUhBObUPiDrIiEUUF110kTieWgjBeYTAjeg73/0BxTqxPHqQ4ziO4zjOMeATHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3F0FBlnGiz86elnYa+lWFxUEYLFvHAonhUWkAfv20mxoQF2GV65gcWJI+vWUWxylsVoLSH0U3LA3Tu4LKeexe68B4R75I++zA7KlQYL2e677kaKPXDn3RTr3cqurRe+8sUUOzzJ51sqsYixIpyR777vXopNlUT9CY3eob0sDgxCtDm6lt1Gq0LMuhRRws+ccOjOCKfgWo3butI6KjlySrinpoUQMQTlisp9ESmhAhUuq0oZnRYC6iCyqAqn2Vq2h2L9Q7yAoCwcWlui70CcW+FEFnfmi1zm3TPsljzczy6re+os0ExJMSYXT7lXZ4TLtRJ450QbWors3c0u6fWGcN7O8L2jJRrOhBB/92VZiDsiFrLUKnytQo3bzY46xwaNF3GkK1wWJfivVlkE/ZMbeVxftZqF1sUBzrc0O0Ix1dZrTR4302KMajaEo7YYcysVXmhTF+kmJ9ndfp8Y//N5vm4q37LItzzHsakpXqjUCX+C4ziO4zhO4vAJjuM4juM4icMnOI7jOI7jJA6f4DiO4ziOkzg6ioxbQkhXyLFQrCEEu8UCi/VqQiTYEpaX+x9g8V+2xSKuPXex8PjQ/Q9S7CnnsotpTYibn/60p1HszjvuoNj6jRsodt/tN1Ps1kPsbpkXwriUEF0NDnG6lavYpXP8wBjFGk2uq/1T4xRr1VgEanluEptP5tfWK6fejBDbVoV4LC2cepVgdinSFMJZCGdfYViKRk0ocYUjblP0JyUS7M2xgK9XiPqQ5eM1WsLxWIiMTQmPVflSQhgduMyVJl/n/hEWGaeFELcs6q8xK8TIKe47O6fYodv6WfA8uoIF8PvGS5yvEHwGIT5VdZoWw246y30ilRFC8CXIpk2bKWZCxD4gxLSHxtixuiwWP1iN21e2xe2rUReLFdRimQF2xa6KdOMHeJHJ6Ci7cff19VHsO9dcw3mI8vUWuB0O9rFTcFPcP8s1Hl+VK3BPD+eREe7j5TIfT7kgK1dltfggn+f5wvr16ym2eg2PAUNDfH8fHuGFRZ3wJziO4ziO4yQOn+A4juM4jpM4fILjOI7jOE7i8AmO4ziO4ziJo7OTcVW4UQq3zqwQHO29906KDW1kl+F7b2Jxbm99mmKzu9kpeOcedoqcnGMhVrUh3CiFUPKnN/HxKk2uA2PtMGrTLOzdcWCCYrk8iyfzeSGwFa6fGeFaeYcQsq3pY0HZvQ2+cLNzLJ7sE+K7bI7dWFtCyLZuzVqKrerj4w32syBvYv8hii1FQp2vVbPCdTtXFo6qUjPK9ZjJivYgxKqVGudRmxGiVtHmmqIsSvytHKZzWT5erofbSCrDxzMhps32cluvHuC2uXvHHooJSTUGL1xBset23EKxjauEq2yahY0H72SBcgM8LvQMcrvuGeAYCizkzOb4mqfTylt96XHhBRdQTC1QEetJ0KjzuLk7dYCPJ/pJS7XhNN/SMmJftdhjbJzHa2U13hKW1YUCt8RqjcXSc2Vu1/WSWJEgxLl9fVzm4hCLkZV4OJvltjQ0zG1dndvsDLsHK0FxViwMUPWySiyW6RcO4qosExPiGnXAn+A4juM4jpM4fILjOI7jOE7i8AmO4ziO4ziJwyc4juM4juMkjo4i4wdv/hnFrvvilyiWzrC46JbrWLDbN8QuhPvFK9a3bOB0a08/iWI7hWALgcV/aSHEqgtH2sMzLEZOC0HlYbAwzjaycKpR4XTlBos2BwrC3VjMPUt1Fq2VdmwT5WMB2M5tOyg2VzpMsS1btlJs9egWim3buY9id4yxsPyUJ51Lsb17WKU9tp1F5C866wyKLTahysK3RpmF7ROHpihWHGQn10KB2xca3Ia7FR4HIcxrCadlobGXLs1Z6VgthMJC8WlCdLj9wHaKzU6xiLE8xgsNclzNWCWcTWer3I8Pz05S7MSN7MB7cCc76z5w070Uy/ayeDI3wOL+1VtYeL96M4ug0eQ6LVeU8/XSo1Li+k4pt/I0x4b6eWxWrrtzJR7rgxhLlfC4LhZY1IU7b024DK8QrsVKOKsWraxYwdd5xagQ9k5z/Z15Jo99qzdvpNjoahbK18TCgG9945sUq9e5Q23dyuO/Evsqt2QVq1aF0HqOz3d6mvt7SVxzJW7uhD/BcRzHcRwncfgEx3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRwdRcYXX3waxa76T3b13D/GYtXRUzZRbK7EottWPwsRZ4xdF+fuZ2FqXYhpzVgQ1RSumpUKu0fOznH5eoqcRyrNIrOUcMvs6WdRaU+KxWhpYfFZF0LTPcJRstFkoVhtJ9dVRjh39mbZPXJ8/xjnUeE8du/jdPfdfhfFfvaDGyg2MsrC0P4iO77+z7/6I4otNq0aC99686IbreC6Vd1NNFekhBtrSqQLQtiYFa7Fyo21KYTHuQL3O9W+mkI4qGxqW8ZteMfenRTbBV5osOJEFneuzrPI0oTT+PUPsjt6/3oWsx4GC4+nhUD5/Gc9iWI54b7czHA9mxCHZ4RDcUUIQ7N94louQaanWFCvRMbZDNdFVSzEKApX27xw9t3xwD0UU0LXQg8fr9kSQn7R73JCKK9Et3NlbjfVae4nzSaf73CKz21mittmaTv3xZkSu2KrBQRT4hopl/JuqYkxYHyc5wGTk3weSqTd18fjv6r7YbGooBP+BMdxHMdxnMThExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdxdBQZ3/ezuylWUarI0UEKpZtCNCRchjecuIFizSYLgK3Kgqh8XQiKhXAqJQSVPUN86n3CadaEoCzVYCFbi0NoiHybLT63XIbnmfWaEA8LYVdaCMpygyzEGhleSbGNG1ZTLJ/iepkVguf7HmQhc36Qr/kJp53MsS1bKDa64tjEY4tFvkvRYSbH9dgUgtim+I5RFwa2qh0ODrBwtiZEwfUKt5uUseCzLDLOsB4WqZRyMeXyVatclmyWD5jOCdfuiujHUtssxiPhyJwSBqj7hJN3rsgCyN4ii1RVmZvgvt0SA0NG1H22ysfLiIULS5HJaRawCq0vslnuEwfHWYQ6NcOutkqYmsqyOBcNvgayGkVfzIoFIK0Gi4IzOW7DtTrnqxyFZ0Rd9Y6yeH7nIa6DMMn77trLiz3UGLVqHd9nIfrsrj17KNYS9+OUcBSulrmuVH8fGRmhmLrnN8R9Z+8eXpDQCX+C4ziO4zhO4vAJjuM4juM4icMnOI7jOI7jJA6f4DiO4ziOkzg6ioy/9NmvUuzHd/ycE/Z250KYKXCsUGShpGWUAJjTZYQCMp3lPNJ5TpcR7rM9PSwyzghHThGCCSfXlLKUFGLR0mEWj+ULLKDLF1kU3BSvvO/t5+uRESIuEUIlLcRyZXbLrNVYUHbJxU+jWFAOt3PsBlwQAu+lSLPJ4rp6nVWMDSUuFca0DSGKrIlYSwh2q8I9NS2cjDPGDbZfOFubEA/nsnz9Gi0WT5pyrs2yMLQpBMCNCveJbIvPo1LiOmgELktaCCALYlyoG9dztcHHqwrBc1q48irxdUGMg6k8p8sKx+MUhIB6CTI1zYJYJRBVdVYR45cSnO4R4tfSHI9LA70sCM+Kaw9wvhD1PTQgXJV7eyh2cGyCyyIWARTE4oOWUMA3RRuWbUTsqxzOZ0vCaVm4SKv7mOrba9etpdiZZ5xBsZxapSD4yU03UezgwUMUq4gyd8Kf4DiO4ziOkzh8guM4juM4TuLwCY7jOI7jOInDJziO4ziO4ySOjiLj6TKLQQeFGHTfOItkpyss4gpCYCtFqEKbmxICtax4vX1KCJSzwvEyLfbNijwyBRZJ5YQjJ4Qbcb6XRZZpJTwWtp854ZaZPizcPIWYddc0C976+lgsl+sXDq1ZPo+qEKilM3weMw0hGCzPUmx6P4vHtt3LrtlLkYxoXyYarBKnl6ssipybE6I50SeUaD+f53adEvkqp+WmyAPi+ikhYkoIByt1TlerCQdUMeQ0A6cLJtyShdNyEG7h6R4hKBaCejVWQLgRi66NmriWeSEgNQjX2zm2ZC6IcUY50i5Fal0KhRvKoV60w4EBvsfURR6tLp3i66Jdq1hGjP914Ubc08MC/SDG4UKO++eqkSGKTU0LAXB5hmIAi5vLQsyt6qUu0uXFfaxXLASaE4sZ5mZ4bnDwwEGKKff9IO53VfGmAjXWKkfrTvgTHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3H4BMdxHMdxnMTRUWR8aIKFTtMVFo0qYZ4JB9RWXbj9CpfVoN5vr8TIIp16tXuzxUKnVIpFVxUhxgSbdELoH9ESQsm0UCc2hVDShOgqIwRgJupKCTRNnIYqsykhp3DzNCGMzgtB3l233E4xJXANFSU843pZipSFg3O9ymVPGddPq8nXL5MWjtVCSBfEdc4IkXEIXN+1Mtd3aZrPIyvag3JeTaeEULjGos3DY4d5X+EqWxQLF+qinltCKDnQz/uK9QNoCPFpsyYEymmug3Sa239B1IFyVm8IdfOccPKGsft4EG1oKaIWhQhNNyDaphiWkM1yPY6MjFBMiZErwt1Yud821LgkyqLGayWcbYl0+/bs5QOu5vNoidrqF4tCVPmqNS6LEmSrOuCaAg4L8bVypZ4Y54UsD95/P8XUWLZixQqKKfFwTw+LqpV4vRP+BMdxHMdxnMThExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdxdBQZ14QYdEK40KZTLByUTqliPqVEskpQ3K3IWAlnLSXEzULE26qLfZU4V7y2Xgl7lWlxRjgUKydj5dxswi0z1eR0TVkHnG1KytYYdWpVJfZSjrmiDbWawn2zy7IsNuMT4xRLGbf/bJqvc0q04XyBhXRK1N1osfivJJzGc8I9tbe/SLGaEAW3lBO1aMOVWV580JPhfIu9Qjgr2nVDuCBn09xge3pZVKrEieUq14sS7Tcboj8JQXEuz9e3IfbNZDhdrcr9JCvGAOWsCyGCXoqYuFYpMaa1RCyIcUQ53SpXW5Wu0Mf9KdfDbSQrrmlZiL97RftqSldl4dxcU+7B3P4npicppiTFe/eyaL+hFq2IcUbFVP0pUbV6A0HF2N1Y3bOGh1lUXSwKAbVauCPyVU7tnfAnOI7jOI7jJA6f4DiO4ziOkzh8guM4juM4TuLwCY7jOI7jOImjo8j4Ba96GcX+/aMfo9jUYRZJtRosEMoIi9GWcLdU6lwTzqFIs6hJzdhaQpjUFG6UShatoinhUCyF0QIlNIUQ6TWDEO4p903h0KrEfFKkLYqi8mgqX1JRPoiyKMGsEnzWqjU+3hJkw7oNFCuXuexNUY/1KotpSyW2ym42xDUVUu9sgYWSOeEUrBR8eeG6W22KdEKMmUtxv8sGvqYDqwYppgTwtTr3xbQaF5RQUtSLKnNdiEDVggTF7CyLTytCPJ/LscAVQYxlYvxQbthNUadLETWkZYUDtnKh7daXNi3GejWOKOFsS7hJZ5QrvFj8UBLjUl30T9WSVNtUzsMbN6yn2FxFOArX2Rm5VutukY4S56qxWTmhqzcLqK7TJxYV9PRyn1ALHNT1VWVW17cT/gTHcRzHcZzE4RMcx3Ecx3ESh09wHMdxHMdJHD7BcRzHcRwncXRUsU2vYgfUrS95NsW238uvSa/MsdNhmBGvdp/jWEaIGIMQdjWbwulQiV+FuE293l7JnTNC3JxSomD1mnkhMAzCuTkd+HxR59KYEHs1hShMOR6bcoBU6aSrsqgDMTcOIo9KjUV6OSFwrQl30KXI4QPsZNxb7KVYsY+FrqGXz3t6Wri7KsdXoa0rlbiP1dMsZE4LMWFF7SsWBqi+0y8FpKJtitjgCI8p05PcF2em2S1ZdCdAtLlURomR1a58HnNlrpdymdtmynjfRorHMiX4TLe4HTRbwpFW2UgvQZoNLntGuNpmxLiZVmO9cthVY70QCksxvrSU52vQI0Sy1Qr3J7m+AlwHEA76Bw/x+FEWQubyXJliDTVGmnC8F/UitblqkYkY65WLtIoph+6KqD/lqpzLifFSjoMuMnYcx3Ec5zjHJziO4ziO4yQOn+A4juM4jpM4fILjOI7jOE7isGMV7TiO4ziO4yx1/AnOMWJmnzKzy7pMe62ZXfow83nY+zrOEby9PnzM7Nlmtr3D9o+Z2du6Seskn6XQ18zs9Wb2vQ77fsvMXnkMeZ1k3b7TZAmy7CY4Zjbb9tcys3Lb564v3PHI0Rq/8+jj7fWRsZQnDiGE14cQ3rfY5XAivK8dnRDCc0MI/77Y5Xi8WB5vc2sjhNB/5P944Ht9COHqhdKbWSaEIEwKHOexx9ur4zw+eF97ZCSxOtZ7dwAAIABJREFUPpbdE5yjYWbvNbMrzewzZjYD4FXzHx3O/1ZoZhvM7EtmdsjMtpnZH3eZ1woz+1q834SZXWVm818Ne7KZ/cTMpuI8htv2f7qZ3WBmk2Z2i5ldfAzn+QdmdreZzZjZHWZ2Thx/u5k9GMd/bma/HsfPAvARABfF32jGus3Leew4HtqrmZ1sZt81s3EzGzOzT5rZYLwtY2bBzLa0pf+UmV0Wp7kKwKa2b+KrzKxgZh82s31mtsfMPmhmufa6MrO3xue518xeZGa/Zmb3mdlhM3tzW14LHqstzTvjsm8zs5fPL+cC5/ywrpHz2HE89LWYlJn97/i4d5nZr7Qd96Gftyx6ov+DuP0fBvB2M0ub2Yfi9v4AgOcfQ75LjsRNcGJeDODTAAYBXNkpoZmlAXwVwE0A1gN4DoA3mdmzusgnBeBfAGwCsBlAHcA/zEvz6vhvHSLvyA/F+W4E8BUA7wIwAuB/APiima0QZTwhbujr4s+vAPB2AK8EMADgJQAOx8nvBfD0+Nz/BsCnzWx1COF2AH8C4IchhP4QwmgX5+c8PiS6vcbHeS+AtQDOALAVwDuOVtgQwhSAFwHYGbfZ/hDCQQDvBPBkAGcDOA9Re39r264b4nNdB+A9AD4O4OVx2ksA/LWZbYrTdnOsYnys1wH4VzM7qVO5H+E1ch5bkt7XAOBpAO4GMIqo/X/JzIYWKOfTANwFYCWAvwXwhwCeC+AcAOcDeFkX57pkSeoE59oQwlUhhFYIgT2vf5mnAhgIIbwvhFALIdyPXwyIHQkhHAohfCmEUA4hTAN4H4Bnzkv2iRDCnSGEEqLB9OVmZoga9ldCCN+My/kNALdCzJhDCNtCCEMhhL1x6PUALg8h3Bwi7g0h7IrTfi6EsC8+5qcBbEc0gDtLl0S317h9XhOX9yCigXx+vsfCKwFcFp/PQQB/DeB32rZXEPWPOoDPIhq8PxRCmA0h3AbgHkQTmm6O1QLwrhBCNYTwHQDfAPDfj1K+h32NnMecRPe1mH0A/r8QQj2+BzwI4FcXKOrOEMI/hRCacX28DFFf2R1CGAdw+dHOdSmz7DQ4XbLrGNJuRvQIfLItlgbwvaPtaGZ9iGblzwVwZIY8/2U77WXZASCPaFa+GcArzOzFbduziAbQo7ERwAMLlOlSAH8RHx8A+hHN5J2lS6Lbq5mtAfBhRE9Hioi+WB062n4dWBuXrb2c7Y//x8IvXvp25CZ2oG17GVG/6OZY4yGEuXnb278tKx72NXIecxLd12J2h1/2f+nUZufXxzpRrmVLUic485e1lQC0vxFxTdv/uwDcF0I4/WHk82YAJwA4P4Sw38yejOhxZjsb2/7fBKCK6OekXQD+LYTwhw8j310ATpwfNLOtAP4JwLMA3BhCaJrZHfjFa9WW7XK/hJP09vq38XHOCiEcNrOXAvgAAIQQGmZWBZ/vkTf4qja7D9FN4J62cu55GOXq5lgrzKyn7dv+JgA/OcoxH8k1ch5bkt7XgOhn1XY2AdirEoLrY58o17IlqT9RzecWAC80s2EzWwvgz9q2XQ+gZmZvjAWHaTM7y8ye1MVxiwDmAEzEv4++U6R5tZmdFs/o3w3gc/Hs+pMAXmxmz4nzLJjZr8z7LXUhPgbgzWZ2nkWcHP9u24+owR4CYGb2egCnte13AMAGM3vota+x0IxfB+8sJklrr0VEN5KpuJ3+1bzttwJ4ZXzcFwJ4Rtu2AwBGzaz92+9nALzTzEbNbCUiPc+nuiiH4mjHSgG4zMxyZnYJokf9XzjKMTteI4uErIlarbKMSVpfA4C1ZvYnFgn4X47oy3C3T38+B+DPzWx9XO63tG+0SKi94Mq0pcbxMsG5ApGQageiC/3ZIxviZXEvQCSo2g5gDMA/IxLvHo0PIhKrjQO4DsDXRZpPIhow9yF6vPnncb7bEQne3oFoQrITwBshromZbbVoBcm6eN/PIPpWfCWAaQBfBDAc6ws+DODHcX6nAbix7VDfBnAfgANmtj+ObQTwoy7O1Xn8uAIJaq+IxJLnA5hCJJ78j3m7/Fl87ElE+pavtJ3vHXH67bGYchWim8GtAG4HcBuiNv7+Ls5fcbRj7UY0OdsH4BOIlh7f1+mAXVwj73NLhyuQrL6GOL8zET0NugzAb4YQJrooMxD9AnANov5wE3gyv6zarr+q4TjHzK4B8IchhHsXuyyOczxgZlcA+GQI4ZrFLovjHAtmdhuAZx7DhGlR8QmO4ziO4ziJ43j5icpxHMdxnOMIn+A4juM4jpM4fILjOI7jOE7i8AmOwMwuMbPdbZ9/Hi8RfazzvcLM3nsM6YMdxTbeOT5ZLm14OWBmW+K+Jn3DzOxtZvaxbtI6yWMp9rX5ZRLbP2pmR31dyrx9lt39Ztl2QjMLiHwGAqLlp1cCeFObg+mjRgjhzGMo08mxpbfjdMTbcPdY9DLObQCyYYm98TiE8L7FLoPTGe9rv0wI4Q2Pd56LwXJ/gnNOCKEfkXPvbwP4vfkJ/JvUseH19bjjbdhxHh+8r3VBkupguU9wAAAhhLsB/BDAEwDAzLab2VviNful2NFxnZn9h/3itfcPOVaaWU/8uG/CzO4E8JT248fHe3b8fzp+JP2Amc2Y2c1mttHMfhAnvzU2XvqtOP2vWfS6+0kzu87Mzm477nlm9tP4OFcCKHQ6TzN7k5ntM7O9ZvbaedvyZvYBM9tpZgfiR5A9bds7lYPq6xiq33kUOB7asEVusV+Nyz8R/7+hbftDZYw/X2ZmR1yFj5RtMi7bhWaWMrO3m9kOMztoZv/HzAbjfY/8VPS7ZrYrzu8NZvYUM7stPpePtOW14LHaeG3c9/aZ2RsXKOf8cx40s4/H++yxyAk2vVAdOY89x0Nfa9vnbWY2FpfplW3xh37esvjnrLgO9gP4tzi+4P1m2RBCWJZ/iB41nhT/fwaA/QBeF3/ejsiCeyOAHkQTuZsR2WXnAGxF9IbV58XpL0fU4Efife5A9MIytB3v2fH/b0Lk8ngqonc8nQNgxfwyxZ+fCOAggAsQOVW+Jj5WPi7HDkQvxswCeCmAOoD3tu0/CeAZ8f/PR2Rb/wQAfQA+Pa8O/h6RA+wIIpvwqwC8/2jlUPW12Nf2ePk7DtvwCgC/iejdP0UAnwfwZVXG+PNlAD4V/78lLlumbftrEb2zaiui15R8EZGBXnv6jyK6ETwX0VvGvwxgFaIXah5EZFrW7bE+g6jvnYXIYfbZRytnnN8/x/utQuQy/geL3faOtz8cf33tEgANRI7KeURvMi8BODXefsWRfdvS/m2ctgdHud8sl79FL8AjbLDTACYQvVn7vQBSbQ3stW1pL0D0Wvj2/d+K6IVmiBvv89u2/X6HBnsPgN84WieKP/8TgPfMS3NP3NguRvQCNGvbdl17g523378CuLzt8ylH8os7TgnAiW3bLwSw7WjlUPXlf96G2z4/am1Y5HUugAlVxvjzZeg8wbkGwB+1fT4V0aCfaUu/vm37OIDfavv8HwD+/BiOdVrb9v8F4OOdyglgNaIXKPa07fcKAN9d7LZ3vP0db30Nv5i09LXFPgfgHfH/V+CXJzg1AIW2tAvebxb7Wh7L33L/KeKJYWGBVvsr3zcDWGf82vsfxv8fyyviNyLqIN2wGcBrzOxP22K5OL8AYE+IW08X+a5D9K1CpV2J6FvxzWZHXhwOQ3SORyvHEdrP33n8OG7asJn1AvgQom+Hw3G4aGbp8PDEnuvm5bcDv5hYHOFA2/9l8bn/GI41v37POkr5NiP6tr2vrV+m4H1tsThu+lrMRAihNC/9Qi/sPBRCqLR97nS/WTYkQoOzAO0NYReipxlDbX/FEMIL4u3H8or4XYjeztoNuwD8zbx8e0P0ssx9ANZb28h3lHw7lXEM0WB9Zls+gyES1B2tHEdory9naZC0NvxGRE9GLgghDCD6VgpEk3EgegrZ25Z+Tdv/qn3uRXRTaM+7gV+exHRLN8eaX797j3LMXYie4Iy21d1A6HKVjfO4krS+BgDDFr2pvD39Qm12fv86lnNcsiR5gtPOjwFMxyKqnlj49QQzOyIO+xyAt8YiyA0A/nThQ+FjAN5jZidbxNkWvVYeiAbDrW1p/wXAG8zsgjhtn5m90MyKAK5HNID+WSxqewmit9YuxOcAXGpmZ8TfhN91ZEMIoRXn9SGL3rYMi153/7wuyuEsD5LQhouIJuKTZjaCtjYccwuAl5tZ1syejEhncIRDAFrzyvYZAH9hZieYWT+A9wG4Mjy8ZeTdHOsdZtZrZmcC+F1ES40XJISwD8C3APy/ZjZgkZD5RDN75sMon/P4kYS+doR3m1nOzC4C8GuIdG/dsOD9ZjlxXExw4sffL0L0m/82RE88PobodfYA8G5Ej+C2IRqQPtnhcB9EdPG/heg33Y8jEmUB0W/xn7BIAf+yEMJPEC1F/Aii337vB3BpXKYagJfEnycA/BYiYeNDWKSuvyhO/3VEQuLvxMf5zrxyvSWO32Bm0wCuRvRtGZ3K4SwPktCGEbXfnrjsNwD4xrxyvQPRt92J+Hw+3Xb+cwD+BsCP4rI9FZFO4JOIVlhtQyQi7nSz6UQ3x/o+ovO/BsAHQgjf6uK4r0b0M8Od8Xl9AcDah1lG53EgIX0NiITUE4ie2vw7gDeEaAVZN3VwtPvNssDfJu44juM4TuI4Lp7gOI7jOI5zfOETHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3H4BMdxHMdxnMTR0cn4zEvX0xKrFYOjlC7d4nfHNatsRVFvciyVM4o1m2WKja4aolg2x/Ozuti3abxSLJREvhOczrKcrrCCq22gf/57+YBQ4fINDA7w8Qr8zrRSqUQxZe9Rq81SbGQF11W9VuM8ZjmPaRGrp3so1gxZilXGp3jfEufbTPG+MzW+brd9+n6u/EXmjHMuokYyOnOQ0qlYFlwX6RS3kaz42pETbTgtYqrCuo21WhxTiyxbYueU2NdEugZ4rCgLD2Nr8gFTPTned/Uqih2qc/tqivOotzjjpijfXI3TVepcvnSO8y1XqhwrzXG+ovJbovInp/YtuT7x4Y+zCaNqN6o9SFS6x+OruMi36zKrw3W5b7d1FYzbSJD+l9yGZR6iTlW+cvzoMl0KYmBocsbdHk+dxx+8SiaN83ccx3Ecx0kYPsFxHMdxHCdx+ATHcRzHcZzE4RMcx3Ecx3ESR0eRcSGfp1g+x0K/jBA1HTg0QbG0EOzWKyy8zLJWDzOT0xQLrTrFCj0s2O0v9lGsocRZ/Vw+E2VJ8ekiZRxsCQFpo8GCRSXuTAshbn+Rhczjh1nEaKKAqTQfL53lazkwwPuWZrmZrFm5hWKzxtd8d30nxQ6WJylWKHJZliLZ8iEOVlnoXRMXNQgtXEZce/W1Q4sOOdit8FgRghDei35SEQfMKZGx6GJjYt+xGgf702LhwhyL7PfsYDH3dIrHrbQ6XpP7YirD40dD1GBVCI9zvXzhWi2OVSgCVBsiqpTRS5BW4DFcvQIopVStqnWK0w5K2S72lf2ky2rUfUz0J5FQnW/3AuUuxcOif6oxBcb3RVVmU/Un8pBF4ZBekCDSpUwsNhI7N2UbUpUqbtJH0i+4xXEcx3EcZ5niExzHcRzHcRKHT3Acx3Ecx0kcPsFxHMdxHCdxdBQZ9+VYcDc6vIJi0+MsGm1UhWtrk4V+fUJRnGtxsVJl3rdRYwlTX46dgjNzLGBNQ4gEhdB1pjpDsbrIt2Z8vHTgMlfLLLAKwn25XOZYoYcdhVUdzEyzU6qllDiLy5c2rvt0Xbg511i4PTDMIujZaRZPjldZMJ5KLQ9B5eAsC6nzQiifaQn3bCE6b4nrooR5TRE1KfVjunUEbYkyC5NhvS83azRE+59Jc/tq9vVSTLkMz05yX5xsssh+JsMFzAmRfV6I7GvC8TudEQJ4IQAuT7ELeE6Mbynh0gzhjKyEq0sR6ULbtStwd+cYxHdxE6LlR+I8rHqeFOdKsbTqtd0RutRep4XQOi9E0KkW3zvqNXaZr1b5vh3EYpm+gZVcmGw/5xG4n7TkPUYIspVI+1Fo//4Ex3Ecx3GcxOETHMdxHMdxEodPcBzHcRzHSRw+wXEcx3EcJ3F0FBkP9BcpNjTAIt7pA+MUGx0aoVgxy2LCjUOrKaakRfkcC2xzIjYwOESxvQd2U6xRYDFhs4djc1Ms2JV6MlHophAizs6yULJbeqbZMTedFsLoKZVOiCwL7PhaLPI1r0xyHYztYwfZtcMbKfaUM59GscIuLvOOsW0UW4rkKyyazirRqApJN1FuI8q0Veh1YWkhstQeo12VTyJ2TYl2XReqyDnR5uaEy3BZCNtRZ5FxEILFjNh1YGCYYg3hPNwUyuhchsvcaLIzLOq8b0YIqFvCubxZZWG0qlNp3rsEUc7u3ftnd0daNtjuGvEjEx4LghL8q3y7y7gZhMu8WLTS1xKLeSZ2UWzfnjsoNjF1H8cm91FsStw78gW+p25YfQrFTt5wMsWGitwX7znM57FjjPNdvW4r57vlNIoBC7vg+xMcx3Ecx3ESh09wHMdxHMdJHD7BcRzHcRwncfgEx3Ecx3GcxNFRZNyTZ0FgvcQuiRnh7Lh5AwtOz1jHwqShCota77rrXorNTnO+K1ey+GlVcR3FCitZjHywuYdiBxp7KYa0EJQ1WTzWrAuRoBCfBvE6+owSNjZY2Dg1yUKs4RF2lJyaYiHzbOkwxU488USKKeH2zBzXS4712OgBC8qG+kcpdsIIi9Emx1iovhRpBRb/1aVQmK9zEOnyQsSbEdrEunAZVl9PcmnlvKpsUYWrsnAYbTW4DStB8axwZN7f4ja8Y4pF2inhZLwqy0OTifI1RR3Uhfi03GBhb18vt/WeIjt0jx1iQX3TWGRsxuerzFhrog1BOY0vE5Gxal7qvB+Z2Hd5uDq383/Ze/MoS86zzPP9btx9yb0yK7OydklV5VJplyXbsmy8LzIYY7DBbdqAZ9o9c6BN0zANw2Kardl7Yc40NF7AjMHGBsZ4vGFhGy9aLKm0VJVU+5aVe97MvPsaMX/cK5zU8+gqJdnWrfDzO0dHVW/Fjfgi4vsivrz5iyc2u7tJw765vvQ41I4d/RTUyktPQK1awtTi5cICaR+OsV2jKPZujeB4Gi7idgvH8GGeNQ/Xl7vuBqgd2nEV1FJJfGsCfaqgB/oGRwghhBChQxMcIYQQQoQOTXCEEEIIETo0wRFCCCFE6Ohp7LCASr+NglyUrGZsBOXS7aPTUJv7+nncyApuo7aMItbjJ45DrbmCMtrVt+6CWimOMmGJpDiWGyhFxlnyZByPwWB2EGrVOsrSjQYau4kEipceEUNrJFnXI4mqQyRZOhbDBEhH1LhEDpdLtrFWL+I5mjmDctvkDjwu24awb/QjTMQlzrm1yHLMI42TnzGCNo689DAes6EdmAI+snUSasOjW6GWIAmjyRSOiRbpm/k1lNhn8stQixNxvDW7BLW5eewj6w0cJ1ny45gj8vXaCgr1LSa9kgtcs4X7W2+hUMykV8fSq0n7yHMGFI8I4/0JE9ufvRTMpPiAWcvfAWhb2HL0s1iLkGJjHfv/Q1/6S6hVyo9B7cA1eN1Mp3ZCLXB4PxmZHscG5nHcpfIo2deq+ODJShSvH2OH7oJaYnwH1PwI3vPZQzrPtFvpGxwhhBBChA5NcIQQQggROjTBEUIIIUTo0ARHCCGEEKGjp2RcLOBrzaNJTDfeOYqJuC/a/z1Q8+dR4t1K0ohTadyGl7yADVxGYXFhHaXl8UXcxuA0Jigz4bNN0lhbRtrn8FCiimwWI+mplShJNvWIpEpSlWt1FHu9GO5HlGhwUSJAsjZnhvD4lUhacmYAxeN4E+XYXIIkUI9uIVvuP1igMKsxosQurRF5MjOO8t/ovgNQ88YGoBZksbb39juhtmP/Iag9fvI01JaW5qE2tB37f6aFfTg7hyLinipKjEeOHoHahccxydXP43iP1HC78Qixh4mdWCdtaRAJNErSU9nDFo0G1tiDAREyFts+MZSvkB8/IxHWUCJXMxV3k/IwTeP+FrPZbbClmDzsk32LkL65uoL3tsXZk1C77XYcsy+/86VQK63itfmRw/dD7dGvobScCFah9pJDeH+3Cr6pYGj0FVCLbcM3C5DnFiyKt1lq7T/TfnCFDCEhhBBCiM2jCY4QQgghQocmOEIIIYQIHZrgCCGEECJ09JSM0zFMP0wZSrIvPPhiqB3aeRPUGmmUllM7MYm36mNyaOLBh6BWf/RhqPkkf7lWIvJflaTFerhvWZIoXCqRZNMEmSsGKN16JM42aKPsyGKk20ZERLIcC0CNERGw3ihDLRLBbWRSWaitk7RY89AeCxwe+2IJZfMgziyz/qNJfHDmSTIXji3nJ1DYzV1zDdS+fAoF4JVv4HiKVFE6v+delAlf/4M/CLWPfOITUDt+HNPC9+zeA7XcAEr7Z8+dg1osTlLAY3gZSmexz3lE7ywsYApsu4l9OBIlDxCQ5TwPB1SEDKhWG0+mT0ThaJTlwW8Oz2PKf//h0bRfkm5MPssUY3KV+478JM7TiIk8zFrDxruP18MIkd0bNUwBH8RnBSxo4voeuf8+bEoDJeMdo3hUb79mH9S2YsC5DQ/gPXBhHdOSF+IoHi+SZPAEOess5T0gPcG5Z9YT9A2OEEIIIUKHJjhCCCGECB2a4AghhBAidGiCI4QQQojQ0VsyNhT99u3ARNUbXnAz1LIeJtgO7ZiAWrOKQpRHRMSWjzJmpYqy0tkzmGTcqKLBFMPNWs5Ds6tqKOK2Sc2IGF0jSamuSWwqkm7pB2izxmIoLTuSeMxkTKY6rq8T6TuehFq9giJ4tEFk0Si2byCL1lpuBPuVRfCz/UiTBq8yoZKIly388L4DB6G241oU9O8+dhZqtSaeg21jY1CbmV+E2umzuD4vjudgrYR9/cxZHGNxknDOzMHFlRWoNYmcOzGK+zGZQdkxO4zLreVxG9EE9uv08AjUqhXc3xZJLeZiOZGRScIzW47JyAMDeA3tR6icS/Vhshy7HD635mwK1roIO6mkxkTXgF2vPXxwwm9i/1qcw9RiR4TiFnk4oxXB5aYnsN+84JpdUBsdxIOfDvA+FtTxGExNEKE+hinIlQJe66sBts+xNOwAx8QzDbTWNzhCCCGECB2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8m4UkBJduvoJNRSJAE46uPcKZVCibdNJDz2Svm4h7LS1MQOqM2eW4JapYLprpkGrm8sNw61Rg3FLiMBozFjQhlKvI06SmHJHJGHDY9L1FDkbDXYHJWlJaMolozgeYu0SZcoo3gWb2BbIgERTX3ct1QOl1tYI9Z3H0I8YWq+BT4umCKJ0C+7681Qu/nVr8X1ZUeh9tj9mO79vXe9EWrLeRwTRx4/ArXp7ZhE6r7xINQSmQwuR4TieBLPfb2F42RyAh8+2LML05KLl+agNhDHPjxEFNJUDq897/3pfw+1SzMzUFsgaclfvPtuqJ0/ewZq0WjPS+w/s3XrVqhNTGCtH2Hp55tTjJ+C5/ThzUFl1U0+QMCvuHgMElF8OOPoY1+E2tIsjsWhYezXWyZQip+awOvC2ChKvOs1vJ/UmjgWczG8uWWTKPez8e418WEGr1CBWiSBD1ZYHMdnwKRvqrQ/NfoGRwghhBChQxMcIYQQQoQOTXCEEEIIETo0wRFCCCFE6OhpwNVJmmK5ikmMxdI61FLZHH62jtJytYGiU4KE2taInHvxzEWoDQ8OQW1hAeWnSAOlsHgDxa7xOErV7SbOC2NEtGsHKJk12XvhG7hvTSPpwQ7l66HkNK6uikKZR804Ip4topDdXsN9a62jeJwYx3OeTqHw1iDpm/kSnqN+hEnGPMEWj9n+a6+F2vTNt+AK0ygj//A7/xXU3vSmN0CtUESpz7uIUvCFOZRpJ3J4/kaGicQ4jlJwOo1JwTVy/UhmcBtTO1FunpreCbVCFLexRgTqg9uwfYeuw+O8bz+ejxtvfCHU8vk81C5emIXa+TOnocbwyGBkQmUyhXJnP+IiLAH4uazwOXyWbHizqdMsedsC8gBBQIRiD6+l8xcehdrjhz+JnzVMAL7qqn1Q27kHr/XWxrbUsSk2NIrCOnuLwBK5z1bJQ0QTWzFBvLlagNr6Al5nvMndUHPxzaV2M7m5F/oGRwghhBChQxMcIYQQQoQOTXCEEEIIETo0wRFCCCFE6OgpGQ9sRdnx0soFqB09g0mM6f0oDfkknbfVRvE4G8fExl1EsJq/gCLikUcfgVoihZbZ2CDuW62GyxUrmOy4Oocip++jLJ2N42eHtmBacjuCwm7MUPZKOWxzOkBp06tiW0prKIBV1rFWWkfxbGEeJbjp7bugNpZDIXvrlimoXSo/BrV6+8pIMm4zA5JYjD6RGFODKOxWfJQET57BRNwtW1DqG92C42QghkJgoYBp3G/7kXdAje3algk8f9his2SKpFhH8OcnL0pSyhPY5kQC15eI4OXK+XhNiSdx3E1OocicSOJ2F+ZRdj969HGoreZxTNCfF5l/S2rVGj5UsLKyQrbRfzBhlyYFE2hYLV9ys63ZVFtYm4lPbD4Rij0f+2Gzhtevw/d/AWr5heNQmxrHhwB27sZ729Q0jsWxMZSHIw7bt0ZE+YFhlNhb5A0ElxaxH9byeA9cmMWk8WZAjGcfPxt3JA2bnbdn+J2MvsERQgghROjQBEcIIYQQoUMTHCGEEEKEDk1whBBCCBE6ekrG03tRGp2/iCJR44kH8bOTu6A2lBiGWrWIAmQuh/JfubwGtfV1TDH1WygtT05iunHER7G3uory04knUPi8SLbr+ShFTg3jdqM13Le6j+nBLSLLLbbwWLXbKI+1GrhvjTIel4RHhM8YxkinDCW4qVGUvicntkHNojiHPn3yFNRKhgnZ/QiTh+lyURxaDz2KcvWBx09AbeskyoT4sbylAAAgAElEQVQLJLG0XMF+Uy9jH94yguNux+6roJbJYFLwwUOY9tsklnGbtI8l9jJYOCkNlSWfJY62tUjcdK2Gx4WlB/ttlCKnt+F10Ivg2GEtjJDl2DaKBRT+q1UUj/sR59h5ZmfrucjIzyUamUHaYnjdNCIZx8nYvnQR5eGLM4ehViDn+babXwS1rZP7oZYj9xNH2lKvYL9xRDxeIaLwAHkTQG5gAGoLs/NQa5bxwZhoG69RtdIlqGWGcIz5AXlwQZKxEEIIIb7b0QRHCCGEEKFDExwhhBBChA5NcIQQQggROnpKxvk1FFhPLZ6FWsrHV6LvvYQS4wv33wm1YgXFpJU1TIVMedjU8a1boJZNYYqpT9IUL8ygLH3/0aNQOzePy1XJtDDqoQR3cWkZamkigfpEyPM8lOAiMRQWq1WUh6sV3EabyZNTKAWniWQ2NIKy19gEJsOmd2Iy5oXVk1C7uHgRatlBFJn7kZq/OSsyEkdZe2wSj/faGibiDgxhCniU2LTtNpFaiYuZJymmjz36KNTGxzFpeXQUa1GSPOwR2bHVxLHt+9jA5yKQbvazjknB5JgOD6GQ/cLbb4faT70XV/cn/wPP+eEHH4Baq43HJSDXqGYDl+tHSGC1BSSGlsvImyMgsu9TLLmJCg3ttiAgD10wUd5HwfbwQ38HtYvnjkHtxbfhPfCmW14DtWwWr6/FwjrUlqrYlgy5By6ShO7BHPb1hWVczkXxaE2MY4Lyah7vE3NLODfIjOJ9gsUWUwE9wvrBU/crfYMjhBBCiNChCY4QQgghQocmOEIIIYQIHZrgCCGEECJ09JSMC3lMl01nUAadO4OvU3/sKApWr7r9zVAbz2Jqq2tjomSamGwHDhyA2vnTKLU+9ugjULv3vnuhNkNErCbZboMIu40Gir01H5OHS3UUqKNE0EwmSIoj+sRWqzehxpKMmaRXLOEKR7ZMQG1yHFOLt0yh4N2M4v4eO3c/1BoBHqtUArfbjzDJ2BEbzjMUFlukxgTgvVdfAzWfyLSVGkmnjmBf8pskKbuN/SaTQXk4m8viZ8kxiJEE7BgRrUlXpynIvo9jLGCxxQxyPgKyDeYnM3HbI8f0ltswfXZoZARqv/c7vw21b9z3NahFSJtbLZKs24dQyZiqvXhON2sAO1YkJ5Btlz8WQJYjomsyhm0+9sjdpPaPUNu1fRfUDh26DWrxDCYFX5jBtN+Iw3GciGO/Xl3FBxc8cpKqLbwGrBGRuVzGWiqLbXZRvFa0k2NQS2aw5kXwWtEmDyRIMhZCCCHEdz2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8m4WiRinqGIuDSL4uz5FCYALy2gxHvDNddDbX1lCT87j9LVOkldnJk5D7VjR49AbWUFP5tOYgIkCRi1ZhPlrIiHAimTLJmQGifLBT4KhiwFdiCNMnIpwPY1m/jZagXPGxO8x8cx3TiRhJLNz2NC8Uoe+8HgELa51kLxuB/xiPwdiRCN0cNzWq2hJLiygpJxrVaD2sgYSt3VGp7nGhGPoyQVu0069r333AO1ffv3Q+3aQzdAbWkJx2x+Ffctm0ERcXQM05JTKexg0RiOzxYZiwwuFOP1rUVqzQYeq2oVz9HEth1Qe9VrXge1I48+jNtt4gMdHulD/Ygj8dluk/YwT6JmIv8mE7CfQyp2hKTRr62dgdqX//FvoBYP8LuCQwdeCDWPpNGfPo/3p3Sc9HVyjayWC9gWDz8bi5Frbh2TspNJbF9xHR8iqs3ieC+3MKF4YGIf1FIZvJaZw6mI84hkT85RL/QNjhBCCCFChyY4QgghhAgdmuAIIYQQInRogiOEEEKI0NFTMk7EUAg8/cQ5qM1dwOTEfeMoVBbWUYgqVbC2MD8LtaWZC1BbnkGp9eTJ41CrNlAIHB1FcbZKRFwjIpZnKEA6D+eKTIocGsbtDg4OQm19nSVKomSWIIIym7XW6ng+KEQ+TSaxmwQBHpcKSWnODaJ4RgJ9Lb+GKcj9SDqFkj1zjCMJ3O/cQA5qq3kcOwsLC1BzUexLg4PDuF0id7ZI/z/yGIqNf/ahD0Jt3z5MVf7F9/0a1HI5TDb9h89h4uvFizhmx4hkvH37NqjdePONUNu9axfU4sSAb7WwX5crKGS3GiQZdm0NasUSSeMmqc+jE5jQPTSM5y2/iG3xojyDt99g/Z/BHrB4Lp4wFY+p3Ey24ZOHQiJ4Th87/A9QO34EU/pvPHA11CIO+9yFMyfIctjmoUG8ziwv432xzCTjGPb/Rh234Rv29WyGPBhDEp4LZbzLjE7dBLXt0zdDzUuSdHSStM9EdXbOe6FvcIQQQggROjTBEUIIIUTo0ARHCCGEEKFDExwhhBBChI6ekvHCRUyhXbiICYbJWAZqo4Mo02ZSKEoWSyjw1RqY6tlqoSjZbqPo6vtYG95CXu2+huJUqo0CU7KKSamVKm6jVEVJMJFEYWt8HFMcPSIo16t4DPw2STb1Uc4aGcFj326TSGZi86VIUm8qgcJblgiVpRVsc6lIEopJr4sw87gPiUax8ZEoaTtJLL10CVO2F5dRMm40UQh/JUnEZeJsLoNjMRrDNp88gTK+38Lk0FMnT0Ptc5/5DNTecNddUNu9ezfUAiKxH3nsEagtLaFQ+ZV/+iLUdhLJ+HWvfz3UxrdshVqhiFJ8fgXPx4XzeN5m5+ahViZJ1eUyPiwQT+A4ZgnKLtLz8tw3MHmYESHXuYDIpQFJbOeJx7g+Jjwz8ZgED1uDPCRx9OH7oTZIxtjY6AjUZi7i2Ekk8eEDj6TgL8/j/a5JkoxbbbJcE1P6WWqxOTwIPhGKvRg+BLN7/4uhduj2V0BtYHw7bpZcA+IO+wHpBua3ntl9Qt/gCCGEECJ0aIIjhBBCiNChCY4QQgghQocmOEIIIYQIHT0ttnoFpathkrp76IXXQ+3FN2GCYSaLct3KEgpRS4uY5FpYy0PNi5E0RSLYJuq43UwWJdm1PO5v1KHElY6jdJsgr7ef3IWC1bbpaajNzszgdslr4dMJPF1jo5gCmyWJuZUyCsCjY2NQm5hAGXPLKNZcCo9pvYptXl0m4jGR+YYHx6HWj3hEEmQp1kw8bjZQWPeJcPeN++4jn0UB+HWvfwPUWoN47qlkSaTNm2+5FWqr+RWoHT78Dahdd90hqDkikLaJQD00iNLm+Dj26wtVTLv+wudReF5bxTa/5I47ofbwQw9D7dixo1BbyeODFWur+HBEgaSPj5KHCtJpvPaYw7Hd9q+Mnz8j5FplNIUW+wNbziJkOZ+JzFhz5Gf2gDx4kkjj+Dx/EWXyJLGW3/r93we1I4dRRp65hKnd2SG8f7LxmYjjwx6ZNN53mm3sN2sFTDdukod0zOG1bGjLHqi98KX4AMELbkShOJbB+4lPpxjk/Do8HxEiQTfIMyu9uDJGkBBCCCHEM0ATHCGEEEKEDk1whBBCCBE6NMERQgghROjoKRlftW8n1OpNFAJf/8pXQ216CNODY3GUixpVlCwd+roWS6CE5JEFkzkUsfIVFAKHhzF5krlPtQrKWVsnJqAWzeJ2x3dsg9rU1BTU6jU0p1ZWUGwcIoL3yDBK1esFlHgntqDEOz6OtbFx3LeJrZNQqzg8bxGfJC238Rw5l8T1lVE+7UcGBvB4O5JE6sWxtjWH5y89gkLg8iKK94cPPwS19VUU79/6w2+HWjKFfdMjEvTw8DDURkaw9tDhB6B26tRJqG2dwL4eI6nKk2RMlEs4JkZHsb8eOEBSYJfwuDxwP4rRn/k0CsqzMyiGuihJd42Qnw2JL1tvoOCaQrfT4nEcE46m9/YfkU2Hy5L9YftISjQsmcURExxJ542RS9XaGib3X7v/GqhFSZtXlsiDMevYD5fXUICPeNi+AXKtj3lETg/w+up5eO8tFPABh1YUaz/wE2+G2jXXoqDf8rETN8q4b8V1rA0M4b231SIp4Gt4HawX8UEDs+8htQ76BkcIIYQQoUMTHCGEEEKEDk1whBBCCBE6NMERQgghROjonWScQAlpbBKlppXq41CbW0J5cjG/G2o7Rg9ALTWE0tV6CWU9L4FpjyszKCdWiMSbG0CJa5Skp5ZKKD/li7hv27dgYmmrgdby6ipJS/aIdFvCY99uoWA1NILi5Z49KMblsphwmyCmXTqOIlu9XcW2kKnx9mFMbj6bOgO1GkmkDTLPMKLy+YKkazbbKFz7GDxsXpukUw9hn9sSJfJ8As/VzMwFqH3w/e+H2qtf+xqosVTli+dRsM2RVOxSCdOpz5/FFNhsGmXHNjtWWLJYDMdENIrHb9s2fOiBUa9jH87lcN9i5Dh7RDIOfGxLs0nEew8vsQFJ4E0QEbxexTb3I8wxDojYy0RhdixYidUC8lQIy0omXrs1qtiHh0i6/azhOT382BGoVVq4HN6xzGpNrLZq2Orx7XgNf+Nb3oXLbSPXejLuCssoQZ89gw8uzF04DbUTR49BjR3oKCmWiGQ8OowPrZAXAVgihteoegXvvZKMhRBCCPFdhSY4QgghhAgdmuAIIYQQInRogiOEEEKI0NFTMi7WUQZ16yhTxeMoTsWInFUvYSpwLomSZSqGwl2xhm3xYyg1RYgkGI/j+hIxFJQ9Dz87NoFJruUSyn+joyRleAumAheLKBlXSlgbHEB5MpnGGkufzRCBen0d05yHBjGVl2Wn+gFLoEYBbB8R4/IlTGTOH5/H9flXhmQ8QmTydSKde8SaixEpPpFBIXCYpICvJ1G6jRNJ/NLsLNT+9hN/B7XxLWNku9iXWNKs72N/mLk0A7Wdu/ChgmYT+83xE8ehVljDMTFMUpUjRPpm7YtE0FJNEbE3IPvbIgIp2YT5RCIfJKmtzJhtEOm73WbKbP9BDq2ZY1cSImuTK44jy7E0Yp+ceyP3nWgEa9UC3k/WSBrxiZMoFM/No4xfrOA9oVbH8xcnD3G86EW3Qm1wDEVcRyKjG028p+bJAz7rc9jmSA3b3CigFOxX8Fg1q3i9LpCHeRp1bN/KDI73OEl+z+TwmlevM3X7qdE3OEIIIYQIHZrgCCGEECJ0aIIjhBBCiNChCY4QQgghQkdPydiRuFom/5EQQhsaRJmqXMLXx8/Mo5zokWzMuQUUU6/eeTXUxreh2JtqYwu3bd0GtWIFRdwiafPoGBE+4yiyDeRQYqxV8fi1GihnjQyjaDo6hqnFqTiKpsvL+Jr5RBIF1+wwysjRDK5vanoaarUyRvWeOYOpxbvG90Btdg2Ft+OzmIbdj4yOoWTsR4joSmRVR5aLRrGWSWO/8Yhg6Dwy8qJ4/uZI4vHaGvb1PbvxXLEE5bExFJQXF7HPreZx7PiG/eb8Oew358+dg9ou0r6A2L5Mb80SmduxaF2CT1KLPQ/PxyCRtLMkCXptFSXQBhGZIzTSt/8g3fq5QU4g24QjDz9EHdacj6Lr6WP3Q+3oQ1+F2sIlTOheJ/3aAuwPW0jK/IEDB6F28y03Qu3Y4yg3f+7jH4DaKOlzmTje1h2RkWsNFHabZDzFSBR0mTwYUyeiPJP7S3X8bIMkPNfO4/oaNaz1Qt/gCCGEECJ0aIIjhBBCiNChCY4QQgghQocmOEIIIYQIHT0lY89HccojqcDFEkmhJX5cbQ1FonNxInGtrEOtUkBJKjuIkuzkMErGUzt3Qm10GBOUyzMoP1kUjbdaG5Mdq6S2vDoHtQpJh06k8TiP5TBlOJ3BJOOACJBDwyg2Joi4Gk/h6R+ZQoE0R2TkwRx+tk0ks+p57Bs37LsJapV2HWr9SKlYhtrgIIp+6wXsw+02SQEnP2LESQpyM0VSckdReHYkQZYlzTaJYFgltcBQOlxfx31LJLB/tX3cX8/Dxkxv3w411teZ3GxEFI6Qn9tiUTymTIweG8VtxBPkMskEZQ+Xm5/Da4AfYPvSWRyzARFD+5FNS8bUmSZFJhkT+TVKTku7jdebB+77HNTu/eInoVZaxQdZkiRVf9fOq6A2QlK2czkU26Mx7IcPPXAvtqWEYyxKZOnFORT0GUyoT5C2sBRw9j1Ifh1F+WaLnCPyEIyL4wMvLkYeqiFjcXJqK2nfU6NvcIQQQggROjTBEUIIIUTo0ARHCCGEEKFDExwhhBBChI6eknFlHYWtsUkUf+pNlIvWyvhZ5pidm2NJkSQlsYgSaoQkLHoHsZYbQ0l2rrwMtaXqKtTWAnylfKVG0pwDlILXFnDf2gGKl4khlK5yY5gEHfNQ2IpbGmqROCYtNw23u1hYgtp4ZApqa3VMvc1GUYrc+wJMlp5ZvQS1+iqeyxSRVPuRUgVFv/EcHos4EfjqDdzvNDmnLME2GkUR3Y/iuMvlSDo16Q8tIqevLWF/SJOw5DxJco3HcD/OnDkJtampSajt2oUJxQcPoYieyaJ4HCPHmUmqly6gjHlpBhO1k8nNiZelArlGNfH8xpPY5nga+4vvk581r4wgY7MIOeCkRIOjSTFKrOWIj9fhZgPPwZHDmEb84Nc+DbWhDCZqj2bwwZNWC9vHUnebVXz44PwyCuYsAbjVxrY0WlhrkrTrah3b0iZp5kZSz1MePhTisU5HBP1YDh9w2LUNx/Edd74SajuvuhbbksH7XTKJ17yjj92H7euBvsERQgghROjQBEcIIYQQoUMTHCGEEEKEDk1whBBCCBE6ekrGAXpOtrqMwml6CEVX5+HcKZVEmbZVJOmppFVzeUyZDIiglsxgW06fOQW1tTzuxzB59XyDpLueO3cBastllOAyKRRnPSJGp5PY5oHUCNSGhlBOTNZQII3G8DgvlzEZ8/xF3I+LeRTjtm2fhtqt190CNS+J52N4ihzTsyi3Ndp4nPuR9TIRzNfxnGbTeE5rFRQRK0VcXyq2uSReI8ux8NvhLfhgQK2CQuypY0ehVl5F8ZiNiXIJ9+2L//gFqA0SCXpwGPtINovJprkckYxJUmqrif1rYX4WakuLC1Cr1vAANpvkQkiIkvMRIRK5R9pnDvtQ0Gapsv2HI1KwRxK1Yx7WWg18GCU/iw8mXDz1MNTOnTyC62tiP5wYwnNQLWGifGEd7wn5NbxuruZJSnmA+8b6TZOkmSdT2NeTGbz+b92xDWrTu3ZDbWwSk8FzgyhQO9LmwSFcbnB0HGrm8HqUX8KHD9otckFyeM9iye9/9ZGPQO1Ld38Kaj/33h/CbXTRNzhCCCGECB2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8k4TUTJoVFM7M1XUC4qNlD2GtmJMqHFUDhi6YyT0yhYjRIhamEVJdmVJZQJ63UU/a5KH4RamSQoP/jEcVyfQyFwaBCPX2ENXzN/1TVXQS1TQ/FsKIrHL0YSiuNRnLfWWnhMj5/C/VgsYJrzxK4JqJ04h5+945aXQm10FGW5aBbbF09hamU/4pMRs7yCqdgRw/1ut7DPFcjxjqVQEvcSKARGPZS640TkjxHhf2kdx2ySCPDVyObOi0e20WrhmFhZWSE1FJm/I9BoXbYYLsfEWvPw+Dly/Kg8HGEiM4kD7kMSEbwG1ct4npfm8GGP008cJrVHodYo4hir1fAeE0TwXBXLuNx6Ea/DjTpe69tEqC9WsRaJowC/42pM7N1/8BDUrj10K9S2TeE9IUfud5ksPnjiSJJxi/S5lo/76xxeP/w23qPNx/6aTZyA2gP3oRT85S98EGonT2Lq+dnTeI+JR59ZvLe+wRFCCCFE6NAERwghhBChQxMcIYQQQoQOTXCEEEIIETp6S8ZDKMhNTWGqYfEkCosJD6WmahXFs3SMyLRMTI1iUnB2BCXeeBzFvPgQpi4uL2JqZTuFAlMDPSyzNK4vQVKGc8NYO3EWRbvoHJ4GD98eb76hpDqd24nrq2N65GoLZdbxnfjK+/ocbmM+vwi1+x96CGqPPvw41F7+ijugtliegdpSBZOq+5G4h2OiVsW+uTiHsjuTgiNE7C2v4TjJkpRtn0ityTSmZ/tEMCyvEck4QRKZc9iH/YAIi0SgrhNpk0mggU8kXttciu/mNdxnJiduJMKSesl5ixC5k8nIfpsIxejpWnCFSMb3feGvoDZDHkIormCadKuGCbaFNbze1OooCtdJInStgbVqA493o4UH3CPnL53F5O00uZ/svXo/1LaRlOHVdUwuP3bkAagtzJ2Dmu/jmGBvDCABxZYncn+5gqJ1NIb31EYL97fRwGtes47XvEvnURg/exoT9Bs1bHQmjsJzu01SwHugb3CEEEIIETo0wRFCCCFE6NAERwghhBChQxMcIYQQQoSOnpJxbhiFo1QCxZ9MkoiNAQqG8TjKSpUKykrWxu0WS7hcvVWB2tR2FJQrDRS7koPYlsBDgSmRRTlx1zXTUIvg6iyexATIqR1TUCtVS1CrN1AUrrbxGMwVUNzL5TBtei6PYu98AROea4ZtKdRQyM4O4TlfI+ml9z34NagNbcPzu7KKbelHIk0UFtkgatTxXLWJcBppYW29RbZB5OYggVIkk1UbNWxLtYRyp0/2LR7HbURZYi9u1gJiOzJR0vex0UxkZvLkZtVhthzxnWlqMZOMI1Ec8NE4jgmPnDd2nI0dF7bDfcjXP/9nUIuQ48jOvZHzXCjjta9CjlmNPAFSJsK/F8V71ujYJNTGx7dCrdXGc5DNYGoxO1VPPIqCLUs9ZwyOkNR/Ip23iWidJ9tYWSaScRGv9YHDbTQMj307IJJ2hFwJySmPxXDseGQ8Ves47q6/8XZcYQ/0DY4QQgghQocmOEIIIYQIHZrgCCGEECJ0aIIjhBBCiNDRUzLOZjFOt1hEYTeTQWk0FsW5U5QIi+uLKDsGRKaNxvCzHhGiqiWSbkmEz/EJTGTOpFB0qlXRkhqfwv2tk4TKRoCy78gWfOV9fZ4IqQ4lriYRt9fKS1ArEim40kahbIGIx14cpciWwxTRaBrPR4bIY80A9y2/jnJ42xpQ60faFRQgWQJwi0iW8SQRU2N4vNtVIvuSxNIW6f/1Fh7H9TVMsa5V8Jy2Sbpri6QMR6LYZkeMXZbiy2zMFktGJunLbSIj87BfUiSfjZDlIkQKZoK3F0NxNUL6P/sJ0rHjR3at1rwyxkSCXDOKpH/lV/G6VCJScJlIxn6dnFOP9KUI1rZN7oHaW37kPVDbvfsFUGPycILI/VGSbO0ckdM3OSYaJAW8WsXj0t6ktL+6isnlDz9wD9Q+/elP4HbXMGU+mcT+75FrXovI4c2AXN8SKFW/4uXfB7W73vKjUOuFvsERQgghROjQBEcIIYQQoUMTHCGEEEKEDk1whBBCCBE6ekrGTOpLEEEuncZkx1iSCFYOxaREgkjBNRSsAsOa84joF0EBLEbaXCyidJUbQHk4G0dxqtHG9dXLKE5VGijaJZJ4yIdGc9iWETym+QImVAYNEhVJ5q21AI9zNIf7VigS6TuKsmMyhamtzQrKw14TBbpmBc9lvXFlCJWFdTwH6TQeCyNyXSzJUoHxo20fz1WkjWOxTZJcvTSee5bO6zsiN5PxHhDZl+UCO5JuHCPppD7ZD2M1EslM94NZoKTmIkQUJuJ2hIihLI04QtYXeQ7JwwFLSybb6EeOnTgPtWodrwXVGpFkSSIuSzxOEImXJcX7hsfsRS99NdRe8ZrvJW3Bcdxub+47gOcSOk0TumnK9ubWt9nlbrz5pVCbnN4NtXvu+QLUHn34fqiVSyiRT07tgNrO3YegNjqGIvidL3sj1PZetR9qvdA3OEIIIYQIHZrgCCGEECJ0aIIjhBBCiNChCY4QQgghQkdPybhawZTJeBxF3FIJZdqhQUxBLhG5dHh4EGrlMoqu1RpKwSx1tE4SINluMrErwpIYicjZbqMQm86i8BZPowSdyWKtWEc5a22dpBFXMI24WsT2xRMoy1VJwm2ZJDxXSHqq77FUZZRoSyVc3+IS7odlibhNEo/7kXYdRckyEaSjRFZtk8TXwCc1cq6K+RWo1Vp4DqxFEoDbuJxHksGbdZSbiU9MLUaP2NJxklxORWEildbreAzaZH8DMpCZFNxmkjapbVZkZm2JE1GYrY9ClmPXt34kNoCJ7cU8Od4OzykZJhYhqdi5YUye9x1K7Pv2XQ+1V73+rVBrGV6vq3Uynqjty4ziTZ5nAuvDPKH72eNI+6IRvBfd9eZ3QO26m2+B2if/9mNQGxrEe/kdd7wcaoHD/nL44SegNjy2FWrxBLa5F/oGRwghhBChQxMcIYQQQoQOTXCEEEIIETo0wRFCCCFE6HBUcBJCCCGEuILRNziX4Zz7C+fc+za57Fedc+96ltt5ys86597tnPvSs1mvEN9qNCZEmOiH/nyl4px7lXPuXI9//1Pn3C9sZtnvBH03wXHOlTb85zvnqhv+js+wCRFyNCZEmFB/fm70w8ThqQiC4N1BEPzm892OJ+mZg/N8EATBPwfodE/iu4MgwLd9fXOZaBCQUBZhZjo+YUBj4luLjs/zi/qz+E7Rd9/gPB3OuV93zn3UOfeXzrmimf2ry79yvHyG65ybds79rXNuyTl31jn3v29yW6POuU93P7fqnPt759y2yxa72jn3gHNuvbuN4Q2ff4lz7l7n3Jpz7mHn3J3PbFfdH3Y/e8Y595rL9udTzrm8c+6kc+7Hn+b43O6ce8g5V3DOLTjnfvdb1EbRB2hMaEyEie+G/uycu9o590Xn3Ipzbtk592Hn3GD336LOucA5t2vD8n/hnHtfd5m/N7Md7pvfeo0755LOuf/mnJtzzl1yzv2Bcy6+8Vg5536+u5+zzrk3Oefu6o6VvHPu5zZs6ynXtWGZX+62/axz7u2Xt/Mp9vlZnaPnwhU3weny/Wb2ETMbNLOP9lrQOeeZ2afM7Btmts3MXm1mP+uce+UmthMxs/9pZjvMbKeZNc3sv162zI92/5uyTpzlH3a3u93MPmlmv2JmI2b2H83sb5xzo6SNu7sDZGpD+cVm9piZjXbX+f4N//ZRMzvb3ebbzOx3nHMv2/Dvlx+f/25mv8+WhhwAACAASURBVBsEwYCZXWVmH3+mbRR9j8aExkSYCHt/dmb262Y2aWYvMLM9ZvZLT9fYIAjWzexNZnYhCIJs979FM/tlM7vFzK4zsxvN7CVm9vMbPjrd3dcpM/s164ydt3eXfbmZ/Sfn3I7usptZV667rp8wsw84567q1e7neI6eNVfqBOerQRD8fRAEfhA8bcb/7WY2EATBbwZB0AiC4JR98+T2JAiCpSAI/jYIgmoQBAUz+00ze9lli/1ZEATHgiAoW6djvN0556wzID4ZBMHnuu38rJk9YmavI9s5GwTBUBAEsxvKp4Mg+EAQBG0z+zMzm3bOjTnndpvZC83sPwZBUAuC4CEz+6CZvbPH8Wla56eQ0SAIikEQ3NddbtNtFH2PxoTGRJgIdX8OguBEEAR3d9u7aJ1J0+XbfSa8w8ze192fRTP7T/Yv+3/NzP5zEARNM/srM9tiZn8YBEEpCIJHzey4dSY0m1mXb2a/EgRBPQiCfzSzz5rZDz5N+571OXou9J2Ds0kuPoNld1rn67yNL0XyzOxLT/dB51zGOrP515jZky/QyPVoy3kzS1hnNr/TzH7YOff9G/49Zp3OsBnmN/y50v1/1jqz5uXuYNu43Wufok1mZj9mZr9qZsedc2es03k//S1oo+gfNCY0JsJEqPuzc26rmf0363w7krPOlw1LT/e5Hkx227axnRt/1bbc/cHAzOzJCePChn+vWmcsbWZdK0EQVC77943ftDKe9Tl6LlypE5zLw3vKZrbxLaAb39J10cxOBkFw4Fls5+fMbLeZvTAIgnnn3C3W+YptI9s3/HmHmdXNLN/d7geDIPi3z2K7vZg1szHnXGbDBX2HmV3asMy/OD5BEBy3zk8dEevMtD/R/T3yt6uN4juPxoTGRJgIe3/+7e56DgVBkHfOvdXMfs/MLAiClnOubri/p7p/ZuF1c9aZRBzf0M5LZLnN8HTrGnXOpTZ8s7bDzB54mnU+l3P0rLlSf0V1OQ+b2Rudc8POuUkz+6kN/3aPmTWccz/Tlac859wh59zNm1hvzjo/Ka52f6/6y2SZH3XO7e/+JPCrZvaxIAgCM/uwmX2/c+7V3W0mnXPfc5lT8IwJguCsdTrTbzrnEs65G6zz0+j/81Sfcc690zk3FgSBb2br1hkg/tO10XVkv6d8ukH0NRoTGhNhImz9OWedSdt61+X5D5f9+yNm9o7uet9oZnds+LcF60zoN37T9Jdm9svdX9lusY7P8xebaAfj6dYVMbP3OefizrmXm9nrreuw9aDnOXIdEfpb/qRcWCY4HzKzx63zVdlnrfM7RjPrzIbN7A3W+R39OTNbNrM/NrOBTaz3D6wjua2Y2dfN7DNkmQ9b5+TPWecrt/d2t3vOOqLcL1nnq8cLZvYzRo65c26P69jwm73Qv83MrrbOV/YfN7NfCILgiz2Wf4OZPe46TyT8npm9rft70Kdr43Yz+9om2yT6iw+ZxoTGRHj4kIWrP/9Kt73r1hGVP3HZR36qu+4163zD+MkN+3uku/w51xGXx60z8XrEOhL+o2Z2n5n91ib2n/F065qxzuRszjou3LuDIDjZa4WbOEfflnGlVzWIp8Q596iZvSwIgtXnuy1C9AMaE0J863HOfcjMPhwEwd3f0vVqgiOEEEKIsBGWX1EJIYQQQvwzmuAIIYQQInRogiOEEEKI0KEJThfn3MudczMb/n60+wjct3u7H3LO/fpm2iTEdxKNCREm+rE/X6k453a5zvuyaJaec+4XnHN/upllv51cUROc7kEqdx+3e/IlYN63Y1tBEBwMguBLm2xTz/dwCPHtQmNChAn1583zfE4cno7uKxne/Xy344qa4HS5PgiCrJm90sx+xMz+l8sX6McT/nyjYxJqNCaeBTomfYv6s/iWcCVOcMzMLAiCJ8zsK9Z934zrvA7+/+jmVJRd55XzU865T7hvvp79n9MvnXOp7leHq865Y2Z268b1d9f3qu6fve5Xbqedc0Xn3IPOue3OuX/qLv5I9yeOt3WXv8s593A3hOnrzrnrNqz3RufcQ931fNTMkk+3r930x0XXeX39j22oDzrn/ry7f+edc7/oOtHz5px7l3Pua865P3TO5a2TPHmVc+7Lzrl159xyd/tPrmu/c+4fnHN559xx59wPPcNTIp5nNCY0JsLEd0N/dp1k5k9127/a/fM0a2P37+9zzj2ZKvxk29a6bXuRcy7S7fPnu+Pjz51zg93PPvmNz4855y52t/ce59ytzrlHu/vyRxu29ZTr2sCPO+dmu+PwZ56inZfv86Bz7v3dz1xynXTwb8u3dBYEwRXzn3Xi1K/q/vkF1kkt/Ynu389ZJ857u5mlrDN5e9A60dtx67yO/oyZvba7/H+2zuAZ6X7miJnNbNjWOTN7VffPP2udVMd91nnN/fVmNnp5m7p/v8nMFs3sNuukXv7r7roS3XacN7Ofts5L2d5qnbca//qGz6+Z2R3dP7/czFrWeZtrzDpJkBUzG+7++5+b2f9rndjvXWZ2YsPxeFf3sz9pnXeOpawTwf1/do9NcsN2MtZ5V8iPdZe9yTpJkwef73Ou/zQmNCa+e/77LuzPo2b2A9Z571TOzP7azP6OtbH79/eZ2V90/7yr27bohn//ceu8s2qPdV6e+TfWCdDbuPz/6Pb111jnLeN/Z2bj1nmh5qJ1giw3u66/7I6VQ9ZJc37V07Wzu70/7n5u3MzuN7N/823pT893h34Wnb9gZqtmdtrMft3MIhs6wo9vWPY2M7tw2ed/3jovRzPrDITXbfi3/7VH5z9uZt/3dAOy+/f/28x+7bJljpvZy8zsTuu8GNBt+Levb+z8l33u5dZ5y+vGDrxonVfPe9Z5WdsLNvzbvzGzL3X//C6y/39uZn9iZtOX1d9mZl+5rPbHZvYrz/c5138aE5d9TmMixP99t/Vnsq0bzGyVtbH79/dZ7wnO3Wb2v234+z7rTLCiG5bftuHfV6zzmpIn//4JM3vvM1jX/g3//jtm9v5e7TSzie4YTW343A+b2Re/Hf3pSvw95k1BEJx6in/b+Fr7nWY25fD17F/p/nnqsuU3vh7+crZbZ7Bthp1m9q+dcz+5oRbvbi8ws0tB96xuYrtmnVfTb3wJWcU6s+kx++ZPCxvXtfG19hv3z6zz5txfM7P7nXOrZvb7QRB8oNvm2y47VlHrvINF9D8aExoTYeK7pj8759Jm9odm9jozG+6Wc845LwiC9ibbs5Epw/7/5MTiSRY2/LlK/p59Buu6/Pgeepr27bTON1tzzrknaxHDcfkt4Uqc4PRiY6e6aGZngyC4+imWnbNOpz7a/fuOHuu9aGZ7rfMV59Nx0cx+IwiC37j8H5xzLzOzbc45t2EA7LDND6yNLFtnNr3TzI5tWNfG19pvPB4WBMG8dYU959wdZvaF7u+YL5rZl4MgePWzaIfobzQmNCbCRNj6889Y55uR24IgmHfO3WBmh63zazKzzkst0xuW37rhz/+iL3eZtU7/f5Id1vm17IKZTZPle7GZdW03syc2/Pvs06zzonW+wRm77IeUbwtXrGS8Ce43s0JXSkt1JbJrnXNPimYfM7Of70pe09b5vfxT8adm9mvOuatdh+ucc6Pdf1uwzu8on+R/mtl7nHO3dZfNOOfe6Dqvtr/HOh3kp7qC3Fus83bVZ0x3dv8xM/sN51zOObfTzP69/cvX2v8LnHM/uEFgW7XOAGmb2afM7Brn3Dudc7Huf7c65w48m7aJvkVj4jI0Jq5owtCfc9b51mTNOTdinbeMb+RhM3t7t//dYh2n50mWzMy/rG1/aWY/7Zzb7ZzLmtlvmtlHn+VkYjPr+iXnXNo5d9A6vtpH2YqeJAiCOTP7vJn9vnNuwHVE5r3dieG3nNBOcLoXuzdZ53eaZ63z092fmtmTFvivWucrtbPWOeC9vnr+A+sMls9b5/fD77eO5GbW+V3jn7mOgf5DQRA8YJ2fCP/IOhfMU9b53b8FQdAws7d0/75qnd/z/83GDbmODf/STe7mT1pnhn/GzL5qZh8xsw/0WP5WM7vPOVcys0+a2b8LguBsEARF6whnb7fODHzezH7bOtKcCAkaExSNiSuUkPTn/9LdzrKZ3Wtmn72sXb9knW+WVrv785EN+18xs98ws69123a7dfr6h63zhNVZ60jEvSZ2vdjMur5snf2/28x+LwiCz29ivT9qnV/pHevu18fNbPJZtrEnepu4EEIIIUJHaL/BEUIIIcR3L5rgCCGEECJ0aIIjhBBCiNChCY4QQgghQocmOEIIIYQIHT2D/v7Lb/0OPGK1XijAcqurq1BrNhtQyw3kyFZ8qNRr+NloDJvabmHQYywRx+V8XC7SakLNkagA33dQa/gxqOWGLn8HmVmUHN1KpQq1HdPbocaebXvo8GGonbuEuUrlWg1qtQbWmm3c30Q6DbVoBI9BNo3vj2uTcx73cA4d83B91TIely9/7X5c8Hnmr4/Mw6nxSL+JGfa5TAqPbXURQ05//+d/GmpnTuJyBw7dBrV/+6u/BLX24AjUkg3sYfEUed+dj+MzgiVLxcnT0+SzFR+PVYt09jbJcPXb2B38Nm7Dkf4VJYMxEsH9ZeGxPtkPNj5ZZ2UPqTryadYWxmsPTvTdmHjze14NOzS5FTPlnMPrQ6KJ182X3no91EYHh6HWTuB5uf/MV6CW8YagFk+loFaoVKCWbIxB7ZrJPVCbnt4Lta8/9g2orZXxPM+WSlAbHcftrtXwnrVjHJ+u3rcNayv5ZahVz2H2oN8uQ+3EDN7zFx324TtvxRDjV9yC8TZRcv1otepQe/TSE1A7fOIY1N73zvc85ZjQNzhCCCGECB2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8k4FkchqtVGUazdwlqzidLQN9/C/k0c0YMcEZgadZRQ2WsmAiIt+z4ul4ySuR1ZLhLBQxQlNbaNVoNIliWUuNbWUdJm++a3UTJLxVF4jkXxvFXr2OZyDY9pjMiYgwMo5CXjuBxxjC2ZIO0j4vF6AY9BP5IlY8KI7J6IoXQ7GMPjaFU8p6+6/gaoPeaRPkekzSzp1vUECuEJIrp6UdwPZ7i/UVKLBDiQA/bj0yYVWUcWjERwhc7hGCOOscVIW5jX6xx7mAHF6DYRj2Me9vWAHOcGEZmNXPOiUVxfPzI+MQ61XBb7f6WIEm+jXoRaljzoMJBDGfn8Oor36018mCLq4RhrFPH+tLKMtdIadpKbrt4CtREiLd9EHgK4sIjX/0lsnm0ZHIDaYgGPX4xI7FvT+FDBcBTPx2IV5eZqHa/DIxUcUDsmp6A2msXtFsgDSAkyjiNkIpAgF5ChLPaDXugbHCGEEEKEDk1whBBCCBE6NMERQgghROjQBEcIIYQQoaOnZNxsov3EEor9gEQTEmIxFLZKxXWoeUSodA4/2yAisxfFJOOIj/O4KJFFo0QwJH6htVtEqGyj7XXhwgWozc/PQS3mYVtSaZTCmGg9NIDidr1BRHBy3hrExmxUUPprEdex3cZjVa+jBOc3cblYHM8Hk8P7kQRJ3IwQYd1jBmuL7GMJj9mL9mJS6v5RlA4fePwU1Br5Jai5EUw2JZqrOZJszSR74hPTa4BP5OtaHcXGCknedg7HMROeHZH7PZaUSsT7KJGCHRHgfRKr3CbHqs1EcHJcWkQoZrRjTDIe3dRnv6MQQZQJxZlEBpcj577dxGM2uXUH1E7OomR86rGLUGtMEll1CEXmxXm8F+WL2A/ZMC6uzkCtUCXX1zLub5ocl3gb2zI+iPeEOPmOolnJQy0gDwI1Wng/aTSwXx+4eh/UDt10I9RSRGRemMHzkRrAZOnhcRS3Uzl8OGJ2cQ1qvdA3OEIIIYQIHZrgCCGEECJ0aIIjhBBCiNChCY4QQgghQkdPyTi/iq9Y930U7oIAxSSWZFwqoTjFROGhLApgjGaBCGok7TcIUDAk3qA5kjrqkzRFcgjMkbliNp2D2mAO5bt6FWUvF5D9IKJwnKSdEi/a0iRReGgI28fSZyPEiWySFOQUSepdL6LwVq6RRE6S/NuP/PWHPwC1F91xJ9Su2XcAaj4ZbVEivzKReXURJcbtaTxZaR/7TZVIkT6NFGbp2djZffJRlu7NErVXzqBkf/rMGajFoygYsnHskYcPIjQtGYssRZ0lkrPl2uS4sKRlEkhOG+jIRjxy/N5wHcq2zzeLi3ifyEbxGp4Zw6TbifExqCViKN3WKiRlu4HHbFsMJeyxLNZcCj87OozXoLKhEDu3iH14+17cj63jmPaby+HYTpME4JEhPAZnZxehFiMPMwwPkPtnHTtipUyS7CMo8Q6k8QEHbJ2ZIw+3LJ7HB22KaewvDXLtyS8uQK1Car3QNzhCCCGECB2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8l4iQg9zRZKjL6PFmO5gomlPhFns0QoDojFm0gwCRXFpNVVlFojEUyjNPJ6+zhJ9jWSqNomIqLfxDZ7JHl15/R2XC6K6yuWcD/iRDpkKchrZZR4E0ncj/3790NtaAhfRz97EaW6lSUUxTySUFypoggeGEnVbLBs3f7j7r//JNQ8IpLuuwbTP6sswTaN8rfXJmnhKzgW0yQ5tLyMScbBTpJQTPp1QFJ3gzaOiUgCLxsRIvvGyHDySDp6u1DGWoyo8gFJVSZjwqcPC2CNjZ02uZYx2HJMFGaCfoQJz2wb3JbuO8aHMIV2agAfYBhJobC77+AhqPk1vI5cnMV+PeZwG3ckcNw1p7E2MYptmU1iintiGaXg6SkUvWNtbHPEw3ubG0BhN0rGjlWwf+VaOGZrTby+PjF7GmqpJGrB0Qi2ZWIQl4uQaPAnHnoEt5HCBwNqFRzbKXIvZwnPVsY5xDTpa73QNzhCCCGECB2a4AghhBAidGiCI4QQQojQoQmOEEIIIUJHT8l4cmoYak0iCbJ043oNJd4sEaxyORTFkkmUlWJxlDFTaWz+0grKr44Ilako1tJxXF+dyK+tAkpX9SbOFR8/eYmsD0W2Ay/YBTWPiJKZDMpZo8N4jlgKcoJI1ekoSoxxh/s2PJCFWpTMjVs+iqFbt0zgNmJ47JdXV6DWj6yvYzuX52ehxuS6FpHYYyk8FqU6Jn4TF9kKZexLp459A2o7bngh1FIpTF6NEqHYRUk6r8Nx0mQPBhD7Ou4RGZPIzRFjEi/L6CaSMfGEI0QyjhADmGm9TFBmy20yQNk8kj7LtsFCkPuRt7z2rVBL1FEQbSyRFPeLmKa72sLldibxOjfdwOthvIr3jhZJMh4mke118vDD/hEiopfwHnOxQhLEFzEVeHByG9TYTXitiNJtjDwsMzyKD4WUiXjcoFI87ls2g2J0bgi3GyF9eGAQ7+9zi5i+HCWp+sUiXvNqRLKfafScsgD6BkcIIYQQoUMTHCGEEEKEDk1whBBCCBE6NMERQgghROjoaezs3YWpgUyGa7eJiNtA6SoSQ7nIEekwGkOBySOCZjqFgtXQCNkGkSIDDKi0dh2LZZIKXC7h+i4t4f7mGyi8LS+jpJq4hKnFe7aj2JuN47EfG0EpbCAzBbU0kccGB/H4Rck5GiWScbmCx4oJ6Nu2ohw4N4dS7vFTV4ZSWaujwHfiCKZ63v25z0Pt5XfdBbV0Go/tpVWUE6trq1AbiKE86a+irJcg0mbUSLpxhIxFR2TMNjlXRKZNJXEck00Y2YQZkd0DY2nXm/sZjVxmzIjc7JNrGRMqA6YAb65kAZE7fdIWlozcj1y14yDU6iR5u1jG69wXH7oXasltW6EWn8XPZgp4dB9Zw75+KCBp+SSNu7CO17RoHMfn6bPzUKvW8LpQItL+m6/B9PgEEeWLLZIAH+CYTWaxfSNt7Ev5Ej6QEHgoDzej+NmFNRSFl4soZJ9ZwGtUrYwS+dLxY1BzaTxHw1MoZJ9dw/3ohb7BEUIIIUTo0ARHCCGEEKFDExwhhBBChA5NcIQQQggROnpKxkEdJSniGFs0gmJq4ONn/TpKUkEExdR6FcUktl0mRbIE4IAkVHo+tjnK3lrvcH2NFopYF2dQxForkDRbkrK6XiIJn0RQThgeq1YLxbgUSccNfPxspVyAWjyO6aAugt2EhMBaOoEHcGhgCGqDadxGOoNJy/3I2BYU75eWl6D2wP2YKPw9r38T1LJxPD5eHOVhjyQPV/LYmdotHDsDDZLG2iLjhCRbOzKemKBcL6IYHSQx2bRNRPQqESCb7IEEIt0GERw7AbkuRMjgJl6vOZ+kL5OHFGhb2AMYLEE2xhKZcX0kyLUvOXvuBNTq5OGMQgVrse3Yry8t47V0YBgfnPhKHfv6Z4jUev7YHNSScTy4xRUi2eNl2FJZbHMziuc+X8P7zl//0xNQ84so5+7YignwO6ZGoFZewmM1lMZr6chABmo1cvtPJfAYNAuYSj08hteoyhyK5VFyT4iS7Toy3tdXMTF6S0ZJxkIIIYT4LkcTHCGEEEKEDk1whBBCCBE6NMERQgghROjoaezkSMoqk+ZiUZIUjMGE1moT2TdKRGEPxSQm9bVaRHYkSbPmo3gWCVDEZZGqNXQirVpDOXd5FZM2fZ8kPLdwhcUiSpaNeg7bEuDxK1VRZI6QY+oTebLp4/GrkOMXj6IUnIhhrU1E5iZJ38yk8djv2b4dav1IhiRupmK4P5NTk/hhktBdIp9NX41pp6skTfRShciJN90EtQiRDmNEsLU2GYtExszPz0Dty1/A5OabbjgEtYcfeAg/+6V/hBoTdplzOzCGknYsRfpmgJ9uE3k4QYT6BPk5MOrhcvEkbne9gNeKto/rY8nIju7xe0jt+WXrIHam+NYdUJsroax6zaFroBY4HBNrZaz9Xx/DfrMY3w21r1/AzybItTlo4v1uhYyxSBOTjD3SbypNFGfzK5jie9UWvC784GtRCk4P4gMO6ST2kbERlPs9cv9skOPskfv74CC2pVzA+0Szhinqq208Br6Px2p0FNOrd+6Zhlo0hfvWC32DI4QQQojQoQmOEEIIIUKHJjhCCCGECB2a4AghhBAidPSUjH2H0mjgSDInkVpLBRRn19ZQiK03UK6rkSTj6e1EOIqiwJRKDULNc8x4JgIfkYwzOTxELR/FQSMSlxfDWr2BclYyiamQkQiKZ/EYyqJBQMTVMm5jeGgYaokESpGlEp6jRhNF4WQSxTMmh8fieI68GB7TOtluP7KwiGmdL3vxS6H20jvvhJofxbGzTsZYYwQTS1tbsI8cvOMVUEseug5qhRSeg7E6jjsvwP7fJpL45z/5N1D73Kc+CbXFmRdDrdHA/T1/9jTUmg2Sem54/N5y0w9AbdtuFE0zQyR91rC/1ioo8o9mUfinEjRNN4YSTUJvtXF/mbjajyxeuAS1Quss1FbJGE+RhxWuv/FlUPuLz38KaruG8Vq/cwITjz/8ma9DrYGn2fbuR9F11148B34Vz/ODD2Oa+XIJx9PEKF6H7/re66E2NoT94eLieai9YNc41OYW8f45mEM5N5lCqdoM7zGew+tHPI7jeGgI7wnJAK9lyTjGQ2fH8frGEpmT2WeWeK9vcIQQQggROjTBEUIIIUTo0ARHCCGEEKFDExwhhBBChI6eFttauQ61NhFOozGSzksk41IF5bpzF/GV6MeOYNpjKoVy0dAwpphOkQTZiTGU0UZyKHsNDaF0FUQw7XFsC4pd8TiKdjMXseYHmKA5NoYC5OjoBLaPuNLZDB6XOjHoWm0iAJM0ZxfB89to4vldL6HI7EVxG2mS2toi/apSxVo/wiRZ1g+v2YcJrbU6yn8xIvqVyRg7e+4c1MZvQrnZZVHWazWxP7B+aHHsD8tLs1C750uYWlwhScuNGu5vggiGUeL7ewm8NOVyKPvOzcxBrVjF4/eu97wOarEMnrf8Gj5AMEbOb5OkqLfIOBkcRsmyTVKk2208H0za70c+/ZUjUJvLY9p1nAjX27fhtbRQxWva0YcegdrP/oefhVokgdf6j3/i02QbeF3av2MP1P7dOw5C7Z8ewdT6uRXs14XzmILsiDi+ZzvKuVdtxX7z9cPnsC3v/zjUiqsXofbOd74Dagd2HICaRx4eMZKy3SJJ0Etra1BLEKE+VmeSPV4rfJLcH5D0/V7oGxwhhBBChA5NcIQQQggROjTBEUIIIUTo0ARHCCGEEKGjp2S8kkcZKCBzooAIi+0WJlSWm7i+FSIelw0/O7+IwlbjAqbKxo5hKupAFu3cyWGU26a3oewbSxJxNrsDaymUzBoNFGdTKRQ50xkUu2570YugVlxCca+wjq+o9yJ4joolFLYKRWyfI58tVypQCzw0Q1kyskeEYg4xTfuQLSRReMfO7WRJHBNxIuYlSWJvM4b9dfECyr4zj6KMf+jqW3F9AR7bNkkFrvvYR5JJ7NcHD94ItfVlHJ/JBI4dnyQjZ8j4XFrEZNioh33z1BNPQK3SPg61PQdugNrNL30l1EpVPAZeBBN4mywKl8qYK1CrEaE4mWAJreznTxSen2+uvvElUNtneJ2r1UpQS5Fz+vjxU1Cr1vE68thpXG4pj6K3T44juyecPI597nf/61ehNovPz9jiGvb1VhvH3WIB9+Prj2Dq80AWr6WxDF57pvfeBLWlObxeP/TwPVDbTu6BI3GUm33Sr6MktXsoi2nJC2R/gxWUkbMZPB/lERTGnYcysg3ick+ib3CEEEIIETo0wRFCCCFE6NAERwghhBChQxMcIYQQQoSOnpJxoYpCT5Ska/oBSkhMsC2VUa4rrKLA10ZPzFIJFJiKaygZr6/h+tZXMZ10cQETlE9fJMm+RII7dAhrI+MoZ2UGMHk1yRKZt2yD2q0vfRXUFs+fhNqnP/bXUCuXUTwuVlDwy+awLSOjuB/VKq4vN4T7lktjd/J9EttKyK+ipNqPJIg4m07ifgcNPGZtMnbqxK1up1Ew3H71Pqhl09hfPZIIGiHbCIzItCSlLXN2YAAAH1dJREFUOR6PQe0tP/xuqNXqOCaSCTz3Q0TSntiG6eMjo5g0fs2eq3A5kgL+5a/dC7VqEc9HOoltbtaIPE/k8OUFlL5Hx1AADmp4HVyYxYcFdu7cC7VoFAXvfuSm/ddC7aFjKHrf9/BRqBXz7NqMSbwNw+N474Nfg1q5hv213UYruF3H+8QChmLbSgn7F0vsrRKRP57dArUWuVd+/n68jxWaeA2YHMF74G3X47F/oIb79sQ5FLJvKxApPor3CUcemGiQVPbFBZS0V+t4fhtFFO+Hk5gi3azjNSpfwLEzObEVak+ib3CEEEIIETo0wRFCCCFE6NAERwghhBChQxMcIYQQQoSOnpLx+BAKW76PYhITSRfLKGK1aygrpeMkZZUkXlbKREYmMlWUSFwxD4UtP0ChstbAzwYBiqGFNRRix7ftgloyicdvZAwly9ffdRcut2UCao5IoOUiSlxnTp2AWhDF47yax31zJOE2hrthxKs155PUyhYm1xYLKP1dPI9iYT9y1R4UYs+dOgK1eh37v0vggYzGUPTeNoLS3MhOTM9eWCdy4v1fgBoT2xNxFJRZIvkY6Ydbx1AIvP4GTDdeWT2D2yXJyNPTKNl/7xvfALWr96CI+9nPfg5qkxMod6Ycjp25E49CLZ/Ha8/izCLUimUcdzfccBBqbSJofvmLd0NtchLTsLNpTGh9w83vhdrzzXgLk3PvfeQhqN3/GD4kMTZI0m/zZOw4vLbsncYxMZ9HMXUglYdahNwn3vvjr4VanqQRf+EwJg8/cuoSbqONfSSdwu8UFs9h7d4aXmDHBy5AbebRx3C7Aa4vlkJpf34VjzNLHm41yQMJDo9fnSTeNxZx7DQvoqD/WAql9HgOE/7jMTwfB6+7BWpPom9whBBCCBE6NMERQgghROjQBEcIIYQQoUMTHCGEEEKEjp6SccqhFFlr1qAWQd/IYj5JdiTTKS+K8l+7haJfAj1hm5ichtr5cyhikeaZZyjYxsh8L5tBQbO5jomNkS2YAOwaKNPm5/D4zZ8/DbW7/z8U477wqU9A7cTDmNoakOTVSJQkfBLRbnzrFNTGBlGMDnzsOvUGSbnGptBE61g0iwv2IZk4HrNSfh5qZ6so8DUM0z8dkdgHbv0eqPkktfjiPKaT1paJ2JvAz6aIeFwlyeU7plEovvk6HCcToyjEDo7th9rJ4yhkH9x/DdT27NoJtTJJIz5zAsfO7m0oLT9yz1ehdt+X/gFqiSQmdB95HI/p1DSK4IUlvPZcmkV5/ihJ+R0dQTHaGV5Df+sX+08yXlpBkXS9gmLq0Aher10M+9LQlt1Q8xw+/HDm2HmoPXYEx8RqEa99W4mI3mijYHucCLGjI7i+a3eT6z9JS/YCHGOrHu5bLo4PsjTJAyWz67i+WBtl32gct3E0g7XxnQegVmdvKijjNuqktmsOj99tERxj96/iZ8+SNxrEPRLL3gN9gyOEEEKI0KEJjhBCCCFChyY4QgghhAgdmuAIIYQQInT0lIxLDZz/+AEmkZpD8SeaQjG1vo4yWmkdRaxsFiWk/ftQWNw2jdLaQ2lM0Dz6BIpnJIjR0nE0mYeJjOkaKEHPnnkCao0SvhZ+vYxS2Af/5I+gliNG9uIllB1TRB7LJbHNfgvF3mINRcBWG9cXieJxaZNE64BIkWYo5GUyQ7hUAs95P1KuoOgXT2HiZrWGyxUqKI4noniujCQKt+M4VGtG0rhr2OcaRP6u11A8bvmYOn1pBvvDvu0oBY9kUBJf8/FaEfewLfv24fpWV5ahNnthBmoVIl6ydO+zJ1DsnZtHOXxodBxq62soQCbIlbNO0tsvXcI2JxN4rFaX8MGFVArPUT/yDyTFd/4UXg9nZ7D/T+y9FmrFIj6IkUjiGHvwCRRY80RqbQd4DZqJ4TXtz+/GBHiXxHOQzuL6EgMonQ8M4Nhu4xCzHPmeoVKcw9ociselIvY5v04eBIpih/3SvXiOdi/hdf3gQRyfUZLK7nmYSr31a5j6PHEJx/bAzZiEXtuNsnm1jiJ4L/QNjhBCCCFChyY4QgghhAgdmuAIIYQQInRogiOEEEKI0NFTMm5EURL0Iygd+j7WLIPJpiuFc1ArE8n4uoPXQ23Hzh1QSxAp+Lbbbodanbzu/fxZbIvn4XwvEkFxtkXSnFfnUaALSDJm0MLPXiJJxlNb8PX2AzmU1lpNPH6OTFuZQN0iCZVrBZTWvCUUQ6PkuHhEIGV9IyDbJaW+JDmAsqMf4AGvNbDPNYiEHW3jMYuQlO02EYqLNZRfmzUm4eHBjUWxfYl0GrdRxf518vHDUGvU8VIyugdTkCe3YoJsmhi7szOYAPzYw7jdVgXTwtcK+BBAkkiqdR/PW9NhW9I5lCcTGVxfwK4VJNF9Sw4fwFheQcm4Wn1mQuXzxWodj2OKSLzVJRSui1EiwBdQlK+QRPT2AKau79yH9461RRSPiwns66fXUSgeiaN03oyOYVuaeJ6X1/AYRMlzGJEIuVY0sI80PVyu6aGwHh9AATgISN8kF93ZEzje821MjH7JLXuhFmnhWIyvoRidOosPy8Sv2QU1Z7iNNhmzvdA3OEIIIYQIHZrgCCGEECJ0aIIjhBBCiNChCY4QQgghQkdPybjlo6wXieCcqNFEGS6fR+l2yziKeZkEphEPD6GcVavh+qpVlKTSRJTcu3cCanOXMGGxQWTkWhvFszqJo4yS1NEBH+XEdgSlOo/IiSQc2lot3G6L+N2NGgqp44MkVTaNxzkSJym6huJZLErSPEnqM+svrBZPYFv6kaYRaZqI1D4TTh0eM99hv3FkuWgEzz07ZtUi2S5JT01n8SGAdA4TpssFFAc9wz6SSeA5XZlDUTg1gNuIx3B/Z2cxHXdhfgFqF89fgFq5iiK/R/YtNYjC8+CW7VAzmtqNY4z9uOiTS2wkhteo6eldUJuZQ7mzH1mvYAdLpTGZfHQQa+U8ntPtY7hcAi+bdur8UWwLuaMNbLsaanEPt5En95j5kySNfgAfAImNYs3YvaOC0m3QxHR7v0meumixh1bY9YgsR1bnyE3GJ+tbnEcpeNsEjh2vjDL3fh/vqdEkkdKzeOIGSK1KH6J4avQNjhBCCCFChyY4QgghhAgdmuAIIYQQInRogiOEEEKI0NFTMs7EUC5NJokATBIW/RqKfoMTKPrFYpgK2SQpsNEYNrXBlouiALm1iUJl1EPBiiUelxpEaiICaYQkoDbbeAzGtmAy5sAACm/rS3NYW0bhLZMm8jBJhs1mcBsNh6JkbgSTMdM53EYqSaRqkvgai2JbWkTSjsVwff3IAJHJY0Q8LtewH7Z8NCWz6GVbJIafTfoo4g4nyXkZ3wa1/Qcw3fvFL3kV1DIDOE7OPXgP1E585mNQi+ycxPaNorB7cWYRamOv2Q21oa3zUPPdcVxuDB9S2DWG+1Hz8Vq2VsSHKII6nqNKG89vgkjfe7fvhNrKBdzfUgFF0wMv2A+1QhGX60dqxGIv1/G6mc6gXD00jteM/7+9c2mO7LzrcJ9L31stqTXSjC4zkubi8S2DQ8qBmIoTO6GoSmJcUCxY8wH4CqypYsWCBSwIBRtIuaBcJJUEMglV2AlO7LHNGHtmbI/nPrpLLalvp885fIDnV6oxC9Acfs/yX+ruc97bedX1vL/uiAMRccA+WK5y8mz1uOZ27/0XakGDz508oCicj2jndvco04Z3uc6VYz4rU/E8GYlE8lysKSWxbspaJmKfc/F+IuE8EGP9lVe5VvzhS19G7aM3f4Rar8Z2OXjyPGrTU5yz5+ZY2+wJ2/wI/A2OMcYYYwqHNzjGGGOMKRze4BhjjDGmcHiDY4wxxpjCcaRkvHiaCcAiEFEmuc6d6qCWyURQIewKiTdTabGilgpJapxR9joxQxlzfZ1JlmMhMouPkNeSJJR4l6YpWi8sLKD23vomagNeSimM+Lm1BhOjxzGFvHTIBM1amZJxOeY152P20aBPObAv+lwl9SZCRj6OVAO2YzmnOLjfZ9tOVdkvFXXbImF0nFOuazYpaL74NQrF3/zmH/AzcsrSoRCoVxq8lo2/+QvU3nzvXdTGl0QKciyk+JCf+7Wvv4zaR1feQW1vS6TA5rzmSEiqsRA0+yK5uSfer9biuJ6bnEGtUeH9PtiieHztBj/jQFzLcSQX/yYvX2B68PPPnEPtrV+9jVpY4VivNLguzYp06qkexfGxEJ7vbbIP1teZbi8fkcLhVYc4vvPK76B2eMhF/Gc/+yVq29u7qAXCr62X2fiNBmsz0xR2Z0X6cq/Pm/vKC1/gZ4jTEU8+8yxqrXmueV0hgremOHdmT7HWHLB2FP4GxxhjjDGFwxscY4wxxhQOb3CMMcYYUzi8wTHGGGNM4TjS7Fx7SIEvTSkhqZ9dH4t0yyBQ+ym+ViUZJ2MKu6kQWMfi70YjCoZTQmra2KBkvNdlrV6jsLW6TIGuKn4WfmaW8nVnltdSrV1C7edvUKDr7nRRC+uUqoOqEPcytn1TJFX3BvzcmQ4Fv3ZLfG7INogjDrsofjz22sMu26w/5Ji7d5up0+efYIJnnlC4y1MhAIs2e+bSl1B77tLzqCnJ8q03L6M2s8BxOC/E45kpyontu2uo/eU/vIbaxaefQu2Ny/+K2m997SXUTszyc3/4zz9GrT/gulWpca0Y9lkbJBuoJeL/wFFIeXh9jenjwxFTagcj9vnGJttvOBSnCo4hWZkJxaeXKRmvTHMMX/mAcvpBn+1TrYvUbp5VKNUmeC2NeabHJ2X2s0pffnif/ZKJ+f7yS99C7c/+5I/52hLX4ddeZwLwn/75X6H21DkejDm/xOfJuQtPoDYxyWfW4SHTvb//IyaX/+PrP0HtRz/5KWrPPs3k8m+/+Guo7RxQnh+J5ObBHv+uWmH/HsXj8VQxxhhjjPkceINjjDHGmMLhDY4xxhhjCoc3OMYYY4wpHEdKxgddCsUqsTeKKE6lKWXMapVWmHptIESsikjiVYnHKms5SSnrjcYU2SoiAbKW85pTkYDaEjLmkxfPoHbm7BJqcYv3+9RzFMpOnqFg9f3XKIB1D5iCWamuoNas8/0WFiitPdxl6ufKKqW16WlKoI8qpUfR47HXvvnpbdQaYhZlYxYPDiichmWKjemYc2zmxAnUlk4z3bV/QHFwf+M6ah+++x+ota5TEu8vP4na1soKas2QY2luTaQHjyhovvcLio2pEE3v37/Fa+lybCaJiBo/5DgMRTTscMz2E8G1pYNDJtd+8hnbOS9zrUhyStBBxr8Ty+CxJEm5No9EvPE4Es+TEsd/XySs9/tcX6M6PzfLuf7v7vCgSDbm51669Axq4zGv+eF9jrnZWYrMccT5NBLv99yv80DJCXHwZKrD5PJWk59x9/Yd1G7d4XxfW6do/cknd1HbXN9DbTzkPAlf/QZqv/Gli6il4tDP/j4PsuQifTwo8VqO4vF4qhhjjDHGfA68wTHGGGNM4fAGxxhjjDGFwxscY4wxxhSOIyXjlbOnWBRpxHHMt1H6b7lCYbcsfu59LMTUUIip6lOynGJSt0u587DPlMTVsxR7JybPonb5xxS2tjYfopauUrqdbFOWK7fZfklICe4rX/0iamv3KV29/k9Mhu31eL+zs7yWLKXgVxP9FgtBMxJjYyBSZVXKdZpSPj2O3N+6h9pcp4nazOl51DYONlGbUH2fse8nRZLrnki/7e/toFbO2H8H20xaPrhD0S/osl92O5Q7g4RrRV0cDBipFNMur/nezU9QGwyY2r14YZbvJyTGMBJrRcq/U+tHKAT4SEjVUcjXtifZ9iszFFKrVc5FteIdR1KRzNwbsH0mJtlX5ZCS7KYY10Egxlwu+lQk2WdC4P7qCy+gVm1yve73uH5tbnLu/OCH/4LaYY/t0mhTTr95h7LvrZsUmW9/xDnx/HNMLV49u4Da089Q9v32q6+g9oPXX0fttb/7LmpBJA7fjDm3NzZFMnifh36GiZDsxSGi6HPOCn+DY4wxxpjC4Q2OMcYYYwqHNzjGGGOMKRze4BhjjDGmcBwpGbenuf8ZDsVP2VcpQOY5ReG8RGErF2mdwlmWQmCSKDGV1/xgjXLn1h4lrotf4M+9P/HEMmrvXLmG2plVphZ3Zpj2u7PNa6llbL+kRBGrPUUJrtGktJYLATgWonBDJIGORf+WA17fzY8foFYqUcAdDnkfSkofJfy748ip80x6roci67bO+2k02d6VCqW5cizkVyGJb+8xVTkvs+/DPufE2t37qH169Spqv/l1CvpT4RRqE3uci5UG2yAKmTSb9ZmKWs75ftMdjvUnLq2gFlaUPI9SKc8pNqpk3XKolkmK5al4v1LA+RSIRS8M2Ucq8fs4cobucKnf5Vp/KMZImHNtyTIKytt77MBBjwnFky3OsYZImT977jxq+0IKPn+B6//Vq++jdvs25+Lf/vV3UZs+xcYKquz7Q5H4rWT877z6R6i9/I3fRm0g1uFMCLsPH3yG2t3PKGTfu8vE43mRcL6xyUMw9Vj8okGFcywVwv+hSDg/Cn+DY4wxxpjC4Q2OMcYYYwqHNzjGGGOMKRze4BhjjDGmcBwpGff6FIT6PZEK3Hu0lOE0VSIp5bFACHeZsASVZByXmYx545MbqJWrvPXJTktcHcXBiUkh9sa833aHMuZel2msW7vi5+gz3tudjKmQ1z+6iVqzRbk5EF0d5CJFeijEbSFtDgccB9RHdQL1WKSN1irst+PIpQvnUMtSim9jkZ5ajSlUtkRaZ7NCCfVwQBE36bEd6w326ebaZ6gNeruoPdiitDlKKDe3c0qg/YjjYWGG9xFklHMnq5zbccJDALMdjuvFZaZ7N5sc65FI6B4r81gcesgTyo5BiaJkELPt84xrXprwWtTkUWvtceTiaY6H3R3K2r94k2m/t+98ilq5xvdLAzbQvXWOVxEyXJo/dQK1vX3Op919JnnPiDG3eoap3d1truGtab62ISToQcq2mj/Ba15+5inUxGOi9N6VD1DrDbhGDcVaH5W4DnfaFKNv7PO58+AO15R6jWL0ZIPPk1gs/yOxNxA/cnAk/gbHGGOMMYXDGxxjjDHGFA5vcIwxxhhTOLzBMcYYY0zhOFIy3tmgEKsShff3KSKmY5VkLGQ9scXKRYKhloxFUqr4Gfe7t5m6u3CKohhfWSrtC3lsYY6vvXebybC3Fpl6WxGpmoFqq5FqZ4pi+3sUESfbbdRqTQrUQZ3S6yijzBeMKcE1G7TCHjV5Nc/5GZUK5bvjyFJMSbbbE+m3VbZPKGTVupgA1YAjcV8liIfsv0bI127fp8g57FP+m19a4PWFvI+NTcruacB7+8LqHK9ljeO1GvHeGiHn3eCAnzE7x/Tx6oD9Mdh+iJpKTE9zFvMhhXEl1EficIRKdC+NKLOqOdGpcq04jsxOUKZt19gH9eAiaqcXmc6+d0AJ++oHTI//sMt1fX2NY3NPHOLYFUnL4sxFKREWb6vNQyZxyLH5nd/9FmqTQh7+++99D7WXXvw6aq98i++38ZAy/vtXmLR86xaTlm/duoPaJzeuo7b1gAn1kzMnUSuV2S4P19j2W6lI/BZDPauwQ2q1z/ec8Dc4xhhjjCkc3uAYY4wxpnB4g2OMMcaYwuENjjHGGGMKx5GS8b1PKTApsXcgUhKVhVcTUmsci+jQRyQTQmAvEQJTQvFy6RQF4GrOa8mGlMdWTi2i9s5dpmpubFImnDszg1ooEp5jJVqnFN6yMe+3IlI/GxMipbkitOpApEPnlIfF5ZUiYW0mIrVYjQ0lrx9HTonxOiOkw7JIIx6JmNxajXOiLuTSwxH7udXgaw92OWfn5ygEfnz9FmpffO5p1E7P896SEQ8f1OuU/8YzHA/vD/i5K2d5fTOTlJtHY77ffpfpqUnI+dQWImxcZn/EkbAdxXiNxDhQByuqon/Vf5WpiGit1R6PdO+JBg81RH0mBS+fWWVteRm1sUiov3j+LGofPsFU8fffZYrvtWtMsk9EX62Kz7j6/tt8rUgArogU6xvXKUafF4dgKmLMXf4xU5/f+Om/odbd5fg/3Of8HA4pfavxmqp1OBfPVDH+c3FgYjBiWyXiGT0WP3KQiTY9PBDPkyPwNzjGGGOMKRze4BhjjDGmcHiDY4wxxpjC4Q2OMcYYYwrHkZKxyqWNIspFEy2RkhsKobJKwSoXP4k+EEmkoUhUTYc0k/Z2mKqcZxSnJpoNfoaQCVORPlsRv+0u/OTSxg4FsM4s26qiPncsUp9F1GYkhOIsYJuKbiv1R5TD201KkU2RHqmuryZE07aKixUoyfI4MnuKSaS1Gsd1GLNfUhHXGQVC1qtwfEVinvT2mRJaF829dJ4y5ub6FmqTM1OoXVg5g1oYctyovt/coGR/7wGv+dyFC6g1yrzfOBPytWi/m/fvotZui7TdFteAihDvH3WsK8plMTZEUnUm1qhHTQb/v6YSs1/CBq89FYckVGp9qcy2bZ6l7L64REH5yaeeQu2dtykK7x/yOTE5PYtaMuAzJhR9NRaHAH7+72+g9vYvr6BWEgnYqTicMRLPxZJKypY5249GXTwXJ6e45p17kmvKiRm+diQS+UeHbKtMpBuPxWuTz/mc8Dc4xhhjjCkc3uAYY4wxpnB4g2OMMcaYwuENjjHGGGMKx5EGaCp+Kl4JTLnQkfOMMpBKte0PKSKOBkIKbjF5uFFrojboM8k1jnjNszPTfD/x8+yDmGJXLoTnyVmmeV4Qcuezy0uo7e5soKYU70qlj9ry0jxqYZn30Z5gW0Uh+6NaFkNCpVKrhFYhRSpPMhCJl3n+Pxfj/jfJ6hQqh6K9QzHmyiJ5Ox9QrgtF30fif5Eg5dic7lDQbFU5Xr/8xUuovfvBR6jt7jN5+/QSpUOVZHz/PkXO5QWmFnc6lJvHQybhVoRfKIZSqd3itRz0KIuemOKaUhJJ7SORxqr+MwwjVlVCd5bxM4biwERdJL8fR5RkHItTDUHA+1EHQMZCJFXrQxxz7qysrqA2Nc21+cp7TBl+51dvoXbjGufE+tpD1A57PdRScW/9AZ93an0tl9mmlRprQci1oiHm4uLiKdRWz62IGg8VqIMV9RbbdDjm+O/1eHMH+3yO7e5zvu+IlObtTa4pR+FvcIwxxhhTOLzBMcYYY0zh8AbHGGOMMYXDGxxjjDHGFI4jJePJNtMjs4zSkBJJVeJxucI/rA2EPDlFKbJRp9SUjFTSJ8Wu02coWLUnKGJFQgqLVSKnMBtXVhdQW5hi+12YmUPt6hbF6G2RMpymFBFXVigZZ6IN0jEluKpIlh4M+HcqbVeJl6lIo1SyXCzSjVXtODLRoJgaiNTiUiAEwz22bUOkmEYh20IEVpcqYo7VRDOWxXhdOMlxuL6zjdqdexTgayIBuLRHcXBzm0Lg+XnOxZq4vq6QJ5Mx50SWsmFOdNhH71y5jtpsk2vKzCRF2DRiXw4HvJa6SPyOIt5HWSRfR/Hj+7+mOjQQiIRidQghE7VYyNr6MygjBxEnwOws2/vSs3y/y5cvo5YkHNe1Bvt5bpHrcO+Q871W47Ot1eJzolLhOKzU+dpag4dHJhoc13MnOScWlyj8t6dE0r44pDAU639FJLBPzLCtJlt8JtRr6mAM+60SOcnYGGOMMf/P8QbHGGOMMYXDGxxjjDHGFA5vcIwxxhhTOI40O5UQpeThTMi5uYhnjIVIV6sxKVX9Inog0l0zIf+dv8AkxqjKa85FSnM+5vsFIkEzF3LbvEgyXqhTkmrtd1Er9VjriZTauEzxrFmhsJWM+dpSTgE4SUS/CVG4xOYrjYR8p0Th4YjXMhg+mqB5HJmbYD+XShwPZTHWHx6yzSpCnoxVYqn4jHHC9M/eAT8jrPG1jZgy4XSH9/bxpx+gNjHD13a7FCrLEcfrVIcJ4smY8nwYce7EDdFWQuaOSmy/06cpVcdlzu00E6J8RVxLzEmRJBzrkUi57osk13JVJNeKuX08UQdPVLq9MOVFKRSS/Sjhej0S40Yl+8YieX5xnodCfv/3XkVtd49puvtdphFfu3EDtav/ybkzHPI+mg3OJ7XWBzHHoXr2xmJNWd/mM+bap7dQm1/gIYCT4kBCq0W5ud3iOjOIOdYjcR+tphCURfr+RFMccDgCf4NjjDHGmMLhDY4xxhhjCoc3OMYYY4wpHN7gGGOMMaZwHCkZK+EuFCmTj5rCmaZC4g3EJYgUzExIloGQjJtNIQSK5NB6TKkpFwLYWMhtSqBWimx9SPGy9JBiXE0Iz7HwC4NMtIEQ92Il2olEVZWEG5ZFLeTF5GJvXK9TjFOJx4lo08eFw+4aav0DJvZWRB/URduWcpGUnVLCjssc65022zsqUeo7HFGKPNjh342FYH7yJFNWT5ygKFyKOJ/2hHi8KQTzKOC4TsRY7yvhX7TVONlDba7DMdzr76AWRBQ+q2XKjqmYs6VArRVc31JxH9VQJL+LFPDjiDp0EYqE7kCsGalIolbmcaiSjB9RZFZuszos02pRso9EOm+7w4MxnTkmBS8sLaL24Ucfo3bn9l3UhiPxnChTsBW+rvxlgYqSlg8539e3uS4k+RZqrSb/bnqSa0Wrwc+tCnlejfWyuLequuEj8Dc4xhhjjCkc3uAYY4wxpnB4g2OMMcaYwuENjjHGGGMKR5AL4c0YY4wx5nHG3+AYY4wxpnB4g2OMMcaYwuENjjHGGGMKhzc4xhhjjCkc3uAYY4wxpnB4g2OMMcaYwvHfP+Y0OvgpJ8MAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#创建图像标签列表\n",
    "category_dict = {0:'airplane',1:'automobile',2:'bird',3:'cat',4:'deer',5:'dog',\n",
    "                 6:'frog',7:'horse',8:'ship',9:'truck'}\n",
    "\n",
    "cifar_ds = get_data('./data/10-verify-bin')\n",
    "df_test = process_dataset(cifar_ds,batch_size=1,status='test')\n",
    "\n",
    "def normalization(data):\n",
    "    _range = np.max(data) - np.min(data)\n",
    "    return (data - np.min(data)) / _range\n",
    "\n",
    "# 设置图像大小\n",
    "plt.figure(figsize=(10,10))\n",
    "i = 1\n",
    "# 打印9张子图\n",
    "for dic in df_test:\n",
    "    # 预测单张图片\n",
    "    input_img = dic[0]    \n",
    "    output = model.predict(Tensor(input_img))\n",
    "    output = nn.Softmax()(output)\n",
    "    # 反馈可能性最大的类别\n",
    "    predicted = np.argmax(output.asnumpy(),axis=1)[0]\n",
    "    \n",
    "    # 可视化\n",
    "    plt.subplot(3,3,i)\n",
    "    # 删除batch维度\n",
    "    input_image = np.squeeze(input_img.asnumpy(),axis=0).transpose(1,2,0)\n",
    "    # 重新归一化，方便可视化\n",
    "    input_image = normalization(input_image)\n",
    "    plt.imshow(input_image)\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.axis('off')\n",
    "    plt.title('True label:%s,\\n Predicted:%s'%(category_dict[dic[1].asnumpy().sum()],category_dict[predicted]))\n",
    "    i +=1\n",
    "    if i > 9 :\n",
    "        break\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "77c8ef61",
   "metadata": {},
   "outputs": [],
   "source": [
    "class LeNet5_2(nn.Cell):\n",
    "    \"\"\"\n",
    "    Lenet network\n",
    "    \n",
    "\n",
    "    Args:\n",
    "        num_class (int): Num classes. Default: 10.\n",
    "\n",
    "    Returns:\n",
    "        Tensor, output tensor\n",
    "    Examples:\n",
    "        >>> LeNet(num_class=10)\n",
    "\n",
    "    \"\"\"\n",
    "    def __init__(self, num_class=10, channel=3):\n",
    "        super(LeNet5_2, self).__init__()\n",
    "        self.num_class = num_class\n",
    "        self.conv1_1 = conv(channel, 8, 3)\n",
    "        self.bn2_1 = nn.BatchNorm2d(num_features=8)\n",
    "        self.conv1_2 = conv(8, 16, 3)\n",
    "        self.bn2_2 = nn.BatchNorm2d(num_features=16)        \n",
    "        self.conv2_1 = conv(16, 32, 3)\n",
    "        self.bn2_3 = nn.BatchNorm2d(num_features=32)        \n",
    "        self.conv2_2 = conv(32, 64, 3)\n",
    "        self.bn2_4 = nn.BatchNorm2d(num_features=64)\n",
    "        self.fc1 = fc_with_initialize(64*8*8, 120)\n",
    "        self.bn1_1 = nn.BatchNorm1d(num_features=120)\n",
    "        self.fc2 = fc_with_initialize(120, 84)\n",
    "        self.bn1_2 = nn.BatchNorm1d(num_features=84)\n",
    "        self.fc3 = fc_with_initialize(84, self.num_class)\n",
    "        self.relu = nn.ReLU()      \n",
    "        self.max_pool2d = nn.MaxPool2d(kernel_size=2, stride=2)\n",
    "        self.flatten = nn.Flatten()\n",
    "        \n",
    "        \n",
    "\n",
    "    def construct(self, x):\n",
    "        x = self.conv1_1(x)\n",
    "        x = self.bn2_1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.conv1_2(x)\n",
    "        x = self.bn2_2(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.max_pool2d(x)\n",
    "        x = self.conv2_1(x)\n",
    "        x = self.bn2_3(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.conv2_2(x)\n",
    "        x = self.bn2_4(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.max_pool2d(x)        \n",
    "        x = self.flatten(x)\n",
    "        x = self.fc1(x)\n",
    "        x = self.bn1_1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.fc2(x)\n",
    "        x = self.bn1_2(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.fc3(x)\n",
    "        return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "63c2c29f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "============== Starting Training ==============\n",
      "epoch: 1 step: 312, loss is 1.2922429\n",
      "{'Accuracy': 0.3954326923076923}\n",
      "epoch: 2 step: 312, loss is 1.2750182\n",
      "{'Accuracy': 0.4872796474358974}\n",
      "epoch: 3 step: 312, loss is 1.0838046\n",
      "{'Accuracy': 0.5557892628205128}\n",
      "epoch: 4 step: 312, loss is 1.1286726\n",
      "{'Accuracy': 0.592948717948718}\n",
      "epoch: 5 step: 312, loss is 1.1604517\n",
      "{'Accuracy': 0.5885416666666666}\n",
      "epoch: 6 step: 312, loss is 1.337981\n",
      "{'Accuracy': 0.649238782051282}\n",
      "epoch: 7 step: 312, loss is 1.055731\n",
      "{'Accuracy': 0.6584535256410257}\n",
      "epoch: 8 step: 312, loss is 1.1750035\n",
      "{'Accuracy': 0.6686698717948718}\n",
      "epoch: 9 step: 312, loss is 1.0010777\n",
      "{'Accuracy': 0.6919070512820513}\n",
      "epoch: 10 step: 312, loss is 0.7225568\n",
      "{'Accuracy': 0.672676282051282}\n",
      "epoch: 11 step: 312, loss is 1.0584934\n",
      "{'Accuracy': 0.6926081730769231}\n",
      "epoch: 12 step: 312, loss is 1.011681\n",
      "{'Accuracy': 0.7020232371794872}\n",
      "epoch: 13 step: 312, loss is 0.98512095\n",
      "{'Accuracy': 0.7444911858974359}\n",
      "epoch: 14 step: 312, loss is 0.59207284\n",
      "{'Accuracy': 0.7182491987179487}\n",
      "epoch: 15 step: 312, loss is 1.0525222\n",
      "{'Accuracy': 0.7546073717948718}\n",
      "epoch: 16 step: 312, loss is 0.97304606\n",
      "{'Accuracy': 0.7320713141025641}\n",
      "epoch: 17 step: 312, loss is 0.8148502\n",
      "{'Accuracy': 0.7602163461538461}\n",
      "epoch: 18 step: 312, loss is 0.7160713\n",
      "{'Accuracy': 0.7518028846153846}\n",
      "epoch: 19 step: 312, loss is 1.0471575\n",
      "{'Accuracy': 0.7700320512820513}\n",
      "epoch: 20 step: 312, loss is 0.9026973\n",
      "{'Accuracy': 0.7863581730769231}\n",
      "epoch: 21 step: 312, loss is 0.6164329\n",
      "{'Accuracy': 0.7955729166666666}\n",
      "epoch: 22 step: 312, loss is 0.8433214\n",
      "{'Accuracy': 0.7872596153846154}\n",
      "epoch: 23 step: 312, loss is 0.59412867\n",
      "{'Accuracy': 0.785957532051282}\n",
      "epoch: 24 step: 312, loss is 0.67253816\n",
      "{'Accuracy': 0.8051883012820513}\n",
      "epoch: 25 step: 312, loss is 0.6700381\n",
      "{'Accuracy': 0.8012820512820513}\n",
      "epoch: 26 step: 312, loss is 0.42357194\n",
      "{'Accuracy': 0.8122996794871795}\n",
      "epoch: 27 step: 312, loss is 0.79985774\n",
      "{'Accuracy': 0.8199118589743589}\n",
      "epoch: 28 step: 312, loss is 0.40744948\n",
      "{'Accuracy': 0.807792467948718}\n",
      "epoch: 29 step: 312, loss is 0.5786352\n",
      "{'Accuracy': 0.8171073717948718}\n",
      "epoch: 30 step: 312, loss is 0.5931815\n",
      "{'Accuracy': 0.8344350961538461}\n",
      "epoch: 31 step: 312, loss is 0.60235614\n",
      "{'Accuracy': 0.8275240384615384}\n",
      "epoch: 32 step: 312, loss is 0.368231\n",
      "{'Accuracy': 0.8303285256410257}\n",
      "epoch: 33 step: 312, loss is 0.94693434\n",
      "{'Accuracy': 0.8308293269230769}\n",
      "epoch: 34 step: 312, loss is 0.47100002\n",
      "{'Accuracy': 0.8423477564102564}\n",
      "epoch: 35 step: 312, loss is 0.6546564\n",
      "{'Accuracy': 0.8365384615384616}\n",
      "epoch: 36 step: 312, loss is 0.46686667\n",
      "{'Accuracy': 0.8471554487179487}\n",
      "epoch: 37 step: 312, loss is 0.3865126\n",
      "{'Accuracy': 0.8457532051282052}\n",
      "epoch: 38 step: 312, loss is 0.40539384\n",
      "{'Accuracy': 0.8322315705128205}\n",
      "epoch: 39 step: 312, loss is 0.44615257\n",
      "{'Accuracy': 0.8483573717948718}\n",
      "epoch: 40 step: 312, loss is 0.45208472\n",
      "{'Accuracy': 0.8486578525641025}\n",
      "epoch: 41 step: 312, loss is 0.4717011\n",
      "{'Accuracy': 0.8510616987179487}\n",
      "epoch: 42 step: 312, loss is 0.484349\n",
      "{'Accuracy': 0.8603766025641025}\n",
      "epoch: 43 step: 312, loss is 0.50808823\n",
      "{'Accuracy': 0.8688902243589743}\n",
      "epoch: 44 step: 312, loss is 0.2444193\n",
      "{'Accuracy': 0.8657852564102564}\n",
      "epoch: 45 step: 312, loss is 0.44857174\n",
      "{'Accuracy': 0.8611778846153846}\n",
      "epoch: 46 step: 312, loss is 0.47406247\n",
      "{'Accuracy': 0.8576722756410257}\n",
      "epoch: 47 step: 312, loss is 0.50422096\n",
      "{'Accuracy': 0.8683894230769231}\n",
      "epoch: 48 step: 312, loss is 0.38480416\n",
      "{'Accuracy': 0.8716947115384616}\n",
      "epoch: 49 step: 312, loss is 0.3262323\n",
      "{'Accuracy': 0.8619791666666666}\n",
      "epoch: 50 step: 312, loss is 0.34584183\n",
      "{'Accuracy': 0.8810096153846154}\n",
      "epoch: 51 step: 312, loss is 0.47978282\n",
      "{'Accuracy': 0.8744991987179487}\n",
      "epoch: 52 step: 312, loss is 0.2703614\n",
      "{'Accuracy': 0.8881209935897436}\n",
      "epoch: 53 step: 312, loss is 0.32773823\n",
      "{'Accuracy': 0.8759014423076923}\n",
      "epoch: 54 step: 312, loss is 0.49481535\n",
      "{'Accuracy': 0.8822115384615384}\n",
      "epoch: 55 step: 312, loss is 0.5637473\n",
      "{'Accuracy': 0.8866185897435898}\n",
      "epoch: 56 step: 312, loss is 0.35517296\n",
      "{'Accuracy': 0.8817107371794872}\n",
      "epoch: 57 step: 312, loss is 0.48494092\n",
      "{'Accuracy': 0.8829126602564102}\n",
      "epoch: 58 step: 312, loss is 0.31654054\n",
      "{'Accuracy': 0.8958333333333334}\n",
      "epoch: 59 step: 312, loss is 0.3120981\n",
      "{'Accuracy': 0.889823717948718}\n",
      "epoch: 60 step: 312, loss is 0.77931935\n",
      "{'Accuracy': 0.8916266025641025}\n",
      "epoch: 61 step: 312, loss is 0.25539473\n",
      "{'Accuracy': 0.8990384615384616}\n",
      "epoch: 62 step: 312, loss is 0.35743648\n",
      "{'Accuracy': 0.8947315705128205}\n",
      "epoch: 63 step: 312, loss is 0.30885863\n",
      "{'Accuracy': 0.8969350961538461}\n",
      "epoch: 64 step: 312, loss is 0.29507613\n",
      "{'Accuracy': 0.8977363782051282}\n",
      "epoch: 65 step: 312, loss is 0.29192814\n",
      "{'Accuracy': 0.901542467948718}\n",
      "epoch: 66 step: 312, loss is 0.2831882\n",
      "{'Accuracy': 0.9042467948717948}\n",
      "epoch: 67 step: 312, loss is 0.27889985\n",
      "{'Accuracy': 0.9067508012820513}\n",
      "epoch: 68 step: 312, loss is 0.48581055\n",
      "{'Accuracy': 0.9071514423076923}\n",
      "epoch: 69 step: 312, loss is 0.5714812\n",
      "{'Accuracy': 0.8902243589743589}\n",
      "epoch: 70 step: 312, loss is 0.34306845\n",
      "{'Accuracy': 0.9028445512820513}\n",
      "epoch: 71 step: 312, loss is 0.4931106\n",
      "{'Accuracy': 0.910957532051282}\n",
      "epoch: 72 step: 312, loss is 0.3540363\n",
      "{'Accuracy': 0.9086538461538461}\n",
      "epoch: 73 step: 312, loss is 0.26690733\n",
      "{'Accuracy': 0.9126602564102564}\n",
      "epoch: 74 step: 312, loss is 0.30992138\n",
      "{'Accuracy': 0.9117588141025641}\n",
      "epoch: 75 step: 312, loss is 0.23805264\n",
      "{'Accuracy': 0.9047475961538461}\n",
      "epoch: 76 step: 312, loss is 0.28517282\n",
      "{'Accuracy': 0.9129607371794872}\n",
      "epoch: 77 step: 312, loss is 0.249605\n",
      "{'Accuracy': 0.9190705128205128}\n",
      "epoch: 78 step: 312, loss is 0.7650362\n",
      "{'Accuracy': 0.9142628205128205}\n",
      "epoch: 79 step: 312, loss is 0.2200245\n",
      "{'Accuracy': 0.921875}\n",
      "epoch: 80 step: 312, loss is 0.39616066\n",
      "{'Accuracy': 0.9200721153846154}\n",
      "epoch: 81 step: 312, loss is 0.22683412\n",
      "{'Accuracy': 0.9127604166666666}\n",
      "epoch: 82 step: 312, loss is 0.15766045\n",
      "{'Accuracy': 0.921875}\n",
      "epoch: 83 step: 312, loss is 0.30624664\n",
      "{'Accuracy': 0.9116586538461539}\n",
      "epoch: 84 step: 312, loss is 0.13320094\n",
      "{'Accuracy': 0.9140625}\n",
      "epoch: 85 step: 312, loss is 0.4179467\n",
      "{'Accuracy': 0.9211738782051282}\n",
      "epoch: 86 step: 312, loss is 0.13844976\n",
      "{'Accuracy': 0.9244791666666666}\n",
      "epoch: 87 step: 312, loss is 0.12879965\n",
      "{'Accuracy': 0.9237780448717948}\n",
      "epoch: 88 step: 312, loss is 0.36271775\n",
      "{'Accuracy': 0.9160657051282052}\n",
      "epoch: 89 step: 312, loss is 0.27919987\n",
      "{'Accuracy': 0.9273838141025641}\n",
      "epoch: 90 step: 312, loss is 0.15769361\n",
      "{'Accuracy': 0.9333934294871795}\n",
      "epoch: 91 step: 312, loss is 0.3243783\n",
      "{'Accuracy': 0.9266826923076923}\n",
      "epoch: 92 step: 312, loss is 0.16334465\n",
      "{'Accuracy': 0.9263822115384616}\n",
      "epoch: 93 step: 312, loss is 0.2720704\n",
      "{'Accuracy': 0.9294871794871795}\n",
      "epoch: 94 step: 312, loss is 0.43370464\n",
      "{'Accuracy': 0.9306891025641025}\n",
      "epoch: 95 step: 312, loss is 0.45572412\n",
      "{'Accuracy': 0.9365985576923077}\n",
      "epoch: 96 step: 312, loss is 0.45010567\n",
      "{'Accuracy': 0.9333934294871795}\n",
      "epoch: 97 step: 312, loss is 0.16789666\n",
      "{'Accuracy': 0.9301883012820513}\n",
      "epoch: 98 step: 312, loss is 0.3021208\n",
      "{'Accuracy': 0.9286858974358975}\n",
      "epoch: 99 step: 312, loss is 0.14036375\n",
      "{'Accuracy': 0.9320913461538461}\n",
      "epoch: 100 step: 312, loss is 0.16794595\n",
      "{'Accuracy': 0.9324919871794872}\n"
     ]
    }
   ],
   "source": [
    "ata_path = os.path.join(current_path, 'data/10-batches-bin')\n",
    "batch_size=32\n",
    "status=\"train\"\n",
    "\n",
    "# 生成训练数据集\n",
    "cifar_ds = get_data(data_path)\n",
    "ds_train = process_dataset(cifar_ds,batch_size =batch_size, status=status)\n",
    "network = LeNet5_2(10)\n",
    "#network = resnet50(10)\n",
    "# 返回当前设备\n",
    "device_target = mindspore.context.get_context('device_target')\n",
    "# 确定图模型是否下沉到芯片上\n",
    "dataset_sink_mode = True if device_target in ['Ascend','GPU'] else False\n",
    "# 设置模型的设备与图的模式\n",
    "context.set_context(mode=context.GRAPH_MODE, device_target=device_target)\n",
    "# 使用交叉熵函数作为损失函数\n",
    "net_loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction=\"mean\")\n",
    "# 优化器为momentum\n",
    "#net_opt = nn.Momentum(params=network.trainable_params(), learning_rate=0.01, momentum=0.9)\n",
    "net_opt = nn.Adam(params=network.trainable_params(), learning_rate=0.001)\n",
    "# 时间监控，反馈每个epoch的运行时间\n",
    "time_cb = TimeMonitor(data_size=ds_train.get_dataset_size())\n",
    "# 设置callback函数。\n",
    "config_ck = CheckpointConfig(save_checkpoint_steps=1562,\n",
    "                             keep_checkpoint_max=10)\n",
    "ckpoint_cb = ModelCheckpoint(prefix=\"checkpoint_lenet_2_verified\",directory='./results', config=config_ck)\n",
    "# 建立可训练模型\n",
    "model = Model(network = network, loss_fn=net_loss,optimizer=net_opt, metrics={\"Accuracy\": Accuracy()})\n",
    "eval_per_epoch = 1\n",
    "epoch_per_eval = {\"epoch\": [], \"acc\": []}\n",
    "eval_cb = EvalCallBack(model, ds_train, eval_per_epoch, epoch_per_eval)\n",
    "print(\"============== Starting Training ==============\")\n",
    "\n",
    "model.train(100, ds_train,callbacks=[ckpoint_cb, LossMonitor(per_print_times=1),eval_cb],dataset_sink_mode=dataset_sink_mode)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8663e0c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test results: {'Accuracy': 0.7110376602564102}\n"
     ]
    }
   ],
   "source": [
    "data_path = os.path.join(current_path, 'data/10-verify-bin')\n",
    "batch_size=32\n",
    "status=\"test\"\n",
    "# 生成测试数据集\n",
    "cifar_ds = ds.Cifar10Dataset(data_path)\n",
    "ds_eval = process_dataset(cifar_ds,batch_size=batch_size,status=status)\n",
    "\n",
    "res = model.eval(ds_eval, dataset_sink_mode=dataset_sink_mode)\n",
    "# 评估测试集\n",
    "print('test results:',res)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "04ef96af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x1000 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#创建图像标签列表\n",
    "category_dict = {0:'airplane',1:'automobile',2:'bird',3:'cat',4:'deer',5:'dog',\n",
    "                 6:'frog',7:'horse',8:'ship',9:'truck'}\n",
    "\n",
    "cifar_ds = get_data('./data/10-verify-bin')\n",
    "df_test = process_dataset(cifar_ds,batch_size=1,status='test')\n",
    "\n",
    "def normalization(data):\n",
    "    _range = np.max(data) - np.min(data)\n",
    "    return (data - np.min(data)) / _range\n",
    "\n",
    "# 设置图像大小\n",
    "plt.figure(figsize=(10,10))\n",
    "i = 1\n",
    "# 打印9张子图\n",
    "for dic in df_test:\n",
    "    # 预测单张图片\n",
    "    input_img = dic[0]    \n",
    "    output = model.predict(Tensor(input_img))\n",
    "    output = nn.Softmax()(output)\n",
    "    # 反馈可能性最大的类别\n",
    "    predicted = np.argmax(output.asnumpy(),axis=1)[0]\n",
    "    \n",
    "    # 可视化\n",
    "    plt.subplot(3,3,i)\n",
    "    # 删除batch维度\n",
    "    input_image = np.squeeze(input_img.asnumpy(),axis=0).transpose(1,2,0)\n",
    "    # 重新归一化，方便可视化\n",
    "    input_image = normalization(input_image)\n",
    "    plt.imshow(input_image)\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.axis('off')\n",
    "    plt.title('True label:%s,\\n Predicted:%s'%(category_dict[dic[1].asnumpy().sum()],category_dict[predicted]))\n",
    "    i +=1\n",
    "    if i > 9 :\n",
    "        break\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e31d6959",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "MindSpore",
   "language": "python",
   "name": "mindspore"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}