FC_MNIST_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import numpy as np\n",
    "from torch import optim\n",
    "from torch.autograd import Variable\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision.datasets import mnist\n",
    "from torchvision import transforms\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Net(nn.Module):\n",
    "    def __init__(self, in_c=784, out_c=10):\n",
    "        super(Net, self).__init__()\n",
    "\n",
    "        self.fc1 = nn.Linear(in_c, 512)\n",
    "        self.act1 = nn.ReLU(inplace=True)\n",
    "\n",
    "        self.fc2 = nn.Linear(512, 256)\n",
    "        self.act2 = nn.ReLU(inplace=True)\n",
    "\n",
    "        self.fc3 = nn.Linear(256, 128)\n",
    "        self.act3 = nn.ReLU(inplace=True)\n",
    "\n",
    "        self.fc4 = nn.Linear(128, out_c)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.act1(self.fc1(x))\n",
    "        x = self.act2(self.fc2(x))\n",
    "        x = self.act3(self.fc3(x))\n",
    "        x = self.fc4(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
      "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./data\\MNIST\\raw\\train-images-idx3-ubyte.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9912422/9912422 [01:47<00:00, 92433.37it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data\\MNIST\\raw\\train-images-idx3-ubyte.gz to ./data\\MNIST\\raw\n",
      "\n",
      "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\n",
      "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./data\\MNIST\\raw\\train-labels-idx1-ubyte.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28881/28881 [00:01<00:00, 27605.27it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data\\MNIST\\raw\\train-labels-idx1-ubyte.gz to ./data\\MNIST\\raw\n",
      "\n",
      "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\n",
      "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./data\\MNIST\\raw\\t10k-images-idx3-ubyte.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1648877/1648877 [00:06<00:00, 237156.12it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data\\MNIST\\raw\\t10k-images-idx3-ubyte.gz to ./data\\MNIST\\raw\n",
      "\n",
      "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\n",
      "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to ./data\\MNIST\\raw\\t10k-labels-idx1-ubyte.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4542/4542 [00:00<00:00, 13493.97it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data\\MNIST\\raw\\t10k-labels-idx1-ubyte.gz to ./data\\MNIST\\raw\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# 构建网络\n",
    "net = Net()\n",
    "\n",
    "# 准备数据集\n",
    "train_set = mnist.MNIST(\n",
    "    \"./data\", train=True, transform=transforms.ToTensor(), download=True\n",
    ")\n",
    "test_set = mnist.MNIST(\n",
    "    \"./data\", train=False, transform=transforms.ToTensor(), download=True\n",
    ")\n",
    "\n",
    "train_data = DataLoader(train_set, batch_size=64, shuffle=True)\n",
    "test_data = DataLoader(test_set, batch_size=128, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import random\n",
    "\n",
    "for i in range(4):\n",
    "    ax = plt.subplot(2, 2, i + 1)\n",
    "    idx = random.randint(0, len(train_set))\n",
    "    digit_0 = train_set[idx][0].numpy()\n",
    "    digit_0_image = digit_0.reshape(28, 28)\n",
    "    ax.imshow(digit_0_image, interpolation=\"nearest\")\n",
    "    ax.set_title(\"label: {}\".format(train_set[idx][1]), fontsize=10, color=\"black\")\n",
    "    ax.axis(\"off\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[INFO] Epoch-1-Batch-200: Train: Loss-2.2834, Accuracy-0.1875\n",
      "[INFO] Epoch-1-Batch-400: Train: Loss-2.2631, Accuracy-0.4531\n",
      "[INFO] Epoch-1-Batch-600: Train: Loss-2.1913, Accuracy-0.6719\n",
      "[INFO] Epoch-1-Batch-800: Train: Loss-2.0199, Accuracy-0.7188\n",
      "[INFO] Epoch-1: Train: Loss-2.1867, Accuracy-0.4446 | Test: Loss-1.7243, Accuracy-0.6717\n",
      "[INFO] Epoch-2-Batch-200: Train: Loss-1.1701, Accuracy-0.6094\n",
      "[INFO] Epoch-2-Batch-400: Train: Loss-0.9219, Accuracy-0.7188\n",
      "[INFO] Epoch-2-Batch-600: Train: Loss-0.7130, Accuracy-0.8281\n",
      "[INFO] Epoch-2-Batch-800: Train: Loss-0.6600, Accuracy-0.7812\n",
      "[INFO] Epoch-2: Train: Loss-0.8484, Accuracy-0.7831 | Test: Loss-0.4873, Accuracy-0.8597\n",
      "[INFO] Epoch-3-Batch-200: Train: Loss-0.5087, Accuracy-0.8594\n",
      "[INFO] Epoch-3-Batch-400: Train: Loss-0.3861, Accuracy-0.8750\n",
      "[INFO] Epoch-3-Batch-600: Train: Loss-0.3928, Accuracy-0.8906\n",
      "[INFO] Epoch-3-Batch-800: Train: Loss-0.4480, Accuracy-0.8750\n",
      "[INFO] Epoch-3: Train: Loss-0.4329, Accuracy-0.8750 | Test: Loss-0.3726, Accuracy-0.8940\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiQAAAHNCAYAAAAuUGCnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABRuklEQVR4nO3deVxU9f4/8NfMAMMOsi+iKIq7gJqEmmKhuESi3bK8pdeyxast8ru3K+WSWfm99/vNrK5meTVt1SylRbOIwt28ghgmKgoKyo7IwLDPnN8fw0ySoAzMzJnl9Xw85vGI4ZyZ9wnn8OLzPufzkQiCIICIiIhIRFKxCyAiIiJiICEiIiLRMZAQERGR6BhIiIiISHQMJERERCQ6BhIiIiISHQMJERERiY6BhIiIiETHQEJERESiYyAhIiIi0TGQENEtSSSSTj3S09O7/V51dXV4+eWXDfJaRGRZ7MQugIjM20cffdTm6w8//BCpqak3PT9o0KBuv1ddXR1WrVoFAIiNje326xGR5WAgIaJbeuSRR9p8fezYMaSmpt70PBFRd7BlQ0TdplarsW7dOgwZMgSOjo7w9/fHU089haqqqjbbnThxAvHx8fDx8YGTkxP69OmDxx57DABw6dIl+Pr6AgBWrVqlawW9/PLLpj4cIhIBR0iIqNueeuopbN26FfPnz8ezzz6L/Px8/Pvf/8bJkydx+PBh2Nvbo6ysDJMnT4avry+WLl0KT09PXLp0Cbt27QIA+Pr64t1338XChQsxc+ZMzJo1CwAwfPhwMQ+NiEyEgYSIuuXQoUP4z3/+g08++QRz5szRPT9x4kRMmTIFO3fuxJw5c3DkyBFUVVXhhx9+wKhRo3TbvfrqqwAAFxcX/OlPf8LChQsxfPhwtoSIbAxbNkTULTt37oSHhwcmTZqEiooK3WPkyJFwdXXFzz//DADw9PQEAHz77bdobm4WsWIiMkcMJETULbm5uaiuroafnx98fX3bPGpra1FWVgYAmDBhAu6//36sWrUKPj4+mDFjBj744AM0NjaKfAREZA7YsiGiblGr1fDz88Mnn3zS7ve1F6pKJBJ88cUXOHbsGL755ht8//33eOyxx/DGG2/g2LFjcHV1NWXZRGRmGEiIqFvCwsLw448/YuzYsXBycrrt9nfeeSfuvPNOvPbaa/j000/x5z//Gdu3b8eCBQsgkUhMUDERmSO2bIioWx588EGoVCqsXr36pu+1tLTg+vXrAICqqioIgtDm+5GRkQCga9s4OzsDgG4fIrIdHCEhom6ZMGECnnrqKaxZswZZWVmYPHky7O3tkZubi507d+Ktt97Cn/70J2zbtg0bNmzAzJkzERYWhpqaGmzatAnu7u6YNm0aAMDJyQmDBw/Gjh07EB4eDi8vLwwdOhRDhw4V+SiJyNgYSIio2zZu3IiRI0fivffew4svvgg7OzuEhobikUcewdixYwFogsvx48exfft2lJaWwsPDA6NHj8Ynn3yCPn366F7rP//5D5555hksWbIETU1NWLlyJQMJkQ2QCH8cQyUiIiIyMV5DQkRERKJjICEiIiLRMZAQERGR6BhIiIiISHQMJERERCQ6BhIiIiISHQMJERERiY6BhIiIiETHQEJERESiYyAhIiIi0TGQEBERkegYSIiIiEh0DCREREQkOgYSIiIiEh0DCREREYmOgYT095e/AImJYldBRBbu6FFAJgOmTxe7EjIHDCRERCSKzZuBZ54BDhwAiorErobExkBChrV/PzB6NCCXA4GBwNKlQEvL79//4gtg2DDAyQnw9gbi4gClUvO99HTNvi4ugKcnMHYscPmyGEdBREZWWwvs2AEsXKgZIdm6te33v/kGuOMOwNER8PEBZs78/XuNjcA//gGEhGhONf36acINWTYGEjKcq1eBadM0Z5FTp4B339WcJV59VfP94mLg4YeBxx4DcnI0AWTWLEAQNKElMRGYMAH49VfNWO6TTwISiZhHRERG8vnnwMCBwIABwCOPAFu2aE4FALBnjyaATJsGnDwJpKVp/lbRmjsX+Owz4O23NaeS994DXF3FOQ4yHIkgaP8JEHXSX/4CXL8OpKS0ff6ll4Avv9ScIbRBYsMGzZ8y1dVAVhYwciRw6RLQu3fbfa9d04yYpKdrQgkRWbWxY4EHHwSee07z90hgILBzJxAbC4wZA/TtC3z88c37nT+vCTGpqZoBVrIeHCEhw8nJAWJi2o5qjB2rGZu9cgWIiADuuUfTsnngAWDTJqCqSrOdl5cm6MTHAwkJwFtvaUZUiMjqnDsHHD+uGTAFADs7YPbs39suWVmaU0V7srI0F8Ly7xbrw0BCpiOTaf6s+e47YPBg4J13NH/q5Odrvv/BB5pWzZgxmuZyeDhw7Ji4NRORwW3erBkVCQrShBE7O02H98svNYOpTk4d73ur75FlYyAhwxk0SBMobuwCHj4MuLkBPXtqvpZINKMmq1ZpmsMODsDu3b9vHxUFJCcDR44AQ4cCn35q2mMgIqNqaQE+/BB44w3NaIf2ceqUJqB89hkwfLjmupH2DBsGqNWa6+fJutiJXQBZKO01ITd68klg3TrNfXyLF2vGZVeuBJKSAKkU+OUXzVlm8mTAz0/zdXm5Jsjk5wPvvw/cd5/mrHTuHJCbq7l6jYisxrffajq1jz8OeHi0/d7992tGT/73fzUtm7Aw4KGHNCFm717N5WihocC8eZpr499+W9MJvnwZKCvTXJNClouBhLomPV0zmnGjxx/XnDX+/nfNWcLLS/PcsmWa77u7ayYcWLcOUCg0F7a+8QYwdSpQWgqcPQts2wZUVmqucFu0CHjqKVMfGREZ0ebNmotR/xhGAE0g+de/NKeOnTuB1auB//kfzalj/Pjft3v3XeDFF4G//lVzuujVS/M1WTbeZUNERESi4zUkREREJDoGEiIiIhIdAwkRERGJjoGEiIiIRMdAQkRERKJjICEiIiLRWcQ8JGq1GkVFRXBzc4OEq78SmZwgCKipqUFQUBCkUsv4O4bnDSLx6XPusIhAUlRUhJCQELHLILJ5hYWF6KldBsDM8bxBZD46c+6wiEDi5uYGQHNA7u7uIldDZHsUCgVCQkJ0n0VLwPMGkfj0OXdYRCDRDre6u7vzxEIkIktqffC8QWQ+OnPusIxmMBGZvfXr1yM0NBSOjo6Ijo7G8ePHO9y2ubkZr7zyCsLCwuDo6IiIiAjs27evW69JRJaNgYSIum3Hjh1ISkrCypUrkZmZiYiICMTHx6OsrKzd7ZctW4b33nsP77zzDs6cOYOnn34aM2fOxMmTJ7v8mkRk2SxicT2FQgEPDw9UV1dz6JVIBLf7DEZHR+OOO+7Av//9bwCaO1xCQkLwzDPPYOnSpTdtHxQUhJdeegmLFi3SPXf//ffDyckJH3/8cZdeU9+aicj49PkccoSEiLqlqakJGRkZiIuL0z0nlUoRFxeHo0ePtrtPY2MjHB0d2zzn5OSEQ4cOdes1FQpFmwcRWQ4GEiLqloqKCqhUKvj7+7d53t/fHyUlJe3uEx8fj7Vr1yI3NxdqtRqpqanYtWsXiouLu/yaa9asgYeHh+7BW36JLAsDCRGZ3FtvvYX+/ftj4MCBcHBwwOLFizF//vxuTbqWnJyM6upq3aOwsNCAFRORsTGQEFG3+Pj4QCaTobS0tM3zpaWlCAgIaHcfX19fpKSkQKlU4vLlyzh79ixcXV3Rt2/fLr+mXC7X3eLLW32JLA8DCRF1i4ODA0aOHIm0tDTdc2q1GmlpaYiJibnlvo6OjggODkZLSwu+/PJLzJgxo9uvSUSWySImRiMi85aUlIR58+Zh1KhRGD16NNatWwelUon58+cDAObOnYvg4GCsWbMGAPDLL7/g6tWriIyMxNWrV/Hyyy9DrVbjhRde6PRrEpF1YSAhom6bPXs2ysvLsWLFCpSUlCAyMhL79u3TXZRaUFDQ5vqQhoYGLFu2DHl5eXB1dcW0adPw0UcfwdPTs9OvSUTWhfOQENFtWeJn0BJrJrI2nIeEiIiILIrVBBJlYwu2Hs7H4k8zYQGDPkRERBbpYnktntt+Egdzyw36ulZzDYlEAvzPvrNoaFbjqfFhGNbTQ+ySiIiIrMalCiXeTstFStZVqAWg8Fod7urva7DXt5pA4uxgh7sH+mFvdgn2ZBczkBARERlAQWUd3vkpF7tOXoVKrelAxA3yw/Nx4QZ9H6sJJAAwfVhQayApwj+mDIBEIhG7JCIiIotUeK0O63++gC8yrqClNYhMHOCL5+PCERHiafD3s6pAMnGgLxztpSi8Vo/TVxUcJSEiItLT1ev1WP/zBew8UYhmlSaIjA/3xZK4/ojq1cNo72tVgYRtGyIioq4prq7Hhp8vYsd/C9GkUgMAxvXzwZJJ/TGyt5fR39+qAgnAtg0REZE+ShUNeDf9Ij49XoCmFk0QienrjSWTwjG6j/GDiJbVBRK2bYiIiG6vrKYBG9Pz8Mkvl9HYGkRGh3rh+Un9MSbMx+T1WF0gYduGiIioYxW1jXhv/0V8dOwyGpo1QWRk7x5ImhSOMWHeonUWrC6QAGzbEBER/dE1ZRPeP5CHbUcuob5ZBQCIDPFE0qRw3NXfR/TflVYZSNi2ISIi0rhe14RNB/Ow9fAlKJs0QWR4Tw8smRSO2HBf0YOIllUGErZtiIjI1lXXNWPzoTxsOXwJtY0tAIAhQe5ImhSOuwf6mU0Q0bLKQAKwbUNERLZJ0dCMLYfysflQPmoaNEFkUKA7no/rj8mD/c3296HVBhK2bYiIyJbUNDRj6+FL2HQwD4rWIDLA3w3Px/VH/JAASKXmGUS0rDaQsG1DRES2QNnYgq1HNEHkel0zAKC/nyuei+uPaUMDzT6IaFltIAHYtiEiIutV19SCD49exvsH8nBN2QQA6Ovrgufu6Y97hwdBZiFBRMuqAwnbNkREZG3qm1T4+NhlvHfgIipqNUEk1NsZz8X1x30RwRYXRLSsOpCwbUNERNaioVmFT34pwLvpF1FR2wgA6OXljGfv6Y/EyCDYyaQiV9g9Vh1IALZtiIjIsjU0q7D9eAE2pF9EWY0miPTs4YRn7+6PmSOCYW/hQUTL6gMJ2zZERGSJGltU+PzEFaz/6QJKFA0AgGBPJyy+ux/uH9ETDnbWEUS0rD6QsG1DRESWpKlFjS8yruDfP+WiqFoTRAI9HLFoYj88MKon5HYykSs0DqsPJADbNkREZP6aVWrsyryCd366gCtV9QAAf3c5/hrbD7PvCIGjvXUGES2bCCRs2xARkblqUamx++RVvPPTBRRcqwMA+LjK8dfYMMyJ7mX1QUTLJgIJ2zZERGRuVGoBX2VdxdtpubhUqQki3i4OWBgbhj9H94aTg20EES2bCCQA2zZERGQeVGoB3/5ahLfScpFXrgQAeLk44KnxffFoTG84O9jMr+Y2bOao2bYhIiIxqdUC9mQX4620XFwoqwUAeDrb48nxfTEvJhQucpv5ldwumzl6tm2IiEgMarWA738rwbofc3GutAYA4O5opwkiY0Lh5mgvcoXmwWYCCcC2DRERmY4gCPjhTCneTD2PsyWaIOLmaIcF4/pi/rhQuDOItKH3rCoHDhxAQkICgoKCIJFIkJKSctt9Ghsb8dJLL6F3796Qy+UIDQ3Fli1bulJvt/yxbUNERGRogiDgxzOluPedQ3jqowycLamBq9wOz97TH4f+cTeei+vPMNIOvUdIlEolIiIi8Nhjj2HWrFmd2ufBBx9EaWkpNm/ejH79+qG4uBhqtVrvYruLbRsiIjIWQRCQfq4cb/54Hr9eqQYAuDjI8JexoXjirr7wdHYQuULzpncgmTp1KqZOndrp7fft24f9+/cjLy8PXl5eAIDQ0FB939ZgtG2bvdnFbNsQEVG3CYKAA7kVeDP1PLIKrwMAnOxlmDcmFE+O7wsvFwaRzjD6NSRff/01Ro0ahX/961/46KOP4OLigvvuuw+rV6+Gk5NTu/s0NjaisbFR97VCYbj2irZtU3CtjnfbEBFRlwmCgMMXKvHmj+eRcbkKAOBoL8XcGE0Q8XGVi1yhZTF6IMnLy8OhQ4fg6OiI3bt3o6KiAn/9619RWVmJDz74oN191qxZg1WrVhmlHrZtiIiou45erMSbqedx/NI1AIDcTopH7uyNpyb0hZ+bo8jVWSajLxWoVqshkUjwySefYPTo0Zg2bRrWrl2Lbdu2ob6+vt19kpOTUV1drXsUFhYatKZpwwIBAHuziyEIgkFfm4iIrNcveZV46P2jeHjTMRy/dA0OdlL8ZUwoDr4wEcvvHcww0g1GHyEJDAxEcHAwPDx+H4kYNGgQBEHAlStX0L9//5v2kcvlkMuNN9R190A/tm2IiKjTTly6hjd/PI/DFyoBAA4yKR4aHYK/xvZDgAdDiCEYPZCMHTsWO3fuRG1tLVxdXQEA58+fh1QqRc+ePY399u1i24aIiDojs6AKb6aex8HcCgCAvUyCB0aFYNHEfgj2bP86SOoavVs2tbW1yMrKQlZWFgAgPz8fWVlZKCgoAKBpt8ydO1e3/Zw5c+Dt7Y358+fjzJkzOHDgAP7+97/jscce6/CiVlNg24aIiDpyqvA6/vLBcczacAQHcytgJ5XgoTtC8NP/i8XrM4cxjBiB3iMkJ06cwMSJE3VfJyUlAQDmzZuHrVu3ori4WBdOAMDV1RWpqal45plnMGrUKHh7e+PBBx/Eq6++aoDyu45tGyIi+qPTV6vxZup5pJ0tAwDIpBLMigrGM3f3Ry9vZ5Grs256B5LY2Nhbjihs3br1pucGDhyI1NRUfd/KqNi2ISIirTNFCqz78Tx+OFMKAJBKgMSoYDx7d3+E+riIXJ1tsKm1bP5o2rBATpJGRGTDzpYo8NaPufjudAkAQCIBZkQE4dl7+qOvr6vI1dkWmw4kbNsQEdmm3NIarEvLxZ5fiwFogsi9w4Pw3D390M/PTeTqbJNNBxK2bYiIbMuFslq8nZaLb34tgvbqg+nDAvFcXH+E+zOIiMmmAwnAtg0RkS3Ir1Di7bRcfJV1FerWIDJlSACei+uPQYHu4hZHABhI2LYhIrJilyuVeDvtAnafvKILIpMG++P5uP4YEsTzvTmx+UDCtg0RkfUpvFaHd37KxZeZV6FqTSL3DPTD83HhPM+bKZsPJADbNkRE1uJKVR3W/3wBO09cQUtrEJkQ7oslk8IRGeIpbnF0SwwkYNuGiMjSFVfXY/3PF7Djv4VoVmmCyF39ffB8XDhG9u4hcnXUGQwkYNuGiMhSlSoasOHnC/jseCGaVGoAwJgwbyyZFI47Qr1Ero70wUDSim0bIiLLUaZowLv7L+KTXwrQ1KIJIqP7eCFpUjju7OstcnXUFQwkrdi2ISIyf+U1jXhv/0V8dOwyGluDyB2hPbAkLhwxYd78Y9KCMZC0YtuGiMh8VdY24v0Defjw6GXUN6sAACN6eWLJpHCM6+fDIGIFGEhuwLYNEZF5qVI24f2Dedh25BLqmjRBJCLEE0vi+mNCuC/P01aEgeQGbNsQEZmP7ccLsPrbM1C2BpGhwe5ImhSOiQP8GESsEAPJDdi2ISIyD1XKJiz/6jSaVQIGB7pjyaRwxA1iELFmUrELMDfThgUCAPZmF0PQrrxEREQmtSe7GM0qAYMC3fHtM+MwabA/w4iVYyD5gz+2bYiIyPR2n7wKAJgVFQyplEHEFjCQ/IG2bQNoEjoREZlWQWUdMi5XQSoB7osMErscMhEGknawbUNEJJ6ULM3oyNh+PvB3dxS5GjIVBpJ23Ni2+a2IbRsiIlMRBEHXrkmMDBa5GjIlBpJ23Ni2+fZXtm2IiEzl1JVq5Fco4WgvRfzQALHLIRNiIOkA2zZERKaX0jo6Ej8kAK5yzkxhSxhIOsC2DRGRaTWr1PjmVBEAIDGK7Rpbw0DSAbZtiIhM62BuOSqVTfBxdcBd/XzELodMjIHkFti2ISIynd0nNaMjCRFBsJPx15Ot4U/8Fti2Ieq89evXIzQ0FI6OjoiOjsbx48dvuf26deswYMAAODk5ISQkBEuWLEFDQ4Pu+y+//DIkEkmbx8CBA419GCSSmoZm/PBbCQBgJts1NomB5BbYtiHqnB07diApKQkrV65EZmYmIiIiEB8fj7Kysna3//TTT7F06VKsXLkSOTk52Lx5M3bs2IEXX3yxzXZDhgxBcXGx7nHo0CFTHA6JYN/pEjS2qNHX1wXDgrmOmC1iILkNtm2Ibm/t2rV44oknMH/+fAwePBgbN26Es7MztmzZ0u72R44cwdixYzFnzhyEhoZi8uTJePjhh28aVbGzs0NAQIDu4ePD6wqslXYytFlRwVyzxkYxkNwG2zZEt9bU1ISMjAzExcXpnpNKpYiLi8PRo0fb3WfMmDHIyMjQBZC8vDzs3bsX06ZNa7Ndbm4ugoKC0LdvX/z5z39GQUFBh3U0NjZCoVC0eZBlKK6ux5GLlQCAGZwMzWYxkNwG2zZEt1ZRUQGVSgV/f/82z/v7+6OkpKTdfebMmYNXXnkF48aNg729PcLCwhAbG9umZRMdHY2tW7di3759ePfdd5Gfn4+77roLNTU17b7mmjVr4OHhoXuEhIQY7iDJqL7OKoIgAHeE9kCIl7PY5ZBIGEg6gW0bIsNKT0/H66+/jg0bNiAzMxO7du3Cnj17sHr1at02U6dOxQMPPIDhw4cjPj4ee/fuxfXr1/H555+3+5rJycmorq7WPQoLC011ONRN2qniZ0b1FLkSEhOnweuEP7ZthvKCKyIdHx8fyGQylJaWtnm+tLQUAQHtT/29fPlyPProo1iwYAEAYNiwYVAqlXjyySfx0ksvQSq9+W8lT09PhIeH48KFC+2+plwuh1wu7+bRkKnlFCtwtqQGDjIpprf+8Ue2iSMkncC2DVHHHBwcMHLkSKSlpemeU6vVSEtLQ0xMTLv71NXV3RQ6ZDIZAHQ4CllbW4uLFy8iMJC/tKyJdqr4iQN94eFsL3I1JCYGkk5i24aoY0lJSdi0aRO2bduGnJwcLFy4EEqlEvPnzwcAzJ07F8nJybrtExIS8O6772L79u3Iz89Hamoqli9fjoSEBF0w+dvf/ob9+/fj0qVLOHLkCGbOnAmZTIaHH35YlGMkw1OpBXyVpZkMje0aYsumk9i2IerY7NmzUV5ejhUrVqCkpASRkZHYt2+f7kLXgoKCNiMiy5Ytg0QiwbJly3D16lX4+voiISEBr732mm6bK1eu4OGHH0ZlZSV8fX0xbtw4HDt2DL6+viY/PjKOY3mVKFE0wN3RDhMH8udq6ySCBfy5r1Ao4OHhgerqari7u4tWx18/ycDe7BI8PSEMS6dyxkiyHebyGdSHJdZsa/628xS+yLiCh0f3wppZw8Quh4xAn88hWzZ6YNuGiMgw6ptU2Hdac1v4rBGce4QYSPTCSdKIiAwjNacUtY0t6NnDCSN79RC7HDIDDCR64N02RESGob27JjEyGFIpp4qnLgSSAwcOICEhAUFBQZBIJEhJSen0vocPH4adnR0iIyP1fVuzwbYNEVH3VNY2Yv/5cgBAIlf2pVZ6BxKlUomIiAisX79er/2uX7+OuXPn4p577tH3Lc0K2zZERN3zzakiqNQChvf0QD8/V7HLITOh922/U6dOxdSpU/V+o6effhpz5syBTCbTa1TF3GjbNnuzS/Dtr8W8/ZeISE+7W+ceSeRCenQDk1xD8sEHHyAvLw8rV67s1Pbmvmon2zZERF2TV16LU4XXIZNKkBARJHY5ZEaMHkhyc3OxdOlSfPzxx7Cz69yAjLmv2sm2DRFR16S0jo7c1d8Hvm5ce4h+Z9RAolKpMGfOHKxatQrh4eGd3s/cV+3k3TZERPoTBEF3d81MXsxKf2DUQFJTU4MTJ05g8eLFsLOzg52dHV555RWcOnUKdnZ2+Omnn9rdTy6Xw93dvc3D3LBtQ0Skn8yCKhRcq4OLgwyTB7e/EjTZLqOuZePu7o7s7Ow2z23YsAE//fQTvvjiC/Tp08eYb29UXNuGiEg/u1tHR+KHBsDJQSZyNWRu9A4ktbW1uHDhgu7r/Px8ZGVlwcvLC7169UJycjKuXr2KDz/8EFKpFEOHDm2zv5+fHxwdHW963tLceLfNnmzebUNEdCtNLWpdi5vtGmqP3i2bEydOICoqClFRUQA0y45HRUVhxYoVAIDi4mIUFBQYtkozpW3b7PmVbRsioltJP1eG63XN8HOTY0yYj9jlkBnSe4QkNjb2lr98t27desv9X375Zbz88sv6vq1ZYtuGiKhzUrI07ZoZkUGQcap4agfXsumGG++22ZPNu22IiNpTXd+MH3PKAHCqeOoYA0k3sW1DRHRr32UXo6lFjQH+bhgcaH53TZJ5YCDpJk6SRkR0a9q7axKjgiGRsF1D7WMg6Sa2bYiIOnalqg6/5F+DRKK5foSoIwwkBsC2DRFR+75qnSr+zj7eCPJ0ErkaMmcMJAbAtg0R0c0EQdC1azj3CN0OA4kBODvYYeIAtm2IiG70W5ECF8pqIbeTYsowThVPt8ZAYiDTh7NtQ0R0I+3oSNxgf7g72otcDZk7BhIDYduGiOh3LSo1vj6luX5kZiTbNXR7DCQGwrYNEdHvDl+sRHlNI3o422N8uK/Y5ZAFYCAxILZtiIg0UlrbNQkRQXCw468auj3+KzEgtm2IiABlYwv2nS4BwKniqfMYSAyIbRsiIuCHMyWob1Yh1NsZUSGeYpdDFoKBxMDYtiEiW7f7pOZiVk4VT/pgIDEwtm2IyJaV1TTgUG45ACCRd9eQHhhIDIxtGyKyZV9nFUEtAFG9PBHq4yJ2OWRBGEiMgG0bIrJVKVmau2tm8WJW0hMDiRGwbUNEtii3tAanrypgJ5Vg+nCu7Ev6YSAxArZtiMgWaaeKjx3gCy8XB5GrIUvDQGIkbNsQkS1RqwV8ldU6VXxUT5GrIUvEQGIkbNsQkS3576VruHq9Hm5yO9wzyE/scsgCMZAYCds2RGRLtO2aqcMC4GgvE7kaskQMJEbEtg0R2YKGZpXuDy+2a6irGEiMiG0bIrIFP58tQ01DCwI9HBHdx0vscshCMZAYEds2RGQLdrW2a2ZEBkMq5VTx1DUMJEambdvszWbbhoisT5WyCennygAAs0ZwMjTqOgYSI9O2bS5Xsm1DRNZnT3YxmlUCBge6I9zfTexyyIIxkBgZ2zZEZM20d9fM5FTx1E0MJCbAtg0RWaOCyjpkXK6CVALcF8mp4ql7GEhMgG0bIrJG2oX0xvbzgb+7o8jVkKVjIDEBtm2IyNoIgqBr1yRGsl1D3cdAYiJs2xCRNTl1pRr5FUo42cswZWiA2OWQFWAgMRG2bYjImqS0jo5MHuIPF7mdyNWQNWAgMRG2bYjIWjSr1PjmlGZl30TeXUMGwkBiQmzbEJE1OJhbjkplE3xcHXBXPx+xyyErwUBiQmzbEJE12H1SMzqSEBEEOxl/jZBh8F+SCbFtQ0SWrqahGT/8VgKAk6GRYTGQmNi0YWzbEJHl2ne6BI0taoT5umBYsIfY5ZAVYSAxMbZtiMiSaSdDmxkVDImEK/uS4egdSA4cOICEhAQEBQVBIpEgJSXlltvv2rULkyZNgq+vL9zd3RETE4Pvv/++q/VaPBc52zZEZJmKq+tx5GIlAGAGJ0MjA9M7kCiVSkRERGD9+vWd2v7AgQOYNGkS9u7di4yMDEycOBEJCQk4efKk3sVaC7ZtiMgSfZ1VBEEARod6IcTLWexyyMroPZvN1KlTMXXq1E5vv27dujZfv/766/jqq6/wzTffICoqSt+3twp/bNsMZR+WiCyAbqp4XsxKRmDya0jUajVqamrg5eXV4TaNjY1QKBRtHtaEbRsisjQ5xQqcLamBg0yK6a2jvESGZPJA8n//93+ora3Fgw8+2OE2a9asgYeHh+4REhJiwgpNg20bIrIk2qni7x7oBw9ne5GrIWtk0kDy6aefYtWqVfj888/h5+fX4XbJycmorq7WPQoLC01YpWnwbhsishQqtYCvsjhVPBmXyQLJ9u3bsWDBAnz++eeIi4u75bZyuRzu7u5tHtaGbRsishTH8ipRomiAh5M9Jg70FbscslImCSSfffYZ5s+fj88++wzTp083xVtaBLZtiMgSaC9mnT48EHI7mcjVkLXS+y6b2tpaXLhwQfd1fn4+srKy4OXlhV69eiE5ORlXr17Fhx9+CEDTppk3bx7eeustREdHo6REM+Wwk5MTPDxs++4S3m1DROauvkmFfac5VTwZn94jJCdOnEBUVJTult2kpCRERUVhxYoVAIDi4mIUFBTotn///ffR0tKCRYsWITAwUPd47rnnDHQIlottGyIyd6k5pahtbEHPHk4Y2auH2OWQFdN7hCQ2NvaW7YWtW7e2+To9PV3ft7Ap04YF4rvTJdibXYwX4gdwKmYiMivau2tmRgVDKuX5iYyHa9mIjHfbkLVYv349QkND4ejoiOjoaBw/fvyW269btw4DBgyAk5MTQkJCsGTJEjQ0NHTrNcmwKmsbsf98OQBOFU/Gx0AiMrZtyBrs2LEDSUlJWLlyJTIzMxEREYH4+HiUlZW1u/2nn36KpUuXYuXKlcjJycHmzZuxY8cOvPjii11+TTK8b04VQaUWMLynB/r5uYpdDlk5BhIzwLttyNKtXbsWTzzxBObPn4/Bgwdj48aNcHZ2xpYtW9rd/siRIxg7dizmzJmD0NBQTJ48GQ8//HCbERB9X5MMb3fr3CO8mJVMgYHEDLBtQ5asqakJGRkZbeYXkkqliIuLw9GjR9vdZ8yYMcjIyNAFkLy8POzduxfTpk3r8muSYeWV1+JU4XXIpBLcOzxI7HLIBuh9USsZnrZt893pEuzJLubtv2RRKioqoFKp4O/v3+Z5f39/nD17tt195syZg4qKCowbNw6CIKClpQVPP/20rmXTlddsbGxEY2Oj7mtrWwPL1LQXs97V3we+bnKRqyFbwBESM8G2DdmS9PR0vP7669iwYQMyMzOxa9cu7NmzB6tXr+7ya9rCGlimIggCdmf9fncNkSkwkJgJtm3IUvn4+EAmk6G0tLTN86WlpQgICGh3n+XLl+PRRx/FggULMGzYMMycOROvv/461qxZA7Va3aXXtIU1sEwls6AKhdfq4eIgw+TB7f//JjI0BhIzwbttyFI5ODhg5MiRSEtL0z2nVquRlpaGmJiYdvepq6uDVNr29COTaaYkFwShS69pC2tgmcquTM3oSPzQADg5cKp4Mg0GEjPCtg1ZqqSkJGzatAnbtm1DTk4OFi5cCKVSifnz5wMA5s6di+TkZN32CQkJePfdd7F9+3bk5+cjNTUVy5cvR0JCgi6Y3O41yTiaWtT49lfNH0WzonqKXA3ZEl7Uaka4tg1ZqtmzZ6O8vBwrVqxASUkJIiMjsW/fPt1FqQUFBW1GRJYtWwaJRIJly5bh6tWr8PX1RUJCAl577bVOvyYZR/q5MlTXN8PPTY6YMG+xyyEbIhEs4E9xhUIBDw8PVFdXW/0w7MKPM/Dd6RIsjA3DP6YMFLscIgCW+Rm0xJrNgfYc9MRdffDS9MFil0MWTp/PIVs2ZoZtGyISS3V9M9JyNDPhzmS7hkyMgcTM8G4bIhLLd9nFaFKpMcDfDYMC3cQuh2wMA4mZ4d02RCSWXa2ToSVGBXPlcTI5BhIzxLYNEZnalao6HM+/BokEmBHJqeLJ9BhIzBDbNkRkal+1LqR3Zx9vBHk6iVwN2SIGEjPEtg0RmZIgCNh9klPFk7gYSMwU2zZEZCq/FSlwoawWcjsppgzjVPEkDgYSM3X3QD/I7di2ISLj046OxA32h7ujvcjVkK1iIDFTLnI73D2QbRsiMq4WlVp3/cjMSLZrSDwMJGaMbRsiMrbDFytRUduIHs72mDDAV+xyyIYxkJgxtm2IyNhSWts1CRFBsJfxVwKJh//6zBjbNkRkTMrGFuw7XQJAMxkakZgYSMwc2zZEZCw/nClBfbMKod7OiArxFLscsnEMJGaObRsiMpbdJzUXs3KqeDIHDCRmjm0bIjKGMkUDDuWWAwASeXcNmQEGEgvAtg0RGdrXp4qgFoARvTwR6uMidjlEDCSWgG0bIjK0lCxOFU/mhYHEArBtQ0SGlFtag9NXFbCTSjB9OFf2JfPAQGIh2LYhIkPRThUfO8APXi4OIldDpMFAYiHYtiEiQ1Crhd+nime7hswIA4mFuLFts5dtGyLqouOXruHq9Xq4ye1wzyA/scsh0mEgsSDats0etm2IqIu0U8VPGxYIR3uZyNUQ/Y6BxIKwbUNE3dHQrNJdGM+p4sncMJBYELZtiKg7fjpbhpqGFgR5OCK6j5fY5RC1wUBiYdi2IaKu0t5dMyMqGFIpp4on88JAYmHYtiGirqhSNiH9XBkA3l1D5omBxMKwbUNEXfFtdjGaVQIGB7oj3N9N7HKIbqJ3IDlw4AASEhIQFBQEiUSClJSU2+6Tnp6OESNGQC6Xo1+/fti6dWsXSiUttm2ISF/au2tmjeDoCJknvQOJUqlEREQE1q9f36nt8/PzMX36dEycOBFZWVl4/vnnsWDBAnz//fd6F0sabNsQkT4KKuuQcbkKUgmQEMGp4sk82em7w9SpUzF16tROb79x40b06dMHb7zxBgBg0KBBOHToEN58803Ex8fr+/aE39s2350uwd7sYgwN9hC7JCIyY9qLWcf284G/u6PI1RC1z+jXkBw9ehRxcXFtnouPj8fRo0eN/dZWjW0bIuoMQRC4si9ZBKMHkpKSEvj7+7d5zt/fHwqFAvX19e3u09jYCIVC0eZBbbFtQ0SdcepKNfIrlHCylyF+SIDY5RB1yCzvslmzZg08PDx0j5CQELFLMju824aIOmN35hUAwOQh/nCR692lJzIZoweSgIAAlJaWtnmutLQU7u7ucHJyanef5ORkVFdX6x6FhYXGLtMisW1DRLfSrFLjm181f7CwXUPmzuhxOSYmBnv37m3zXGpqKmJiYjrcRy6XQy6XG7s0i/fHtg0vbiWiGx3MLcc1ZRN8XB0wrp+P2OUQ3ZLeIyS1tbXIyspCVlYWAM1tvVlZWSgoKACgGd2YO3eubvunn34aeXl5eOGFF3D27Fls2LABn3/+OZYsWWKYI7BhLnI7TBzAtg0RtW9XpuZi1oSIINjJzLJDT6Sj97/QEydOICoqClFRUQCApKQkREVFYcWKFQCA4uJiXTgBgD59+mDPnj1ITU1FREQE3njjDfznP//hLb8GMn042zZEdLOahmakntG0y2dF9RS5GqLb07tlExsbe8tffO3NwhobG4uTJ0/q+1bUCWzbEFF79p0uQWOLGmG+Lhga7C52OUS3xTE8C8e2DRG1RzsZ2syoYEgkXNmXzB8DiRVg24aIblRcXY+jeZUAgBmRvLuGLAMDiRXgJGlEdKOvs4ogCMDoUC+EeDmLXQ5RpzCQWAG2bYjoRtp2TSLnHiELwkBiJdi2ISIAyClW4GxJDRxkUkxvnTyRyBIwkFgJtm2ICABSWkdH7h7oBw9ne5GrIeo8BhIrwbYNEanUv6/sy3YNWRoGEivCtg2RbTuWV4lSRSM8nOwxcaCv2OUQ6YWBxIqwbUNk27QXs04fHgi5nUzkaoj0w0BiRdi2IbJd9U0q7DtdAoAr+5JlYiCxMtq2zV62bYhsSmpOKWobW9CzhxNG9e4hdjlEemMgsTLats0ltm2IbEoKp4onC8dAYmXYtiGyPZW1jdh/vhwA764hy8VAYoXYtiGyLd+cKoJKLSCipwfCfF3FLoeoSxhIrBDbNkS2ZXdWEQCOjpBlYyCxQmzbENmOvPJanCq8DplUgoSIILHLIeoyBhIrxbYNkW3QXsw6vr8PfFzlIldD1HUMJFaKbRsi6ycIAnZzqniyEgwkVoptGyLrl1lQhcJr9XBxkGHy4ACxyyHqFgYSK8a2DZF125WpGR2ZMjQQTg6cKp4sGwOJFWPbhsh6NbWo8e2vmtFPThVP1oCBxIqxbUNkvdLPlaG6vhn+7nLEhHmLXQ5RtzGQWDm2bYisk3Zl3xmRwZBJOVU8WT4GEivHtg2R9amub0ZaThkAIDGS7RqyDgwkVo5tGyLr8112MZpUagzwd8OgQDexyyEyCAYSG8C2DZF12aVd2XcEV/Yl68FAYgPYtiGyHleq6nA8/xokEuA+ThVPVoSBxAawbUOmsH79eoSGhsLR0RHR0dE4fvx4h9vGxsZCIpHc9Jg+fbpum7/85S83fX/KlCmmOBSz9lXrQnp39vFGkKeTyNUQGQ4DiY2YxrYNGdGOHTuQlJSElStXIjMzExEREYiPj0dZWVm72+/atQvFxcW6x+nTpyGTyfDAAw+02W7KlClttvvss89McThmSxAE3d01M0fwYlayLgwkNuIetm3IiNauXYsnnngC8+fPx+DBg7Fx40Y4Oztjy5Yt7W7v5eWFgIAA3SM1NRXOzs43BRK5XN5mux49epjicMzWb0UKXCirhdxOiilDOVU8WRcGEhvBtg0ZS1NTEzIyMhAXF6d7TiqVIi4uDkePHu3Ua2zevBkPPfQQXFxc2jyfnp4OPz8/DBgwAAsXLkRlZWWHr9HY2AiFQtHmYW20oyNxg/3h7mgvcjVEhsVAYkPYtiFjqKiogEqlgr+/f5vn/f39UVJSctv9jx8/jtOnT2PBggVtnp8yZQo+/PBDpKWl4Z///Cf279+PqVOnQqVStfs6a9asgYeHh+4REhLS9YMyQy0qte76kVmcKp6skJ3YBZDp/LFtMzTYQ+ySiLB582YMGzYMo0ePbvP8Qw89pPvvYcOGYfjw4QgLC0N6ejruueeem14nOTkZSUlJuq8VCoVVhZLDFytRUduIHs72GB/uK3Y5RAbHERIbwrYNGYOPjw9kMhlKS0vbPF9aWoqAgFtf56BUKrF9+3Y8/vjjt32fvn37wsfHBxcuXGj3+3K5HO7u7m0e1iSltV2TEBEEexlP3WR9+K/axrBtQ4bm4OCAkSNHIi0tTfecWq1GWloaYmJibrnvzp070djYiEceeeS273PlyhVUVlYiMDCw2zVbGmVjC/ad1rS/uLIvWSsGEhvDu23IGJKSkrBp0yZs27YNOTk5WLhwIZRKJebPnw8AmDt3LpKTk2/ab/PmzUhMTIS3d9vVamtra/H3v/8dx44dw6VLl5CWloYZM2agX79+iI+PN8kxmZMfzpSgvlmFUG9nRIZ4il0OkVHwGhIbo23b7PutBHuzi3kdCRnE7NmzUV5ejhUrVqCkpASRkZHYt2+f7kLXgoICSKVt//45d+4cDh06hB9++OGm15PJZPj111+xbds2XL9+HUFBQZg8eTJWr14NuVxukmMyJ7tPai5mTYziVPFkvSSCBYzbKxQKeHh4oLq62ur6wmL4+lQRnv3sJEK9nfHz32J5gqPbssTPoCXW3J4yRQPuXJMGtQDs/3ssenu73H4nIjOhz+eQLRsbdGPb5kwx2zZE5uzrU0VQC8CIXp4MI2TVuhRI9FmzAgDWrVuHAQMGwMnJCSEhIViyZAkaGhq6VDB134132+z5lXfbEJmzlKzWqeJ5MStZOb0Dib5rVnz66adYunQpVq5ciZycHGzevBk7duzAiy++2O3iqet4tw2R+cstrcHpqwrYSSW4dzhX9iXrpncg0XfNiiNHjmDs2LGYM2cOQkNDMXnyZDz88MO3HVUh42Lbhsj8aaeKjx3ghx4uDiJXQ2RcegWSrqxZMWbMGGRkZOgCSF5eHvbu3Ytp06Z1+D62sCaF2Ni2ITJvarWgmyqe7RqyBXoFkq6sWTFnzhy88sorGDduHOzt7REWFobY2NhbtmysfU0Kc8G2DZH5On7pGq5er4eb3A73DPITuxwiozP6XTbp6el4/fXXsWHDBmRmZmLXrl3Ys2cPVq9e3eE+ycnJqK6u1j0KCwuNXaZNYtuGyHxpp4qfNiwQjvYykashMj69JkbrypoVy5cvx6OPPqpbyXPYsGFQKpV48skn8dJLL900WRKgWZPCFic/MrUbJ0nb82sxhgRxkjQic9DQrMKe1vWmEtmuIRuh1whJV9asqKuruyl0yGSatM82gfjYtiEyPz+dLUNNQwuCPBwR3cdL7HKITELvqeOTkpIwb948jBo1CqNHj8a6detuWrMiODgYa9asAQAkJCRg7dq1iIqKQnR0NC5cuIDly5cjISFBF0xIPH9s23CUhEh82rtrZkQFQyrlTMpkG/QOJPquWbFs2TJIJBIsW7YMV69eha+vLxISEvDaa68Z7iioy9i2ITIvVcompJ/TzOvEu2vIlnAtG+LaNnRblvgZtMSaAeCjY5exPOU0hgS5Y8+zd4ldDlG3cC0b0gvvtiEyH9q7azg6QraGgYQ4SRqRmSiorEPG5SpIJcB9EZwqnmwLAwkB4N02ROZAezHr2H4+8HN3FLkaItNiICEAbNsQiU0QBK7sSzaNgYQAsG1DJLZTV6qRX6GEk70M8UPan2iSyJoxkJAO2zZE4tmdeQUAED/EHy5yvWdkILJ4DCSkw7YNkTiaVWp88yuniifbxkBCOmzbEInjYG45rimb4OMqx7h+PmKXQyQKBhJqg20bItPblam5mPW+iCDYyXhaJtvEf/nUBts2RKZV09CM1DOaFdR5dw3ZMgYSaoNtGyLT2ne6BI0taoT5umBosOVMcU9kaAwkdBO2bYhMRzsZ2qwRPbmOFNk0BhK6Cds2RKZRXF2Po3mVADhVPBEDCd2EbRsi0/g6qwiCAIwO9UKIl7PY5RCJioGE2sW2DZHxads1M0fwYlYiBhJqF9s2RMaVU6zA2ZIaOMikmDY0UOxyiETHQELturFtszebbRsiQ0tpHR25e6AfPJztRa6GSHwMJNQhbdtmz69s2xAZkkp9w8q+bNcQAWAgoVtg24bIOI7lVaJU0QgPJ3vEDvAVuxwis8BAQh1i24bIOLQXs04fHgi5nUzkaojMAwMJ3RLbNkSGVd+kwnetAX8Wp4on0mEgoVti24bIsFJzSqFsUqFnDyeM7N1D7HKIzAYDCd0S2zZEhqW9u2ZmVDCniie6AQMJ3RbbNkSGUVHbiP3nywEAiWzXELXBQEK3xbYNkWF8e6oIKrWAiJ4eCPN1FbscIrPCQEK3xbYNkWHszioCwNERovYwkFCnsG1D1D0Xy2txqvA6ZFIJEriyL9FNGEioU9i2Ieqer1ovZh3f3wc+rnKRqyEyPwwk1Cls2xB1nSAI2N06VTzbNUTtYyChTmPbhqhrMi5XofBaPVwcZJg8OEDscojMEgMJdRrbNkRdo50qfsrQQDg5cKp4ovYwkFCnsW1DpL+mFjW+/VXzeZnJdg1RhxhISC9s2xDp5+dzZaiub4a/uxwxYd5il0NkthhISC9s2xDpRztV/IzIYMiknCqeqCMMJKQXtm2IOq+6vhlpOWUAgMRItmuIboWBhPTGtg1R5+zNLkaTSo2BAW4YHOQudjlEZo2BhPTGtg1R52jvruHcI0S3x0BCemPbhuj2rlTV4Xj+NUgkwIxIThVPdDsMJNQlbNsQ3dpXrQvpxfT1RqCHk8jVEJm/LgWS9evXIzQ0FI6OjoiOjsbx48dvuf3169exaNEiBAYGQi6XIzw8HHv37u1SwWQe2LYh6pggCNiVeQUA2zVEnaV3INmxYweSkpKwcuVKZGZmIiIiAvHx8SgrK2t3+6amJkyaNAmXLl3CF198gXPnzmHTpk0IDuaH1JK5yO0QO8AXANs2RH/0W5ECF8uVkNtJMXUop4on6gy9A8natWvxxBNPYP78+Rg8eDA2btwIZ2dnbNmypd3tt2zZgmvXriElJQVjx45FaGgoJkyYgIiIiG4XT+KaPlzTF2fbhqitXZmai1knDfaHm6O9yNUQWQa9AklTUxMyMjIQFxf3+wtIpYiLi8PRo0fb3efrr79GTEwMFi1aBH9/fwwdOhSvv/46VCpVh+/T2NgIhULR5kHmh20bopu1qNT4+pTm+hFOFU/UeXoFkoqKCqhUKvj7+7d53t/fHyUlJe3uk5eXhy+++AIqlQp79+7F8uXL8cYbb+DVV1/t8H3WrFkDDw8P3SMkJESfMslE2LYhutnhi5WoqG2El4sDxof7il0OkcUw+l02arUafn5+eP/99zFy5EjMnj0bL730EjZu3NjhPsnJyaiurtY9CgsLjV0mdZG2bbM3u4RtGyIAu1svZk0YHgh7GW9kJOosO3029vHxgUwmQ2lpaZvnS0tLERDQ/oVbgYGBsLe3h0z2+5LbgwYNQklJCZqamuDg4HDTPnK5HHK5XJ/SSCTatk1+hRJnihUYEuQhdklEolE2tuD73zTnR95dQ6QfveK7g4MDRo4cibS0NN1zarUaaWlpiImJaXefsWPH4sKFC1Cr1brnzp8/j8DAwHbDCFkWtm2IfvfDmRLUN6vQx8cFkSGeYpdDZFH0Hk9MSkrCpk2bsG3bNuTk5GDhwoVQKpWYP38+AGDu3LlITk7Wbb9w4UJcu3YNzz33HM6fP489e/bg9ddfx6JFiwx3FCQqtm2INLR31yRGBkMi4cq+RPrQq2UDALNnz0Z5eTlWrFiBkpISREZGYt++fboLXQsKCiCV/p5zQkJC8P3332PJkiUYPnw4goOD8dxzz+Ef//iH4Y6CRMW2DRFQpmjA4QsVAIDEKE4VT6QvvQMJACxevBiLFy9u93vp6ek3PRcTE4Njx4515a3IAmjbNt//Voq92cUMJGSTvj5VBLUAjOjlid7eLmKXQ2RxeAk4GQTbNmTrtCv7zhzRU+RKiCwTAwkZxB/bNkS25HxpDX4rUsBOKsG9wwLFLofIIjGQkEHwbhvSZ9HN2NhYSCSSmx7Tp0/XbSMIAlasWIHAwEA4OTkhLi4Oubm5pjgUvaW0jo7EDvBDDxfePUjUFQwkZDBs29gufRfd3LVrF4qLi3WP06dPQyaT4YEHHtBt869//Qtvv/02Nm7ciF9++QUuLi6Ij49HQ0ODqQ6rU9RqAV9laaaKnzWCc48QdRUDCRkM2za2S99FN728vBAQEKB7pKamwtnZWRdIBEHAunXrsGzZMsyYMQPDhw/Hhx9+iKKiIqSkpJjwyG7v+KVruHq9Hm5yO9w90E/scogsFgMJGQzbNrapK4tu/tHmzZvx0EMPwcVFc3dKfn4+SkpK2rymh4cHoqOjO/2apqJt10wbFghHe9lttiaijjCQkEGxbWN7urLo5o2OHz+O06dPY8GCBbrntPvp85pirBLe0KzCntbwPZPtGqJuYSAhg2LbhvS1efNmDBs2DKNHj+7W64ixSvhPZ8tQ09CCIA9HjA71Mvr7EVkzBhIyKLZtbE9XFt3UUiqV2L59Ox5//PE2z2v30+c1xVglXDv3yIyoYEilnCqeqDsYSMjg2LaxLV1ZdFNr586daGxsxCOPPNLm+T59+iAgIKDNayoUCvzyyy8dvqZcLoe7u3ubhzFVKZuQfk5zF9EsruxL1G0MJGRwbNvYHn0X3dTavHkzEhMT4e3t3eZ5iUSC559/Hq+++iq+/vprZGdnY+7cuQgKCkJiYqIpDum2vs0uRrNKwJAgd/T3dxO7HCKL16W1bIhuhWvb2B59F90EgHPnzuHQoUP44Ycf2n3NF154AUqlEk8++SSuX7+OcePGYd++fXB0dDT68XSG9u6amRwdITIIiWABY+oKhQIeHh6orq42+jAsGcbXp4rw7Gcn0cfHBT/9vwlcit3CWeJn0Jg1X65UYsL/pkMqAY4l3wM/d/MISUTmRp/PIVs2ZBRs25A1SzmpmZl1bD8fhhEiA2EgIaPg3TZkrQRBQEoW2zVEhsZAQkbDu23IGmUVXkd+hRJO9jLED7n1bc1E1HkMJGQ0bNuQNdJezBo/xB8uct4XQGQoDCRkNGzbkLVpVqnxza+af8uJbNcQGRQDCRkV2zZkTQ6cL8c1ZRN8XOUY189H7HKIrAoDCRkV2zZkTbRTxd8XEQQ7GU+fRIbETxQZ1Y1tmwXbTuCl3dn4/rcS1DQ0i1wZkX5qGpqRekaztg7vriEyPF6RRUY3LyYU+8+Xo7i6AZ/8UoBPfimAnVSCEb16YMIAX4zv74shQe5cnIzM2nenS9DYokY/P1cMDbaMyeGILAkDCRndmH4+yFg2CcfyKnHgfDkO5FYgv0KJ45eu4fila/jf78/B28UB4/r7YHx/X9wV7gM/N042ReblxqniOfMwkeExkJBJuMjtcM8gf9wzqHVtk8o67M8tx4Hz5ThyoQKVyiZ8lVWEr7I0M2AODnTH+HBfTAj3xcjePeBgx+4iiae4uh5H8yoBaK4fISLDYyAhUfTydsaj3r3x6J290axSI/NyFfafL8eB3HKcvqrAmWLNY+P+i3BxkCEmzFsXUHp7u4hdPtmYr7KKIAjA6D5eCPFyFrscIqvEQEKis5dJEd3XG9F9vfHClIGoqG3EodwK7D9fjoO55aiobcKPOWX4MacMANDb2xnj+/tifLgvYsK84crJqcjIuLIvkfHxTE5mx8dVjsSoYCRGBUOtFnCmWIEDueXYf64cGZercLmyDh9VXsZHxy7DXibByN49MD5cc3Hs4EBeHEuGlVOswNmSGjjIpJg2NFDscoisFgMJmTWpVIKhwR4YGuyBv8b2Q21jC45erMT+82U4cL4CBdfqcCzvGo7lXcO/9p2Dj6sc4/v7YHy4L8b194GPq1zsQyALpx0duWeQHzyc7UWuhsh6MZCQRXGV22HSYH9MGqy5OPZShVI3enI0rxIVtY3YdfIqdrX+EhkW7IHx4Zq7d0b07gF7TmZFelCpf1/Zl1PFExkXAwlZtFAfF4T6uGBuTCgaW1TIuFyFA+crcOB8Oc4UK5B9tRrZV6ux/ueLcJXbYcwNF8fy4kS6nWN5lShVNMLDyV43wR8RGQcDCVkNuZ0MY8J8MCbMB0unDkRZTQMOnq/AgdxyHMytwDVlE344U4ofWmfb7OPjggnhvhgf7oM7+3rD2YEfB2pLO1X8vcMDIbeTiVwNkXXjGZislp+bI+4f2RP3j+wJtVrA6aJqzcRs5yuQUVCF/Aol8iuU2HrkEhxkUowK7dEaUHwxMMCNk1/ZuPomFb5rXaWad9cQGR8DCdkEqVSC4T09MbynJxbf3R+KhmYcuVCJA62Ts12pqseRi5U4crESa747Cz83Oe7qrxk9uau/L7xcHMQ+BDKx1JxSKJtUCPFywsjePcQuh8jqMZCQTXJ3tMeUoQGYMjQAgiAgr0LZOnpSjmN511BW04gvM6/gy8wrkEiA4cEemluLw30RFeLJlV5tgG7ukUhOFU9kCgwkZPMkEgnCfF0R5uuK+WP7oKFZhROXqnSjJ2dLanDqSjVOXanGOz9dgJujHcaG+bQGFB/07MGLY61NRW0j9p8vBwDMYLuGyCQYSIj+wNFehnH9fTCuvw9enDYIpYoGHDhfjv3ny3HoQgWu1zVj328l2PdbCQAgzNdFN3pyZx9vODnw4kdL9+2pIqjUAiJ6eiDM11XscohsAgMJ0W34uzvigVEheGBUCFRqAdlXq3UB5WRBFS6WK3GxXIkPDl+Cg50U0X28dFPbh/u7crjfAu1uXeSRF7MSmQ4DCZEeZFIJIkM8ERniiWfv6Y/q+mYcuaBZd+fA+XIUVTfgYG4FDuZW4LW9OQhwd9RMzBbui3H9fODpzItjzd3F8lqcKrwOmVSCe7myL5HJdOnKvPXr1yM0NBSOjo6Ijo7G8ePHO7Xf9u3bIZFIkJiY2JW3JTI7Hk72mDosEP9z/3AcXno3fkwaj+X3Dsb4cF/I7aQoUTTg8xNXsPjTkxixOhWJ6w/jzdTzyLhchRaVWuzyqR1ftV7MOp5LDxCZlN4jJDt27EBSUhI2btyI6OhorFu3DvHx8Th37hz8/Pw63O/SpUv429/+hrvuuqtbBROZK4lEgn5+bujn54bHx2kujj2ef01z905uOc6X1iKr8DqyCq/jrbRcuDva6W4tHh/ui0APJ7EPweYJgoDdrVPFzxzRU+RqiGyLRBAEQZ8doqOjcccdd+Df//43AECtViMkJATPPPMMli5d2u4+KpUK48ePx2OPPYaDBw/i+vXrSElJ6fR7KhQKeHh4oLq6Gu7u7vqUS2Q2iq7X42CuZmK2g7nlUDS0tPl+fz9X3cRso/t4wdHefC6OtcTPYFdqPnHpGv608ShcHGQ4sWwSL1Am6iZ9Pod6jZA0NTUhIyMDycnJuuekUini4uJw9OjRDvd75ZVX4Ofnh8cffxwHDx7U5y2JrEaQpxNm39ELs+/ohRaVGqeuVOtGT04VXkduWS1yy2rxn0P5kNtJEd3XGxPCfTEh3Adhvrw41hS0U8VPGRrIMEJkYnoFkoqKCqhUKvj7+7d53t/fH2fPnm13n0OHDmHz5s3Iysrq9Ps0NjaisbFR97VCodCnTCKzZyeTYmTvHhjZuweWTArH9bomHLpQoZvavqT1VuMD58uxGkCQh6Pu1uKx/Xzg4WQv9iFYnaYWNb79VTNV/KwRvLuGyNSMepdNTU0NHn30UWzatAk+Pj6d3m/NmjVYtWqVESsjMi+ezg64d3gQ7h0eBEEQcL60Vjd68kv+NRRVN2D7fwux/b+Fujt9xrdefzK8pydkUo6edNfP58pQXd8Mf3c57uzrLXY5RDZHr0Di4+MDmUyG0tLSNs+XlpYiICDgpu0vXryIS5cuISEhQfecWq25s8DOzg7nzp1DWFjYTfslJycjKSlJ97VCoUBISIg+pRJZLIlEggEBbhgQ4IYnxvdFfZMKx/IrdSMmF8uVyLhchYzLVXjzx/PwdLbHuH6aC2MnhPvC391R7EOwSNqp4mdEBjPgEYlAr0Di4OCAkSNHIi0tTXfrrlqtRlpaGhYvXnzT9gMHDkR2dnab55YtW4aamhq89dZbHYYMuVwOuZy32xEBgJODDBMH+GHiAM1dbFeq6nAwtwL7z5Xj8EXNzLHf/lqsazcMDHDTtHf6++KOPj0gt+O1ELdTXd+MtJwyAJwMjUgserdskpKSMG/ePIwaNQqjR4/GunXroFQqMX/+fADA3LlzERwcjDVr1sDR0RFDhw5ts7+npycA3PQ8EXVOzx7OeHh0Lzw8WnNxbFbhdd3Msb9ercbZkhqcLanB+wfy4GQvw519vXTXn/T1ceHFse3Ym12MJpUaAwPcMCjQMu4iIrI2egeS2bNno7y8HCtWrEBJSQkiIyOxb98+3YWuBQUFkEq5EiqRKdjJpBgV6oVRoV5ImjwA15Sai2P3n9Ncf1Je04ifz5Xj53OaheJ69nDSjZ6M6ecNd0deHAv8fndNIkdHiESj9zwkYrDEORCIxCYIAs6W1OhGT05cqkLTDbPDyqQSjOjliQnhvph9Ry/4unXcJrXEz2Bna75SVYdx//wZEglwZOndnKCOyICMNg8JEVkOiUSCQYHuGBTojqcmhKGuqQXH8ipx4Lzm9uK8CiX+e6kK/71UhRmRtjsy8FXrQnoxfb0ZRohExEBCZCOcHexw90B/3D1Q014tvFaH/efLkVOsQIiXs8jViSd+SAAUDc2I7OkpdilENo2BhMhGhXg545E7e4tdhuj6+bkieeogscsgsnm8+pSIiIhEx0BCREREomMgISIiItExkBAREZHoGEiIiIhIdAwkREREJDoGEiIiIhIdAwkRERGJjoGEiIiIRMdAQkRERKJjICEiIiLRMZAQERGR6BhIiIiISHQWsdqvIAgAAIVCIXIlRLZJ+9nTfhYtAc8bROLT59xhEYGkpqYGABASEiJyJUS2raamBh4eHmKX0Sk8bxCZj86cOySCBfzJo1arUVRUBDc3N0gkkg63UygUCAkJQWFhIdzd3U1YoeFZ07EA1nU8tngsgiCgpqYGQUFBkEoto9Pb2fMGYJs/U0vAYzFfxjh3WMQIiVQqRc+ePTu9vbu7u1X8wAHrOhbAuo7H1o7FUkZGtPQ9bwC29zO1FDwW82XIc4dl/KlDREREVo2BhIiIiERnVYFELpdj5cqVkMvlYpfSbdZ0LIB1HQ+PxfpY0/8HHot5sqZjAYxzPBZxUSsRERFZN6saISEiIiLLxEBCREREomMgISIiItExkBAREZHoLC6QrF+/HqGhoXB0dER0dDSOHz9+y+137tyJgQMHwtHREcOGDcPevXtNVOnt6XMsW7duhUQiafNwdHQ0YbUdO3DgABISEhAUFASJRIKUlJTb7pOeno4RI0ZALpejX79+2Lp1q9Hr7Ax9jyU9Pf2mn4tEIkFJSYlpCr6FNWvW4I477oCbmxv8/PyQmJiIc+fO3XY/c/7MdAfPHTx3GBPPHd3/zFhUINmxYweSkpKwcuVKZGZmIiIiAvHx8SgrK2t3+yNHjuDhhx/G448/jpMnTyIxMRGJiYk4ffq0iSu/mb7HAmhmxCsuLtY9Ll++bMKKO6ZUKhEREYH169d3avv8/HxMnz4dEydORFZWFp5//nksWLAA33//vZErvT19j0Xr3LlzbX42fn5+Rqqw8/bv349Fixbh2LFjSE1NRXNzMyZPngylUtnhPub8mekOnjt47jA2njsM8JkRLMjo0aOFRYsW6b5WqVRCUFCQsGbNmna3f/DBB4Xp06e3eS46Olp46qmnjFpnZ+h7LB988IHg4eFhouq6DoCwe/fuW27zwgsvCEOGDGnz3OzZs4X4+HgjVqa/zhzLzz//LAAQqqqqTFJTd5SVlQkAhP3793e4jTl/ZrqD5w4PE1XXdTx3mC9TnTssZoSkqakJGRkZiIuL0z0nlUoRFxeHo0ePtrvP0aNH22wPAPHx8R1ubypdORYAqK2tRe/evRESEoIZM2bgt99+M0W5BmeuP5fuiIyMRGBgICZNmoTDhw+LXU67qqurAQBeXl4dbmONPxueO3juMGc8d/zOYgJJRUUFVCoV/P392zzv7+/fYc+tpKREr+1NpSvHMmDAAGzZsgVfffUVPv74Y6jVaowZMwZXrlwxRckG1dHPRaFQoL6+XqSquiYwMBAbN27El19+iS+//BIhISGIjY1FZmam2KW1oVar8fzzz2Ps2LEYOnRoh9uZ62emO3ju4LnDHPHccTOLWO2XgJiYGMTExOi+HjNmDAYNGoT33nsPq1evFrEy2zZgwAAMGDBA9/WYMWNw8eJFvPnmm/joo49ErKytRYsW4fTp0zh06JDYpZCJ8dxhnnjuuJnFjJD4+PhAJpOhtLS0zfOlpaUICAhod5+AgAC9tjeVrhzLH9nb2yMqKgoXLlwwRolG1dHPxd3dHU5OTiJVZTijR482q5/L4sWL8e233+Lnn39Gz549b7mtuX5muoPnjrZ47jBftn7usJhA4uDggJEjRyItLU33nFqtRlpaWpv0f6OYmJg22wNAampqh9ubSleO5Y9UKhWys7MRGBhorDKNxlx/LoaSlZVlFj8XQRCwePFi7N69Gz/99BP69Olz232s8WfDc0dbPHeYL5s/d3TxoltRbN++XZDL5cLWrVuFM2fOCE8++aTg6ekplJSUCIIgCI8++qiwdOlS3faHDx8W7OzshP/7v/8TcnJyhJUrVwr29vZCdna2WIego++xrFq1Svj++++FixcvChkZGcJDDz0kODo6Cr/99ptYh6BTU1MjnDx5Ujh58qQAQFi7dq1w8uRJ4fLly4IgCMLSpUuFRx99VLd9Xl6e4OzsLPz9738XcnJyhPXr1wsymUzYt2+fWIego++xvPnmm0JKSoqQm5srZGdnC88995wglUqFH3/8UaxD0Fm4cKHg4eEhpKenC8XFxbpHXV2dbhtL+sx0B88dPHcYG88d3f/MWFQgEQRBeOedd4RevXoJDg4OwujRo4Vjx47pvjdhwgRh3rx5bbb//PPPhfDwcMHBwUEYMmSIsGfPHhNX3DF9juX555/Xbevv7y9MmzZNyMzMFKHqm2lvX/vjQ1v/vHnzhAkTJty0T2RkpODg4CD07dtX+OCDD0xed3v0PZZ//vOfQlhYmODo6Ch4eXkJsbGxwk8//SRO8X/Q3nEAaPP/2tI+M93BcwfPHcbEc0f3PzOS1jcnIiIiEo3FXENCRERE1ouBhIiIiETHQEJERESiYyAhIiIi0TGQEBERkegYSIiIiEh0DCREREQkOgYSIiIiEh0DCREREYmOgYSIiIhEx0BCREREomMgISIiItH9f1n9+P06j0UyAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 定义损失函数\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "# 定义优化器\n",
    "optimizer = optim.SGD(net.parameters(), lr=1e-2, weight_decay=5e-4)\n",
    "\n",
    "# 开始训练\n",
    "losses = []\n",
    "acces = []\n",
    "eval_losses = []\n",
    "eval_acces = []\n",
    "num_epoch = 3\n",
    "\n",
    "for epoch in range(num_epoch):\n",
    "    train_loss = 0\n",
    "    train_acc = 0\n",
    "    net = net.train()\n",
    "    for batch, (img, label) in enumerate(train_data):\n",
    "        img = img.reshape(img.size(0), -1)\n",
    "        img = Variable(img)\n",
    "        label = Variable(label)\n",
    "\n",
    "        # 前向传播\n",
    "        out = net(img)\n",
    "        loss = criterion(out, label)\n",
    "\n",
    "        # 反向传播\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        # 记录误差\n",
    "        train_loss += loss.item()\n",
    "        # 计算分类的准确率\n",
    "        _, pred = out.max(1)\n",
    "        num_correct = (pred == label).sum().item()\n",
    "        acc = num_correct / img.shape[0]\n",
    "\n",
    "        if (batch + 1) % 200 == 0:\n",
    "            print(\n",
    "                \"[INFO] Epoch-{}-Batch-{}: Train: Loss-{:.4f}, Accuracy-{:.4f}\".format(\n",
    "                    epoch + 1, batch + 1, loss.item(), acc\n",
    "                )\n",
    "            )\n",
    "        train_acc += acc\n",
    "\n",
    "    losses.append(train_loss / len(train_data))\n",
    "    acces.append(train_acc / len(train_data))\n",
    "\n",
    "    eval_loss = 0\n",
    "    eval_acc = 0\n",
    "    # 测试集不训练\n",
    "    for img, label in test_data:\n",
    "        img = img.reshape(img.size(0), -1)\n",
    "        img = Variable(img)\n",
    "        label = Variable(label)\n",
    "\n",
    "        out = net(img)\n",
    "        loss = criterion(out, label)\n",
    "        # 记录误差\n",
    "        eval_loss += loss.item()\n",
    "\n",
    "        _, pred = out.max(1)\n",
    "        num_correct = (pred == label).sum().item()\n",
    "        acc = num_correct / img.shape[0]\n",
    "\n",
    "        eval_acc += acc\n",
    "    eval_losses.append(eval_loss / len(test_data))\n",
    "    eval_acces.append(eval_acc / len(test_data))\n",
    "\n",
    "    print(\n",
    "        \"[INFO] Epoch-{}: Train: Loss-{:.4f}, Accuracy-{:.4f} | Test: Loss-{:.4f}, Accuracy-{:.4f}\".format(\n",
    "            epoch + 1,\n",
    "            train_loss / len(train_data),\n",
    "            train_acc / len(train_data),\n",
    "            eval_loss / len(test_data),\n",
    "            eval_acc / len(test_data),\n",
    "        )\n",
    "    )\n",
    "\n",
    "plt.figure()\n",
    "plt.suptitle(\"Test\", fontsize=12)\n",
    "ax1 = plt.subplot(1, 2, 1)\n",
    "ax1.plot(eval_losses)\n",
    "ax1.set_title(\"Loss\", fontsize=10, color=\"r\")\n",
    "ax2 = plt.subplot(1, 2, 2)\n",
    "ax2.plot(eval_acces)\n",
    "ax2.set_title(\"Acc\", fontsize=10, color=\"b\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}