19_网络模型保存与读取.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. 网络模型保存(方式一)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "第一种方法是保存vgg16整个模型，这会同时保存“模型结果+模型参数”"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=1000, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torchvision\n",
    "import torch\n",
    "\n",
    "vgg16 = torchvision.models.vgg16(pretrained=False)\n",
    "torch.save(vgg16, \"./model/vgg16_method1.pth\")  # 保存方式一：模型结构 + 模型参数  \n",
    "\n",
    "print(vgg16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. 网络模型导入(方式一)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "因为我们同时将模型结构和模型参数都保存在了pth文件中了，所以直接torch.load()即可获得model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=1000, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "\n",
    "model = torch.load(\"./model/vgg16_method1.pth\")  # 保存方式一对应的加载模型\n",
    "\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. 网络模型保存(方式二)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "第二种方法则是vgg16.state_dict()，这样的话仅会保存vgg16模型的参数，而不保存模型的网络结构。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=1000, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torchvision\n",
    "import torch\n",
    "\n",
    "vgg16 = torchvision.models.vgg16(pretrained=False)\n",
    "torch.save(vgg16.state_dict(), \"./model/vgg16_method2.pth\")  # 保存方式二：仅保存模型参数（官方推荐），不再保存网络模型结构  \n",
    "\n",
    "print(vgg16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. 网络模型导入(方式二)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "这里，我们仅在pth文件中保存了模型参数，所以直接torch.load()的话，只会得到一个OrderedDict的权重值。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "OrderedDict([('features.0.weight', tensor([[[[-0.0614,  0.0060,  0.0894],\n",
      "          [-0.0568, -0.0307, -0.0981],\n",
      "          [ 0.0724,  0.0579,  0.0302]],\n",
      "\n",
      "         [[-0.0610, -0.0309, -0.0323],\n",
      "          [ 0.0390,  0.0306,  0.0067],\n",
      "          [-0.1225,  0.0532,  0.0606]],\n",
      "\n",
      "         [[-0.1396, -0.0278,  0.0215],\n",
      "          [ 0.0327,  0.0603, -0.0905],\n",
      "          [-0.1055,  0.0174, -0.1139]]],\n",
      "\n",
      "\n",
      "        [[[-0.0140,  0.0979, -0.0988],\n",
      "          [ 0.0031,  0.0544, -0.0463],\n",
      "          [ 0.0785, -0.0141, -0.0283]],\n",
      "\n",
      "         [[ 0.0606,  0.0677, -0.0176],\n",
      "          [ 0.0286, -0.0782, -0.0021],\n",
      "          [-0.0144,  0.0246, -0.0092]],\n",
      "\n",
      "         [[-0.0033, -0.0167,  0.0914],\n",
      "          [ 0.0647, -0.0615, -0.0128],\n",
      "          [ 0.0430, -0.1075,  0.0017]]],\n",
      "\n",
      "\n",
      "        [[[-0.0006,  0.0023, -0.0590],\n",
      "          [-0.0143, -0.0473,  0.0593],\n",
      "          [-0.0142, -0.0404,  0.0908]],\n",
      "\n",
      "         [[ 0.0876, -0.0393, -0.1139],\n",
      "          [ 0.0368,  0.0460,  0.0621],\n",
      "          [-0.1057, -0.0183,  0.0261]],\n",
      "\n",
      "         [[ 0.0274, -0.0516, -0.0282],\n",
      "          [ 0.0988, -0.0437,  0.0534],\n",
      "          [-0.0337, -0.0539,  0.0141]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 0.0551, -0.0077, -0.0288],\n",
      "          [ 0.0267, -0.0330,  0.0205],\n",
      "          [ 0.0228,  0.0010,  0.0490]],\n",
      "\n",
      "         [[-0.0139, -0.0133,  0.0091],\n",
      "          [ 0.1062,  0.0290,  0.0103],\n",
      "          [ 0.0103,  0.0564, -0.1358]],\n",
      "\n",
      "         [[ 0.0161, -0.0144, -0.0666],\n",
      "          [ 0.0335, -0.0277, -0.0445],\n",
      "          [ 0.0686,  0.0136,  0.0296]]],\n",
      "\n",
      "\n",
      "        [[[-0.0540,  0.0866, -0.0431],\n",
      "          [ 0.0699,  0.0204,  0.0223],\n",
      "          [ 0.1328,  0.0580,  0.0384]],\n",
      "\n",
      "         [[-0.0285, -0.0540,  0.0649],\n",
      "          [ 0.0021,  0.1402,  0.0720],\n",
      "          [ 0.0187, -0.0471, -0.0790]],\n",
      "\n",
      "         [[-0.0389, -0.0592,  0.0387],\n",
      "          [-0.0548, -0.1043,  0.0604],\n",
      "          [-0.0742,  0.0096,  0.1155]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0050,  0.0215,  0.0573],\n",
      "          [ 0.1764, -0.0329,  0.0293],\n",
      "          [ 0.0443,  0.0240,  0.0320]],\n",
      "\n",
      "         [[ 0.1018, -0.0583,  0.0317],\n",
      "          [ 0.0101, -0.0515,  0.0305],\n",
      "          [ 0.0259, -0.0284, -0.0683]],\n",
      "\n",
      "         [[ 0.0462,  0.0730,  0.0049],\n",
      "          [ 0.0822, -0.0131, -0.0078],\n",
      "          [ 0.0312, -0.0870,  0.0461]]]])), ('features.0.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])), ('features.2.weight', tensor([[[[-1.0438e-02, -8.0732e-02,  5.0754e-02],\n",
      "          [-1.1811e-01, -9.8694e-02,  4.0122e-02],\n",
      "          [ 4.1571e-03, -7.8649e-02, -1.0402e-01]],\n",
      "\n",
      "         [[-6.3048e-02,  3.8256e-02, -1.2794e-01],\n",
      "          [ 5.6840e-02, -7.9274e-02,  4.2867e-02],\n",
      "          [ 2.2111e-02, -6.7725e-02, -5.4401e-02]],\n",
      "\n",
      "         [[-4.4558e-03, -1.7692e-02,  5.4239e-03],\n",
      "          [ 1.2000e-02, -1.0282e-01,  1.0429e-02],\n",
      "          [-9.8378e-02, -8.1808e-02, -1.2515e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-7.3623e-02,  2.0766e-02, -6.7971e-02],\n",
      "          [-6.0199e-02,  9.6938e-02, -3.2427e-02],\n",
      "          [ 7.2536e-02, -2.0183e-02,  6.5014e-02]],\n",
      "\n",
      "         [[ 1.2090e-01,  2.1355e-03, -9.8730e-02],\n",
      "          [-4.6611e-02, -1.2176e-01,  7.9178e-02],\n",
      "          [ 3.8749e-02,  2.1356e-02,  9.4775e-03]],\n",
      "\n",
      "         [[ 2.2491e-02,  1.4424e-01,  5.7507e-02],\n",
      "          [ 1.4222e-02, -2.7642e-02,  4.5109e-02],\n",
      "          [ 3.4716e-02, -6.8796e-02,  1.0750e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 4.5092e-03, -4.5713e-02,  2.2665e-02],\n",
      "          [-4.7407e-02,  5.0780e-02, -9.2941e-02],\n",
      "          [-8.8743e-03,  9.3759e-02,  5.8964e-02]],\n",
      "\n",
      "         [[-2.6128e-02, -7.2335e-02, -2.9258e-02],\n",
      "          [ 8.4940e-03,  7.8007e-03,  2.5752e-02],\n",
      "          [ 3.8606e-02, -1.1406e-01,  7.3967e-02]],\n",
      "\n",
      "         [[ 4.8105e-02, -3.0651e-02,  7.7044e-03],\n",
      "          [-3.7829e-02, -2.7380e-02,  5.5843e-03],\n",
      "          [-8.5098e-04, -1.2554e-02, -6.4706e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-5.5227e-02,  4.1530e-03,  6.3645e-02],\n",
      "          [ 8.7502e-02,  4.8963e-02,  9.6855e-02],\n",
      "          [-1.5784e-02,  1.7384e-02, -1.3886e-02]],\n",
      "\n",
      "         [[ 1.5684e-02,  3.8221e-02, -7.8788e-03],\n",
      "          [ 1.1988e-01,  4.8851e-02, -1.3548e-01],\n",
      "          [-1.8424e-02, -5.0903e-02,  1.5859e-01]],\n",
      "\n",
      "         [[ 7.2154e-03, -8.6638e-02, -1.0730e-02],\n",
      "          [ 3.4196e-02,  2.3963e-02, -4.4345e-02],\n",
      "          [-9.7798e-05, -2.2516e-02, -1.4288e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 8.2239e-02,  5.8219e-02,  5.4547e-02],\n",
      "          [ 1.9185e-03, -1.1384e-01, -5.6295e-02],\n",
      "          [ 5.7953e-03, -1.9896e-02, -7.4209e-02]],\n",
      "\n",
      "         [[-2.8172e-02, -2.6805e-02,  2.4495e-02],\n",
      "          [ 8.7137e-02, -3.3627e-02, -1.8612e-02],\n",
      "          [ 6.4783e-02,  5.3625e-02, -7.1994e-04]],\n",
      "\n",
      "         [[ 5.7662e-02,  1.9493e-02, -5.2792e-02],\n",
      "          [-4.1552e-02, -3.9442e-03,  2.9851e-02],\n",
      "          [-5.3896e-02,  1.3661e-01, -4.6758e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-7.2621e-03, -9.2882e-02, -1.3084e-01],\n",
      "          [ 2.5774e-02,  4.0615e-02, -3.0916e-02],\n",
      "          [-2.0266e-02, -7.8342e-02,  2.3430e-02]],\n",
      "\n",
      "         [[-6.7708e-03,  6.9817e-02,  5.5482e-02],\n",
      "          [-3.4737e-02, -6.2787e-02, -4.7039e-02],\n",
      "          [-5.4059e-03, -3.9252e-02, -2.5724e-02]],\n",
      "\n",
      "         [[ 4.6938e-02,  1.5468e-02,  5.4797e-02],\n",
      "          [ 2.7087e-02,  4.1380e-02, -7.1708e-02],\n",
      "          [-4.3504e-02, -9.1200e-02,  3.5527e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 1.3021e-02,  1.5430e-02,  4.0084e-02],\n",
      "          [ 1.0204e-01, -9.0138e-02, -1.0940e-01],\n",
      "          [ 1.0610e-01,  1.2099e-02, -5.0952e-02]],\n",
      "\n",
      "         [[ 6.6791e-02, -8.0456e-02,  1.6968e-02],\n",
      "          [ 1.0021e-01,  8.4245e-02,  2.0019e-02],\n",
      "          [-1.6165e-01, -8.4779e-02, -6.9368e-03]],\n",
      "\n",
      "         [[-7.1679e-02,  2.0895e-03,  7.3763e-02],\n",
      "          [ 8.3415e-02,  8.7670e-02,  1.2888e-01],\n",
      "          [ 5.8871e-02, -5.3122e-04, -7.9133e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.7889e-02,  5.7664e-03,  4.8184e-02],\n",
      "          [-1.7938e-02,  4.6703e-02,  3.9104e-02],\n",
      "          [ 1.6849e-02, -5.8203e-02, -1.0343e-02]],\n",
      "\n",
      "         [[-3.4487e-02, -5.9653e-02, -2.5021e-02],\n",
      "          [-9.5713e-02,  6.1472e-02,  8.1000e-02],\n",
      "          [-2.7402e-02, -1.0759e-01,  2.0316e-02]],\n",
      "\n",
      "         [[ 5.3532e-02,  1.3266e-01, -6.2331e-02],\n",
      "          [ 1.4057e-01,  1.4090e-01, -4.5542e-02],\n",
      "          [-1.1490e-01,  4.5959e-02, -6.6521e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 2.1232e-02,  3.7311e-02, -4.1042e-02],\n",
      "          [ 5.8319e-03,  8.6589e-02,  3.7541e-03],\n",
      "          [-1.2128e-02, -9.3297e-02,  5.9055e-02]],\n",
      "\n",
      "         [[-2.0930e-02,  1.3990e-02,  1.9287e-02],\n",
      "          [ 2.0195e-03,  1.3117e-01, -8.6856e-02],\n",
      "          [ 1.4110e-02,  4.2012e-02,  9.8515e-02]],\n",
      "\n",
      "         [[-2.3256e-03,  2.4146e-02,  2.9970e-02],\n",
      "          [-2.5784e-02, -1.0404e-01, -1.3573e-01],\n",
      "          [ 1.8545e-02,  3.8366e-03,  3.6214e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.9950e-02,  3.4884e-02, -2.0946e-02],\n",
      "          [ 1.0503e-01, -6.1659e-02, -8.2877e-02],\n",
      "          [ 3.5696e-02,  1.4700e-02, -1.8578e-02]],\n",
      "\n",
      "         [[-6.2949e-02,  1.5283e-02, -2.2848e-02],\n",
      "          [-3.7103e-03, -1.1567e-02,  2.7467e-02],\n",
      "          [-1.1176e-01, -2.2966e-02, -2.6886e-02]],\n",
      "\n",
      "         [[-2.4639e-02, -3.7366e-02, -2.9362e-04],\n",
      "          [ 1.4726e-02, -7.1988e-02, -1.0335e-02],\n",
      "          [-3.9200e-02,  3.8141e-02, -4.1050e-03]]],\n",
      "\n",
      "\n",
      "        [[[-7.3425e-02, -4.7063e-02, -1.9058e-02],\n",
      "          [ 1.6823e-01,  4.1547e-03,  7.2022e-02],\n",
      "          [-3.1054e-02, -1.9766e-02, -4.4654e-02]],\n",
      "\n",
      "         [[ 2.6949e-02, -1.4510e-01, -1.5947e-02],\n",
      "          [-2.7152e-02,  3.3382e-02,  2.0227e-02],\n",
      "          [ 6.7002e-02,  8.1422e-02,  3.0193e-02]],\n",
      "\n",
      "         [[ 3.7065e-02,  4.2310e-02,  3.3473e-02],\n",
      "          [-5.1048e-02, -4.4232e-02,  1.1889e-01],\n",
      "          [ 3.6891e-02, -1.4724e-01,  8.3119e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.6302e-03, -2.0383e-02, -1.0366e-02],\n",
      "          [-1.3936e-02, -9.1748e-02, -6.3679e-02],\n",
      "          [ 1.6309e-02,  7.4978e-02,  5.6538e-02]],\n",
      "\n",
      "         [[-8.1874e-02, -2.1968e-02,  5.6978e-02],\n",
      "          [-4.5703e-02, -1.0755e-01,  1.2476e-01],\n",
      "          [-8.7407e-02,  3.7783e-02, -6.9608e-02]],\n",
      "\n",
      "         [[ 8.6459e-02,  7.5576e-02,  4.3240e-02],\n",
      "          [-2.5504e-04,  2.4490e-02,  1.3635e-01],\n",
      "          [-2.9601e-02, -6.1316e-02,  1.8531e-02]]]])), ('features.2.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])), ('features.5.weight', tensor([[[[ 0.0198, -0.0143,  0.0788],\n",
      "          [ 0.0237, -0.0299,  0.0129],\n",
      "          [-0.0049, -0.0630, -0.0340]],\n",
      "\n",
      "         [[ 0.0454,  0.0202, -0.0043],\n",
      "          [-0.0029,  0.0154, -0.0348],\n",
      "          [-0.0159, -0.0285,  0.0289]],\n",
      "\n",
      "         [[-0.0289,  0.0388, -0.0463],\n",
      "          [ 0.0106, -0.0117,  0.0686],\n",
      "          [ 0.0145, -0.0545,  0.0278]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0135,  0.0384, -0.0518],\n",
      "          [-0.0076,  0.0417, -0.0365],\n",
      "          [ 0.0570, -0.0113,  0.0829]],\n",
      "\n",
      "         [[-0.0055, -0.0235, -0.0655],\n",
      "          [-0.0013, -0.0572,  0.0155],\n",
      "          [-0.0409,  0.0327,  0.0546]],\n",
      "\n",
      "         [[ 0.0275, -0.0633,  0.0006],\n",
      "          [ 0.0307,  0.1056, -0.0598],\n",
      "          [-0.0196,  0.0358, -0.0216]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0213,  0.0279, -0.0737],\n",
      "          [-0.0336, -0.0392, -0.0183],\n",
      "          [ 0.0294,  0.0827,  0.0105]],\n",
      "\n",
      "         [[ 0.0444,  0.0084, -0.0171],\n",
      "          [-0.0296,  0.0450,  0.0618],\n",
      "          [ 0.0105,  0.0085, -0.0095]],\n",
      "\n",
      "         [[-0.0408, -0.0826, -0.0265],\n",
      "          [ 0.0225, -0.0007, -0.0856],\n",
      "          [-0.0462, -0.0261,  0.0599]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.1333, -0.0659,  0.0171],\n",
      "          [-0.0404,  0.0060, -0.0033],\n",
      "          [-0.0397,  0.0007,  0.0318]],\n",
      "\n",
      "         [[ 0.0492,  0.0564,  0.0169],\n",
      "          [ 0.0059,  0.0318, -0.0109],\n",
      "          [-0.0226,  0.0076, -0.0018]],\n",
      "\n",
      "         [[-0.0334, -0.0677, -0.0444],\n",
      "          [ 0.0643,  0.0024, -0.0395],\n",
      "          [-0.0389,  0.0143,  0.0393]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0768, -0.0402,  0.0682],\n",
      "          [ 0.0036,  0.0562, -0.0302],\n",
      "          [-0.0375, -0.0196, -0.0078]],\n",
      "\n",
      "         [[-0.0436, -0.0011,  0.0682],\n",
      "          [ 0.0029, -0.0719, -0.0603],\n",
      "          [-0.0019,  0.0304,  0.0402]],\n",
      "\n",
      "         [[-0.0347,  0.0697,  0.0202],\n",
      "          [ 0.0659, -0.0062,  0.0445],\n",
      "          [-0.0550,  0.0312,  0.0807]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0306,  0.0075,  0.0711],\n",
      "          [ 0.0263, -0.0229, -0.0703],\n",
      "          [ 0.0134, -0.0126, -0.0002]],\n",
      "\n",
      "         [[ 0.0388, -0.0094,  0.0256],\n",
      "          [ 0.0721,  0.0299, -0.0562],\n",
      "          [ 0.0075, -0.0336,  0.0690]],\n",
      "\n",
      "         [[ 0.0175,  0.0518,  0.0592],\n",
      "          [-0.0393,  0.0146,  0.0031],\n",
      "          [-0.0025,  0.0314,  0.0114]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 0.0539, -0.0433,  0.0408],\n",
      "          [-0.0158,  0.0473, -0.0762],\n",
      "          [-0.0350,  0.0002, -0.0721]],\n",
      "\n",
      "         [[-0.0390, -0.0127,  0.0109],\n",
      "          [ 0.0107,  0.0413, -0.0092],\n",
      "          [ 0.0219,  0.0522,  0.0207]],\n",
      "\n",
      "         [[ 0.0057, -0.0926, -0.0457],\n",
      "          [ 0.0048,  0.0168, -0.0403],\n",
      "          [ 0.0483, -0.0443, -0.0134]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0289,  0.0268,  0.0769],\n",
      "          [-0.0872, -0.0599,  0.0205],\n",
      "          [-0.0117,  0.0062, -0.0387]],\n",
      "\n",
      "         [[-0.0436, -0.0445, -0.0307],\n",
      "          [ 0.0041, -0.0605, -0.0590],\n",
      "          [ 0.0369,  0.0248,  0.0029]],\n",
      "\n",
      "         [[ 0.0318, -0.0735,  0.0204],\n",
      "          [ 0.0354,  0.0033,  0.0367],\n",
      "          [-0.0643,  0.0138, -0.0005]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0263,  0.0605, -0.0910],\n",
      "          [ 0.0030,  0.0376,  0.0660],\n",
      "          [ 0.0579, -0.0397,  0.0485]],\n",
      "\n",
      "         [[-0.0410, -0.0134, -0.0509],\n",
      "          [ 0.0072,  0.0576, -0.0315],\n",
      "          [-0.0010, -0.0178,  0.0403]],\n",
      "\n",
      "         [[-0.0356, -0.0068, -0.0091],\n",
      "          [-0.0236, -0.0357, -0.0111],\n",
      "          [-0.0673,  0.0169,  0.0202]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0620, -0.0473, -0.0124],\n",
      "          [-0.0508, -0.0307, -0.0514],\n",
      "          [ 0.0359, -0.0254,  0.0932]],\n",
      "\n",
      "         [[ 0.0257, -0.0169,  0.0215],\n",
      "          [ 0.0724, -0.0151,  0.0267],\n",
      "          [-0.0082,  0.0359, -0.0108]],\n",
      "\n",
      "         [[-0.0831, -0.0081,  0.0195],\n",
      "          [ 0.0009,  0.0223, -0.0890],\n",
      "          [ 0.0283,  0.0686, -0.0434]]],\n",
      "\n",
      "\n",
      "        [[[-0.0232, -0.0446,  0.0050],\n",
      "          [ 0.0056, -0.0369, -0.0010],\n",
      "          [ 0.0489,  0.0543,  0.0240]],\n",
      "\n",
      "         [[ 0.0015, -0.0415,  0.0050],\n",
      "          [ 0.0266,  0.0390, -0.0120],\n",
      "          [-0.0112, -0.0280,  0.0131]],\n",
      "\n",
      "         [[-0.0240,  0.0912,  0.0281],\n",
      "          [-0.0070,  0.0581, -0.0384],\n",
      "          [-0.0266,  0.0534,  0.0248]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0094, -0.0546,  0.0015],\n",
      "          [-0.0513,  0.0360, -0.0677],\n",
      "          [-0.0151, -0.0696, -0.0212]],\n",
      "\n",
      "         [[-0.0429, -0.0058, -0.0226],\n",
      "          [-0.0177, -0.0193, -0.0368],\n",
      "          [-0.0034, -0.0521,  0.0427]],\n",
      "\n",
      "         [[ 0.0303,  0.0298, -0.0043],\n",
      "          [-0.0184, -0.0021, -0.0776],\n",
      "          [ 0.0201, -0.0849, -0.0043]]]])), ('features.5.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.7.weight', tensor([[[[-0.0179,  0.0020,  0.0204],\n",
      "          [ 0.0598, -0.0437, -0.0534],\n",
      "          [-0.0417,  0.0167, -0.0165]],\n",
      "\n",
      "         [[-0.0301,  0.0632,  0.0258],\n",
      "          [-0.0005, -0.0119,  0.0150],\n",
      "          [ 0.0106, -0.0163,  0.0022]],\n",
      "\n",
      "         [[-0.0187, -0.0025, -0.0616],\n",
      "          [-0.0029, -0.0348,  0.0112],\n",
      "          [ 0.0051,  0.0100, -0.0664]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0705, -0.0391, -0.0433],\n",
      "          [-0.0657, -0.1079,  0.0261],\n",
      "          [-0.0296, -0.0077,  0.0261]],\n",
      "\n",
      "         [[ 0.0728,  0.0161, -0.0507],\n",
      "          [-0.0175, -0.0749, -0.0095],\n",
      "          [ 0.0277,  0.0308,  0.0161]],\n",
      "\n",
      "         [[-0.0144, -0.0596,  0.0505],\n",
      "          [-0.0180,  0.0407,  0.0145],\n",
      "          [-0.0337, -0.0276,  0.0179]]],\n",
      "\n",
      "\n",
      "        [[[-0.0199, -0.0251,  0.0577],\n",
      "          [ 0.0710, -0.0452, -0.0479],\n",
      "          [ 0.0789,  0.0433,  0.0216]],\n",
      "\n",
      "         [[ 0.0230,  0.0019,  0.0480],\n",
      "          [ 0.0339, -0.0199, -0.0177],\n",
      "          [-0.0262,  0.0931,  0.0277]],\n",
      "\n",
      "         [[-0.0363, -0.0273, -0.0213],\n",
      "          [-0.0717,  0.0206, -0.1049],\n",
      "          [ 0.0126, -0.0066, -0.0068]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0362,  0.0492,  0.0423],\n",
      "          [ 0.0715, -0.0255, -0.0133],\n",
      "          [ 0.0363, -0.0036, -0.0710]],\n",
      "\n",
      "         [[-0.0834, -0.0153,  0.0761],\n",
      "          [ 0.0261, -0.0344,  0.0402],\n",
      "          [-0.0009,  0.0018, -0.0720]],\n",
      "\n",
      "         [[-0.0291, -0.0271,  0.0304],\n",
      "          [ 0.0276,  0.0156, -0.0281],\n",
      "          [-0.0289, -0.0088, -0.0567]]],\n",
      "\n",
      "\n",
      "        [[[-0.0667, -0.0898, -0.0169],\n",
      "          [-0.0741, -0.0294,  0.0584],\n",
      "          [-0.0088, -0.0159, -0.0489]],\n",
      "\n",
      "         [[-0.0530,  0.0366, -0.0544],\n",
      "          [-0.0460,  0.0257, -0.0225],\n",
      "          [ 0.0217, -0.0757, -0.0082]],\n",
      "\n",
      "         [[ 0.0357,  0.0707,  0.0278],\n",
      "          [-0.0625,  0.0286,  0.0019],\n",
      "          [-0.0550, -0.0100, -0.0216]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0131,  0.0490, -0.0190],\n",
      "          [-0.0429,  0.0095, -0.0328],\n",
      "          [ 0.0838, -0.0517, -0.0121]],\n",
      "\n",
      "         [[-0.0179, -0.0687, -0.0292],\n",
      "          [-0.0619,  0.0564, -0.0164],\n",
      "          [-0.0423, -0.0691, -0.0106]],\n",
      "\n",
      "         [[-0.0813,  0.0057,  0.0011],\n",
      "          [-0.0075, -0.0276,  0.0079],\n",
      "          [ 0.0364,  0.0054,  0.0438]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 0.0136, -0.0330, -0.0422],\n",
      "          [-0.0246,  0.0434,  0.0103],\n",
      "          [-0.0734,  0.0460,  0.1119]],\n",
      "\n",
      "         [[ 0.0045, -0.0079, -0.0098],\n",
      "          [ 0.0176, -0.0139,  0.0384],\n",
      "          [ 0.0470,  0.0504,  0.0285]],\n",
      "\n",
      "         [[ 0.0195,  0.0144,  0.0318],\n",
      "          [-0.0082, -0.0281,  0.0801],\n",
      "          [ 0.0179,  0.0157,  0.0288]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0640,  0.0182, -0.0141],\n",
      "          [ 0.0271, -0.0030,  0.0487],\n",
      "          [ 0.0377,  0.0179, -0.0501]],\n",
      "\n",
      "         [[ 0.0412, -0.0082,  0.0208],\n",
      "          [ 0.0278, -0.0299,  0.1116],\n",
      "          [-0.0045, -0.0207,  0.0122]],\n",
      "\n",
      "         [[ 0.0180, -0.0730, -0.0239],\n",
      "          [-0.0621,  0.0025, -0.0234],\n",
      "          [-0.0192,  0.0305, -0.0008]]],\n",
      "\n",
      "\n",
      "        [[[-0.0117,  0.0513, -0.0840],\n",
      "          [ 0.0125, -0.0281,  0.0192],\n",
      "          [ 0.0326,  0.0487,  0.0220]],\n",
      "\n",
      "         [[-0.1284,  0.0048,  0.0082],\n",
      "          [ 0.0398,  0.0144, -0.0100],\n",
      "          [-0.0508, -0.0766, -0.0657]],\n",
      "\n",
      "         [[-0.0727,  0.0639,  0.0356],\n",
      "          [-0.0191,  0.0845,  0.0567],\n",
      "          [-0.0198, -0.0606, -0.0239]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0334, -0.0136,  0.0190],\n",
      "          [ 0.0698, -0.0066,  0.0334],\n",
      "          [ 0.0458,  0.0087,  0.0083]],\n",
      "\n",
      "         [[-0.0270, -0.0466,  0.0153],\n",
      "          [ 0.0077, -0.0456, -0.0262],\n",
      "          [-0.0020, -0.0214,  0.0211]],\n",
      "\n",
      "         [[ 0.0937,  0.0456, -0.0904],\n",
      "          [-0.0302,  0.0222,  0.0320],\n",
      "          [-0.0353,  0.0623, -0.0161]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0092, -0.0170,  0.0209],\n",
      "          [ 0.0587, -0.0206,  0.0666],\n",
      "          [ 0.0247,  0.0288, -0.0073]],\n",
      "\n",
      "         [[ 0.0014,  0.0216,  0.0330],\n",
      "          [ 0.0975, -0.0287, -0.0847],\n",
      "          [-0.0070, -0.0078,  0.0142]],\n",
      "\n",
      "         [[-0.0615, -0.0259, -0.0657],\n",
      "          [-0.0497,  0.0310,  0.0403],\n",
      "          [-0.0141,  0.0089,  0.0463]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0884, -0.0276,  0.0537],\n",
      "          [-0.0316,  0.0593,  0.0401],\n",
      "          [ 0.0411,  0.0204, -0.0375]],\n",
      "\n",
      "         [[-0.0676,  0.0689, -0.0399],\n",
      "          [ 0.0101,  0.0524,  0.0342],\n",
      "          [-0.0167,  0.0060, -0.0364]],\n",
      "\n",
      "         [[ 0.0578, -0.0256,  0.0055],\n",
      "          [-0.0130,  0.0202, -0.0312],\n",
      "          [-0.0212, -0.0006,  0.0582]]]])), ('features.7.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.10.weight', tensor([[[[-5.0728e-02, -1.7108e-02, -1.2431e-02],\n",
      "          [ 3.8660e-02, -2.0634e-02,  8.4864e-03],\n",
      "          [-1.3596e-02, -1.3351e-02,  1.8935e-02]],\n",
      "\n",
      "         [[-2.6075e-02,  4.4814e-03, -5.7448e-02],\n",
      "          [-5.4025e-03, -1.7040e-03,  3.2107e-02],\n",
      "          [-5.6801e-03,  1.8247e-02, -5.8293e-02]],\n",
      "\n",
      "         [[-3.4310e-02,  2.7729e-02, -1.5182e-02],\n",
      "          [ 2.4536e-02,  5.2430e-02,  8.5617e-02],\n",
      "          [-1.3779e-02,  5.1793e-02, -3.8497e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 3.0967e-02,  3.8846e-02, -2.0589e-02],\n",
      "          [ 1.1386e-02, -9.3430e-04, -2.3165e-02],\n",
      "          [-3.2127e-02, -1.1117e-02, -5.9535e-03]],\n",
      "\n",
      "         [[-8.6784e-03, -5.6213e-02, -5.3420e-02],\n",
      "          [ 2.6674e-02,  1.8395e-02,  1.9045e-02],\n",
      "          [ 5.5717e-02, -1.6232e-02, -8.2629e-03]],\n",
      "\n",
      "         [[ 3.4560e-02, -6.0152e-02, -2.0904e-02],\n",
      "          [ 1.6136e-02, -3.7779e-02,  4.6297e-03],\n",
      "          [ 5.9874e-02, -1.8214e-02, -9.7496e-03]]],\n",
      "\n",
      "\n",
      "        [[[-4.1261e-02, -1.2283e-02,  4.5996e-03],\n",
      "          [-2.1172e-02,  3.1135e-02, -3.1670e-02],\n",
      "          [ 3.1775e-02, -3.0435e-02,  4.3926e-02]],\n",
      "\n",
      "         [[ 2.3291e-02,  5.4029e-02,  9.7152e-03],\n",
      "          [ 6.5058e-03, -1.1811e-02, -2.7091e-02],\n",
      "          [-3.0916e-03, -3.8577e-02, -1.3920e-02]],\n",
      "\n",
      "         [[-2.7659e-02, -2.6991e-03,  5.0054e-03],\n",
      "          [ 1.0475e-02,  1.5573e-02, -2.7041e-02],\n",
      "          [-9.2047e-03,  9.6607e-03,  5.3533e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-4.5700e-02, -3.6779e-02,  3.9766e-02],\n",
      "          [ 1.3928e-02,  2.8469e-02, -1.4947e-02],\n",
      "          [-1.1571e-02,  2.0989e-02, -5.6874e-03]],\n",
      "\n",
      "         [[ 2.7825e-02, -1.9116e-02, -2.3352e-03],\n",
      "          [ 6.4976e-03, -1.2660e-02, -2.6424e-03],\n",
      "          [-3.6476e-02,  1.0098e-02, -2.9593e-02]],\n",
      "\n",
      "         [[ 2.4187e-02, -2.6923e-03, -2.8982e-02],\n",
      "          [-1.2791e-02, -1.9913e-02, -7.6459e-03],\n",
      "          [-1.7129e-02,  3.2724e-02, -6.6450e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 3.2745e-02, -3.0847e-03, -2.3039e-02],\n",
      "          [-4.2012e-02, -1.0737e-02, -1.2865e-02],\n",
      "          [ 8.9466e-02, -2.3234e-02, -2.4460e-02]],\n",
      "\n",
      "         [[-1.1765e-02, -4.1363e-02,  1.8264e-02],\n",
      "          [ 1.2940e-02,  3.7620e-02, -6.2298e-03],\n",
      "          [-1.8549e-02, -1.3401e-02, -1.0866e-02]],\n",
      "\n",
      "         [[ 2.8781e-02, -1.9674e-02, -2.5169e-02],\n",
      "          [-5.5550e-02,  6.3253e-02, -2.0357e-03],\n",
      "          [-2.6246e-02,  1.2429e-03,  3.6821e-05]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.9539e-02, -1.4941e-02, -3.1366e-02],\n",
      "          [-9.0841e-03, -4.6973e-03,  2.6057e-02],\n",
      "          [-6.3784e-03,  3.9437e-02,  3.6896e-02]],\n",
      "\n",
      "         [[-3.4673e-02,  1.4955e-02,  3.7908e-03],\n",
      "          [ 1.4995e-03, -1.0287e-02,  2.3906e-02],\n",
      "          [-1.7831e-02, -1.5603e-02, -2.0531e-02]],\n",
      "\n",
      "         [[-1.6050e-02, -1.9391e-02,  9.9876e-04],\n",
      "          [ 6.9745e-03, -2.2201e-02, -2.0886e-02],\n",
      "          [-2.6979e-02, -2.6555e-02, -1.2621e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 5.2519e-03, -4.3720e-02, -3.8265e-02],\n",
      "          [-1.5415e-04, -4.9476e-02, -5.0327e-02],\n",
      "          [ 1.6534e-02, -1.2704e-02,  1.9462e-03]],\n",
      "\n",
      "         [[ 3.0514e-02,  4.1127e-03, -2.0615e-02],\n",
      "          [ 2.2808e-02,  3.9030e-02,  2.2970e-02],\n",
      "          [-1.6600e-02,  2.4024e-02,  4.1612e-04]],\n",
      "\n",
      "         [[-2.9390e-03,  1.4730e-02, -5.8987e-03],\n",
      "          [ 1.0378e-02, -3.2351e-02, -4.6606e-02],\n",
      "          [ 5.1378e-03, -1.5672e-02,  1.9980e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 3.5689e-02,  5.7881e-02,  2.3657e-02],\n",
      "          [-2.2713e-03, -1.9203e-02,  2.3403e-02],\n",
      "          [-3.5671e-02,  8.6337e-05, -4.6627e-02]],\n",
      "\n",
      "         [[-1.4890e-02,  3.4385e-03,  3.8877e-02],\n",
      "          [-1.8392e-02, -4.4461e-02, -3.5726e-02],\n",
      "          [-3.5855e-03,  2.9852e-02,  3.2744e-02]],\n",
      "\n",
      "         [[ 1.0990e-02,  3.4887e-02,  3.1713e-02],\n",
      "          [ 1.3434e-02, -5.9027e-03,  2.2930e-02],\n",
      "          [ 5.9324e-02,  5.8546e-03,  5.5969e-02]]],\n",
      "\n",
      "\n",
      "        [[[-4.1047e-02, -6.3498e-02, -3.9575e-02],\n",
      "          [-1.5297e-02, -9.3860e-03,  9.5833e-03],\n",
      "          [ 3.0794e-02,  1.5145e-03, -3.3490e-03]],\n",
      "\n",
      "         [[-1.7605e-02, -4.2189e-02,  8.3832e-02],\n",
      "          [-2.2616e-04, -3.7770e-02, -6.3806e-03],\n",
      "          [ 1.9286e-02,  1.2264e-02,  4.2856e-03]],\n",
      "\n",
      "         [[ 6.2195e-02,  1.4930e-02,  5.8531e-02],\n",
      "          [-5.6018e-03,  2.6160e-03,  1.2444e-02],\n",
      "          [ 1.2881e-02, -1.0514e-02,  9.8994e-04]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-7.7630e-02, -1.1334e-02, -2.5648e-02],\n",
      "          [ 1.9293e-02,  1.3706e-02, -2.9314e-02],\n",
      "          [ 3.1384e-02,  4.3326e-02,  2.0578e-02]],\n",
      "\n",
      "         [[-5.2145e-03, -2.1698e-02,  3.4683e-02],\n",
      "          [-2.3567e-03, -7.1217e-03, -4.4763e-03],\n",
      "          [-3.3698e-03,  6.9897e-02, -1.2648e-02]],\n",
      "\n",
      "         [[ 4.3271e-02, -2.3940e-02, -1.9062e-02],\n",
      "          [-4.1852e-02, -1.2821e-02,  1.4785e-03],\n",
      "          [ 2.6176e-02,  1.1000e-02, -2.9053e-04]]],\n",
      "\n",
      "\n",
      "        [[[ 6.3701e-02,  1.8061e-02,  4.6117e-02],\n",
      "          [-1.0261e-01, -2.1930e-02, -2.9473e-02],\n",
      "          [-3.8681e-02,  2.2733e-03, -1.0466e-02]],\n",
      "\n",
      "         [[-2.7260e-03, -3.2273e-03,  1.7933e-02],\n",
      "          [-3.5209e-02, -1.8307e-02, -4.8862e-02],\n",
      "          [ 5.2072e-03,  1.0847e-02,  3.8667e-03]],\n",
      "\n",
      "         [[-2.3782e-02,  3.9720e-02,  6.3642e-02],\n",
      "          [ 2.6771e-02,  2.4970e-02,  4.7713e-02],\n",
      "          [ 4.6140e-03,  5.8765e-03, -5.4634e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 6.5235e-02, -4.0641e-02,  1.0302e-02],\n",
      "          [-2.5287e-02, -2.0474e-02, -7.5008e-05],\n",
      "          [-3.5796e-04, -1.2892e-02,  3.6186e-03]],\n",
      "\n",
      "         [[ 7.5217e-02, -3.5593e-02,  1.2196e-02],\n",
      "          [ 4.7095e-02, -2.9569e-02, -4.2409e-02],\n",
      "          [ 3.5647e-02, -1.6836e-02,  2.9756e-02]],\n",
      "\n",
      "         [[-2.9744e-02, -4.1802e-02,  2.4000e-02],\n",
      "          [ 3.8848e-02, -3.7880e-02,  5.4153e-04],\n",
      "          [ 3.3881e-03,  3.4924e-02, -6.8246e-03]]]])), ('features.10.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])), ('features.12.weight', tensor([[[[ 0.0172,  0.0443, -0.0104],\n",
      "          [ 0.0266,  0.0317,  0.0110],\n",
      "          [-0.0059,  0.0141, -0.0114]],\n",
      "\n",
      "         [[ 0.0012, -0.0438,  0.0078],\n",
      "          [-0.0363,  0.0395, -0.0348],\n",
      "          [ 0.0069, -0.0319, -0.0247]],\n",
      "\n",
      "         [[-0.0330, -0.0048, -0.0180],\n",
      "          [ 0.0245, -0.0126,  0.0025],\n",
      "          [-0.0159, -0.0390, -0.0255]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0279, -0.0046, -0.0024],\n",
      "          [-0.0162, -0.0079, -0.0343],\n",
      "          [-0.0109, -0.0183,  0.0119]],\n",
      "\n",
      "         [[ 0.0294,  0.0085, -0.0162],\n",
      "          [-0.0134,  0.0465, -0.0365],\n",
      "          [-0.0249, -0.0312,  0.0423]],\n",
      "\n",
      "         [[ 0.0201, -0.0422,  0.0098],\n",
      "          [-0.0124,  0.0291,  0.0300],\n",
      "          [-0.0391,  0.0070,  0.0335]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0140,  0.0233,  0.0577],\n",
      "          [-0.0507, -0.0212,  0.0256],\n",
      "          [ 0.0393,  0.0085,  0.0275]],\n",
      "\n",
      "         [[ 0.0050, -0.0197,  0.0075],\n",
      "          [-0.0144, -0.0145, -0.0064],\n",
      "          [ 0.0353, -0.0057,  0.0098]],\n",
      "\n",
      "         [[ 0.0011, -0.0099, -0.0225],\n",
      "          [-0.0071, -0.0247, -0.0469],\n",
      "          [-0.0070,  0.0150, -0.0150]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0206, -0.0395,  0.0603],\n",
      "          [-0.0044,  0.0015,  0.0481],\n",
      "          [-0.0244,  0.0061, -0.0386]],\n",
      "\n",
      "         [[ 0.0041, -0.0491, -0.0213],\n",
      "          [ 0.0574, -0.0101, -0.0032],\n",
      "          [-0.0329,  0.0234,  0.0288]],\n",
      "\n",
      "         [[ 0.0214,  0.0300, -0.0346],\n",
      "          [ 0.0057, -0.0249, -0.0013],\n",
      "          [-0.0332, -0.0187, -0.0011]]],\n",
      "\n",
      "\n",
      "        [[[-0.0254,  0.0041,  0.0026],\n",
      "          [ 0.0320, -0.0532,  0.0467],\n",
      "          [-0.0189,  0.0154,  0.0213]],\n",
      "\n",
      "         [[ 0.0045, -0.0216,  0.0211],\n",
      "          [ 0.0075,  0.0053,  0.0056],\n",
      "          [ 0.0376, -0.0108,  0.0014]],\n",
      "\n",
      "         [[-0.0100,  0.0118,  0.0592],\n",
      "          [-0.0269, -0.0373, -0.0268],\n",
      "          [ 0.0688,  0.0119, -0.0165]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0496,  0.0349, -0.0324],\n",
      "          [ 0.0006,  0.0171, -0.0291],\n",
      "          [ 0.0093, -0.0207, -0.0088]],\n",
      "\n",
      "         [[ 0.0020, -0.0428,  0.0158],\n",
      "          [-0.0388, -0.0381, -0.0418],\n",
      "          [-0.0011,  0.0332, -0.0099]],\n",
      "\n",
      "         [[-0.0004,  0.0370,  0.0281],\n",
      "          [ 0.0309,  0.0107, -0.0118],\n",
      "          [ 0.0323, -0.0062, -0.0197]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-0.0182, -0.0140, -0.0314],\n",
      "          [-0.0059,  0.0248,  0.0336],\n",
      "          [-0.0053,  0.0424,  0.0295]],\n",
      "\n",
      "         [[-0.0127,  0.0249,  0.0254],\n",
      "          [ 0.0288,  0.0009, -0.0457],\n",
      "          [ 0.0263, -0.0296,  0.0282]],\n",
      "\n",
      "         [[-0.0564, -0.0179, -0.0319],\n",
      "          [ 0.0545, -0.0175, -0.0065],\n",
      "          [ 0.0268,  0.0396, -0.0235]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0169, -0.0426, -0.0198],\n",
      "          [ 0.0579,  0.0553,  0.0870],\n",
      "          [ 0.0211, -0.0046, -0.0026]],\n",
      "\n",
      "         [[ 0.0364, -0.0570,  0.0189],\n",
      "          [-0.0222,  0.0162,  0.0649],\n",
      "          [-0.0031, -0.0405, -0.0051]],\n",
      "\n",
      "         [[ 0.0062,  0.0109,  0.0284],\n",
      "          [-0.0339,  0.0391,  0.0293],\n",
      "          [ 0.0163, -0.0099, -0.0289]]],\n",
      "\n",
      "\n",
      "        [[[-0.0238, -0.0667, -0.0395],\n",
      "          [-0.0387, -0.0274, -0.0386],\n",
      "          [ 0.0130,  0.0413, -0.0551]],\n",
      "\n",
      "         [[ 0.0835,  0.0176,  0.0132],\n",
      "          [-0.0253, -0.0015,  0.0425],\n",
      "          [-0.0210,  0.0390,  0.0186]],\n",
      "\n",
      "         [[ 0.0204,  0.0001,  0.0141],\n",
      "          [ 0.0289,  0.0277,  0.0005],\n",
      "          [ 0.0217,  0.0143, -0.0190]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0268,  0.0097, -0.0008],\n",
      "          [-0.0100,  0.0517, -0.0061],\n",
      "          [-0.0464,  0.0004,  0.0598]],\n",
      "\n",
      "         [[-0.0087, -0.0123,  0.0300],\n",
      "          [ 0.0123, -0.0286,  0.0119],\n",
      "          [-0.0100,  0.0710, -0.0044]],\n",
      "\n",
      "         [[ 0.0687,  0.0186, -0.0118],\n",
      "          [ 0.0448,  0.0085, -0.0273],\n",
      "          [ 0.0052, -0.0164, -0.0106]]],\n",
      "\n",
      "\n",
      "        [[[-0.0232,  0.0096, -0.0058],\n",
      "          [ 0.0200,  0.0363, -0.0480],\n",
      "          [ 0.0242,  0.0377,  0.0045]],\n",
      "\n",
      "         [[ 0.0612, -0.0501,  0.0262],\n",
      "          [-0.0057,  0.0018,  0.0391],\n",
      "          [ 0.0109,  0.0356, -0.0293]],\n",
      "\n",
      "         [[ 0.0071,  0.0363, -0.0005],\n",
      "          [-0.0536, -0.0072,  0.0114],\n",
      "          [ 0.0818, -0.0035,  0.0465]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0024, -0.0250,  0.0119],\n",
      "          [ 0.0105, -0.0003, -0.0265],\n",
      "          [ 0.0207,  0.0319,  0.0003]],\n",
      "\n",
      "         [[ 0.0483, -0.0100,  0.0420],\n",
      "          [-0.0246, -0.0281,  0.0038],\n",
      "          [ 0.0429,  0.0403, -0.0461]],\n",
      "\n",
      "         [[ 0.0512, -0.0404, -0.0360],\n",
      "          [-0.0466, -0.0028, -0.0436],\n",
      "          [-0.0289,  0.0153, -0.0048]]]])), ('features.12.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])), ('features.14.weight', tensor([[[[ 3.7990e-03,  1.2749e-02, -5.4047e-03],\n",
      "          [-4.5647e-02,  5.2605e-02, -4.4494e-02],\n",
      "          [-4.9945e-02, -4.5833e-02,  2.3300e-02]],\n",
      "\n",
      "         [[-2.2144e-02,  3.8238e-02, -4.5147e-02],\n",
      "          [ 4.0281e-02, -1.0734e-02,  2.6067e-02],\n",
      "          [ 9.2010e-03,  2.2655e-02,  1.1610e-02]],\n",
      "\n",
      "         [[-8.3604e-03, -3.1750e-02,  2.7814e-02],\n",
      "          [ 1.1826e-02,  1.9279e-02,  7.7513e-03],\n",
      "          [-1.7540e-02,  9.5859e-03,  3.5723e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.2839e-02, -5.2271e-02,  1.4241e-02],\n",
      "          [-4.9002e-02, -1.3865e-02,  1.8473e-02],\n",
      "          [ 1.0640e-02, -2.0067e-02, -1.8935e-02]],\n",
      "\n",
      "         [[-1.8829e-02, -3.7436e-02, -3.1213e-02],\n",
      "          [-2.8192e-03, -2.3482e-02,  1.1922e-02],\n",
      "          [ 5.9035e-03,  6.0204e-05,  6.2223e-02]],\n",
      "\n",
      "         [[-4.2984e-02, -2.6945e-02,  1.1575e-02],\n",
      "          [-1.9041e-02, -1.8767e-02,  7.7000e-03],\n",
      "          [-3.1595e-02,  2.8739e-03, -3.3372e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 1.1093e-02,  4.2177e-02,  4.0686e-02],\n",
      "          [ 3.8714e-02, -2.1996e-02, -3.4866e-02],\n",
      "          [-1.7526e-03, -5.2953e-02,  2.6350e-02]],\n",
      "\n",
      "         [[-4.4562e-02,  2.2737e-02,  1.0994e-02],\n",
      "          [-1.2116e-02,  3.4346e-02,  1.8354e-02],\n",
      "          [-2.1520e-02,  2.8972e-02,  2.3985e-02]],\n",
      "\n",
      "         [[-4.5689e-02,  2.7650e-02,  3.4155e-02],\n",
      "          [-4.7179e-02, -4.9325e-02, -9.3212e-03],\n",
      "          [ 4.2177e-02, -4.9274e-02,  1.9284e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.3627e-02,  3.4774e-02,  5.8519e-03],\n",
      "          [-4.6150e-02,  6.0198e-02,  2.5496e-02],\n",
      "          [ 1.7021e-02,  1.6084e-02,  4.1039e-02]],\n",
      "\n",
      "         [[ 7.4745e-03,  3.9859e-03, -4.4189e-02],\n",
      "          [-1.2501e-03, -4.4239e-02,  1.8005e-02],\n",
      "          [ 1.5001e-02,  3.7488e-03,  2.7370e-02]],\n",
      "\n",
      "         [[-4.4105e-02,  1.5098e-02, -2.3949e-02],\n",
      "          [-1.0426e-02,  1.8618e-03, -5.5903e-02],\n",
      "          [-3.8754e-02,  8.2089e-04, -3.9853e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 2.4894e-04,  4.0526e-02, -3.9592e-02],\n",
      "          [ 1.1283e-02, -4.6168e-02, -5.8365e-02],\n",
      "          [-2.2778e-02,  3.7312e-02,  1.2398e-02]],\n",
      "\n",
      "         [[-9.7269e-03, -2.3275e-02, -9.7151e-03],\n",
      "          [ 1.6911e-02,  1.2109e-02,  3.7863e-03],\n",
      "          [ 5.4404e-02, -5.0686e-03,  6.3785e-02]],\n",
      "\n",
      "         [[ 3.3390e-02,  1.1906e-02,  4.2577e-02],\n",
      "          [ 4.7698e-02,  1.1337e-02,  3.4378e-02],\n",
      "          [ 1.4335e-02,  5.0533e-04,  2.8273e-04]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.3428e-02, -1.9538e-02, -7.2461e-02],\n",
      "          [ 2.4406e-03, -2.6865e-03,  1.7779e-02],\n",
      "          [-2.0090e-02,  1.3574e-02, -1.9948e-02]],\n",
      "\n",
      "         [[ 4.4362e-02,  6.9409e-02, -2.8973e-02],\n",
      "          [ 2.7026e-02,  6.6470e-03,  7.3094e-02],\n",
      "          [ 3.0157e-02,  8.7623e-03,  9.8786e-03]],\n",
      "\n",
      "         [[-1.4355e-02,  1.7284e-02, -1.6912e-02],\n",
      "          [ 2.7417e-02, -4.0205e-02, -1.8404e-02],\n",
      "          [ 2.8721e-04,  6.7504e-03,  5.8366e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-1.3221e-02, -2.1935e-02,  4.0496e-02],\n",
      "          [ 2.0233e-02, -3.1296e-02, -3.9653e-02],\n",
      "          [-6.0594e-03, -1.5637e-02, -3.5512e-02]],\n",
      "\n",
      "         [[-2.4412e-02,  7.2781e-02,  1.1587e-02],\n",
      "          [-4.3854e-02,  1.3033e-02,  2.8101e-02],\n",
      "          [-3.6836e-02,  6.3739e-04, -4.3617e-02]],\n",
      "\n",
      "         [[-1.4466e-02, -2.9441e-02,  3.2261e-02],\n",
      "          [-2.9080e-02,  2.5448e-02, -2.6080e-02],\n",
      "          [ 1.4058e-02, -3.5805e-02, -1.3369e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.7417e-02, -3.7416e-02,  2.7727e-02],\n",
      "          [-2.2788e-02, -3.1497e-02, -6.6583e-02],\n",
      "          [-1.2223e-03, -4.5345e-02,  3.1210e-02]],\n",
      "\n",
      "         [[-3.5176e-02, -3.3811e-02,  2.7674e-02],\n",
      "          [ 3.4242e-03, -3.3408e-02, -2.5055e-02],\n",
      "          [-4.2352e-02, -2.5077e-02,  1.8738e-02]],\n",
      "\n",
      "         [[ 1.3914e-02, -2.3251e-02, -5.9595e-02],\n",
      "          [-2.5528e-02,  1.1255e-02,  9.9272e-03],\n",
      "          [-2.0097e-02,  4.8980e-02,  1.2132e-02]]],\n",
      "\n",
      "\n",
      "        [[[-1.5663e-02,  6.7241e-03,  4.4443e-02],\n",
      "          [ 4.9165e-02,  1.9743e-02, -3.6399e-02],\n",
      "          [ 5.6171e-02,  8.4329e-02, -4.1378e-02]],\n",
      "\n",
      "         [[ 1.4452e-03,  1.4609e-02, -9.3590e-03],\n",
      "          [ 6.5260e-03,  3.8684e-02,  8.4658e-03],\n",
      "          [-1.9520e-03,  2.5115e-02,  3.5677e-02]],\n",
      "\n",
      "         [[-1.2420e-02,  3.8337e-02, -3.9295e-02],\n",
      "          [-1.2806e-03,  1.0609e-02,  2.9888e-02],\n",
      "          [ 4.8978e-02, -2.9099e-02, -3.2014e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.6175e-02, -2.0378e-02, -3.0795e-04],\n",
      "          [-7.5269e-03, -5.2249e-02, -1.5760e-02],\n",
      "          [-1.1851e-02, -5.2358e-03,  2.4697e-02]],\n",
      "\n",
      "         [[-2.4994e-03,  3.3424e-02, -1.0450e-02],\n",
      "          [-1.0161e-02,  1.4692e-02, -5.2341e-02],\n",
      "          [-5.0429e-02, -1.1051e-02,  2.0386e-02]],\n",
      "\n",
      "         [[-1.3251e-02, -4.2457e-03, -1.3579e-02],\n",
      "          [ 2.9507e-03, -2.4081e-03, -2.1879e-02],\n",
      "          [ 4.4858e-02, -2.9229e-02, -4.1031e-02]]],\n",
      "\n",
      "\n",
      "        [[[-2.8219e-02,  1.8013e-02,  1.6290e-02],\n",
      "          [-1.9128e-02, -6.8713e-03, -1.0671e-02],\n",
      "          [ 4.6443e-03, -1.3427e-02, -7.8816e-03]],\n",
      "\n",
      "         [[-5.9415e-03, -2.2998e-03, -1.4953e-02],\n",
      "          [-9.3470e-03, -1.6780e-02,  4.9198e-02],\n",
      "          [-1.5918e-02, -2.3066e-02, -3.3413e-02]],\n",
      "\n",
      "         [[ 2.9235e-02,  4.2089e-03,  6.2047e-03],\n",
      "          [-1.0609e-02,  2.9642e-02,  1.1674e-02],\n",
      "          [-1.1115e-02,  1.9964e-02,  5.2756e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.0481e-03,  1.9553e-02,  8.2139e-03],\n",
      "          [-1.3483e-02, -6.5390e-03, -1.5748e-02],\n",
      "          [ 2.2896e-02,  2.0985e-02, -5.7292e-02]],\n",
      "\n",
      "         [[-3.7572e-02, -7.9109e-03,  3.1165e-02],\n",
      "          [ 5.8786e-02,  3.6436e-02,  5.4326e-03],\n",
      "          [-3.8395e-02,  4.9357e-03, -2.2643e-03]],\n",
      "\n",
      "         [[ 3.8761e-02, -1.5272e-02,  3.4247e-02],\n",
      "          [ 1.6973e-02,  3.0678e-02,  2.0344e-03],\n",
      "          [-1.4769e-02, -2.3724e-02,  1.2544e-02]]]])), ('features.14.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])), ('features.17.weight', tensor([[[[-3.3797e-02,  1.9428e-02, -8.8944e-03],\n",
      "          [ 1.8384e-02, -8.5795e-04,  2.6184e-03],\n",
      "          [ 1.2201e-02,  2.5525e-03, -3.7828e-03]],\n",
      "\n",
      "         [[ 2.3036e-03,  4.1825e-02,  1.2860e-02],\n",
      "          [-1.4790e-02, -3.4668e-02, -2.5799e-03],\n",
      "          [-7.2855e-03,  4.2147e-03, -3.6675e-02]],\n",
      "\n",
      "         [[-5.2634e-03, -4.9214e-03, -1.7965e-02],\n",
      "          [-1.8695e-02, -1.7579e-02, -1.7842e-02],\n",
      "          [-1.3087e-02,  3.5091e-02,  5.6798e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 1.0644e-02, -7.2521e-03,  8.6163e-03],\n",
      "          [-8.5319e-03,  1.9024e-02,  4.6620e-02],\n",
      "          [-2.7423e-02,  2.2656e-02,  1.3122e-02]],\n",
      "\n",
      "         [[-1.1668e-02,  1.4561e-02, -3.3384e-02],\n",
      "          [-2.2356e-02,  3.4192e-02, -3.7918e-02],\n",
      "          [ 9.2916e-03, -8.8806e-03,  3.9609e-03]],\n",
      "\n",
      "         [[-3.4623e-02,  1.4903e-02, -2.0391e-02],\n",
      "          [ 1.2714e-02, -5.1271e-03, -7.3591e-03],\n",
      "          [ 1.5941e-02, -2.4092e-02,  2.3126e-02]]],\n",
      "\n",
      "\n",
      "        [[[-2.3412e-02,  3.1946e-02, -1.5469e-02],\n",
      "          [ 2.7107e-02, -1.9036e-02,  2.8333e-02],\n",
      "          [ 1.6352e-02, -3.2736e-02,  1.5011e-02]],\n",
      "\n",
      "         [[ 7.3284e-03, -9.0941e-03,  3.0097e-02],\n",
      "          [-1.1288e-03, -1.0780e-02, -2.6984e-02],\n",
      "          [ 3.5234e-03, -1.9793e-02,  1.4693e-02]],\n",
      "\n",
      "         [[-1.5648e-02,  1.2541e-02,  1.8623e-02],\n",
      "          [-2.3017e-03, -2.6709e-02,  1.4825e-02],\n",
      "          [ 2.2554e-02, -1.1400e-02,  4.1865e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-4.3871e-02,  1.1503e-02, -2.7594e-02],\n",
      "          [ 6.3295e-02, -7.5141e-03,  2.0983e-02],\n",
      "          [ 1.1981e-02,  4.0431e-02, -1.5824e-02]],\n",
      "\n",
      "         [[-1.2683e-02,  1.8869e-02, -6.2342e-03],\n",
      "          [ 1.3240e-02,  9.1860e-03, -1.8904e-02],\n",
      "          [-1.9828e-02,  4.1392e-02,  3.5791e-03]],\n",
      "\n",
      "         [[ 4.2740e-03,  1.2731e-02, -1.7823e-02],\n",
      "          [-5.5597e-03, -2.7505e-02,  2.4826e-02],\n",
      "          [-1.9754e-02,  2.2208e-02, -7.7114e-03]]],\n",
      "\n",
      "\n",
      "        [[[-3.6805e-03,  4.6080e-03,  5.5469e-04],\n",
      "          [-4.7147e-03, -2.1764e-02,  1.8577e-02],\n",
      "          [-2.1697e-02,  1.7766e-02, -1.6204e-02]],\n",
      "\n",
      "         [[ 9.0511e-03,  1.1391e-02, -2.8430e-03],\n",
      "          [ 4.2742e-03, -1.3951e-02,  2.7765e-02],\n",
      "          [-8.5753e-03, -3.2767e-03,  9.0087e-03]],\n",
      "\n",
      "         [[-1.8234e-02, -6.7848e-03,  1.9646e-02],\n",
      "          [ 2.6426e-02,  1.6720e-04,  1.8292e-03],\n",
      "          [-4.6752e-03,  1.0315e-02, -4.4858e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.3267e-02,  3.6214e-02, -1.0276e-02],\n",
      "          [ 9.6159e-03, -2.3365e-02,  1.6609e-02],\n",
      "          [ 5.0534e-02, -2.1281e-02, -1.8567e-02]],\n",
      "\n",
      "         [[ 1.7993e-03,  1.6062e-02,  9.8113e-03],\n",
      "          [-1.0189e-03,  9.1161e-03, -5.7486e-03],\n",
      "          [ 2.0758e-03,  1.7131e-02,  3.0668e-02]],\n",
      "\n",
      "         [[-7.6456e-03,  1.3166e-02,  2.0536e-03],\n",
      "          [ 2.2847e-02, -1.2913e-02,  2.7351e-02],\n",
      "          [-4.9152e-03,  2.8072e-02,  1.1845e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 1.1578e-02,  3.4811e-02, -2.4089e-02],\n",
      "          [ 5.3931e-03, -1.5719e-02, -1.9249e-02],\n",
      "          [-8.3535e-03, -1.4478e-02,  1.1905e-02]],\n",
      "\n",
      "         [[-2.2133e-02,  1.6893e-02,  3.6446e-03],\n",
      "          [-9.1980e-04,  1.3328e-02, -2.4034e-03],\n",
      "          [ 2.6148e-02, -5.0865e-03, -5.8180e-02]],\n",
      "\n",
      "         [[-6.1777e-02,  5.4675e-03,  7.4628e-03],\n",
      "          [ 9.0439e-05, -8.7669e-03, -9.7921e-03],\n",
      "          [-1.6945e-02,  5.8867e-03,  2.7113e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 1.3303e-04,  9.4807e-03, -8.5024e-03],\n",
      "          [ 6.7900e-03, -1.5502e-02, -7.9969e-03],\n",
      "          [ 9.9207e-03,  4.1237e-02, -3.9001e-03]],\n",
      "\n",
      "         [[-8.9462e-03, -5.0820e-03, -2.2374e-02],\n",
      "          [-4.7101e-02, -1.1938e-02, -5.5470e-02],\n",
      "          [-4.2945e-03, -4.4701e-03,  1.6548e-02]],\n",
      "\n",
      "         [[ 3.5235e-02, -4.2451e-03, -2.4367e-02],\n",
      "          [ 1.9934e-02,  3.9649e-03, -3.8741e-02],\n",
      "          [-5.6985e-03, -1.3507e-02, -1.8549e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 5.8560e-03, -5.2733e-03,  4.6975e-02],\n",
      "          [ 6.0746e-03, -1.8620e-03,  1.2580e-02],\n",
      "          [ 6.0772e-04,  8.0086e-03, -6.4337e-02]],\n",
      "\n",
      "         [[ 1.4952e-02,  3.1515e-02,  3.9892e-03],\n",
      "          [ 1.0265e-02, -4.1535e-03,  1.1264e-02],\n",
      "          [ 2.7382e-02,  8.5708e-03,  7.0595e-03]],\n",
      "\n",
      "         [[ 4.8214e-03,  1.5990e-02,  8.3233e-03],\n",
      "          [ 4.5168e-02,  7.2268e-04, -1.3044e-03],\n",
      "          [-3.7123e-02, -2.5354e-03,  2.5854e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.0859e-02,  6.8170e-02, -1.2308e-02],\n",
      "          [ 1.9642e-03,  3.8276e-02,  1.4362e-02],\n",
      "          [ 3.6024e-02, -2.5531e-03,  2.1016e-02]],\n",
      "\n",
      "         [[ 1.7181e-02,  1.5277e-02,  2.2581e-02],\n",
      "          [-3.7499e-02, -6.2870e-03,  9.5484e-03],\n",
      "          [ 1.1467e-02, -2.2792e-02,  5.2908e-03]],\n",
      "\n",
      "         [[-1.5388e-02,  4.7932e-02, -1.3893e-03],\n",
      "          [-1.0617e-02,  4.9928e-03, -3.3003e-03],\n",
      "          [ 3.4894e-03,  1.6918e-02, -1.6071e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 7.4810e-04, -3.4681e-02,  2.6725e-03],\n",
      "          [ 2.4404e-02, -2.4063e-02,  1.4105e-02],\n",
      "          [-3.9017e-02,  1.9258e-02,  1.4539e-02]],\n",
      "\n",
      "         [[ 6.1188e-02, -1.9034e-02, -3.3710e-03],\n",
      "          [ 1.3391e-02, -1.1827e-02,  4.4561e-03],\n",
      "          [-1.8688e-03,  6.6168e-03, -2.1776e-02]],\n",
      "\n",
      "         [[ 9.2083e-03, -3.9734e-03, -3.0878e-02],\n",
      "          [-6.8264e-03, -2.8882e-02,  1.0448e-03],\n",
      "          [ 2.1320e-02,  2.7241e-03,  2.0947e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.9056e-02, -2.3291e-02, -2.9870e-02],\n",
      "          [ 1.0383e-02,  1.9747e-02,  1.4074e-02],\n",
      "          [-9.5454e-03,  5.1451e-03, -1.8557e-02]],\n",
      "\n",
      "         [[ 4.1168e-02, -4.7596e-03, -2.4948e-02],\n",
      "          [-2.8801e-02,  1.0753e-02, -2.6829e-02],\n",
      "          [-4.6759e-03, -2.3469e-02, -2.1854e-02]],\n",
      "\n",
      "         [[-5.2090e-02,  7.1620e-03,  3.7034e-02],\n",
      "          [ 5.2398e-03, -1.8808e-02,  4.3432e-03],\n",
      "          [ 2.2484e-02,  4.0225e-03, -2.3067e-02]]]])), ('features.17.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.19.weight', tensor([[[[ 1.6682e-02, -4.8286e-03, -5.7655e-03],\n",
      "          [-9.1116e-03,  8.0941e-03, -4.4255e-03],\n",
      "          [-2.1678e-02, -1.0327e-04, -1.2437e-02]],\n",
      "\n",
      "         [[-1.7155e-02,  1.8404e-02,  3.0327e-03],\n",
      "          [ 2.9715e-02,  1.0640e-02,  9.0750e-03],\n",
      "          [ 2.9649e-02, -3.3763e-02,  1.7941e-02]],\n",
      "\n",
      "         [[-5.8155e-03,  9.9259e-03,  3.2548e-02],\n",
      "          [-9.1324e-03, -1.3650e-02,  3.5464e-02],\n",
      "          [-2.1410e-02, -9.9691e-03, -3.0668e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.5891e-02,  2.5794e-02,  2.1239e-02],\n",
      "          [ 1.9271e-02,  2.7600e-02, -1.0427e-02],\n",
      "          [-1.3865e-02, -2.2471e-02,  8.4914e-03]],\n",
      "\n",
      "         [[ 8.8838e-03,  1.3080e-02,  2.2605e-02],\n",
      "          [ 8.6915e-03,  5.1386e-03,  2.3485e-02],\n",
      "          [ 1.1814e-02, -8.7799e-03, -6.3472e-03]],\n",
      "\n",
      "         [[-2.8965e-04,  4.0621e-02, -7.4183e-03],\n",
      "          [ 2.5115e-02, -2.7635e-04,  2.1317e-02],\n",
      "          [-3.0889e-02,  3.4192e-02, -9.7632e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 1.9419e-02,  2.6263e-03,  1.0851e-02],\n",
      "          [-1.3461e-03, -2.6680e-03, -1.2104e-02],\n",
      "          [ 1.2360e-02,  6.7555e-04, -1.0292e-04]],\n",
      "\n",
      "         [[ 3.4805e-02, -2.4773e-03, -8.1447e-03],\n",
      "          [ 3.2683e-02,  4.3852e-03,  2.1389e-03],\n",
      "          [ 1.0637e-02,  1.8334e-02, -1.8036e-02]],\n",
      "\n",
      "         [[-2.1327e-02, -2.6901e-02, -7.5727e-03],\n",
      "          [ 1.3017e-02, -2.8609e-02,  4.1433e-02],\n",
      "          [-1.3124e-02, -3.5407e-02, -2.3913e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.5642e-03, -1.9734e-03, -1.5938e-02],\n",
      "          [-1.8151e-02,  3.1467e-02,  3.2463e-02],\n",
      "          [-4.5488e-02,  7.1415e-03, -1.8882e-02]],\n",
      "\n",
      "         [[-5.0685e-02, -1.7718e-02,  2.8716e-02],\n",
      "          [ 2.2850e-02, -1.6682e-02,  4.2901e-02],\n",
      "          [-8.9641e-03, -1.0563e-02, -9.9440e-03]],\n",
      "\n",
      "         [[-4.0926e-02, -2.5386e-02,  1.6183e-02],\n",
      "          [ 9.8235e-03,  3.7604e-04,  1.7556e-03],\n",
      "          [ 1.2976e-02, -3.6117e-03,  1.1292e-02]]],\n",
      "\n",
      "\n",
      "        [[[-9.1958e-03, -6.1140e-03,  3.9071e-03],\n",
      "          [ 3.7421e-03,  6.1215e-03,  2.0372e-03],\n",
      "          [-1.7799e-02,  3.8393e-02,  1.0931e-02]],\n",
      "\n",
      "         [[ 3.0733e-02, -1.7921e-03,  2.3083e-02],\n",
      "          [ 2.5795e-02, -1.8241e-03,  1.0359e-02],\n",
      "          [-4.4803e-03, -3.4364e-02, -8.7264e-03]],\n",
      "\n",
      "         [[-1.6523e-02,  1.0271e-02, -7.6502e-03],\n",
      "          [-3.1639e-02, -2.0269e-02, -2.0609e-03],\n",
      "          [-1.4977e-03, -2.1486e-02, -3.9217e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-7.6280e-03,  1.0616e-02,  1.5230e-02],\n",
      "          [-2.5661e-02,  4.8391e-03, -1.5761e-02],\n",
      "          [-1.6956e-02,  1.6303e-02, -4.0219e-02]],\n",
      "\n",
      "         [[-9.7940e-03, -2.0915e-02, -6.4425e-02],\n",
      "          [ 2.1594e-02,  2.4288e-02,  1.0682e-02],\n",
      "          [ 6.8693e-03, -1.3814e-02, -1.5151e-02]],\n",
      "\n",
      "         [[-2.9991e-02, -4.0975e-02, -2.5856e-02],\n",
      "          [-2.1552e-02,  1.7369e-02, -1.8010e-02],\n",
      "          [ 5.0218e-02, -1.2003e-03,  1.6518e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-1.0320e-02, -1.0765e-02,  1.9044e-02],\n",
      "          [-9.4786e-04,  2.6097e-02,  2.0181e-02],\n",
      "          [ 2.0905e-02,  1.6875e-02, -1.1826e-03]],\n",
      "\n",
      "         [[-3.8486e-02,  1.1957e-02,  9.6352e-03],\n",
      "          [ 2.3352e-02,  1.6294e-02,  3.0671e-03],\n",
      "          [ 8.1283e-03, -1.5216e-02,  7.2308e-03]],\n",
      "\n",
      "         [[-2.5199e-02,  8.3713e-03,  7.0088e-03],\n",
      "          [-3.0299e-02, -2.3253e-02, -3.7497e-02],\n",
      "          [-1.3641e-02,  2.4876e-02,  3.5553e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-4.5047e-02,  8.6116e-04, -3.1774e-02],\n",
      "          [-4.8246e-03, -8.0411e-03, -2.6311e-02],\n",
      "          [-7.6639e-03,  1.3780e-03, -4.4704e-03]],\n",
      "\n",
      "         [[-7.4062e-04,  2.1555e-02, -5.9455e-03],\n",
      "          [ 3.8568e-03, -3.9770e-02,  1.2948e-03],\n",
      "          [-5.7530e-03,  8.1819e-04,  1.3021e-02]],\n",
      "\n",
      "         [[-4.5193e-03,  1.4317e-02, -7.9586e-03],\n",
      "          [-2.3228e-02,  9.2547e-03,  8.7560e-03],\n",
      "          [-2.1431e-02, -1.2320e-02, -5.4140e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 2.0905e-02,  2.5207e-02,  2.5936e-05],\n",
      "          [ 7.8624e-03, -2.2031e-02, -6.2387e-03],\n",
      "          [-2.0816e-02, -2.7207e-03,  1.0189e-02]],\n",
      "\n",
      "         [[-4.4057e-03,  1.5556e-02, -8.7288e-03],\n",
      "          [-1.8830e-02,  6.2471e-03,  4.7682e-02],\n",
      "          [-8.8828e-03, -6.1424e-03, -6.5875e-03]],\n",
      "\n",
      "         [[-1.6442e-02,  1.9436e-03,  6.9444e-03],\n",
      "          [ 3.8780e-02,  2.4497e-02,  5.8671e-04],\n",
      "          [ 6.7259e-04, -8.4784e-03,  2.0812e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 6.4268e-03,  2.0607e-03, -4.1370e-03],\n",
      "          [-1.7716e-02, -3.1898e-02, -1.6816e-02],\n",
      "          [ 3.6192e-03, -2.4250e-02, -1.2757e-02]],\n",
      "\n",
      "         [[ 4.2087e-02,  1.5035e-02, -1.4428e-04],\n",
      "          [ 2.2718e-02, -4.9407e-02,  3.7624e-03],\n",
      "          [-1.7070e-02, -9.4667e-04,  1.4883e-02]],\n",
      "\n",
      "         [[ 1.0807e-02, -1.9698e-02, -3.4649e-03],\n",
      "          [-2.0951e-02,  2.6890e-02, -8.9816e-03],\n",
      "          [ 8.0066e-03, -1.2305e-02,  2.5397e-02]]],\n",
      "\n",
      "\n",
      "        [[[-1.6995e-02,  8.3505e-03, -8.4297e-03],\n",
      "          [-4.2716e-02,  1.8539e-02, -4.7730e-03],\n",
      "          [ 4.7995e-03, -3.3454e-02,  1.5789e-02]],\n",
      "\n",
      "         [[-2.6742e-02, -2.2377e-02, -1.1211e-02],\n",
      "          [ 3.0506e-02, -1.1437e-02, -1.3139e-02],\n",
      "          [-1.9996e-02, -2.7721e-02, -1.2886e-02]],\n",
      "\n",
      "         [[-1.9726e-02, -4.3408e-03, -3.9814e-03],\n",
      "          [-3.7861e-03,  8.2684e-03, -4.1457e-03],\n",
      "          [ 2.5249e-04, -1.1302e-02,  7.2957e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.0646e-02,  1.4476e-02,  4.6316e-02],\n",
      "          [ 4.2249e-03,  6.7192e-03, -1.6377e-02],\n",
      "          [ 9.1256e-03, -1.0832e-03, -1.0103e-02]],\n",
      "\n",
      "         [[ 7.5438e-03, -5.9867e-03, -2.5296e-02],\n",
      "          [-3.8453e-03, -1.6856e-02, -2.8444e-03],\n",
      "          [ 1.8320e-02,  1.7574e-03, -2.3611e-02]],\n",
      "\n",
      "         [[-8.3585e-03,  2.8815e-03, -2.2396e-02],\n",
      "          [-1.2036e-02, -1.4054e-02, -1.8215e-02],\n",
      "          [-1.3515e-03,  1.7600e-02, -1.9858e-02]]]])), ('features.19.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.21.weight', tensor([[[[ 6.7913e-04,  3.7883e-02, -1.0069e-02],\n",
      "          [ 3.7696e-02, -2.7948e-02,  1.9376e-03],\n",
      "          [ 2.4415e-03, -7.1744e-03, -1.6950e-02]],\n",
      "\n",
      "         [[-2.1079e-02, -3.0271e-02, -5.2252e-03],\n",
      "          [ 5.9825e-03,  4.0957e-02,  1.8459e-02],\n",
      "          [-4.9365e-03,  2.7527e-03, -1.0275e-02]],\n",
      "\n",
      "         [[ 5.4309e-02, -1.7159e-02, -1.8440e-02],\n",
      "          [ 3.2204e-02, -2.8033e-02, -5.0004e-03],\n",
      "          [ 1.9455e-03,  7.4202e-04, -2.0191e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 6.4143e-03,  1.7203e-02,  1.0126e-02],\n",
      "          [ 3.7318e-03, -2.6249e-02, -3.7371e-03],\n",
      "          [-3.7861e-02, -4.0444e-02,  1.1054e-02]],\n",
      "\n",
      "         [[-1.3824e-02,  1.3983e-02,  1.4000e-02],\n",
      "          [-1.7951e-03, -1.6711e-02, -3.7917e-03],\n",
      "          [ 5.0237e-02,  2.2256e-03, -1.0776e-02]],\n",
      "\n",
      "         [[-8.3325e-03,  1.9664e-02, -1.6993e-03],\n",
      "          [-1.5396e-02,  1.2803e-02, -1.5842e-02],\n",
      "          [ 1.4056e-02, -7.8327e-04, -3.5096e-02]]],\n",
      "\n",
      "\n",
      "        [[[-1.3266e-02,  1.8257e-03, -4.7886e-03],\n",
      "          [ 3.0293e-02, -1.0867e-03,  1.1360e-02],\n",
      "          [ 2.5141e-02, -4.9371e-03,  5.1707e-02]],\n",
      "\n",
      "         [[ 3.5178e-02,  1.4698e-02,  3.4075e-04],\n",
      "          [-3.7041e-03,  2.2784e-02, -9.0666e-03],\n",
      "          [-5.1989e-03, -3.5266e-02, -1.5797e-02]],\n",
      "\n",
      "         [[ 1.7003e-02, -3.4403e-03,  8.6588e-06],\n",
      "          [-8.9758e-03,  3.4249e-03, -4.8161e-03],\n",
      "          [-4.8937e-02, -4.6984e-03,  2.2361e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 1.1639e-02,  3.9580e-03, -2.4139e-03],\n",
      "          [-1.3134e-02,  2.5664e-02, -1.5432e-02],\n",
      "          [-1.2670e-02, -3.2207e-02, -7.7309e-03]],\n",
      "\n",
      "         [[-3.8441e-02, -9.0577e-03, -1.0537e-02],\n",
      "          [-1.1785e-02,  2.2206e-04, -3.3475e-02],\n",
      "          [ 1.1359e-02,  1.5746e-02, -3.8821e-02]],\n",
      "\n",
      "         [[-2.0172e-02, -2.3242e-02, -1.2824e-02],\n",
      "          [-1.0743e-02, -8.4829e-03, -9.3224e-03],\n",
      "          [ 1.4635e-02,  2.2530e-02, -1.3620e-02]]],\n",
      "\n",
      "\n",
      "        [[[-3.4982e-02,  6.4684e-03, -4.3797e-03],\n",
      "          [-4.8759e-02, -8.7256e-03,  8.7891e-03],\n",
      "          [-2.8683e-02,  2.1213e-02,  2.3256e-02]],\n",
      "\n",
      "         [[-1.9122e-02,  6.1323e-03, -4.9622e-02],\n",
      "          [-2.0853e-02,  2.5149e-02,  1.1820e-02],\n",
      "          [ 1.4159e-02,  1.0226e-03,  1.1662e-02]],\n",
      "\n",
      "         [[-1.4970e-02, -4.9139e-03,  7.8696e-03],\n",
      "          [-8.8358e-03, -1.3553e-02,  9.5734e-03],\n",
      "          [-1.7812e-02, -3.0135e-02, -1.7485e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.1965e-02, -5.7162e-02, -1.1648e-02],\n",
      "          [-9.4672e-03, -3.6087e-02,  5.5106e-02],\n",
      "          [-2.2539e-03,  2.0752e-02, -2.0052e-03]],\n",
      "\n",
      "         [[-6.8742e-02, -1.9565e-02, -1.2837e-02],\n",
      "          [ 6.1161e-03,  2.4728e-03,  1.7426e-02],\n",
      "          [ 1.0395e-02, -2.6503e-02, -2.1520e-03]],\n",
      "\n",
      "         [[-4.5066e-03, -2.0193e-03, -2.2242e-02],\n",
      "          [-5.1360e-02, -3.1500e-02, -2.1070e-02],\n",
      "          [ 1.3474e-02, -3.1736e-02, -2.3859e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-1.9527e-02,  3.3411e-04,  1.9802e-02],\n",
      "          [-1.1761e-02,  2.6157e-02, -8.2287e-03],\n",
      "          [ 2.4748e-02,  1.5359e-02, -9.9618e-03]],\n",
      "\n",
      "         [[-3.4784e-02,  1.3025e-02, -2.8655e-03],\n",
      "          [ 3.0120e-04,  1.2455e-02, -1.3957e-02],\n",
      "          [ 1.0012e-02,  8.3676e-03, -4.1081e-03]],\n",
      "\n",
      "         [[ 2.0982e-02, -1.8684e-03,  4.6839e-02],\n",
      "          [-1.6798e-02, -1.8333e-02, -1.6985e-02],\n",
      "          [-3.4370e-03,  1.2304e-02,  1.1886e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 9.1729e-03,  2.2695e-02, -1.0652e-02],\n",
      "          [ 3.3410e-02,  2.5597e-02, -5.4330e-02],\n",
      "          [-1.4923e-02, -3.5785e-03, -3.9114e-04]],\n",
      "\n",
      "         [[-1.3692e-02,  9.0915e-04,  1.4682e-02],\n",
      "          [-3.3662e-03, -8.0015e-03,  7.9627e-03],\n",
      "          [-3.3838e-02, -1.3590e-02, -2.0434e-02]],\n",
      "\n",
      "         [[ 2.1995e-02,  3.3335e-02, -8.9438e-03],\n",
      "          [-2.1710e-02,  1.0612e-02,  1.3747e-02],\n",
      "          [-2.9752e-03,  2.3690e-02, -1.0659e-02]]],\n",
      "\n",
      "\n",
      "        [[[-1.9775e-02,  1.6155e-02,  2.1342e-03],\n",
      "          [ 3.2210e-02, -2.0765e-02,  2.3513e-03],\n",
      "          [ 1.0781e-03, -2.6103e-02, -6.6850e-03]],\n",
      "\n",
      "         [[ 7.1004e-03,  5.7949e-03, -1.9414e-02],\n",
      "          [-8.8707e-03, -1.4804e-02,  9.7878e-03],\n",
      "          [-2.1196e-02, -4.0492e-02, -2.2145e-02]],\n",
      "\n",
      "         [[ 1.7322e-02, -3.7201e-02, -1.6349e-02],\n",
      "          [ 3.5522e-02,  1.2576e-02, -2.2611e-02],\n",
      "          [ 2.3540e-04,  9.1097e-03,  1.5394e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 5.2979e-03, -1.5541e-02, -1.6889e-03],\n",
      "          [ 1.3413e-02, -1.4310e-02,  1.2549e-02],\n",
      "          [-1.8398e-02, -2.7379e-03, -1.0650e-03]],\n",
      "\n",
      "         [[ 3.6801e-02,  1.9516e-04,  3.0216e-02],\n",
      "          [-3.1944e-03, -1.9570e-02, -8.7997e-03],\n",
      "          [-2.1287e-02,  1.7163e-02, -4.3893e-03]],\n",
      "\n",
      "         [[-2.9992e-02, -2.1959e-02,  4.3976e-02],\n",
      "          [-1.4159e-02, -2.7641e-02,  7.4165e-03],\n",
      "          [-2.2940e-02, -2.8233e-02, -1.6026e-02]]],\n",
      "\n",
      "\n",
      "        [[[-4.1821e-03, -2.7790e-02, -4.6386e-02],\n",
      "          [-1.6926e-02, -1.0875e-02,  3.8525e-03],\n",
      "          [-1.4909e-02,  8.0921e-03,  3.9320e-03]],\n",
      "\n",
      "         [[ 1.1873e-02,  8.5355e-03,  2.1286e-02],\n",
      "          [-5.8336e-03, -8.2942e-03, -1.9853e-02],\n",
      "          [ 1.7961e-03, -4.0590e-03,  2.5403e-02]],\n",
      "\n",
      "         [[-6.4134e-03,  7.6504e-03, -4.1213e-03],\n",
      "          [ 1.5786e-02,  2.7674e-04, -5.4896e-03],\n",
      "          [ 3.5755e-03, -1.3752e-02, -2.0332e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-3.6798e-02, -1.3672e-02, -6.4134e-03],\n",
      "          [ 2.5003e-02, -1.4994e-02, -1.0799e-02],\n",
      "          [-5.6315e-03,  8.2930e-03,  2.1430e-02]],\n",
      "\n",
      "         [[ 1.0245e-02, -4.6340e-02,  4.2320e-02],\n",
      "          [ 5.7954e-03,  1.1107e-02, -8.8707e-03],\n",
      "          [-2.2425e-03, -9.9925e-03,  2.3391e-02]],\n",
      "\n",
      "         [[ 1.2318e-02, -9.4089e-03, -5.3406e-03],\n",
      "          [-3.7463e-02, -3.8024e-03,  1.6953e-02],\n",
      "          [ 3.4084e-02, -1.1958e-02,  8.8512e-03]]]])), ('features.21.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.24.weight', tensor([[[[ 0.0392,  0.0345,  0.0296],\n",
      "          [-0.0043,  0.0086,  0.0185],\n",
      "          [ 0.0104,  0.0210,  0.0046]],\n",
      "\n",
      "         [[-0.0563, -0.0421,  0.0193],\n",
      "          [-0.0173, -0.0087,  0.0141],\n",
      "          [ 0.0254, -0.0066,  0.0167]],\n",
      "\n",
      "         [[-0.0336,  0.0210,  0.0071],\n",
      "          [ 0.0056, -0.0119, -0.0065],\n",
      "          [ 0.0250, -0.0157,  0.0164]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0052,  0.0006,  0.0122],\n",
      "          [ 0.0080, -0.0040, -0.0022],\n",
      "          [-0.0234, -0.0094,  0.0294]],\n",
      "\n",
      "         [[-0.0027,  0.0230, -0.0002],\n",
      "          [-0.0308, -0.0086, -0.0017],\n",
      "          [ 0.0129,  0.0076, -0.0187]],\n",
      "\n",
      "         [[ 0.0276,  0.0287,  0.0075],\n",
      "          [-0.0081, -0.0291,  0.0075],\n",
      "          [ 0.0127,  0.0116,  0.0287]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0126,  0.0303, -0.0041],\n",
      "          [-0.0204,  0.0204,  0.0146],\n",
      "          [ 0.0172, -0.0322, -0.0120]],\n",
      "\n",
      "         [[ 0.0021,  0.0121, -0.0168],\n",
      "          [-0.0138, -0.0102,  0.0226],\n",
      "          [ 0.0055,  0.0031,  0.0245]],\n",
      "\n",
      "         [[-0.0073, -0.0100, -0.0262],\n",
      "          [ 0.0052, -0.0134,  0.0106],\n",
      "          [-0.0201, -0.0057, -0.0341]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0055,  0.0380, -0.0071],\n",
      "          [ 0.0307, -0.0462,  0.0089],\n",
      "          [-0.0129, -0.0270, -0.0220]],\n",
      "\n",
      "         [[-0.0066, -0.0316, -0.0118],\n",
      "          [-0.0042,  0.0280, -0.0154],\n",
      "          [-0.0210,  0.0271,  0.0116]],\n",
      "\n",
      "         [[ 0.0040, -0.0195,  0.0101],\n",
      "          [ 0.0127,  0.0227, -0.0183],\n",
      "          [-0.0337, -0.0206, -0.0263]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0198,  0.0215,  0.0286],\n",
      "          [-0.0238,  0.0127, -0.0227],\n",
      "          [-0.0023,  0.0139, -0.0036]],\n",
      "\n",
      "         [[ 0.0260, -0.0020,  0.0087],\n",
      "          [ 0.0206, -0.0181,  0.0041],\n",
      "          [-0.0017,  0.0125, -0.0078]],\n",
      "\n",
      "         [[ 0.0082, -0.0151, -0.0165],\n",
      "          [-0.0046, -0.0052,  0.0105],\n",
      "          [ 0.0421,  0.0233, -0.0168]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0132,  0.0117, -0.0643],\n",
      "          [-0.0712, -0.0014, -0.0039],\n",
      "          [ 0.0012, -0.0587, -0.0283]],\n",
      "\n",
      "         [[ 0.0516, -0.0079,  0.0127],\n",
      "          [-0.0292,  0.0014,  0.0119],\n",
      "          [-0.0156,  0.0176,  0.0177]],\n",
      "\n",
      "         [[ 0.0494, -0.0119,  0.0202],\n",
      "          [-0.0026,  0.0277, -0.0072],\n",
      "          [ 0.0057, -0.0162, -0.0105]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[ 0.0072, -0.0251, -0.0234],\n",
      "          [ 0.0142, -0.0156, -0.0199],\n",
      "          [ 0.0077,  0.0027,  0.0006]],\n",
      "\n",
      "         [[-0.0081,  0.0207,  0.0203],\n",
      "          [ 0.0327, -0.0215, -0.0239],\n",
      "          [-0.0262,  0.0331, -0.0116]],\n",
      "\n",
      "         [[-0.0009, -0.0111, -0.0226],\n",
      "          [-0.0022, -0.0277,  0.0080],\n",
      "          [ 0.0120, -0.0298, -0.0339]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 0.0296, -0.0191, -0.0207],\n",
      "          [ 0.0163,  0.0331,  0.0377],\n",
      "          [-0.0207, -0.0277, -0.0099]],\n",
      "\n",
      "         [[ 0.0359, -0.0292,  0.0314],\n",
      "          [-0.0107, -0.0036, -0.0320],\n",
      "          [ 0.0027,  0.0027,  0.0143]],\n",
      "\n",
      "         [[ 0.0082, -0.0265,  0.0552],\n",
      "          [ 0.0242,  0.0009,  0.0005],\n",
      "          [ 0.0077,  0.0049,  0.0129]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0114, -0.0072,  0.0012],\n",
      "          [-0.0177,  0.0276, -0.0185],\n",
      "          [-0.0131, -0.0049, -0.0212]],\n",
      "\n",
      "         [[-0.0030,  0.0257,  0.0010],\n",
      "          [-0.0205, -0.0342, -0.0014],\n",
      "          [ 0.0220,  0.0212,  0.0171]],\n",
      "\n",
      "         [[-0.0199,  0.0008, -0.0083],\n",
      "          [ 0.0268, -0.0314,  0.0198],\n",
      "          [ 0.0427,  0.0108, -0.0054]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0158,  0.0100,  0.0068],\n",
      "          [-0.0218, -0.0003, -0.0210],\n",
      "          [-0.0182,  0.0557,  0.0328]],\n",
      "\n",
      "         [[-0.0041,  0.0243, -0.0666],\n",
      "          [ 0.0161, -0.0283, -0.0211],\n",
      "          [-0.0403,  0.0046,  0.0063]],\n",
      "\n",
      "         [[-0.0067, -0.0158,  0.0186],\n",
      "          [ 0.0356,  0.0149, -0.0341],\n",
      "          [-0.0288, -0.0136, -0.0047]]],\n",
      "\n",
      "\n",
      "        [[[ 0.0683,  0.0009,  0.0421],\n",
      "          [ 0.0177, -0.0044, -0.0033],\n",
      "          [-0.0361, -0.0032, -0.0250]],\n",
      "\n",
      "         [[ 0.0034,  0.0005,  0.0443],\n",
      "          [-0.0431, -0.0123, -0.0036],\n",
      "          [ 0.0098,  0.0170, -0.0062]],\n",
      "\n",
      "         [[ 0.0083, -0.0038,  0.0167],\n",
      "          [ 0.0240,  0.0175,  0.0012],\n",
      "          [ 0.0027,  0.0070,  0.0161]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-0.0006,  0.0172, -0.0002],\n",
      "          [ 0.0085,  0.0225, -0.0144],\n",
      "          [ 0.0051, -0.0017,  0.0084]],\n",
      "\n",
      "         [[-0.0289,  0.0474, -0.0357],\n",
      "          [ 0.0098, -0.0063,  0.0298],\n",
      "          [-0.0081,  0.0439, -0.0114]],\n",
      "\n",
      "         [[-0.0056, -0.0190,  0.0015],\n",
      "          [ 0.0132,  0.0109, -0.0512],\n",
      "          [ 0.0167,  0.0205,  0.0227]]]])), ('features.24.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.26.weight', tensor([[[[-1.6113e-02, -3.1643e-02, -8.9414e-03],\n",
      "          [-3.0833e-02, -2.4871e-03,  2.1722e-02],\n",
      "          [-3.6739e-03, -1.7219e-02,  1.0381e-02]],\n",
      "\n",
      "         [[ 1.1289e-02, -1.5219e-02, -9.2427e-03],\n",
      "          [ 2.6984e-02, -2.6399e-02,  1.7534e-02],\n",
      "          [-1.3972e-02,  2.6627e-02, -5.4636e-03]],\n",
      "\n",
      "         [[-3.5469e-03, -4.2022e-02,  1.6063e-02],\n",
      "          [ 1.3006e-02, -5.5555e-05,  1.1058e-02],\n",
      "          [ 2.8691e-03, -8.1209e-03,  3.1208e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.7651e-04,  2.2797e-02,  1.7886e-03],\n",
      "          [-4.8038e-03,  1.6033e-02,  1.0965e-02],\n",
      "          [ 1.8229e-02, -1.9613e-03, -8.6358e-03]],\n",
      "\n",
      "         [[ 2.0921e-02, -3.5178e-03, -6.8566e-03],\n",
      "          [-1.0536e-02,  4.2675e-02,  4.1752e-02],\n",
      "          [-1.2348e-02, -7.3629e-03,  9.7740e-03]],\n",
      "\n",
      "         [[ 3.4843e-02, -4.8805e-03, -9.2775e-03],\n",
      "          [ 1.7957e-02,  1.7350e-02, -4.6127e-02],\n",
      "          [-9.7791e-03, -1.2861e-03,  7.0591e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 2.3066e-02, -5.7661e-03,  1.3359e-02],\n",
      "          [-4.0767e-02, -1.5051e-02, -1.1035e-02],\n",
      "          [-1.5424e-02, -5.4685e-03, -5.8271e-03]],\n",
      "\n",
      "         [[ 4.0120e-02, -2.0359e-02,  6.8472e-03],\n",
      "          [-6.3965e-03,  2.6133e-02, -3.0684e-02],\n",
      "          [ 1.3124e-02, -2.4845e-02, -1.0697e-02]],\n",
      "\n",
      "         [[-1.8724e-02,  1.9447e-02,  3.4954e-02],\n",
      "          [-7.0367e-03, -1.8307e-02, -9.3814e-03],\n",
      "          [-2.9639e-02, -2.7211e-02, -1.4595e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.2860e-02, -1.1758e-02, -1.5146e-02],\n",
      "          [-1.5549e-04,  3.0116e-02,  1.9603e-02],\n",
      "          [ 1.0912e-02,  2.5546e-03,  3.8748e-02]],\n",
      "\n",
      "         [[ 1.7312e-02, -1.2116e-02,  4.0106e-03],\n",
      "          [-2.9217e-03, -2.5432e-02, -8.6465e-03],\n",
      "          [ 6.2906e-03,  2.7120e-02,  1.4674e-02]],\n",
      "\n",
      "         [[ 3.3391e-03,  1.5759e-02, -1.0101e-02],\n",
      "          [-1.9282e-02, -1.0581e-02,  2.3090e-02],\n",
      "          [ 4.7505e-03,  6.5524e-03, -3.4457e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 1.6999e-02,  4.2478e-03, -3.1130e-03],\n",
      "          [-1.2922e-02,  1.8181e-03,  5.5340e-03],\n",
      "          [ 2.8363e-03, -9.8430e-03, -1.6130e-02]],\n",
      "\n",
      "         [[ 3.1433e-04,  2.4754e-03, -2.3053e-02],\n",
      "          [-2.1913e-02,  8.6636e-03,  3.4677e-03],\n",
      "          [-2.7932e-02, -5.1641e-03,  2.5979e-02]],\n",
      "\n",
      "         [[-4.7695e-03,  1.6996e-02, -2.1125e-02],\n",
      "          [-3.7504e-02,  2.5140e-02,  9.9471e-04],\n",
      "          [-2.1335e-03, -8.7493e-03,  6.8224e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.9965e-03,  2.7332e-04, -2.0970e-02],\n",
      "          [-5.2889e-03,  1.4541e-02,  3.8752e-03],\n",
      "          [ 1.3597e-02, -1.6221e-02, -7.0303e-03]],\n",
      "\n",
      "         [[-6.4427e-03,  5.8118e-03, -8.3713e-03],\n",
      "          [-2.3980e-02,  6.8634e-03, -4.8325e-03],\n",
      "          [ 9.0454e-04, -5.2553e-03,  2.4248e-02]],\n",
      "\n",
      "         [[ 1.3376e-02,  1.3216e-02,  7.0154e-03],\n",
      "          [-2.6011e-02,  3.4105e-03, -6.2273e-03],\n",
      "          [ 3.9411e-02, -9.6214e-03,  3.6158e-03]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-4.5335e-03, -1.0164e-02,  3.1238e-02],\n",
      "          [ 2.9447e-02,  4.2972e-02,  3.4138e-02],\n",
      "          [ 2.5550e-02,  6.9810e-03,  2.9446e-02]],\n",
      "\n",
      "         [[-5.8083e-03, -1.7616e-02,  1.2783e-03],\n",
      "          [ 1.1269e-02,  4.1715e-03, -1.7200e-02],\n",
      "          [ 2.9605e-02,  4.3536e-03,  4.8746e-03]],\n",
      "\n",
      "         [[-2.6903e-02,  5.8314e-04, -1.5813e-03],\n",
      "          [-4.6498e-03, -1.8734e-02,  1.3307e-02],\n",
      "          [ 4.6437e-03,  1.0477e-02,  3.3219e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 6.5461e-03,  3.2539e-02,  7.1562e-03],\n",
      "          [-7.3586e-03,  4.5247e-03, -2.6854e-02],\n",
      "          [ 2.0586e-02, -1.2486e-02,  2.5954e-02]],\n",
      "\n",
      "         [[-1.0268e-02, -1.7653e-02,  2.3557e-04],\n",
      "          [-1.0831e-02, -6.8212e-03,  9.5876e-03],\n",
      "          [ 2.4509e-02,  9.4460e-03, -1.3900e-02]],\n",
      "\n",
      "         [[-9.5878e-03,  2.7034e-02,  6.9945e-03],\n",
      "          [-1.9462e-02,  1.5376e-02, -5.4774e-03],\n",
      "          [-7.6628e-04, -4.3698e-02,  2.3133e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 3.9448e-02,  1.2482e-03,  4.5816e-02],\n",
      "          [-9.5891e-03,  1.4354e-04, -1.5883e-02],\n",
      "          [ 1.7720e-02, -3.0719e-03,  2.9139e-03]],\n",
      "\n",
      "         [[-2.2877e-03,  9.3075e-03,  3.2416e-02],\n",
      "          [ 2.5772e-02,  2.0950e-03,  2.1766e-02],\n",
      "          [ 3.8733e-02, -4.6185e-02,  1.6779e-02]],\n",
      "\n",
      "         [[ 5.1422e-03, -1.9640e-02, -2.2203e-02],\n",
      "          [ 6.6071e-03, -2.6707e-03, -2.1203e-02],\n",
      "          [ 6.9031e-03, -1.5620e-02, -3.2459e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-3.4424e-02,  1.7513e-02,  2.7454e-02],\n",
      "          [-2.1728e-02,  1.3778e-02, -5.7812e-03],\n",
      "          [ 3.4845e-03, -2.2221e-02,  3.1509e-02]],\n",
      "\n",
      "         [[-2.7943e-02,  1.1031e-03,  4.0349e-04],\n",
      "          [-4.7038e-02, -9.6642e-03,  1.3429e-02],\n",
      "          [-1.9471e-02,  1.5958e-02,  4.5608e-03]],\n",
      "\n",
      "         [[ 2.2295e-04,  1.2617e-02, -1.6024e-02],\n",
      "          [-2.8492e-02, -1.6713e-03,  2.9203e-04],\n",
      "          [ 5.5706e-03,  3.4496e-02,  9.1851e-03]]],\n",
      "\n",
      "\n",
      "        [[[ 6.3155e-02,  3.9283e-02,  7.7750e-02],\n",
      "          [ 1.1010e-02,  1.8840e-02, -5.8392e-03],\n",
      "          [ 3.5478e-02, -7.8551e-03,  8.8507e-03]],\n",
      "\n",
      "         [[ 4.0149e-03, -1.9943e-02, -2.7410e-02],\n",
      "          [-4.8233e-02,  1.3717e-03,  9.0000e-04],\n",
      "          [ 6.9609e-03,  1.5190e-04, -2.6530e-02]],\n",
      "\n",
      "         [[-1.9012e-02, -6.8783e-04,  4.8842e-04],\n",
      "          [ 1.3499e-02,  2.9359e-02, -8.7118e-03],\n",
      "          [-3.1522e-02, -8.9710e-03,  5.3872e-03]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 1.2918e-02,  1.1378e-02,  1.6370e-02],\n",
      "          [-6.6914e-03, -3.8365e-02, -1.0900e-02],\n",
      "          [-1.4432e-02, -2.0509e-02, -1.2176e-02]],\n",
      "\n",
      "         [[-4.9055e-02, -9.1158e-03, -1.5665e-02],\n",
      "          [ 2.9165e-02, -1.9346e-02,  1.0260e-02],\n",
      "          [-3.2069e-03,  1.9321e-03, -2.3949e-02]],\n",
      "\n",
      "         [[ 3.7106e-02, -4.2898e-03, -3.8835e-02],\n",
      "          [ 5.3973e-04,  5.0226e-02,  1.9207e-02],\n",
      "          [-3.1604e-02,  1.2607e-02, -1.1266e-02]]]])), ('features.26.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('features.28.weight', tensor([[[[ 1.3294e-02, -1.0723e-02,  2.9744e-02],\n",
      "          [ 2.8903e-02,  3.5660e-02, -1.0562e-02],\n",
      "          [-7.2417e-03, -1.3068e-02, -3.0763e-02]],\n",
      "\n",
      "         [[ 2.5755e-04,  2.8229e-02,  1.2125e-02],\n",
      "          [ 1.1119e-02, -2.8119e-04,  3.7761e-03],\n",
      "          [-9.1151e-03, -3.0420e-02, -1.7987e-02]],\n",
      "\n",
      "         [[-3.0279e-03,  3.6247e-02,  1.8744e-02],\n",
      "          [ 2.5128e-02, -1.1578e-02, -2.0302e-03],\n",
      "          [-9.8241e-03,  1.1639e-03,  1.1370e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 2.9948e-02,  1.9768e-02, -2.0515e-02],\n",
      "          [ 2.9377e-02, -4.3536e-02, -2.0580e-02],\n",
      "          [ 2.4591e-02,  1.1973e-02, -4.8728e-02]],\n",
      "\n",
      "         [[-7.4338e-03,  2.6912e-02, -6.2521e-03],\n",
      "          [ 1.8102e-02,  2.2358e-02, -1.5016e-02],\n",
      "          [ 8.0082e-03, -3.0602e-03,  2.8319e-04]],\n",
      "\n",
      "         [[-2.9842e-02,  1.9564e-02,  1.3212e-02],\n",
      "          [ 8.4526e-03, -2.7503e-02,  1.9999e-02],\n",
      "          [ 1.0340e-02,  2.2322e-02, -3.0661e-02]]],\n",
      "\n",
      "\n",
      "        [[[-2.8789e-02, -2.3858e-02,  7.6929e-03],\n",
      "          [ 4.9500e-02, -6.5701e-03, -2.1801e-02],\n",
      "          [ 1.8393e-02, -7.5672e-03, -2.7840e-02]],\n",
      "\n",
      "         [[-1.7074e-02,  1.3450e-02, -2.0333e-02],\n",
      "          [-2.9383e-02, -3.7348e-03, -1.9601e-02],\n",
      "          [-1.3827e-02,  4.0025e-03, -1.6497e-02]],\n",
      "\n",
      "         [[ 2.4285e-02,  3.2758e-02, -1.3479e-02],\n",
      "          [ 1.7020e-02, -7.5953e-03,  1.5601e-02],\n",
      "          [ 8.3546e-03, -3.0462e-03,  2.3758e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-1.6864e-02, -3.3992e-02,  4.1263e-02],\n",
      "          [-9.1143e-04,  3.8641e-03, -6.7985e-03],\n",
      "          [-9.0212e-03, -3.2898e-02,  6.3832e-05]],\n",
      "\n",
      "         [[ 1.1394e-02, -6.0643e-03,  4.1277e-03],\n",
      "          [ 9.8712e-03, -7.3746e-03,  4.8501e-03],\n",
      "          [ 2.0374e-02,  2.5832e-02,  7.3732e-03]],\n",
      "\n",
      "         [[-2.9328e-03,  2.7516e-02,  3.0441e-02],\n",
      "          [ 2.8656e-03,  9.8491e-03,  1.1178e-02],\n",
      "          [-1.3448e-03, -2.4091e-02, -1.9475e-02]]],\n",
      "\n",
      "\n",
      "        [[[-1.2761e-02,  1.6289e-02, -1.8452e-03],\n",
      "          [ 2.4663e-02,  1.3251e-02, -1.3973e-02],\n",
      "          [-1.4091e-02,  1.8897e-02, -8.5916e-03]],\n",
      "\n",
      "         [[-2.3741e-04,  3.8926e-02,  1.6559e-02],\n",
      "          [ 1.6918e-02, -2.7996e-03, -1.9736e-02],\n",
      "          [-1.7108e-02,  3.4969e-02,  2.6052e-02]],\n",
      "\n",
      "         [[ 1.9889e-02,  8.0235e-03, -3.0879e-03],\n",
      "          [-2.0893e-03,  2.6518e-02,  1.7839e-02],\n",
      "          [-4.1641e-02,  3.4477e-03, -1.2353e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-2.9984e-03, -1.2647e-02,  1.5694e-02],\n",
      "          [ 7.8857e-03, -1.8574e-02, -2.6935e-02],\n",
      "          [-2.8747e-02,  5.4362e-02,  2.9189e-02]],\n",
      "\n",
      "         [[ 1.0434e-02,  1.3153e-02,  3.5004e-02],\n",
      "          [ 6.7382e-04,  9.5156e-03,  7.0734e-03],\n",
      "          [-1.2039e-03,  5.8025e-03,  1.9250e-02]],\n",
      "\n",
      "         [[ 6.3616e-03,  1.3821e-02,  7.6097e-03],\n",
      "          [ 4.1193e-03, -2.9957e-02, -1.0766e-02],\n",
      "          [ 1.6936e-02,  1.5415e-02, -1.1101e-02]]],\n",
      "\n",
      "\n",
      "        ...,\n",
      "\n",
      "\n",
      "        [[[-3.8537e-03, -3.7667e-02,  3.0051e-02],\n",
      "          [ 1.5097e-02,  6.4089e-03, -1.6207e-02],\n",
      "          [ 1.4222e-03,  1.1623e-02, -2.6380e-02]],\n",
      "\n",
      "         [[ 1.1068e-02,  2.5740e-02,  4.0970e-03],\n",
      "          [-2.1168e-02, -2.3213e-02, -1.0292e-02],\n",
      "          [ 6.8297e-03, -2.9867e-02,  2.0869e-02]],\n",
      "\n",
      "         [[ 5.2141e-03,  1.8055e-02,  2.0633e-02],\n",
      "          [ 4.2736e-03,  1.4931e-02,  1.2443e-02],\n",
      "          [ 1.8632e-04,  3.3759e-02, -1.4888e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 3.8311e-02,  2.1952e-02, -9.2052e-03],\n",
      "          [ 1.2227e-02, -5.4735e-04,  1.2155e-02],\n",
      "          [-1.5917e-02, -2.3659e-02,  2.9862e-03]],\n",
      "\n",
      "         [[ 5.1840e-03,  4.2812e-04,  5.8954e-03],\n",
      "          [-6.6519e-03,  1.0446e-02, -2.7579e-03],\n",
      "          [ 5.3675e-03,  4.3602e-03,  1.1339e-02]],\n",
      "\n",
      "         [[-4.5833e-03, -5.1103e-03,  1.3782e-02],\n",
      "          [-3.5696e-02, -3.8571e-02,  1.2824e-02],\n",
      "          [ 9.8456e-03,  1.2443e-02, -2.2173e-02]]],\n",
      "\n",
      "\n",
      "        [[[-9.4143e-04, -2.3261e-03,  3.3037e-02],\n",
      "          [-2.5762e-03, -1.0751e-02, -2.1031e-02],\n",
      "          [-3.1595e-03, -1.3291e-02,  6.9542e-03]],\n",
      "\n",
      "         [[ 8.6266e-04, -5.2262e-03, -4.2945e-03],\n",
      "          [-3.0937e-03,  7.5974e-03,  1.0244e-02],\n",
      "          [-3.3212e-02, -2.9974e-02,  2.0553e-02]],\n",
      "\n",
      "         [[-2.9643e-02,  1.7775e-02, -2.2697e-02],\n",
      "          [ 2.7335e-02,  2.2941e-02, -3.9827e-02],\n",
      "          [ 2.2169e-02,  8.3535e-02,  1.1410e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[ 1.5241e-02, -3.9138e-02, -2.8165e-02],\n",
      "          [ 1.0179e-02,  1.6887e-02, -9.7327e-03],\n",
      "          [ 5.3969e-03, -2.6010e-02, -2.5985e-03]],\n",
      "\n",
      "         [[-3.6294e-02,  2.7138e-02,  2.1312e-02],\n",
      "          [-4.5966e-02, -8.1846e-03, -2.2565e-03],\n",
      "          [-1.9215e-02,  4.2630e-02, -3.5719e-02]],\n",
      "\n",
      "         [[-4.1432e-02,  5.0816e-02, -6.5246e-03],\n",
      "          [-7.0435e-03,  1.3991e-03,  1.7577e-02],\n",
      "          [ 1.1973e-02, -2.5753e-02,  3.7631e-02]]],\n",
      "\n",
      "\n",
      "        [[[ 1.1429e-02,  1.0403e-02, -1.0900e-03],\n",
      "          [-1.9854e-02, -9.3881e-03,  1.9029e-02],\n",
      "          [ 2.8609e-03,  1.3377e-02,  1.0098e-02]],\n",
      "\n",
      "         [[-9.4272e-03,  1.2461e-02,  2.0667e-02],\n",
      "          [-2.5731e-02,  3.0918e-03, -2.4821e-02],\n",
      "          [-8.0219e-03, -2.3782e-03,  1.9467e-02]],\n",
      "\n",
      "         [[ 2.9004e-02, -1.2569e-03,  2.9710e-02],\n",
      "          [-2.2040e-03,  1.3430e-02,  2.8936e-03],\n",
      "          [ 2.5642e-02,  2.2777e-02,  1.1646e-02]],\n",
      "\n",
      "         ...,\n",
      "\n",
      "         [[-4.7712e-03, -2.9191e-02,  1.8980e-02],\n",
      "          [ 2.7791e-02, -1.8423e-02,  1.3739e-02],\n",
      "          [ 3.2074e-03,  1.5439e-03, -2.6040e-02]],\n",
      "\n",
      "         [[ 1.9905e-02,  2.5032e-02,  2.0747e-02],\n",
      "          [ 1.3369e-02, -3.0880e-02, -3.2563e-03],\n",
      "          [-1.8468e-02, -1.3404e-02, -1.4254e-02]],\n",
      "\n",
      "         [[ 3.0152e-02,  7.8657e-03,  8.2863e-03],\n",
      "          [ 6.5479e-03, -1.2454e-02, -1.1695e-02],\n",
      "          [ 9.1283e-03,  3.1459e-02,  2.9980e-02]]]])), ('features.28.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0.])), ('classifier.0.weight', tensor([[ 0.0096,  0.0037,  0.0145,  ..., -0.0190, -0.0147,  0.0065],\n",
      "        [ 0.0031, -0.0156, -0.0074,  ..., -0.0091, -0.0011,  0.0053],\n",
      "        [ 0.0150,  0.0061,  0.0215,  ..., -0.0011,  0.0038,  0.0104],\n",
      "        ...,\n",
      "        [ 0.0054,  0.0172, -0.0011,  ...,  0.0087, -0.0072,  0.0051],\n",
      "        [-0.0051,  0.0115, -0.0127,  ..., -0.0044,  0.0157, -0.0020],\n",
      "        [ 0.0094, -0.0104,  0.0084,  ...,  0.0011, -0.0174, -0.0065]])), ('classifier.0.bias', tensor([0., 0., 0.,  ..., 0., 0., 0.])), ('classifier.3.weight', tensor([[-3.5545e-03, -1.2687e-02,  1.7251e-03,  ..., -5.1235e-03,\n",
      "          1.4304e-02,  1.3995e-02],\n",
      "        [ 1.4689e-02,  7.9987e-03, -1.7486e-02,  ...,  6.3619e-04,\n",
      "         -2.8975e-02, -1.9667e-03],\n",
      "        [ 2.7256e-03, -4.6652e-03, -1.0054e-02,  ..., -8.2107e-03,\n",
      "          2.2397e-02,  1.9811e-02],\n",
      "        ...,\n",
      "        [ 1.2933e-02,  8.3498e-03,  9.9967e-04,  ..., -2.3078e-03,\n",
      "          9.2060e-03,  4.9232e-03],\n",
      "        [-4.8290e-03, -1.5238e-03,  9.0799e-03,  ..., -9.7217e-04,\n",
      "         -1.2381e-03, -3.7515e-05],\n",
      "        [-4.4716e-03,  1.7077e-02,  1.2799e-03,  ...,  4.1426e-03,\n",
      "          9.4201e-03, -4.2960e-03]])), ('classifier.3.bias', tensor([0., 0., 0.,  ..., 0., 0., 0.])), ('classifier.6.weight', tensor([[-1.0420e-02,  1.1247e-02,  4.4231e-03,  ..., -1.2903e-02,\n",
      "          2.1508e-02, -4.7858e-04],\n",
      "        [ 8.0994e-03, -3.2156e-03,  1.2291e-02,  ...,  2.5068e-02,\n",
      "         -7.5288e-03, -1.2252e-03],\n",
      "        [ 2.1609e-03, -1.8573e-02,  2.3817e-03,  ..., -1.7167e-03,\n",
      "          4.6937e-03,  9.4094e-03],\n",
      "        ...,\n",
      "        [-1.2712e-02,  5.8916e-03, -1.8256e-02,  ..., -1.9786e-02,\n",
      "         -1.0236e-02, -2.4478e-03],\n",
      "        [-9.5590e-05, -9.5327e-03, -5.7350e-04,  ..., -3.3392e-03,\n",
      "         -1.1314e-02, -3.9965e-03],\n",
      "        [-2.8020e-03,  8.2193e-03, -3.0818e-03,  ...,  1.6548e-02,\n",
      "         -2.1522e-02, -1.3440e-02]])), ('classifier.6.bias', tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
      "        0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.]))])\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "\n",
    "model = torch.load(\"./model/vgg16_method2.pth\")  # 导入模型参数\n",
    "\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "而为了得到模型的话，我们则需要先得到一个模型，然后再使用load_state_dict()方法，将权重值赋予给这个模型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=1000, bias=True)\n",
      "  )\n",
      ")\n",
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=1000, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "\n",
    "vgg16 = torchvision.models.vgg16(pretrained=False)\n",
    "print(vgg16)\n",
    "\n",
    "vgg16.load_state_dict(torch.load(\"./model/vgg16_method2.pth\"))  # 将模型参数导入到模型结构中   \n",
    "print(vgg16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. 网络陷阱-创建模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "\n",
    "class Tudui(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Tudui,self).__init__()\n",
    "        self.conv1 = nn.Conv2d(3, 64, kernel_size=3)\n",
    "        \n",
    "    def forward(self,x):\n",
    "        x = self.conv1(x)\n",
    "        return x\n",
    "\n",
    "tudui = Tudui()\n",
    "torch.save(tudui, \"./model/tudui_method1.pth\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. 网络陷阱-失败加载模型"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "① 点击 Kernel，再点击 Restart。"
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "② 再运行下面的代码，即下面为第1个代码块运行，无法直接导入网络模型。（即使pth中保存了模型结构&模型参数，在没看到模型结构的定义的情况下，仍然无法读到具体的结构？）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "Can't get attribute 'Tudui' on <module '__main__'>",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_2812403/1581248595.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"./model/tudui_method1.pth\"\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# 无法直接加载方式一保存的网络结构\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.9/site-packages/torch/serialization.py\u001b[0m in \u001b[0;36mload\u001b[0;34m(f, map_location, pickle_module, **pickle_load_args)\u001b[0m\n\u001b[1;32m    710\u001b[0m                     \u001b[0mopened_file\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mseek\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0morig_position\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    711\u001b[0m                     \u001b[0;32mreturn\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjit\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mopened_file\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 712\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0m_load\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mopened_zipfile\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmap_location\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpickle_module\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mpickle_load_args\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    713\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0m_legacy_load\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mopened_file\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmap_location\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpickle_module\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mpickle_load_args\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    714\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.9/site-packages/torch/serialization.py\u001b[0m in \u001b[0;36m_load\u001b[0;34m(zip_file, map_location, pickle_module, pickle_file, **pickle_load_args)\u001b[0m\n\u001b[1;32m   1047\u001b[0m     \u001b[0munpickler\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mUnpicklerWrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata_file\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mpickle_load_args\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1048\u001b[0m     \u001b[0munpickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpersistent_load\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpersistent_load\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1049\u001b[0;31m     \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0munpickler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1050\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1051\u001b[0m     \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_utils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_validate_loaded_sparse_tensors\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.9/site-packages/torch/serialization.py\u001b[0m in \u001b[0;36mfind_class\u001b[0;34m(self, mod_name, name)\u001b[0m\n\u001b[1;32m   1040\u001b[0m                     \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1041\u001b[0m             \u001b[0mmod_name\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload_module_mapping\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmod_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmod_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1042\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind_class\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmod_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1043\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1044\u001b[0m     \u001b[0;31m# Load the data (which may in turn use `persistent_load` to load tensors)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAttributeError\u001b[0m: Can't get attribute 'Tudui' on <module '__main__'>"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "\n",
    "model = torch.load(\"./model/tudui_method1.pth\")  # 无法直接加载方式一保存的网络结构\n",
    "\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7. 网络陷阱-成功加载模型(方式一)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tudui(\n",
      "  (conv1): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1))\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "\n",
    "# 确保网络模型是我们想要的网络模型,要在加载前还写明网络模型\n",
    "class Tudui(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Tudui,self).__init__()\n",
    "        self.conv1 = nn.Conv2d(3,64,kernel_size=3)\n",
    "        \n",
    "    def forward(self,x):\n",
    "        x = self.conv1(x)\n",
    "        return x\n",
    "\n",
    "# tudui = Tudui()  # 不需要写这一步，不需要创建网络模型    \n",
    "model = torch.load(\"./model/tudui_method1.pth\")  # 无法直接加载方式一保存的网络结构    \n",
    "print(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8. 网络陷阱-成功加载模型(方式二)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from model_save import *  # 它就相当于把 model_save.py 里的 Tudui 的网络模型定义写到这里了\n",
    "\n",
    "#tudui = Tudui() # 不需要写这一步，不需要创建网络模型    \n",
    "model = torch.load(\"tudui_method1.pth\")\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "350px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}