cifar-100.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pytorchcv.model_provider import get_model as ptcv_get_model\n",
    "import torch\n",
    "import torch.nn.utils.prune as prune\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch.autograd import Variable\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "net = ptcv_get_model(\"resnet20_cifar100\", pretrained=True)\n",
    "net = net.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n",
      "Files already downloaded and verified\n"
     ]
    }
   ],
   "source": [
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "transform = transforms.Compose(\n",
    "    [transforms.ToTensor(),\n",
    "     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])\n",
    "\n",
    "trainset = torchvision.datasets.CIFAR100(root='./data', train=True,\n",
    "                                        download=True, transform=transform)\n",
    "trainloader = torch.utils.data.DataLoader(trainset, batch_size=20,\n",
    "                                          shuffle=True, num_workers=2)\n",
    "\n",
    "testset = torchvision.datasets.CIFAR100(root='./data', train=False,\n",
    "                                       download=True, transform=transform)\n",
    "testloader = torch.utils.data.DataLoader(testset, batch_size=4,\n",
    "                                         shuffle=False, num_workers=2)\n",
    "\n",
    "classes = (\"beaver\", \"dolphin\", \"otter\", \"seal\", \"whale\", \"aquarium fish\", \"flatfish\", \"ray\", \"shark\", \"trout\", \"orchids\", \\\n",
    "           \"poppies\", \"roses\", \"sunflowers\", \"tulips\", \"bottles\", \"bowls\", \"cans\", \"cups\", \"plates\", \"apples\", \"mushrooms\", \\\n",
    "           \"oranges\", \"pears\", \"sweet peppers\", \"clock\", \"computer keyboard\", \"lamp\", \"telephone\", \"television\", \"bed\", \\\n",
    "           \"chair\", \"couch\", \"table\", \"wardrobe\", \"bee\", \"beetle\", \"butterfly\", \"caterpillar\", \"cockroach\", \"bear\", \"leopard\", \\\n",
    "           \"lion\", \"tiger\", \"wolf\", \"bridge\", \"castle\", \"house\", \"road\", \"skyscraper\", \"cloud\", \"forest\", \"mountain\", \"plain\", \"sea\", \\\n",
    "           \"camel\", \"cattle\", \"chimpanzee\", \"elephant\", \"kangaroo\", \"fox\", \"porcupine\", \"possum\", \"raccoon\", \"skunk\", \"crab\", \\\n",
    "           \"lobster\", \"snail\", \"spider\", \"worm\", \"baby\", \"boy\", \"girl\", \"man\", \"woman\", \"crocodile\", \"dinosaur\", \"lizard\", \"snake\",\\\n",
    "           \"turtle\", \"hamster\", \"mouse\", \"rabbit\", \"shrew\", \"squirrel\", \"maple\", \"oak\", \"palm\", \"pine\", \"willow\", \"bicycle\", \"bus\",\\\n",
    "           \"motorcycle\", \"pickup truck\", \"train\", \"lawn-mower\", \"rocket\", \"streetcar\", \"tank\", \"tractor\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " worm couch shrew bicycle\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# functions to show an image\n",
    "\n",
    "\n",
    "def imshow(img):\n",
    "    img = img / 2 + 0.5     # unnormalize\n",
    "    npimg = img.numpy()\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "# get some random training images\n",
    "dataiter = iter(trainloader)\n",
    "images, labels = dataiter.next()\n",
    "\n",
    "# show images\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "# print labels\n",
    "print(' '.join('%5s' % classes[labels[j]] for j in range(4)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "parameters_to_prune = (\n",
    "    (net.features.init_block.conv, \"weight\"),\n",
    "    (net.features.stage1.unit1.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage1.unit1.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage1.unit2.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage1.unit2.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage1.unit3.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage1.unit3.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage2.unit1.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage2.unit1.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage2.unit1.identity_conv.conv, \"weight\"),\n",
    "    (net.features.stage2.unit2.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage2.unit2.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage2.unit3.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage2.unit3.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage3.unit1.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage3.unit1.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage3.unit1.identity_conv.conv, \"weight\"),\n",
    "    (net.features.stage3.unit2.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage3.unit2.body.conv2.conv, \"weight\"),\n",
    "    (net.features.stage3.unit3.body.conv1.conv, \"weight\"),\n",
    "    (net.features.stage3.unit3.body.conv2.conv, \"weight\"),\n",
    "    (net.output, \"weight\")\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.optim as optim\n",
    "\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.8, weight_decay=1e-4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test():\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for i, data in enumerate(testloader, 0):\n",
    "            if i > 25:\n",
    "                break\n",
    "            images, labels = data[0].to(device), data[1].to(device)\n",
    "            outputs = net(images)\n",
    "            _, predicted = torch.max(outputs.data, 1)\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "\n",
    "    print('Accuracy of the network on 100 test images: %d %%' % (\n",
    "        100 * correct / total))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on 100 test images: 55 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 56 %\n",
      "Accuracy of the network on 100 test images: 56 %\n",
      "Accuracy of the network on 100 test images: 53 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "Accuracy of the network on 100 test images: 52 %\n",
      "Accuracy of the network on 100 test images: 51 %\n",
      "Accuracy of the network on 100 test images: 53 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 49 %\n",
      "Accuracy of the network on 100 test images: 49 %\n",
      "Accuracy of the network on 100 test images: 51 %\n",
      "Accuracy of the network on 100 test images: 51 %\n",
      "Accuracy of the network on 100 test images: 56 %\n",
      "Accuracy of the network on 100 test images: 52 %\n",
      "Accuracy of the network on 100 test images: 52 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "Accuracy of the network on 100 test images: 56 %\n",
      "Accuracy of the network on 100 test images: 52 %\n",
      "Accuracy of the network on 100 test images: 55 %\n",
      "Accuracy of the network on 100 test images: 51 %\n",
      "Accuracy of the network on 100 test images: 52 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 46 %\n",
      "Accuracy of the network on 100 test images: 46 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "Accuracy of the network on 100 test images: 48 %\n",
      "Accuracy of the network on 100 test images: 48 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 34 %\n",
      "Accuracy of the network on 100 test images: 34 %\n",
      "Accuracy of the network on 100 test images: 42 %\n",
      "Accuracy of the network on 100 test images: 46 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 45 %\n",
      "Accuracy of the network on 100 test images: 45 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 41 %\n",
      "Accuracy of the network on 100 test images: 40 %\n",
      "Accuracy of the network on 100 test images: 43 %\n",
      "Accuracy of the network on 100 test images: 44 %\n",
      "Accuracy of the network on 100 test images: 42 %\n",
      "Accuracy of the network on 100 test images: 48 %\n",
      "Accuracy of the network on 100 test images: 50 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 27 %\n",
      "Accuracy of the network on 100 test images: 27 %\n",
      "Accuracy of the network on 100 test images: 35 %\n",
      "Accuracy of the network on 100 test images: 37 %\n",
      "Accuracy of the network on 100 test images: 34 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 40 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 24 %\n",
      "Accuracy of the network on 100 test images: 24 %\n",
      "Accuracy of the network on 100 test images: 37 %\n",
      "Accuracy of the network on 100 test images: 35 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 37 %\n",
      "Accuracy of the network on 100 test images: 33 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 18 %\n",
      "Accuracy of the network on 100 test images: 19 %\n",
      "Accuracy of the network on 100 test images: 29 %\n",
      "Accuracy of the network on 100 test images: 34 %\n",
      "Accuracy of the network on 100 test images: 35 %\n",
      "Accuracy of the network on 100 test images: 36 %\n",
      "Accuracy of the network on 100 test images: 36 %\n",
      "prune\n",
      "Accuracy of the network on 100 test images: 17 %\n",
      "Accuracy of the network on 100 test images: 17 %\n",
      "Accuracy of the network on 100 test images: 28 %\n",
      "Accuracy of the network on 100 test images: 31 %\n",
      "Accuracy of the network on 100 test images: 32 %\n",
      "Accuracy of the network on 100 test images: 32 %\n"
     ]
    }
   ],
   "source": [
    "n_prune = 10\n",
    "for p in range(n_prune):\n",
    "    test()\n",
    "    print(\"prune\")\n",
    "    prune.global_unstructured(\n",
    "                parameters_to_prune,\n",
    "                pruning_method=prune.L1Unstructured,\n",
    "                amount=0.2,\n",
    "            )\n",
    "    test()\n",
    "    for i, data in enumerate(trainloader, 0):\n",
    "        # get the inputs; data is a list of [inputs, labels]\n",
    "        inputs, labels = data[0].to(device), data[1].to(device)\n",
    "\n",
    "        # zero the parameter gradients\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        # forward + backward + optimize\n",
    "        outputs = net(inputs)\n",
    "        loss = criterion(outputs, labels)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        # print statistics\n",
    "        if i % 500 == 0:\n",
    "            test()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 40 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 39 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 37 %\n",
      "Accuracy of the network on 100 test images: 38 %\n",
      "Accuracy of the network on 100 test images: 36 %\n",
      "Accuracy of the network on 100 test images: 37 %\n",
      "Accuracy of the network on 100 test images: 38 %\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-10-a019d24aef0f>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     10\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m         \u001b[1;31m# forward + backward + optimize\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 12\u001b[1;33m         \u001b[0moutputs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnet\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     13\u001b[0m         \u001b[0mloss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlabels\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m         \u001b[0mloss\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\pytorchcv\\models\\resnet_cifar.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m     85\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     86\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 87\u001b[1;33m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfeatures\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     88\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mview\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     89\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0moutput\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\container.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, input)\u001b[0m\n\u001b[0;32m    115\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    116\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 117\u001b[1;33m             \u001b[0minput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    118\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    119\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\container.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, input)\u001b[0m\n\u001b[0;32m    115\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    116\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 117\u001b[1;33m             \u001b[0minput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    118\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    119\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\pytorchcv\\models\\resnet.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m    180\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    181\u001b[0m             \u001b[0midentity\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 182\u001b[1;33m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbody\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    183\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mx\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0midentity\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    184\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mactiv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\pytorchcv\\models\\resnet.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m     54\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     55\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconv1\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 56\u001b[1;33m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconv2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     57\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     58\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\pytorchcv\\models\\common.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m    449\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0muse_pad\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    450\u001b[0m             \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpad\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 451\u001b[1;33m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    452\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0muse_bn\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    453\u001b[0m             \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    726\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 727\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    728\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\conv.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, input)\u001b[0m\n\u001b[0;32m    421\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    422\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 423\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_conv_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mweight\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    424\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    425\u001b[0m \u001b[1;32mclass\u001b[0m \u001b[0mConv3d\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m_ConvNd\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\miniconda3\\lib\\site-packages\\torch\\nn\\modules\\conv.py\u001b[0m in \u001b[0;36m_conv_forward\u001b[1;34m(self, input, weight)\u001b[0m\n\u001b[0;32m    417\u001b[0m                             \u001b[0mweight\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbias\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstride\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    418\u001b[0m                             _pair(0), self.dilation, self.groups)\n\u001b[1;32m--> 419\u001b[1;33m         return F.conv2d(input, weight, self.bias, self.stride,\n\u001b[0m\u001b[0;32m    420\u001b[0m                         self.padding, self.dilation, self.groups)\n\u001b[0;32m    421\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "optimizer = optim.SGD(net.parameters(), lr=0.0005)\n",
    "n_epoch = 5\n",
    "for p in range(n_epoch):\n",
    "    for i, data in enumerate(trainloader, 0):\n",
    "        # get the inputs; data is a list of [inputs, labels]\n",
    "        inputs, labels = data[0].to(device), data[1].to(device)\n",
    "\n",
    "        # zero the parameter gradients\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        # forward + backward + optimize\n",
    "        outputs = net(inputs)\n",
    "        loss = criterion(outputs, labels)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        # print statistics\n",
    "        if i % 500 == 0:\n",
    "            test()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on 10000 test images: 39 %\n"
     ]
    }
   ],
   "source": [
    "if True: \n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for i, data in enumerate(testloader, 0):\n",
    "            images, labels = data[0].to(device), data[1].to(device)\n",
    "            outputs = net(images)\n",
    "            _, predicted = torch.max(outputs.data, 1)\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "\n",
    "    print('Accuracy of the network on 10000 test images: %d %%' % (\n",
    "        100 * correct / total))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "262.0  /  432\n",
      "568.0  /  2304\n",
      "524.0  /  2304\n",
      "493.0  /  2304\n",
      "479.0  /  2304\n",
      "673.0  /  2304\n",
      "426.0  /  2304\n",
      "1036.0  /  4608\n",
      "1407.0  /  9216\n",
      "196.0  /  512\n",
      "1203.0  /  9216\n",
      "799.0  /  9216\n",
      "1317.0  /  9216\n",
      "711.0  /  9216\n",
      "1905.0  /  18432\n",
      "3099.0  /  36864\n",
      "420.0  /  2048\n",
      "2891.0  /  36864\n",
      "2430.0  /  36864\n",
      "3412.0  /  36864\n",
      "1497.0  /  36864\n",
      "3957.0  /  6400\n"
     ]
    }
   ],
   "source": [
    "for (p, _) in parameters_to_prune:\n",
    "    print(torch.sum(p.weight_mask).item(), \" / \", torch.sum(1*p.weight_mask >= 0).item())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}