CNN/NiN.ipynb

{
 "cells": [
  {
   "attachments": {
    "image-2.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "卷积层的输入和输出由四维张量组成，张量的每个轴分别对应样本、通道、高度和宽度。 另外，全连接层的输入和输出通常是分别对应于样本和特征的二维张量。 NiN的想法是在每个像素位置（针对每个高度和宽度）应用一个全连接层。 如果我们将权重连接到每个空间位置，我们可以将其视为\n",
    "卷积层，或作为在每个像素位置上独立作用的全连接层。 从另一个角度看，即将空间维度中的每个像素视为单个样本，将通道维度视为不同特征（feature）。\n",
    "\n",
    "下图说明了VGG和NiN及它们的块之间主要架构差异。 NiN块以一个普通卷积层开始，后面是两个1x1的卷积层。这两个1x1卷积层充当带有ReLU激活函数的逐像素全连接层。 第一层的卷积窗口形状通常由用户设置。 随后的卷积窗口形状固定为1x1。\n",
    "\n",
    "![image-2.png](attachment:image-2.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "import sys\n",
    "sys.path.append('..')\n",
    "from tools import tool_pytorch_017 as ty\n",
    "\n",
    "\n",
    "def nin_block(in_channels, out_channels, kernel_size, strides, padding):\n",
    "    return nn.Sequential(\n",
    "        nn.Conv2d(in_channels, out_channels, kernel_size, strides, padding),\n",
    "        nn.ReLU(),\n",
    "        nn.Conv2d(out_channels, out_channels, kernel_size=1), nn.ReLU(),\n",
    "        nn.Conv2d(out_channels, out_channels, kernel_size=1), nn.ReLU())\n",
    "\n",
    "net = nn.Sequential(\n",
    "    nin_block(1, 96, kernel_size=11, strides=4, padding=0),\n",
    "    nn.MaxPool2d(3, stride=2),\n",
    "    nin_block(96, 256, kernel_size=5, strides=1, padding=2),\n",
    "    nn.MaxPool2d(3, stride=2),\n",
    "    nin_block(256, 384, kernel_size=3, strides=1, padding=1),\n",
    "    nn.MaxPool2d(3, stride=2),\n",
    "    nn.Dropout(0.5),\n",
    "    # 标签类别数是10\n",
    "    nin_block(384, 10, kernel_size=3, strides=1, padding=1),\n",
    "    nn.AdaptiveAvgPool2d((1, 1)),\n",
    "    # 将四维的输出转成二维的输出，其形状为(批量大小,10)\n",
    "    nn.Flatten())\n",
    "\n",
    "# 查看AlexNet的分层结构输出\n",
    "X = torch.rand(size=(1, 1, 224, 224))\n",
    "for layer in net:\n",
    "    X = layer(X)\n",
    "    print(layer.__class__.__name__,'output shape:\\t', X.shape)\n",
    "    \n",
    "batch_size = 128\n",
    "train_iter, test_iter = ty.load_data_fashion_mnist(batch_size, resize=224)\n",
    "\n",
    "def evaluate_accuracy_gpu(net, data_iter, device=None): #@save\n",
    "    \"\"\"使用GPU计算模型在数据集上的精度\"\"\"\n",
    "    if isinstance(net, nn.Module):\n",
    "        net.eval()  # 设置为评估模式\n",
    "        if not device:\n",
    "            device = next(iter(net.parameters())).device\n",
    "    # 正确预测的数量，总预测的数量\n",
    "    metric = ty.Accumulator(2)\n",
    "    with torch.no_grad():\n",
    "        for X, y in data_iter:\n",
    "            if isinstance(X, list):\n",
    "                # BERT微调所需的（之后将介绍）\n",
    "                X = [x.to(device) for x in X]\n",
    "            else:\n",
    "                X = X.to(device)\n",
    "            y = y.to(device)\n",
    "            metric.add(ty.accuracy(net(X), y), y.numel())\n",
    "    return metric[0] / metric[1]\n",
    "\n",
    "\n",
    "#@save\n",
    "def train_AlexNet(net, train_iter, test_iter, num_epochs, lr, device):\n",
    "    \"\"\"训练AlexNet模型\"\"\"\n",
    "    def init_weights(m):\n",
    "        if type(m) == nn.Linear or type(m) == nn.Conv2d:\n",
    "            nn.init.xavier_uniform_(m.weight)\n",
    "    net.apply(init_weights)\n",
    "    print('training on', device)\n",
    "    net.to(device)\n",
    "    optimizer = torch.optim.SGD(net.parameters(), lr=lr)\n",
    "    loss = nn.CrossEntropyLoss()\n",
    "    animator = ty.Animator(xlabel='epoch', xlim=[1, num_epochs],\n",
    "                            legend=['train loss', 'train acc', 'test acc'])\n",
    "    timer, num_batches = ty.Timer(), len(train_iter)\n",
    "    for epoch in range(num_epochs):\n",
    "        # 训练损失之和，训练准确率之和，样本数\n",
    "        metric = ty.Accumulator(3)\n",
    "        net.train()\n",
    "        for i, (X, y) in enumerate(train_iter):\n",
    "            timer.start()\n",
    "            optimizer.zero_grad()\n",
    "            X, y = X.to(device), y.to(device)\n",
    "            y_hat = net(X)\n",
    "            l = loss(y_hat, y)\n",
    "            l.backward()\n",
    "            optimizer.step()\n",
    "            with torch.no_grad():\n",
    "                metric.add(l * X.shape[0], ty.accuracy(y_hat, y), X.shape[0])\n",
    "            timer.stop()\n",
    "            train_l = metric[0] / metric[2]\n",
    "            train_acc = metric[1] / metric[2]\n",
    "            if (i + 1) % (num_batches // 5) == 0 or i == num_batches - 1:\n",
    "                animator.add(epoch + (i + 1) / num_batches,\n",
    "                             (train_l, train_acc, None))\n",
    "        test_acc = evaluate_accuracy_gpu(net, test_iter)\n",
    "        animator.add(epoch + 1, (None, None, test_acc))\n",
    "    print(f'loss {train_l:.3f}, train acc {train_acc:.3f}, '\n",
    "          f'test acc {test_acc:.3f}')\n",
    "    print(f'{metric[2] * num_epochs / timer.sum():.1f} examples/sec '\n",
    "          f'on {str(device)}')\n",
    "    \n",
    "lr, num_epochs = 0.1, 10\n",
    "train_AlexNet(net, train_iter, test_iter, num_epochs, lr, ty.try_gpu())"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "python3.8",
   "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": "3.8.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}