{ "cells": [ { "cell_type": "code", "execution_count": 1, "id": "858273c0-29b6-4600-a936-da5f3ebdd6e0", "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Epoch: 1 \t Loss: 2.301505\n", "Epoch 1 time: 18.78 seconds\n", "Test -- Average loss: 0.0007, Accuracy: 98.600\n", "\n", "Train Epoch: 2 \t Loss: 0.072409\n", "Epoch 2 time: 19.17 seconds\n", "Test -- Average loss: 0.0006, Accuracy: 98.730\n", "\n", "Train Epoch: 3 \t Loss: 0.023574\n", "Epoch 3 time: 18.58 seconds\n", "Test -- Average loss: 0.0006, Accuracy: 98.730\n", "\n", "Train Epoch: 4 \t Loss: 0.029821\n", "Epoch 4 time: 19.15 seconds\n", "Test -- Average loss: 0.0006, Accuracy: 98.790\n", "\n", "Train Epoch: 5 \t Loss: 0.061575\n", "Epoch 5 time: 17.77 seconds\n", "Test -- Average loss: 0.0005, Accuracy: 98.940\n", "\n", "Train Epoch: 6 \t Loss: 0.000245\n", "Epoch 6 time: 17.83 seconds\n", "Test -- Average loss: 0.0005, Accuracy: 99.090\n", "\n", "Train Epoch: 7 \t Loss: 0.000922\n", "Epoch 7 time: 17.68 seconds\n", "Test -- Average loss: 0.0006, Accuracy: 98.930\n", "\n", "Train Epoch: 8 \t Loss: 0.044371\n", "Epoch 8 time: 17.54 seconds\n", "Test -- Average loss: 0.0008, Accuracy: 98.910\n", "\n", "Train Epoch: 9 \t Loss: 0.000014\n", "Epoch 9 time: 17.84 seconds\n", "Test -- Average loss: 0.0007, Accuracy: 98.870\n", "\n", "Train Epoch: 10 \t Loss: 0.144589\n", "Epoch 10 time: 18.12 seconds\n", "Test -- Average loss: 0.0007, Accuracy: 98.980\n", "\n", "Total training time: 182.44 seconds\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import torch\n", "import torch.nn as nn\n", "import torch.nn.functional as F\n", "import torch.optim as optim\n", "from torchvision import datasets, transforms\n", "from torch.utils.data import DataLoader\n", "import matplotlib.pyplot as plt\n", "import time\n", "\n", "# 超参数\n", "BATCH_SIZE = 64\n", "DEVICE = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", "EPOCHS = 10\n", "\n", "# 图像处理\n", "pipeline = transforms.Compose(\n", " [transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))]\n", ")\n", "\n", "# 下载加载数据集\n", "train_set = datasets.MNIST(\"data\", train=True, download=True, transform=pipeline)\n", "test_set = datasets.MNIST(\"data\", train=False, download=True, transform=pipeline)\n", "\n", "# 加载数据\n", "train_loader = DataLoader(train_set, batch_size=BATCH_SIZE, shuffle=True)\n", "test_loader = DataLoader(test_set, batch_size=BATCH_SIZE, shuffle=True)\n", "\n", "\n", "# 构建网络模型\n", "class Digit(nn.Module):\n", " def __init__(self):\n", " super().__init__()\n", " self.conv1 = nn.Conv2d(1, 10, 5)\n", " self.conv2 = nn.Conv2d(10, 20, 3)\n", " self.fc1 = nn.Linear(20 * 10 * 10, 500)\n", " self.fc2 = nn.Linear(500, 10)\n", "\n", " def forward(self, x):\n", " input_size = x.size(0)\n", " x = self.conv1(x)\n", " x = F.relu(x)\n", " x = F.max_pool2d(x, 2, 2)\n", "\n", " x = self.conv2(x)\n", " x = F.relu(x)\n", "\n", " x = x.view(input_size, -1)\n", "\n", " x = self.fc1(x)\n", " x = F.relu(x)\n", "\n", " x = self.fc2(x)\n", "\n", " output = F.log_softmax(x, dim=1)\n", "\n", " return output\n", "\n", "\n", "# 优化器\n", "model = Digit().to(DEVICE)\n", "optimizer = optim.Adam(model.parameters())\n", "\n", "\n", "# 训练方法\n", "def train_model(model, device, train_loader, optimizer, epoch):\n", " model.train()\n", " train_losses = []\n", " train_accuracy = []\n", " start_time = time.time()\n", " for batch_index, (data, target) in enumerate(train_loader):\n", " data, target = data.to(device), target.to(device)\n", " optimizer.zero_grad()\n", " output = model(data)\n", " loss = F.cross_entropy(output, target)\n", " loss.backward()\n", " optimizer.step()\n", " if batch_index % 1000 == 0:\n", " print(\"Train Epoch: {} \\t Loss: {:.6f}\".format(epoch, loss.item()))\n", "\n", " # 计算准确率\n", " model.eval()\n", " correct = 0.0\n", " with torch.no_grad():\n", " for data, target in train_loader:\n", " data, target = data.to(device), target.to(device)\n", " output = model(data)\n", " pred = output.max(1, keepdim=True)[1]\n", " correct += pred.eq(target.view_as(pred)).sum().item()\n", " accuracy = 100.0 * correct / len(train_loader.dataset)\n", " train_accuracy.append(accuracy)\n", "\n", " train_losses.append(loss.item())\n", " end_time = time.time()\n", " epoch_time = end_time - start_time\n", "\n", " return train_losses, train_accuracy, epoch_time\n", "\n", "\n", "# 测试方法\n", "def test_model(model, device, test_loader):\n", " model.eval()\n", " correct = 0.0\n", " test_loss = 0.0\n", " with torch.no_grad():\n", " for data, target in test_loader:\n", " data, target = data.to(device), target.to(device)\n", " output = model(data)\n", " test_loss += F.cross_entropy(output, target).item()\n", " pred = output.max(1, keepdim=True)[1]\n", " correct += pred.eq(target.view_as(pred)).sum().item()\n", " test_loss /= len(test_loader.dataset)\n", " print(\n", " \"Test -- Average loss: {:.4f}, Accuracy: {:.3f}\\n\".format(\n", " test_loss, 100.0 * correct / len(test_loader.dataset)\n", " )\n", " )\n", "\n", "\n", "# 调用\n", "train_losses = []\n", "train_accuracy = []\n", "total_time = 0.0\n", "for epoch in range(1, EPOCHS + 1):\n", " losses, accuracy, epoch_time = train_model(model, DEVICE, train_loader, optimizer, epoch)\n", " train_losses.extend(losses)\n", " train_accuracy.extend(accuracy)\n", " total_time += epoch_time\n", " print(\"Epoch {} time: {:.2f} seconds\".format(epoch, epoch_time))\n", " test_model(model, DEVICE, test_loader)\n", "\n", "print(\"Total training time: {:.2f} seconds\".format(total_time))\n", "\n", "# 绘制损失曲线\n", "plt.plot(range(1, EPOCHS + 1), train_losses, label=\"Training Loss\")\n", "plt.xlabel(\"Epochs\")\n", "plt.ylabel(\"Loss\")\n", "plt.title(\"Training Loss\")\n", "plt.legend()\n", "plt.show()\n", "\n", "# 绘制准确率曲线\n", "plt.plot(range(1, EPOCHS + 1), train_accuracy, label=\"Training Accuracy\")\n", "plt.xlabel(\"Epochs\")\n", "plt.ylabel(\"Accuracy\")\n", "plt.title(\"Training Accuracy\")\n", "plt.legend()\n", "plt.show()\n" ] }, { "cell_type": "code", "execution_count": null, "id": "6ed815e9-735b-459d-a43c-ae90f7ad34f2", "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": "3.11.5" } }, "nbformat": 4, "nbformat_minor": 5 }