Cifar10直接练.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a06895fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import torch\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "from torchvision.datasets import CIFAR10\n",
    "from torch.utils.data import DataLoader\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# from rich import print\n",
    "import time \n",
    "# 全局取消证书验证\n",
    "import ssl\n",
    "ssl._create_default_https_context = ssl._create_unverified_context\n",
    "\n",
    "\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c2d514c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def timeSince(since: float) -> str:\n",
    "    s = time.time() - since\n",
    "    return f\"{int(s/60):2}m {int(s%60):02}s\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8e8d3ca",
   "metadata": {},
   "source": [
    "## 加载并归一化数据集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6190e466",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset CIFAR10\n",
      "    Number of datapoints: 50000\n",
      "    Root location: /home/ma-user/data\n",
      "    Split: Train\n",
      "    StandardTransform\n",
      "Transform: Compose(\n",
      "               RandomHorizontalFlip(p=0.9)\n",
      "               ToTensor()\n",
      "               Normalize(mean=[0.4915 0.4823 0.4468], std=[0.247  0.2435 0.2616])\n",
      "           )\n"
     ]
    }
   ],
   "source": [
    "myTransforms = transforms.Compose([\n",
    "    transforms.Resize((32, 32)),\n",
    "    transforms.RandomHorizontalFlip(p=0.5),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])\n",
    "\n",
    "batch_size = 128\n",
    "\n",
    "trainset = CIFAR10(root='./cifar-10-python', train=True, download=True, transform=myTransforms)\n",
    "trainloader = DataLoader(trainset, batch_size=batch_size, shuffle=True, num_workers=0)\n",
    "\n",
    "testset = CIFAR10(root='./cifar-10-python', train=False, download=True, transform=myTransforms)\n",
    "testloader = DataLoader(testset, batch_size=batch_size, shuffle=True, num_workers=0)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecbd8011",
   "metadata": {},
   "source": [
    "# 定义网络，直接调用我自己的就ok"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "801caaef",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1be11657",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 10])\n"
     ]
    }
   ],
   "source": [
    "# 定义自己的网络\n",
    "from coatnet import coatnet_0\n",
    "from coatnet import coatnet_1\n",
    "from coatnet import coatnet_2\n",
    "from coatnet import coatnet_3\n",
    "from coatnet import coatnet_4\n",
    "from coatnet import coatnet_5\n",
    "\n",
    "model = coatnet_3()\n",
    "print(model(torch.randn(1,3,32,32,)).shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d75fb87b",
   "metadata": {},
   "source": [
    "## 模型参数统计"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b4f652d4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "            Conv2d-1        [1024, 192, 16, 16]           5,184\n",
      "       BatchNorm2d-2        [1024, 192, 16, 16]             384\n",
      "              GELU-3        [1024, 192, 16, 16]               0\n",
      "            Conv2d-4        [1024, 192, 16, 16]         331,776\n",
      "       BatchNorm2d-5        [1024, 192, 16, 16]             384\n",
      "              GELU-6        [1024, 192, 16, 16]               0\n",
      "         MaxPool2d-7          [1024, 192, 8, 8]               0\n",
      "            Conv2d-8          [1024, 192, 8, 8]          36,864\n",
      "       BatchNorm2d-9        [1024, 192, 16, 16]             384\n",
      "           Conv2d-10          [1024, 768, 8, 8]         147,456\n",
      "      BatchNorm2d-11          [1024, 768, 8, 8]           1,536\n",
      "             GELU-12          [1024, 768, 8, 8]               0\n",
      "           Conv2d-13          [1024, 768, 8, 8]           6,912\n",
      "      BatchNorm2d-14          [1024, 768, 8, 8]           1,536\n",
      "             GELU-15          [1024, 768, 8, 8]               0\n",
      "AdaptiveAvgPool2d-16          [1024, 768, 1, 1]               0\n",
      "           Linear-17                 [1024, 48]          36,864\n",
      "             GELU-18                 [1024, 48]               0\n",
      "           Linear-19                [1024, 768]          36,864\n",
      "          Sigmoid-20                [1024, 768]               0\n",
      "               SE-21          [1024, 768, 8, 8]               0\n",
      "           Conv2d-22          [1024, 192, 8, 8]         147,456\n",
      "      BatchNorm2d-23          [1024, 192, 8, 8]             384\n",
      "          PreNorm-24          [1024, 192, 8, 8]               0\n",
      "           MBConv-25          [1024, 192, 8, 8]               0\n",
      "      BatchNorm2d-26          [1024, 192, 8, 8]             384\n",
      "           Conv2d-27          [1024, 768, 8, 8]         147,456\n",
      "      BatchNorm2d-28          [1024, 768, 8, 8]           1,536\n",
      "             GELU-29          [1024, 768, 8, 8]               0\n",
      "           Conv2d-30          [1024, 768, 8, 8]           6,912\n",
      "      BatchNorm2d-31          [1024, 768, 8, 8]           1,536\n",
      "             GELU-32          [1024, 768, 8, 8]               0\n",
      "AdaptiveAvgPool2d-33          [1024, 768, 1, 1]               0\n",
      "           Linear-34                 [1024, 48]          36,864\n",
      "             GELU-35                 [1024, 48]               0\n",
      "           Linear-36                [1024, 768]          36,864\n",
      "          Sigmoid-37                [1024, 768]               0\n",
      "               SE-38          [1024, 768, 8, 8]               0\n",
      "           Conv2d-39          [1024, 192, 8, 8]         147,456\n",
      "      BatchNorm2d-40          [1024, 192, 8, 8]             384\n",
      "          PreNorm-41          [1024, 192, 8, 8]               0\n",
      "           MBConv-42          [1024, 192, 8, 8]               0\n",
      "        MaxPool2d-43          [1024, 192, 4, 4]               0\n",
      "           Conv2d-44          [1024, 384, 4, 4]          73,728\n",
      "      BatchNorm2d-45          [1024, 192, 8, 8]             384\n",
      "           Conv2d-46          [1024, 768, 4, 4]         147,456\n",
      "      BatchNorm2d-47          [1024, 768, 4, 4]           1,536\n",
      "             GELU-48          [1024, 768, 4, 4]               0\n",
      "           Conv2d-49          [1024, 768, 4, 4]           6,912\n",
      "      BatchNorm2d-50          [1024, 768, 4, 4]           1,536\n",
      "             GELU-51          [1024, 768, 4, 4]               0\n",
      "AdaptiveAvgPool2d-52          [1024, 768, 1, 1]               0\n",
      "           Linear-53                 [1024, 48]          36,864\n",
      "             GELU-54                 [1024, 48]               0\n",
      "           Linear-55                [1024, 768]          36,864\n",
      "          Sigmoid-56                [1024, 768]               0\n",
      "               SE-57          [1024, 768, 4, 4]               0\n",
      "           Conv2d-58          [1024, 384, 4, 4]         294,912\n",
      "      BatchNorm2d-59          [1024, 384, 4, 4]             768\n",
      "          PreNorm-60          [1024, 384, 4, 4]               0\n",
      "           MBConv-61          [1024, 384, 4, 4]               0\n",
      "      BatchNorm2d-62          [1024, 384, 4, 4]             768\n",
      "           Conv2d-63         [1024, 1536, 4, 4]         589,824\n",
      "      BatchNorm2d-64         [1024, 1536, 4, 4]           3,072\n",
      "             GELU-65         [1024, 1536, 4, 4]               0\n",
      "           Conv2d-66         [1024, 1536, 4, 4]          13,824\n",
      "      BatchNorm2d-67         [1024, 1536, 4, 4]           3,072\n",
      "             GELU-68         [1024, 1536, 4, 4]               0\n",
      "AdaptiveAvgPool2d-69         [1024, 1536, 1, 1]               0\n",
      "           Linear-70                 [1024, 96]         147,456\n",
      "             GELU-71                 [1024, 96]               0\n",
      "           Linear-72               [1024, 1536]         147,456\n",
      "          Sigmoid-73               [1024, 1536]               0\n",
      "               SE-74         [1024, 1536, 4, 4]               0\n",
      "           Conv2d-75          [1024, 384, 4, 4]         589,824\n",
      "      BatchNorm2d-76          [1024, 384, 4, 4]             768\n",
      "          PreNorm-77          [1024, 384, 4, 4]               0\n",
      "           MBConv-78          [1024, 384, 4, 4]               0\n",
      "      BatchNorm2d-79          [1024, 384, 4, 4]             768\n",
      "           Conv2d-80         [1024, 1536, 4, 4]         589,824\n",
      "      BatchNorm2d-81         [1024, 1536, 4, 4]           3,072\n",
      "             GELU-82         [1024, 1536, 4, 4]               0\n",
      "           Conv2d-83         [1024, 1536, 4, 4]          13,824\n",
      "      BatchNorm2d-84         [1024, 1536, 4, 4]           3,072\n",
      "             GELU-85         [1024, 1536, 4, 4]               0\n",
      "AdaptiveAvgPool2d-86         [1024, 1536, 1, 1]               0\n",
      "           Linear-87                 [1024, 96]         147,456\n",
      "             GELU-88                 [1024, 96]               0\n",
      "           Linear-89               [1024, 1536]         147,456\n",
      "          Sigmoid-90               [1024, 1536]               0\n",
      "               SE-91         [1024, 1536, 4, 4]               0\n",
      "           Conv2d-92          [1024, 384, 4, 4]         589,824\n",
      "      BatchNorm2d-93          [1024, 384, 4, 4]             768\n",
      "          PreNorm-94          [1024, 384, 4, 4]               0\n",
      "           MBConv-95          [1024, 384, 4, 4]               0\n",
      "      BatchNorm2d-96          [1024, 384, 4, 4]             768\n",
      "           Conv2d-97         [1024, 1536, 4, 4]         589,824\n",
      "      BatchNorm2d-98         [1024, 1536, 4, 4]           3,072\n",
      "             GELU-99         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-100         [1024, 1536, 4, 4]          13,824\n",
      "     BatchNorm2d-101         [1024, 1536, 4, 4]           3,072\n",
      "            GELU-102         [1024, 1536, 4, 4]               0\n",
      "AdaptiveAvgPool2d-103         [1024, 1536, 1, 1]               0\n",
      "          Linear-104                 [1024, 96]         147,456\n",
      "            GELU-105                 [1024, 96]               0\n",
      "          Linear-106               [1024, 1536]         147,456\n",
      "         Sigmoid-107               [1024, 1536]               0\n",
      "              SE-108         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-109          [1024, 384, 4, 4]         589,824\n",
      "     BatchNorm2d-110          [1024, 384, 4, 4]             768\n",
      "         PreNorm-111          [1024, 384, 4, 4]               0\n",
      "          MBConv-112          [1024, 384, 4, 4]               0\n",
      "     BatchNorm2d-113          [1024, 384, 4, 4]             768\n",
      "          Conv2d-114         [1024, 1536, 4, 4]         589,824\n",
      "     BatchNorm2d-115         [1024, 1536, 4, 4]           3,072\n",
      "            GELU-116         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-117         [1024, 1536, 4, 4]          13,824\n",
      "     BatchNorm2d-118         [1024, 1536, 4, 4]           3,072\n",
      "            GELU-119         [1024, 1536, 4, 4]               0\n",
      "AdaptiveAvgPool2d-120         [1024, 1536, 1, 1]               0\n",
      "          Linear-121                 [1024, 96]         147,456\n",
      "            GELU-122                 [1024, 96]               0\n",
      "          Linear-123               [1024, 1536]         147,456\n",
      "         Sigmoid-124               [1024, 1536]               0\n",
      "              SE-125         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-126          [1024, 384, 4, 4]         589,824\n",
      "     BatchNorm2d-127          [1024, 384, 4, 4]             768\n",
      "         PreNorm-128          [1024, 384, 4, 4]               0\n",
      "          MBConv-129          [1024, 384, 4, 4]               0\n",
      "     BatchNorm2d-130          [1024, 384, 4, 4]             768\n",
      "          Conv2d-131         [1024, 1536, 4, 4]         589,824\n",
      "     BatchNorm2d-132         [1024, 1536, 4, 4]           3,072\n",
      "            GELU-133         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-134         [1024, 1536, 4, 4]          13,824\n",
      "     BatchNorm2d-135         [1024, 1536, 4, 4]           3,072\n",
      "            GELU-136         [1024, 1536, 4, 4]               0\n",
      "AdaptiveAvgPool2d-137         [1024, 1536, 1, 1]               0\n",
      "          Linear-138                 [1024, 96]         147,456\n",
      "            GELU-139                 [1024, 96]               0\n",
      "          Linear-140               [1024, 1536]         147,456\n",
      "         Sigmoid-141               [1024, 1536]               0\n",
      "              SE-142         [1024, 1536, 4, 4]               0\n",
      "          Conv2d-143          [1024, 384, 4, 4]         589,824\n",
      "     BatchNorm2d-144          [1024, 384, 4, 4]             768\n",
      "         PreNorm-145          [1024, 384, 4, 4]               0\n",
      "          MBConv-146          [1024, 384, 4, 4]               0\n",
      "       MaxPool2d-147          [1024, 384, 2, 2]               0\n",
      "          Conv2d-148          [1024, 768, 2, 2]         294,912\n",
      "       MaxPool2d-149          [1024, 384, 2, 2]               0\n",
      "       Rearrange-150             [1024, 4, 384]               0\n",
      "       LayerNorm-151             [1024, 4, 384]             768\n",
      "          Linear-152             [1024, 4, 768]         294,912\n",
      "         Softmax-153            [1024, 8, 4, 4]               0\n",
      "          Linear-154             [1024, 4, 768]         197,376\n",
      "         Dropout-155             [1024, 4, 768]               0\n",
      "       Attention-156             [1024, 4, 768]               0\n",
      "         PreNorm-157             [1024, 4, 768]               0\n",
      "       Rearrange-158          [1024, 768, 2, 2]               0\n",
      "       Rearrange-159             [1024, 4, 768]               0\n",
      "       LayerNorm-160             [1024, 4, 768]           1,536\n",
      "          Linear-161            [1024, 4, 1536]       1,181,184\n",
      "            GELU-162            [1024, 4, 1536]               0\n",
      "         Dropout-163            [1024, 4, 1536]               0\n",
      "          Linear-164             [1024, 4, 768]       1,180,416\n",
      "         Dropout-165             [1024, 4, 768]               0\n",
      "     FeedForward-166             [1024, 4, 768]               0\n",
      "         PreNorm-167             [1024, 4, 768]               0\n",
      "       Rearrange-168          [1024, 768, 2, 2]               0\n",
      "     Transformer-169          [1024, 768, 2, 2]               0\n",
      "       Rearrange-170             [1024, 4, 768]               0\n",
      "       LayerNorm-171             [1024, 4, 768]           1,536\n",
      "          Linear-172             [1024, 4, 768]         589,824\n",
      "         Softmax-173            [1024, 8, 4, 4]               0\n",
      "          Linear-174             [1024, 4, 768]         197,376\n",
      "         Dropout-175             [1024, 4, 768]               0\n",
      "       Attention-176             [1024, 4, 768]               0\n",
      "         PreNorm-177             [1024, 4, 768]               0\n",
      "       Rearrange-178          [1024, 768, 2, 2]               0\n",
      "       Rearrange-179             [1024, 4, 768]               0\n",
      "       LayerNorm-180             [1024, 4, 768]           1,536\n",
      "          Linear-181            [1024, 4, 3072]       2,362,368\n",
      "            GELU-182            [1024, 4, 3072]               0\n",
      "         Dropout-183            [1024, 4, 3072]               0\n",
      "          Linear-184             [1024, 4, 768]       2,360,064\n",
      "         Dropout-185             [1024, 4, 768]               0\n",
      "     FeedForward-186             [1024, 4, 768]               0\n",
      "         PreNorm-187             [1024, 4, 768]               0\n",
      "       Rearrange-188          [1024, 768, 2, 2]               0\n",
      "     Transformer-189          [1024, 768, 2, 2]               0\n",
      "       Rearrange-190             [1024, 4, 768]               0\n",
      "       LayerNorm-191             [1024, 4, 768]           1,536\n",
      "          Linear-192             [1024, 4, 768]         589,824\n",
      "         Softmax-193            [1024, 8, 4, 4]               0\n",
      "          Linear-194             [1024, 4, 768]         197,376\n",
      "         Dropout-195             [1024, 4, 768]               0\n",
      "       Attention-196             [1024, 4, 768]               0\n",
      "         PreNorm-197             [1024, 4, 768]               0\n",
      "       Rearrange-198          [1024, 768, 2, 2]               0\n",
      "       Rearrange-199             [1024, 4, 768]               0\n",
      "       LayerNorm-200             [1024, 4, 768]           1,536\n",
      "          Linear-201            [1024, 4, 3072]       2,362,368\n",
      "            GELU-202            [1024, 4, 3072]               0\n",
      "         Dropout-203            [1024, 4, 3072]               0\n",
      "          Linear-204             [1024, 4, 768]       2,360,064\n",
      "         Dropout-205             [1024, 4, 768]               0\n",
      "     FeedForward-206             [1024, 4, 768]               0\n",
      "         PreNorm-207             [1024, 4, 768]               0\n",
      "       Rearrange-208          [1024, 768, 2, 2]               0\n",
      "     Transformer-209          [1024, 768, 2, 2]               0\n",
      "       Rearrange-210             [1024, 4, 768]               0\n",
      "       LayerNorm-211             [1024, 4, 768]           1,536\n",
      "          Linear-212             [1024, 4, 768]         589,824\n",
      "         Softmax-213            [1024, 8, 4, 4]               0\n",
      "          Linear-214             [1024, 4, 768]         197,376\n",
      "         Dropout-215             [1024, 4, 768]               0\n",
      "       Attention-216             [1024, 4, 768]               0\n",
      "         PreNorm-217             [1024, 4, 768]               0\n",
      "       Rearrange-218          [1024, 768, 2, 2]               0\n",
      "       Rearrange-219             [1024, 4, 768]               0\n",
      "       LayerNorm-220             [1024, 4, 768]           1,536\n",
      "          Linear-221            [1024, 4, 3072]       2,362,368\n",
      "            GELU-222            [1024, 4, 3072]               0\n",
      "         Dropout-223            [1024, 4, 3072]               0\n",
      "          Linear-224             [1024, 4, 768]       2,360,064\n",
      "         Dropout-225             [1024, 4, 768]               0\n",
      "     FeedForward-226             [1024, 4, 768]               0\n",
      "         PreNorm-227             [1024, 4, 768]               0\n",
      "       Rearrange-228          [1024, 768, 2, 2]               0\n",
      "     Transformer-229          [1024, 768, 2, 2]               0\n",
      "       Rearrange-230             [1024, 4, 768]               0\n",
      "       LayerNorm-231             [1024, 4, 768]           1,536\n",
      "          Linear-232             [1024, 4, 768]         589,824\n",
      "         Softmax-233            [1024, 8, 4, 4]               0\n",
      "          Linear-234             [1024, 4, 768]         197,376\n",
      "         Dropout-235             [1024, 4, 768]               0\n",
      "       Attention-236             [1024, 4, 768]               0\n",
      "         PreNorm-237             [1024, 4, 768]               0\n",
      "       Rearrange-238          [1024, 768, 2, 2]               0\n",
      "       Rearrange-239             [1024, 4, 768]               0\n",
      "       LayerNorm-240             [1024, 4, 768]           1,536\n",
      "          Linear-241            [1024, 4, 3072]       2,362,368\n",
      "            GELU-242            [1024, 4, 3072]               0\n",
      "         Dropout-243            [1024, 4, 3072]               0\n",
      "          Linear-244             [1024, 4, 768]       2,360,064\n",
      "         Dropout-245             [1024, 4, 768]               0\n",
      "     FeedForward-246             [1024, 4, 768]               0\n",
      "         PreNorm-247             [1024, 4, 768]               0\n",
      "       Rearrange-248          [1024, 768, 2, 2]               0\n",
      "     Transformer-249          [1024, 768, 2, 2]               0\n",
      "       Rearrange-250             [1024, 4, 768]               0\n",
      "       LayerNorm-251             [1024, 4, 768]           1,536\n",
      "          Linear-252             [1024, 4, 768]         589,824\n",
      "         Softmax-253            [1024, 8, 4, 4]               0\n",
      "          Linear-254             [1024, 4, 768]         197,376\n",
      "         Dropout-255             [1024, 4, 768]               0\n",
      "       Attention-256             [1024, 4, 768]               0\n",
      "         PreNorm-257             [1024, 4, 768]               0\n",
      "       Rearrange-258          [1024, 768, 2, 2]               0\n",
      "       Rearrange-259             [1024, 4, 768]               0\n",
      "       LayerNorm-260             [1024, 4, 768]           1,536\n",
      "          Linear-261            [1024, 4, 3072]       2,362,368\n",
      "            GELU-262            [1024, 4, 3072]               0\n",
      "         Dropout-263            [1024, 4, 3072]               0\n",
      "          Linear-264             [1024, 4, 768]       2,360,064\n",
      "         Dropout-265             [1024, 4, 768]               0\n",
      "     FeedForward-266             [1024, 4, 768]               0\n",
      "         PreNorm-267             [1024, 4, 768]               0\n",
      "       Rearrange-268          [1024, 768, 2, 2]               0\n",
      "     Transformer-269          [1024, 768, 2, 2]               0\n",
      "       Rearrange-270             [1024, 4, 768]               0\n",
      "       LayerNorm-271             [1024, 4, 768]           1,536\n",
      "          Linear-272             [1024, 4, 768]         589,824\n",
      "         Softmax-273            [1024, 8, 4, 4]               0\n",
      "          Linear-274             [1024, 4, 768]         197,376\n",
      "         Dropout-275             [1024, 4, 768]               0\n",
      "       Attention-276             [1024, 4, 768]               0\n",
      "         PreNorm-277             [1024, 4, 768]               0\n",
      "       Rearrange-278          [1024, 768, 2, 2]               0\n",
      "       Rearrange-279             [1024, 4, 768]               0\n",
      "       LayerNorm-280             [1024, 4, 768]           1,536\n",
      "          Linear-281            [1024, 4, 3072]       2,362,368\n",
      "            GELU-282            [1024, 4, 3072]               0\n",
      "         Dropout-283            [1024, 4, 3072]               0\n",
      "          Linear-284             [1024, 4, 768]       2,360,064\n",
      "         Dropout-285             [1024, 4, 768]               0\n",
      "     FeedForward-286             [1024, 4, 768]               0\n",
      "         PreNorm-287             [1024, 4, 768]               0\n",
      "       Rearrange-288          [1024, 768, 2, 2]               0\n",
      "     Transformer-289          [1024, 768, 2, 2]               0\n",
      "       Rearrange-290             [1024, 4, 768]               0\n",
      "       LayerNorm-291             [1024, 4, 768]           1,536\n",
      "          Linear-292             [1024, 4, 768]         589,824\n",
      "         Softmax-293            [1024, 8, 4, 4]               0\n",
      "          Linear-294             [1024, 4, 768]         197,376\n",
      "         Dropout-295             [1024, 4, 768]               0\n",
      "       Attention-296             [1024, 4, 768]               0\n",
      "         PreNorm-297             [1024, 4, 768]               0\n",
      "       Rearrange-298          [1024, 768, 2, 2]               0\n",
      "       Rearrange-299             [1024, 4, 768]               0\n",
      "       LayerNorm-300             [1024, 4, 768]           1,536\n",
      "          Linear-301            [1024, 4, 3072]       2,362,368\n",
      "            GELU-302            [1024, 4, 3072]               0\n",
      "         Dropout-303            [1024, 4, 3072]               0\n",
      "          Linear-304             [1024, 4, 768]       2,360,064\n",
      "         Dropout-305             [1024, 4, 768]               0\n",
      "     FeedForward-306             [1024, 4, 768]               0\n",
      "         PreNorm-307             [1024, 4, 768]               0\n",
      "       Rearrange-308          [1024, 768, 2, 2]               0\n",
      "     Transformer-309          [1024, 768, 2, 2]               0\n",
      "       Rearrange-310             [1024, 4, 768]               0\n",
      "       LayerNorm-311             [1024, 4, 768]           1,536\n",
      "          Linear-312             [1024, 4, 768]         589,824\n",
      "         Softmax-313            [1024, 8, 4, 4]               0\n",
      "          Linear-314             [1024, 4, 768]         197,376\n",
      "         Dropout-315             [1024, 4, 768]               0\n",
      "       Attention-316             [1024, 4, 768]               0\n",
      "         PreNorm-317             [1024, 4, 768]               0\n",
      "       Rearrange-318          [1024, 768, 2, 2]               0\n",
      "       Rearrange-319             [1024, 4, 768]               0\n",
      "       LayerNorm-320             [1024, 4, 768]           1,536\n",
      "          Linear-321            [1024, 4, 3072]       2,362,368\n",
      "            GELU-322            [1024, 4, 3072]               0\n",
      "         Dropout-323            [1024, 4, 3072]               0\n",
      "          Linear-324             [1024, 4, 768]       2,360,064\n",
      "         Dropout-325             [1024, 4, 768]               0\n",
      "     FeedForward-326             [1024, 4, 768]               0\n",
      "         PreNorm-327             [1024, 4, 768]               0\n",
      "       Rearrange-328          [1024, 768, 2, 2]               0\n",
      "     Transformer-329          [1024, 768, 2, 2]               0\n",
      "       Rearrange-330             [1024, 4, 768]               0\n",
      "       LayerNorm-331             [1024, 4, 768]           1,536\n",
      "          Linear-332             [1024, 4, 768]         589,824\n",
      "         Softmax-333            [1024, 8, 4, 4]               0\n",
      "          Linear-334             [1024, 4, 768]         197,376\n",
      "         Dropout-335             [1024, 4, 768]               0\n",
      "       Attention-336             [1024, 4, 768]               0\n",
      "         PreNorm-337             [1024, 4, 768]               0\n",
      "       Rearrange-338          [1024, 768, 2, 2]               0\n",
      "       Rearrange-339             [1024, 4, 768]               0\n",
      "       LayerNorm-340             [1024, 4, 768]           1,536\n",
      "          Linear-341            [1024, 4, 3072]       2,362,368\n",
      "            GELU-342            [1024, 4, 3072]               0\n",
      "         Dropout-343            [1024, 4, 3072]               0\n",
      "          Linear-344             [1024, 4, 768]       2,360,064\n",
      "         Dropout-345             [1024, 4, 768]               0\n",
      "     FeedForward-346             [1024, 4, 768]               0\n",
      "         PreNorm-347             [1024, 4, 768]               0\n",
      "       Rearrange-348          [1024, 768, 2, 2]               0\n",
      "     Transformer-349          [1024, 768, 2, 2]               0\n",
      "       Rearrange-350             [1024, 4, 768]               0\n",
      "       LayerNorm-351             [1024, 4, 768]           1,536\n",
      "          Linear-352             [1024, 4, 768]         589,824\n",
      "         Softmax-353            [1024, 8, 4, 4]               0\n",
      "          Linear-354             [1024, 4, 768]         197,376\n",
      "         Dropout-355             [1024, 4, 768]               0\n",
      "       Attention-356             [1024, 4, 768]               0\n",
      "         PreNorm-357             [1024, 4, 768]               0\n",
      "       Rearrange-358          [1024, 768, 2, 2]               0\n",
      "       Rearrange-359             [1024, 4, 768]               0\n",
      "       LayerNorm-360             [1024, 4, 768]           1,536\n",
      "          Linear-361            [1024, 4, 3072]       2,362,368\n",
      "            GELU-362            [1024, 4, 3072]               0\n",
      "         Dropout-363            [1024, 4, 3072]               0\n",
      "          Linear-364             [1024, 4, 768]       2,360,064\n",
      "         Dropout-365             [1024, 4, 768]               0\n",
      "     FeedForward-366             [1024, 4, 768]               0\n",
      "         PreNorm-367             [1024, 4, 768]               0\n",
      "       Rearrange-368          [1024, 768, 2, 2]               0\n",
      "     Transformer-369          [1024, 768, 2, 2]               0\n",
      "       Rearrange-370             [1024, 4, 768]               0\n",
      "       LayerNorm-371             [1024, 4, 768]           1,536\n",
      "          Linear-372             [1024, 4, 768]         589,824\n",
      "         Softmax-373            [1024, 8, 4, 4]               0\n",
      "          Linear-374             [1024, 4, 768]         197,376\n",
      "         Dropout-375             [1024, 4, 768]               0\n",
      "       Attention-376             [1024, 4, 768]               0\n",
      "         PreNorm-377             [1024, 4, 768]               0\n",
      "       Rearrange-378          [1024, 768, 2, 2]               0\n",
      "       Rearrange-379             [1024, 4, 768]               0\n",
      "       LayerNorm-380             [1024, 4, 768]           1,536\n",
      "          Linear-381            [1024, 4, 3072]       2,362,368\n",
      "            GELU-382            [1024, 4, 3072]               0\n",
      "         Dropout-383            [1024, 4, 3072]               0\n",
      "          Linear-384             [1024, 4, 768]       2,360,064\n",
      "         Dropout-385             [1024, 4, 768]               0\n",
      "     FeedForward-386             [1024, 4, 768]               0\n",
      "         PreNorm-387             [1024, 4, 768]               0\n",
      "       Rearrange-388          [1024, 768, 2, 2]               0\n",
      "     Transformer-389          [1024, 768, 2, 2]               0\n",
      "       Rearrange-390             [1024, 4, 768]               0\n",
      "       LayerNorm-391             [1024, 4, 768]           1,536\n",
      "          Linear-392             [1024, 4, 768]         589,824\n",
      "         Softmax-393            [1024, 8, 4, 4]               0\n",
      "          Linear-394             [1024, 4, 768]         197,376\n",
      "         Dropout-395             [1024, 4, 768]               0\n",
      "       Attention-396             [1024, 4, 768]               0\n",
      "         PreNorm-397             [1024, 4, 768]               0\n",
      "       Rearrange-398          [1024, 768, 2, 2]               0\n",
      "       Rearrange-399             [1024, 4, 768]               0\n",
      "       LayerNorm-400             [1024, 4, 768]           1,536\n",
      "          Linear-401            [1024, 4, 3072]       2,362,368\n",
      "            GELU-402            [1024, 4, 3072]               0\n",
      "         Dropout-403            [1024, 4, 3072]               0\n",
      "          Linear-404             [1024, 4, 768]       2,360,064\n",
      "         Dropout-405             [1024, 4, 768]               0\n",
      "     FeedForward-406             [1024, 4, 768]               0\n",
      "         PreNorm-407             [1024, 4, 768]               0\n",
      "       Rearrange-408          [1024, 768, 2, 2]               0\n",
      "     Transformer-409          [1024, 768, 2, 2]               0\n",
      "       Rearrange-410             [1024, 4, 768]               0\n",
      "       LayerNorm-411             [1024, 4, 768]           1,536\n",
      "          Linear-412             [1024, 4, 768]         589,824\n",
      "         Softmax-413            [1024, 8, 4, 4]               0\n",
      "          Linear-414             [1024, 4, 768]         197,376\n",
      "         Dropout-415             [1024, 4, 768]               0\n",
      "       Attention-416             [1024, 4, 768]               0\n",
      "         PreNorm-417             [1024, 4, 768]               0\n",
      "       Rearrange-418          [1024, 768, 2, 2]               0\n",
      "       Rearrange-419             [1024, 4, 768]               0\n",
      "       LayerNorm-420             [1024, 4, 768]           1,536\n",
      "          Linear-421            [1024, 4, 3072]       2,362,368\n",
      "            GELU-422            [1024, 4, 3072]               0\n",
      "         Dropout-423            [1024, 4, 3072]               0\n",
      "          Linear-424             [1024, 4, 768]       2,360,064\n",
      "         Dropout-425             [1024, 4, 768]               0\n",
      "     FeedForward-426             [1024, 4, 768]               0\n",
      "         PreNorm-427             [1024, 4, 768]               0\n",
      "       Rearrange-428          [1024, 768, 2, 2]               0\n",
      "     Transformer-429          [1024, 768, 2, 2]               0\n",
      "       MaxPool2d-430          [1024, 768, 1, 1]               0\n",
      "          Conv2d-431         [1024, 1536, 1, 1]       1,179,648\n",
      "       MaxPool2d-432          [1024, 768, 1, 1]               0\n",
      "       Rearrange-433             [1024, 1, 768]               0\n",
      "       LayerNorm-434             [1024, 1, 768]           1,536\n",
      "          Linear-435             [1024, 1, 768]         589,824\n",
      "         Softmax-436            [1024, 8, 1, 1]               0\n",
      "          Linear-437            [1024, 1, 1536]         394,752\n",
      "         Dropout-438            [1024, 1, 1536]               0\n",
      "       Attention-439            [1024, 1, 1536]               0\n",
      "         PreNorm-440            [1024, 1, 1536]               0\n",
      "       Rearrange-441         [1024, 1536, 1, 1]               0\n",
      "       Rearrange-442            [1024, 1, 1536]               0\n",
      "       LayerNorm-443            [1024, 1, 1536]           3,072\n",
      "          Linear-444            [1024, 1, 3072]       4,721,664\n",
      "            GELU-445            [1024, 1, 3072]               0\n",
      "         Dropout-446            [1024, 1, 3072]               0\n",
      "          Linear-447            [1024, 1, 1536]       4,720,128\n",
      "         Dropout-448            [1024, 1, 1536]               0\n",
      "     FeedForward-449            [1024, 1, 1536]               0\n",
      "         PreNorm-450            [1024, 1, 1536]               0\n",
      "       Rearrange-451         [1024, 1536, 1, 1]               0\n",
      "     Transformer-452         [1024, 1536, 1, 1]               0\n",
      "       Rearrange-453            [1024, 1, 1536]               0\n",
      "       LayerNorm-454            [1024, 1, 1536]           3,072\n",
      "          Linear-455             [1024, 1, 768]       1,179,648\n",
      "         Softmax-456            [1024, 8, 1, 1]               0\n",
      "          Linear-457            [1024, 1, 1536]         394,752\n",
      "         Dropout-458            [1024, 1, 1536]               0\n",
      "       Attention-459            [1024, 1, 1536]               0\n",
      "         PreNorm-460            [1024, 1, 1536]               0\n",
      "       Rearrange-461         [1024, 1536, 1, 1]               0\n",
      "       Rearrange-462            [1024, 1, 1536]               0\n",
      "       LayerNorm-463            [1024, 1, 1536]           3,072\n",
      "          Linear-464            [1024, 1, 6144]       9,443,328\n",
      "            GELU-465            [1024, 1, 6144]               0\n",
      "         Dropout-466            [1024, 1, 6144]               0\n",
      "          Linear-467            [1024, 1, 1536]       9,438,720\n",
      "         Dropout-468            [1024, 1, 1536]               0\n",
      "     FeedForward-469            [1024, 1, 1536]               0\n",
      "         PreNorm-470            [1024, 1, 1536]               0\n",
      "       Rearrange-471         [1024, 1536, 1, 1]               0\n",
      "     Transformer-472         [1024, 1536, 1, 1]               0\n",
      "       AvgPool2d-473         [1024, 1536, 1, 1]               0\n",
      "          Linear-474               [1024, 1000]       1,536,000\n",
      "          Linear-475                 [1024, 10]          10,000\n",
      "================================================================\n",
      "Total params: 117,650,128\n",
      "Trainable params: 117,650,128\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 12.00\n",
      "Forward/backward pass size (MB): 28237.77\n",
      "Params size (MB): 448.80\n",
      "Estimated Total Size (MB): 28698.57\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "from torchsummary import summary\n",
    "summary(model.to(device), input_size=(3, 32, 32), batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7532a003",
   "metadata": {},
   "source": [
    "## 训练网络"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2f8c4b35",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score\n",
    "from datetime import datetime\n",
    "\n",
    "\n",
    "def fwd_pass(net, loader, loss_function, optimizer, train=False):\n",
    "    if train: net.train()\n",
    "    else:     net.eval()\n",
    "\n",
    "    total_loss = 0\n",
    "    y_pred, y = [], []\n",
    "    for X, labels in loader: #加载\n",
    "        X, labels = X.to(device), labels.to(device)\n",
    "        # ===================前向传播=====================\n",
    "        outputs = model(X)\n",
    "        loss = loss_function(outputs, labels) \n",
    "        total_loss += loss.item()\n",
    "        # ===================反向传播====================\n",
    "        if train:\n",
    "            optimizer.zero_grad()# 归零\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "        # ===============预测===============\n",
    "        with torch.no_grad():\n",
    "            y_hat = outputs.argmax(dim=1)\n",
    "            y_pred.extend(y_hat.cpu().numpy().tolist())\n",
    "            y.extend(labels.cpu().numpy().tolist())\n",
    "\n",
    "    # =================返回=====================\n",
    "    with torch.no_grad():\n",
    "        total_loss /= len(loader.dataset)\n",
    "        acc = accuracy_score(y, y_pred)*100\n",
    "        f1 = f1_score(y, y_pred, average=\"weighted\")*100\n",
    "\n",
    "    return total_loss, acc, f1\n",
    "\n",
    "\n",
    "logs = dict( trainlosses=list(), testlosses=list(), \n",
    "            trainaccs=list(), testaccs=list() )\n",
    "    \n",
    "    \n",
    "#TRAINING THE NETWORK\n",
    "def train(model, train_loader, test_loader,\n",
    "          optimizer, loss_function, epoch_num=1, device=device):\n",
    "    model.to(device)\n",
    "    torch.manual_seed(14)\n",
    "    \n",
    "    start = time.time()\n",
    "    print(f\"Training Process Starts at {datetime.now().strftime('%H:%M:%S')} ... \")\n",
    "    \n",
    "    epoch = 1\n",
    "\n",
    "    # loop over the dataset multiple times\n",
    "    while epoch<=epoch_num:\n",
    "        # Training\n",
    "        train_loss, train_acc, train_f1 = fwd_pass(model, train_loader, loss_function, \n",
    "                                                       optimizer, train=True)\n",
    "            \n",
    "        # Testing\n",
    "        with torch.no_grad():\n",
    "            test_loss, test_acc, test_f1 = fwd_pass(model, test_loader, loss_function, \n",
    "                                                    optimizer, train=False)\n",
    "        \n",
    "        \n",
    "        reset='\\n' if epoch <= 5 or epoch%5==0 else '\\r'\n",
    "        print(f\"Epoch[{epoch:2d}]>>>\",\n",
    "            f\"Train/Test loss: {train_loss:.7f}/{test_loss:.7f},\",\n",
    "            f\"Acc.: {train_acc:.2f}/{test_acc:.2f}\",\n",
    "            f\"[[{timeSince(start)}]]\", end=reset)\n",
    "            \n",
    "        # ===================log========================\n",
    "        logs['trainlosses'].append(train_loss)\n",
    "        logs['trainaccs'].append(train_acc)\n",
    "        logs['testlosses'].append(test_loss)\n",
    "        logs['testaccs'].append(test_acc)\n",
    "                \n",
    "        epoch += 1\n",
    "\n",
    "    print(f'测试集F1: {test_f1}')\n",
    "    print(f'Finished in {timeSince(start)}')\n",
    "    return logs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "fd4e34b8",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Process Starts at 13:32:45 ... \n",
      "Epoch[ 1]>>> Train/Test loss: 0.0190010/0.0047902, Acc.: 15.29/19.49 [[ 0m 25s]]\n",
      "Epoch[ 2]>>> Train/Test loss: 0.0020776/0.0016095, Acc.: 31.46/41.30 [[ 0m 51s]]\n",
      "Epoch[ 3]>>> Train/Test loss: 0.0014769/0.0013909, Acc.: 44.35/49.05 [[ 1m 17s]]\n",
      "Epoch[ 4]>>> Train/Test loss: 0.0012931/0.0012708, Acc.: 51.73/53.27 [[ 1m 43s]]\n",
      "Epoch[ 5]>>> Train/Test loss: 0.0011182/0.0011487, Acc.: 58.75/58.54 [[ 2m 08s]]\n",
      "Epoch[10]>>> Train/Test loss: 0.0003639/0.0011040, Acc.: 87.27/68.62 [[ 4m 17s]]\n",
      "Epoch[15]>>> Train/Test loss: 0.0001497/0.0010627, Acc.: 95.14/73.40 [[ 6m 26s]]\n",
      "Epoch[20]>>> Train/Test loss: 0.0000992/0.0010881, Acc.: 96.93/74.10 [[ 8m 35s]]\n",
      "Epoch[25]>>> Train/Test loss: 0.0000733/0.0010397, Acc.: 97.65/76.37 [[10m 51s]]\n",
      "Epoch[30]>>> Train/Test loss: 0.0000613/0.0010711, Acc.: 98.13/76.54 [[13m 06s]]\n",
      "Epoch[35]>>> Train/Test loss: 0.0000595/0.0010494, Acc.: 98.09/77.26 [[15m 15s]]\n",
      "Epoch[40]>>> Train/Test loss: 0.0000546/0.0010395, Acc.: 98.26/77.14 [[17m 36s]]\n",
      "Epoch[45]>>> Train/Test loss: 0.0000461/0.0010607, Acc.: 98.55/78.56 [[19m 50s]]\n",
      "Epoch[50]>>> Train/Test loss: 0.0000403/0.0011473, Acc.: 98.70/77.67 [[22m 53s]]\n",
      "Epoch[55]>>> Train/Test loss: 0.0000351/0.0010466, Acc.: 98.87/78.62 [[25m 54s]]\n",
      "Epoch[60]>>> Train/Test loss: 0.0000356/0.0010498, Acc.: 98.84/78.96 [[29m 08s]]\n",
      "测试集F1: 78.92454089951684\n",
      "Finished in 29m 08s\n"
     ]
    }
   ],
   "source": [
    "model = coatnet_3()\n",
    "\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "learning_rate = 1e-3\n",
    "optimizer = optim.Adam( model.parameters(), lr=learning_rate, weight_decay=1e-5,)\n",
    "\n",
    "logs = train(model, train_loader=trainloader, test_loader=testloader,\n",
    "             optimizer=optimizer, loss_function=criterion, \n",
    "             epoch_num=60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "27acbe7f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Process Starts at 14:01:55 ... \n",
      "Epoch[ 1]>>> Train/Test loss: 0.0000271/0.0010124, Acc.: 99.14/79.54 [[ 0m 38s]]\n",
      "Epoch[ 2]>>> Train/Test loss: 0.0000270/0.0010116, Acc.: 99.09/79.54 [[ 1m 18s]]\n",
      "Epoch[ 3]>>> Train/Test loss: 0.0000261/0.0010126, Acc.: 99.15/79.52 [[ 1m 56s]]\n",
      "Epoch[ 4]>>> Train/Test loss: 0.0000284/0.0010129, Acc.: 99.07/79.41 [[ 2m 35s]]\n",
      "Epoch[ 5]>>> Train/Test loss: 0.0000276/0.0010119, Acc.: 99.07/79.52 [[ 3m 14s]]\n",
      "Epoch[10]>>> Train/Test loss: 0.0000261/0.0010131, Acc.: 99.13/79.46 [[ 6m 28s]]\n",
      "Epoch[15]>>> Train/Test loss: 0.0000284/0.0010120, Acc.: 99.06/79.51 [[ 9m 36s]]\n",
      "Epoch[20]>>> Train/Test loss: 0.0000263/0.0010135, Acc.: 99.15/79.45 [[12m 51s]]\n",
      "Epoch[25]>>> Train/Test loss: 0.0000269/0.0010125, Acc.: 99.09/79.40 [[15m 57s]]\n",
      "Epoch[30]>>> Train/Test loss: 0.0000250/0.0010124, Acc.: 99.18/79.50 [[18m 56s]]\n",
      "测试集F1: 79.54241748422825\n",
      "Finished in 18m 56s\n"
     ]
    }
   ],
   "source": [
    "model_name = f'cifar10.h5'\n",
    "torch.save(model, model_name)\n",
    "\n",
    "#用保存了的模型继续跑\n",
    "model = torch.load('./cifar10.h5')\n",
    "learning_rate = 1e-4\n",
    "logs = train(model, train_loader=trainloader, test_loader=testloader,\n",
    "             optimizer=optimizer, loss_function=criterion, \n",
    "             epoch_num=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "01a6a822",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Process Starts at 14:20:53 ... \n",
      "Epoch[ 1]>>> Train/Test loss: 0.0000271/0.0010123, Acc.: 99.14/79.55 [[ 0m 26s]]\n",
      "Epoch[ 2]>>> Train/Test loss: 0.0000270/0.0010116, Acc.: 99.09/79.54 [[ 0m 58s]]\n",
      "Epoch[ 3]>>> Train/Test loss: 0.0000261/0.0010126, Acc.: 99.15/79.52 [[ 1m 37s]]\n",
      "Epoch[ 4]>>> Train/Test loss: 0.0000284/0.0010129, Acc.: 99.07/79.41 [[ 2m 11s]]\n",
      "Epoch[ 5]>>> Train/Test loss: 0.0000276/0.0010119, Acc.: 99.07/79.52 [[ 2m 40s]]\n",
      "Epoch[10]>>> Train/Test loss: 0.0000261/0.0010131, Acc.: 99.13/79.46 [[ 4m 53s]]\n",
      "测试集F1: 79.4997832955029\n",
      "Finished in  4m 53s\n"
     ]
    }
   ],
   "source": [
    "model_name = f'cifar10.h5'\n",
    "torch.save(model, model_name)\n",
    "\n",
    "#用保存了的模型继续跑\n",
    "model = torch.load('./cifar10.h5')\n",
    "learning_rate = 1e-5\n",
    "logs = train(model, train_loader=trainloader, test_loader=testloader,\n",
    "             optimizer=optimizer, loss_function=criterion, \n",
    "             epoch_num=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "db4928c4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1700x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1700x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def res_plot(plotdata, xlabel='epoch', legend=['Train','Test'], title='Loss'):\n",
    "    fig, axes = plt.subplots(1,2, figsize = (17,6))\n",
    "    axes[0].set_ylabel(title)\n",
    "    epochs = len(plotdata[0])-1\n",
    "    for index, ax in enumerate(axes):\n",
    "        n = 10 if index==1 else 0\n",
    "        x = range(epochs-n+1, epochs+1) if index==1 else range(epochs+1)\n",
    "        title = f\"{title} ({n} last values)\" if index==1 else title\n",
    "#         ax.set_xlabel(xlabel)\n",
    "        ax.set_title(title)\n",
    "        ax.plot(x, plotdata[0][-n:], '-o', label=legend[0])\n",
    "        ax.plot(x, plotdata[1][-n:], '-o', label=legend[1])\n",
    "        p = 7 if 'loss'in title.lower() else 2\n",
    "        if index==1:\n",
    "            for i,j in list(zip(x,plotdata[1][-n:]))+list(zip(x, plotdata[0][-n:])):\n",
    "                ax.annotate(f\"{j:.{p}f}\", xy=(i,j), rotation=90, va='top')\n",
    "        ax.annotate(f\"{plotdata[0][-1]:.{p}f}\", xy=(len(plotdata[0])-1,plotdata[0][-1]),\n",
    "                   ha=\"right\")\n",
    "        ax.annotate(f\"{plotdata[1][-1]:.{p}f}\", xy=(len(plotdata[1])-1,plotdata[1][-1]),\n",
    "                   ha=\"right\")\n",
    "        ax.legend()\n",
    "\n",
    "    plt.show()\n",
    "    \n",
    "res_plot((logs['trainlosses'], logs['testlosses']), title='Loss')\n",
    "res_plot((logs['trainaccs'], logs['testaccs']), title='Accuracy' )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71e4de9c",
   "metadata": {},
   "source": [
    "## 每个类的精度"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "22221752",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acuracy of plane: 82%\n",
      "Acuracy of car  : 90%\n",
      "Acuracy of bird : 75%\n",
      "Acuracy of cat  : 64%\n",
      "Acuracy of deer : 83%\n",
      "Acuracy of dog  : 65%\n",
      "Acuracy of frog : 82%\n",
      "Acuracy of horse: 81%\n",
      "Acuracy of ship : 88%\n",
      "Acuracy of truck: 85%\n"
     ]
    }
   ],
   "source": [
    "classes = ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck')\n",
    "\n",
    "class_correct = [0.] * 10\n",
    "class_total = [0.] * 10 \n",
    "y_test, y_pred = [] , []\n",
    "X_test = []\n",
    "\n",
    "with torch.no_grad():\n",
    "    for images, labels in testloader:\n",
    "        X_test.extend([_ for _ in images])\n",
    "        outputs = model(images.to(device))\n",
    "        _, predicted = torch.max(outputs, 1)\n",
    "        predicted = predicted.cpu()\n",
    "        c = (predicted == labels).squeeze()\n",
    "        for i, label in enumerate(labels):\n",
    "            class_correct[label] += c[i].item()\n",
    "            class_total[label] += 1\n",
    "        y_pred.extend(predicted.numpy())\n",
    "        y_test.extend(labels.cpu().numpy())\n",
    "\n",
    "for i in range(10):\n",
    "    print(f\"Acuracy of {classes[i]:5s}: {100*class_correct[i]/class_total[i]:2.0f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "ee65db06",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy is : 0.7946\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "       plane       0.84      0.82      0.83      1000\n",
      "         car       0.89      0.90      0.90      1000\n",
      "        bird       0.71      0.75      0.73      1000\n",
      "         cat       0.60      0.64      0.62      1000\n",
      "        deer       0.73      0.83      0.77      1000\n",
      "         dog       0.76      0.65      0.70      1000\n",
      "        frog       0.83      0.82      0.82      1000\n",
      "       horse       0.86      0.81      0.83      1000\n",
      "        ship       0.88      0.88      0.88      1000\n",
      "       truck       0.88      0.85      0.86      1000\n",
      "\n",
      "    accuracy                           0.79     10000\n",
      "   macro avg       0.80      0.79      0.79     10000\n",
      "weighted avg       0.80      0.79      0.79     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "\n",
    "ac = accuracy_score(y_test, y_pred)\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "cr = classification_report(y_test, y_pred, target_names=classes)\n",
    "print(\"Accuracy is :\",ac) \n",
    "print(cr)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a48e7f2",
   "metadata": {},
   "source": [
    "### 混淆矩阵"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "207465ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 700x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 混淆矩阵\n",
    "\n",
    "import seaborn as sns, pandas as pd\n",
    "\n",
    "labels = pd.DataFrame(cm).applymap(lambda v: f\"{v}\" if v!=0 else f\"\")\n",
    "plt.figure(figsize=(7,5))\n",
    "sns.heatmap(cm, annot=labels, fmt='s', xticklabels=classes, yticklabels=classes, linewidths=0.1 )\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3830dd0f",
   "metadata": {},
   "source": [
    "## 保存模型的参数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c514692f",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_name = f'cifar10[1-82%].h5'\n",
    "torch.save(model, model_name)\n",
    "!cp {model_name} cifar10.h5\n",
    "#!ls -sh cifar10.h5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "31ee91a9",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd3e85fe",
   "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": "3.8.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}