Resnet18.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"patience5-PlateauLR-Resnet18-TinyImagenet.ipynb","provenance":[],"collapsed_sections":[],"mount_file_id":"15WsmW5WAkqHyHn1sFTueELgDEr8aF-fy","authorship_tag":"ABX9TyNc5g4lgiA8TV/+ZC5zYLjs"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"markdown","metadata":{"id":"3L8_c-rYCeJN"},"source":["# Load Tiny ImageNet"]},{"cell_type":"code","metadata":{"id":"Oie-kYHLCbGj","executionInfo":{"status":"ok","timestamp":1628779251552,"user_tz":240,"elapsed":188,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["#use sys.path.append to call resnet model"],"execution_count":1,"outputs":[]},{"cell_type":"code","metadata":{"id":"IQOp2_zfDhZP","executionInfo":{"status":"ok","timestamp":1628779254252,"user_tz":240,"elapsed":2555,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["import torch\n","import torch.nn as nn\n","import torch.optim as optim\n","import torch.nn.functional as F\n","import torchvision\n","import torchvision.transforms as transforms\n","from torch.utils.data import DataLoader\n","from torch.utils.data import Dataset\n","import torchvision.datasets as datasets\n","import matplotlib.pyplot as plt\n","import os, glob\n","from PIL import Image"],"execution_count":2,"outputs":[]},{"cell_type":"code","metadata":{"id":"d2yREX6yNTX9","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1628779298968,"user_tz":240,"elapsed":44718,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}},"outputId":"e2ca3030-8774-4f13-c077-bde420581993"},"source":["if not os.path.exists('/content/tiny-imagenet-200/'):\n","    !wget http://cs231n.stanford.edu/tiny-imagenet-200.zip\n","    !unzip -q tiny-imagenet-200.zip\n","\n","batch_size = 128\n","\n","id_dict = {}\n","for i, line in enumerate(open('tiny-imagenet-200/wnids.txt', 'r')):\n","  id_dict[line.replace('\\n', '')] = i"],"execution_count":3,"outputs":[{"output_type":"stream","text":["--2021-08-12 14:40:54--  http://cs231n.stanford.edu/tiny-imagenet-200.zip\n","Resolving cs231n.stanford.edu (cs231n.stanford.edu)... 171.64.68.10\n","Connecting to cs231n.stanford.edu (cs231n.stanford.edu)|171.64.68.10|:80... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 248100043 (237M) [application/zip]\n","Saving to: ‘tiny-imagenet-200.zip’\n","\n","tiny-imagenet-200.z 100%[===================>] 236.61M  7.19MB/s    in 36s     \n","\n","2021-08-12 14:41:30 (6.51 MB/s) - ‘tiny-imagenet-200.zip’ saved [248100043/248100043]\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"PwRVG5YxEH24","executionInfo":{"status":"ok","timestamp":1628779299382,"user_tz":240,"elapsed":416,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["class TrainTinyImageNetDataset(Dataset):\n","    def __init__(self, id, transform=None):\n","        self.filenames = glob.glob(\"tiny-imagenet-200/train/*/*/*.JPEG\")\n","        self.transform = transform\n","        self.id_dict = id\n","\n","    def __len__(self):\n","        return len(self.filenames)\n","\n","    def __getitem__(self, idx):\n","        img_path = self.filenames[idx]\n","        image = Image.open(img_path).convert('RGB')\n","        label = self.id_dict[img_path.split('/')[2]]\n","        if self.transform:\n","            image = self.transform(image)\n","        return image, label\n","\n","class TestTinyImageNetDataset(Dataset):\n","    def __init__(self, id, transform=None):\n","        self.filenames = glob.glob(\"tiny-imagenet-200/val/images/*.JPEG\")\n","        self.transform = transform\n","        self.id_dict = id\n","        self.cls_dic = {}\n","        for i, line in enumerate(open('tiny-imagenet-200/val/val_annotations.txt', 'r')):\n","            a = line.split('\\t')\n","            img, cls_id = a[0],a[1]\n","            self.cls_dic[img] = self.id_dict[cls_id]\n","\n","    def __len__(self):\n","        return len(self.filenames)\n","\n","    def __getitem__(self, idx):\n","        img_path = self.filenames[idx]\n","        image = Image.open(img_path).convert('RGB')\n","        label = self.cls_dic[img_path.split('/')[-1]]\n","        if self.transform:\n","            image = self.transform(image)\n","        return image, label\n","\n","\n","transform_train = transforms.Compose([\n","    transforms.RandomAffine(degrees=30, translate=(0.05, 0.05), scale=(1.0, 1.2)),\n","    transforms.RandomHorizontalFlip(0.5),\n","    transforms.ToTensor(),\n","    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n","])\n","\n","transform_test = transforms.Compose([\n","    transforms.ToTensor(),\n","    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n","])\n","\n","trainset = TrainTinyImageNetDataset(id=id_dict, transform=transform_train)\n","trainloader = torch.utils.data.DataLoader(trainset, batch_size=batch_size, shuffle=True)\n","\n","testset = TestTinyImageNetDataset(id=id_dict, transform=transform_test)\n","testloader = torch.utils.data.DataLoader(testset, batch_size=batch_size, shuffle=False)"],"execution_count":4,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"Xj7hxqiH5TbJ"},"source":["#Training ResNet-18\n","\n","Source - https://github.com/kuangliu/pytorch-cifar"]},{"cell_type":"code","metadata":{"id":"qLYAZde55Xtz","executionInfo":{"status":"ok","timestamp":1628779299383,"user_tz":240,"elapsed":7,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["#from vonenet import *\n","\n","# parser = argparse.ArgumentParser(description='PyTorch CIFAR10 Training')\n","# parser.add_argument('--lr', default=0.1, type=float, help='learning rate')\n","# parser.add_argument('--resume', '-r', action='store_true',\n","#                     help='resume from checkpoint')\n","# args = parser.parse_args()"],"execution_count":5,"outputs":[]},{"cell_type":"code","metadata":{"id":"FAzi0Bypj9xn","executionInfo":{"status":"ok","timestamp":1628779299383,"user_tz":240,"elapsed":6,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["# Device configuration\n","device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')"],"execution_count":6,"outputs":[]},{"cell_type":"code","metadata":{"id":"rVw5Ba2J52jM","executionInfo":{"status":"ok","timestamp":1628779299384,"user_tz":240,"elapsed":6,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["best_acc = 0  # best test accuracy\n","start_epoch = 0  # start from epoch 0 or last checkpoint epoch"],"execution_count":7,"outputs":[]},{"cell_type":"code","metadata":{"id":"5kkqve-A6LPi","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1628779306243,"user_tz":240,"elapsed":6864,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}},"outputId":"53eb5140-bb24-412c-b0b0-57d70ef38618"},"source":["# Model\n","print('==> Building model..')\n","import torchvision.models as models\n","net = models.resnet18()\n","net = net.to(device)\n","\n","# if device == 'cuda':\n","#     net = torch.nn.DataParallel(net)\n","#     cudnn.benchmark = True\n","\n","# if args.resume:\n","#     # Load checkpoint.\n","#     print('==> Resuming from checkpoint..')\n","#     assert os.path.isdir('checkpoint'), 'Error: no checkpoint directory found!'\n","#     checkpoint = torch.load('./checkpoint/ckpt.pth')\n","#     net.load_state_dict(checkpoint['net'])\n","#     best_acc = checkpoint['acc']\n","#     start_epoch = checkpoint['epoch']"],"execution_count":8,"outputs":[{"output_type":"stream","text":["==> Building model..\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"8hP6z2jX6aWy","executionInfo":{"status":"ok","timestamp":1628779306244,"user_tz":240,"elapsed":7,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["criterion = nn.CrossEntropyLoss()\n","# replace 0.01 as argument for lr with args.lr if using argparse\n","optimizer = optim.SGD(net.parameters(), lr=0.1,\n","                      momentum=0.9, weight_decay=5e-4)\n","\n","scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, 'min', patience = 5)"],"execution_count":9,"outputs":[]},{"cell_type":"code","metadata":{"id":"9rbnRz056hpp","executionInfo":{"status":"ok","timestamp":1628779306245,"user_tz":240,"elapsed":6,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["# Training\n","train_loss_vec = []\n","train_acc_vec = []\n","def train(epoch):\n","    print('\\nEpoch: %d' % epoch)\n","    net.train()\n","    train_loss = 0\n","    correct = 0\n","    total = 0\n","    for batch_idx, (inputs, targets) in enumerate(trainloader):\n","        inputs, targets = inputs.to(device), targets.to(device)\n","        optimizer.zero_grad()\n","        outputs = net(inputs)\n","        loss = criterion(outputs, targets)\n","        loss.backward()\n","        optimizer.step()\n","        \n","        train_loss += loss.item()\n","        _, predicted = outputs.max(1)\n","        total += targets.size(0)\n","        correct += predicted.eq(targets).sum().item()\n","\n","        #progress_bar(batch_idx, len(trainloader))\n","    print(f'Training Loss: {train_loss/(batch_idx+1)} | Training Acc: {100.*correct/total} ({correct}/{total})')\n","    train_loss_vec.append(train_loss/(batch_idx+1))\n","    train_acc_vec.append(100.*correct/total)"],"execution_count":10,"outputs":[]},{"cell_type":"code","metadata":{"id":"Qi_yEJIz6lR8","executionInfo":{"status":"ok","timestamp":1628779306246,"user_tz":240,"elapsed":6,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["# Testing\n","test_loss_vec = []\n","test_acc_vec = []\n","def test(epoch):\n","    global best_acc\n","    net.eval()\n","    test_loss = 0\n","    correct = 0\n","    total = 0\n","    with torch.no_grad():\n","        for batch_idx, (inputs, targets) in enumerate(testloader):\n","            inputs, targets = inputs.to(device), targets.to(device)\n","            outputs = net(inputs)\n","            loss = criterion(outputs, targets)\n","\n","            test_loss += loss.item()\n","            _, predicted = outputs.max(1)\n","            total += targets.size(0)\n","            correct += predicted.eq(targets).sum().item()\n","        \n","        scheduler.step(test_loss)\n","\n","        print(f'Validation Loss: {test_loss/(batch_idx+1)} | Validation Acc: {100.*correct/total} ({correct}/{total})')\n","\n","    test_loss_vec.append(test_loss/(batch_idx+1))\n","    test_acc_vec.append(100.*correct/total)\n","\n","    # Save checkpoint.\n","    acc = 100.*correct/total\n","    if acc > best_acc:\n","        print('Saving..')\n","        state = {\n","            'net': net.state_dict(),\n","            'acc': acc,\n","            'epoch': epoch,\n","        }\n","        if not os.path.isdir('checkpoint'):\n","            os.mkdir('checkpoint')\n","        torch.save(state, f'./checkpoint/resnet18_{epoch}.pth')\n","        best_acc = acc"],"execution_count":11,"outputs":[]},{"cell_type":"code","metadata":{"id":"WHym29C36pzS","colab":{"base_uri":"https://localhost:8080/","height":1000},"executionInfo":{"status":"error","timestamp":1628785394218,"user_tz":240,"elapsed":6087978,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}},"outputId":"6a899585-e58a-4eef-c1d9-ac74d9e815e0"},"source":["for epoch in range(start_epoch, start_epoch+80):\n","    train(epoch)\n","    test(epoch)"],"execution_count":12,"outputs":[{"output_type":"stream","text":["\n","Epoch: 0\n"],"name":"stdout"},{"output_type":"stream","text":["/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:718: UserWarning: Named tensors and all their associated APIs are an experimental feature and subject to change. Please do not use them for anything important until they are released as stable. (Triggered internally at  /pytorch/c10/core/TensorImpl.h:1156.)\n","  return torch.max_pool2d(input, kernel_size, stride, padding, dilation, ceil_mode)\n"],"name":"stderr"},{"output_type":"stream","text":["Training Loss: 4.667828205601333 | Training Acc: 6.764 (6764/100000)\n","Validation Loss: 4.2185799049425725 | Validation Acc: 11.27 (1127/10000)\n","Saving..\n","\n","Epoch: 1\n","Training Loss: 3.9195700713130823 | Training Acc: 15.222 (15222/100000)\n","Validation Loss: 3.836131527454038 | Validation Acc: 16.53 (1653/10000)\n","Saving..\n","\n","Epoch: 2\n","Training Loss: 3.613932890965201 | Training Acc: 19.942 (19942/100000)\n","Validation Loss: 3.6090437732165372 | Validation Acc: 20.24 (2024/10000)\n","Saving..\n","\n","Epoch: 3\n","Training Loss: 3.4402677954920113 | Training Acc: 22.872 (22872/100000)\n","Validation Loss: 3.482808116116101 | Validation Acc: 22.59 (2259/10000)\n","Saving..\n","\n","Epoch: 4\n","Training Loss: 3.324504895893204 | Training Acc: 24.708 (24708/100000)\n","Validation Loss: 3.4749277151083646 | Validation Acc: 22.55 (2255/10000)\n","\n","Epoch: 5\n","Training Loss: 3.2313811175353693 | Training Acc: 26.505 (26505/100000)\n","Validation Loss: 3.3150622512720807 | Validation Acc: 25.01 (2501/10000)\n","Saving..\n","\n","Epoch: 6\n","Training Loss: 3.1555750985889484 | Training Acc: 27.751 (27751/100000)\n","Validation Loss: 3.438648604139497 | Validation Acc: 23.69 (2369/10000)\n","\n","Epoch: 7\n","Training Loss: 3.0943720654758344 | Training Acc: 28.714 (28714/100000)\n","Validation Loss: 3.2639962661115427 | Validation Acc: 26.47 (2647/10000)\n","Saving..\n","\n","Epoch: 8\n","Training Loss: 3.040148105462799 | Training Acc: 29.65 (29650/100000)\n","Validation Loss: 3.1918957625763325 | Validation Acc: 27.61 (2761/10000)\n","Saving..\n","\n","Epoch: 9\n","Training Loss: 2.983339206032131 | Training Acc: 30.968 (30968/100000)\n","Validation Loss: 3.1892893284182007 | Validation Acc: 27.49 (2749/10000)\n","\n","Epoch: 10\n","Training Loss: 2.9515322718169075 | Training Acc: 31.422 (31422/100000)\n","Validation Loss: 3.2968373268465454 | Validation Acc: 26.19 (2619/10000)\n","\n","Epoch: 11\n","Training Loss: 2.9165741958276694 | Training Acc: 32.113 (32113/100000)\n","Validation Loss: 3.2545514136930054 | Validation Acc: 26.91 (2691/10000)\n","\n","Epoch: 12\n","Training Loss: 2.8947353707555004 | Training Acc: 32.472 (32472/100000)\n","Validation Loss: 3.2155701389795617 | Validation Acc: 27.63 (2763/10000)\n","Saving..\n","\n","Epoch: 13\n","Training Loss: 2.872956728691335 | Training Acc: 32.98 (32980/100000)\n","Validation Loss: 3.0631329500222506 | Validation Acc: 29.95 (2995/10000)\n","Saving..\n","\n","Epoch: 14\n","Training Loss: 2.849668665005423 | Training Acc: 33.278 (33278/100000)\n","Validation Loss: 3.0734697414349905 | Validation Acc: 30.41 (3041/10000)\n","Saving..\n","\n","Epoch: 15\n","Training Loss: 2.8334669899147795 | Training Acc: 33.618 (33618/100000)\n","Validation Loss: 3.342131780672677 | Validation Acc: 26.41 (2641/10000)\n","\n","Epoch: 16\n","Training Loss: 2.816183459118504 | Training Acc: 34.172 (34172/100000)\n","Validation Loss: 3.171908982192414 | Validation Acc: 27.81 (2781/10000)\n","\n","Epoch: 17\n","Training Loss: 2.800822526597611 | Training Acc: 34.282 (34282/100000)\n","Validation Loss: 3.10844316965417 | Validation Acc: 29.36 (2936/10000)\n","\n","Epoch: 18\n","Training Loss: 2.7821340667622168 | Training Acc: 34.806 (34806/100000)\n","Validation Loss: 3.124394006366971 | Validation Acc: 30.13 (3013/10000)\n","\n","Epoch: 19\n","Training Loss: 2.7682580682627687 | Training Acc: 34.871 (34871/100000)\n","Validation Loss: 3.0017142235478267 | Validation Acc: 31.3 (3130/10000)\n","Saving..\n","\n","Epoch: 20\n","Training Loss: 2.7632614044887025 | Training Acc: 34.995 (34995/100000)\n","Validation Loss: 2.903002476390404 | Validation Acc: 32.97 (3297/10000)\n","Saving..\n","\n","Epoch: 21\n","Training Loss: 2.740574509286515 | Training Acc: 35.409 (35409/100000)\n","Validation Loss: 3.04853242560278 | Validation Acc: 30.26 (3026/10000)\n","\n","Epoch: 22\n","Training Loss: 2.7363522555821995 | Training Acc: 35.375 (35375/100000)\n","Validation Loss: 3.0275608497329904 | Validation Acc: 30.64 (3064/10000)\n","\n","Epoch: 23\n","Training Loss: 2.7347592036132617 | Training Acc: 35.597 (35597/100000)\n","Validation Loss: 3.0695261683645128 | Validation Acc: 30.81 (3081/10000)\n","\n","Epoch: 24\n","Training Loss: 2.7182122255530197 | Training Acc: 35.909 (35909/100000)\n","Validation Loss: 2.9335791793050645 | Validation Acc: 32.83 (3283/10000)\n","\n","Epoch: 25\n","Training Loss: 2.7151557576016088 | Training Acc: 35.877 (35877/100000)\n","Validation Loss: 2.982343552987787 | Validation Acc: 32.14 (3214/10000)\n","\n","Epoch: 26\n","Training Loss: 2.700272882381059 | Training Acc: 36.114 (36114/100000)\n","Validation Loss: 3.0612786359424833 | Validation Acc: 30.35 (3035/10000)\n","\n","Epoch: 27\n","Training Loss: 2.0859154296653046 | Training Acc: 48.847 (48847/100000)\n","Validation Loss: 2.2277337388147282 | Validation Acc: 46.16 (4616/10000)\n","Saving..\n","\n","Epoch: 28\n","Training Loss: 1.8705846792291803 | Training Acc: 53.222 (53222/100000)\n","Validation Loss: 2.200805538817297 | Validation Acc: 47.15 (4715/10000)\n","Saving..\n","\n","Epoch: 29\n","Training Loss: 1.7697147934333137 | Training Acc: 55.26 (55260/100000)\n","Validation Loss: 2.1867316976378235 | Validation Acc: 47.48 (4748/10000)\n","Saving..\n","\n","Epoch: 30\n","Training Loss: 1.6937950912034114 | Training Acc: 56.91 (56910/100000)\n","Validation Loss: 2.1930348858048645 | Validation Acc: 47.49 (4749/10000)\n","Saving..\n","\n","Epoch: 31\n","Training Loss: 1.6247712459100787 | Training Acc: 58.389 (58389/100000)\n","Validation Loss: 2.2020501637760597 | Validation Acc: 47.69 (4769/10000)\n","Saving..\n","\n","Epoch: 32\n","Training Loss: 1.5694980112183126 | Training Acc: 59.458 (59458/100000)\n","Validation Loss: 2.2120158295088177 | Validation Acc: 47.71 (4771/10000)\n","Saving..\n","\n","Epoch: 33\n","Training Loss: 1.5190541352457403 | Training Acc: 60.634 (60634/100000)\n","Validation Loss: 2.219949286195296 | Validation Acc: 47.91 (4791/10000)\n","Saving..\n","\n","Epoch: 34\n","Training Loss: 1.4651046584329337 | Training Acc: 61.695 (61695/100000)\n","Validation Loss: 2.2692275122751164 | Validation Acc: 47.32 (4732/10000)\n","\n","Epoch: 35\n","Training Loss: 1.4351091390985358 | Training Acc: 62.391 (62391/100000)\n","Validation Loss: 2.2666819684113126 | Validation Acc: 46.69 (4669/10000)\n","\n","Epoch: 36\n","Training Loss: 1.1518103583999302 | Training Acc: 69.95 (69950/100000)\n","Validation Loss: 2.119887080373643 | Validation Acc: 49.98 (4998/10000)\n","Saving..\n","\n","Epoch: 37\n","Training Loss: 1.0536502640113197 | Training Acc: 72.502 (72502/100000)\n","Validation Loss: 2.120866774003717 | Validation Acc: 50.31 (5031/10000)\n","Saving..\n","\n","Epoch: 38\n","Training Loss: 1.0072435496560752 | Training Acc: 73.758 (73758/100000)\n","Validation Loss: 2.135293649721749 | Validation Acc: 50.24 (5024/10000)\n","\n","Epoch: 39\n","Training Loss: 0.9694870326220227 | Training Acc: 74.672 (74672/100000)\n","Validation Loss: 2.141123054902765 | Validation Acc: 50.11 (5011/10000)\n","\n","Epoch: 40\n","Training Loss: 0.9360592402613072 | Training Acc: 75.493 (75493/100000)\n","Validation Loss: 2.1485941138448594 | Validation Acc: 50.28 (5028/10000)\n","\n","Epoch: 41\n","Training Loss: 0.9048781304255776 | Training Acc: 76.44 (76440/100000)\n","Validation Loss: 2.1733720589287673 | Validation Acc: 50.21 (5021/10000)\n","\n","Epoch: 42\n","Training Loss: 0.8741256197547669 | Training Acc: 76.986 (76986/100000)\n","Validation Loss: 2.1836090389686293 | Validation Acc: 50.24 (5024/10000)\n","\n","Epoch: 43\n","Training Loss: 0.8227658590392384 | Training Acc: 78.649 (78649/100000)\n","Validation Loss: 2.178741251366048 | Validation Acc: 50.19 (5019/10000)\n","\n","Epoch: 44\n","Training Loss: 0.8132189899454336 | Training Acc: 78.843 (78843/100000)\n","Validation Loss: 2.170776186110098 | Validation Acc: 50.18 (5018/10000)\n","\n","Epoch: 45\n","Training Loss: 0.8062965261661793 | Training Acc: 79.005 (79005/100000)\n","Validation Loss: 2.1848188261442547 | Validation Acc: 50.21 (5021/10000)\n","\n","Epoch: 46\n","Training Loss: 0.8034302755008877 | Training Acc: 79.231 (79231/100000)\n","Validation Loss: 2.176003804689721 | Validation Acc: 50.31 (5031/10000)\n","\n","Epoch: 47\n","Training Loss: 0.8002840277484006 | Training Acc: 79.093 (79093/100000)\n","Validation Loss: 2.1794526939150654 | Validation Acc: 50.46 (5046/10000)\n","Saving..\n","\n","Epoch: 48\n","Training Loss: 0.7977188096174499 | Training Acc: 79.204 (79204/100000)\n","Validation Loss: 2.180691780923288 | Validation Acc: 50.12 (5012/10000)\n","\n","Epoch: 49\n","Training Loss: 0.7923688236862192 | Training Acc: 79.379 (79379/100000)\n","Validation Loss: 2.185761899887761 | Validation Acc: 50.22 (5022/10000)\n","\n","Epoch: 50\n","Training Loss: 0.7872002260840457 | Training Acc: 79.535 (79535/100000)\n","Validation Loss: 2.18209543258329 | Validation Acc: 50.52 (5052/10000)\n","Saving..\n","\n","Epoch: 51\n","Training Loss: 0.7890527527350599 | Training Acc: 79.518 (79518/100000)\n","Validation Loss: 2.182986528058595 | Validation Acc: 50.4 (5040/10000)\n","\n","Epoch: 52\n","Training Loss: 0.7847036014661155 | Training Acc: 79.79 (79790/100000)\n","Validation Loss: 2.180211943916128 | Validation Acc: 50.41 (5041/10000)\n","\n","Epoch: 53\n","Training Loss: 0.7915754958491801 | Training Acc: 79.607 (79607/100000)\n","Validation Loss: 2.176482709148262 | Validation Acc: 50.48 (5048/10000)\n","\n","Epoch: 54\n","Training Loss: 0.7891375172092482 | Training Acc: 79.587 (79587/100000)\n","Validation Loss: 2.189361753343027 | Validation Acc: 50.24 (5024/10000)\n","\n","Epoch: 55\n","Training Loss: 0.7869351754712937 | Training Acc: 79.72 (79720/100000)\n","Validation Loss: 2.1739159143423734 | Validation Acc: 50.47 (5047/10000)\n","\n","Epoch: 56\n","Training Loss: 0.78510682185745 | Training Acc: 79.746 (79746/100000)\n","Validation Loss: 2.1876171842406067 | Validation Acc: 50.14 (5014/10000)\n","\n","Epoch: 57\n","Training Loss: 0.7846062927675979 | Training Acc: 79.672 (79672/100000)\n","Validation Loss: 2.1825512134576144 | Validation Acc: 50.29 (5029/10000)\n","\n","Epoch: 58\n","Training Loss: 0.7854484225935339 | Training Acc: 79.618 (79618/100000)\n","Validation Loss: 2.1875817458840867 | Validation Acc: 50.21 (5021/10000)\n","\n","Epoch: 59\n","Training Loss: 0.7906306757951331 | Training Acc: 79.374 (79374/100000)\n","Validation Loss: 2.1795206175574773 | Validation Acc: 50.48 (5048/10000)\n","\n","Epoch: 60\n","Training Loss: 0.7858011108987472 | Training Acc: 79.689 (79689/100000)\n","Validation Loss: 2.1793795024292377 | Validation Acc: 50.35 (5035/10000)\n","\n","Epoch: 61\n","Training Loss: 0.7873096618506 | Training Acc: 79.659 (79659/100000)\n","Validation Loss: 2.1779340973383263 | Validation Acc: 50.56 (5056/10000)\n","Saving..\n","\n","Epoch: 62\n","Training Loss: 0.787163971939965 | Training Acc: 79.56 (79560/100000)\n","Validation Loss: 2.1782432538044603 | Validation Acc: 50.6 (5060/10000)\n","Saving..\n","\n","Epoch: 63\n","Training Loss: 0.7882407836596984 | Training Acc: 79.606 (79606/100000)\n","Validation Loss: 2.181396516063545 | Validation Acc: 50.59 (5059/10000)\n","\n","Epoch: 64\n","Training Loss: 0.7868881764466805 | Training Acc: 79.564 (79564/100000)\n","Validation Loss: 2.1793033442919767 | Validation Acc: 50.48 (5048/10000)\n","\n","Epoch: 65\n","Training Loss: 0.7874617259520704 | Training Acc: 79.592 (79592/100000)\n","Validation Loss: 2.1828060708468473 | Validation Acc: 50.1 (5010/10000)\n","\n","Epoch: 66\n"],"name":"stdout"},{"output_type":"error","ename":"KeyboardInterrupt","evalue":"ignored","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)","\u001b[0;32m<ipython-input-12-70dae1dc8e34>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mepoch\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstart_epoch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstart_epoch\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m80\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mepoch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m     \u001b[0mtest\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mepoch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m<ipython-input-10-a9fc0704fdf8>\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(epoch)\u001b[0m\n\u001b[1;32m     16\u001b[0m         \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m         \u001b[0mtrain_loss\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     19\u001b[0m         \u001b[0m_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredicted\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0moutputs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m         \u001b[0mtotal\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mtargets\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mKeyboardInterrupt\u001b[0m: "]}]},{"cell_type":"code","metadata":{"id":"utAOuioM9WTw","executionInfo":{"status":"aborted","timestamp":1628785393714,"user_tz":240,"elapsed":288,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":["# Checking Best Accuracy\n","def check_accuracy():\n","    net.eval()\n","    test_loss = 0\n","    correct = 0\n","    total = 0\n","    with torch.no_grad():\n","        for batch_idx, (inputs, targets) in enumerate(testloader):\n","            inputs, targets = inputs.to(device), targets.to(device)\n","            outputs = net(inputs)\n","            loss = criterion(outputs, targets)\n","\n","            test_loss += loss.item()\n","            _, predicted = outputs.max(1)\n","            total += targets.size(0)\n","            correct += predicted.eq(targets).sum().item()\n","\n","        print(f'Loss: {test_loss/(batch_idx+1)} | Acc: {100.*correct/total} ({correct}/{total})')\n","\n","check_accuracy()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"RM2QwQ4DJL5i","colab":{"base_uri":"https://localhost:8080/","height":638},"executionInfo":{"status":"ok","timestamp":1628785403295,"user_tz":240,"elapsed":986,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}},"outputId":"1a3dbe9d-597c-4963-fc17-837caf6fe37a"},"source":["# Training Plots\n","\n","plt.figure(figsize=(10,10))\n","plt.subplot(2,2,1)\n","plt.title('training loss')\n","plt.plot(train_loss_vec)\n","plt.subplot(2,2,2)\n","plt.title('testing loss')\n","plt.plot(test_loss_vec)\n","plt.subplot(2,2,3)\n","plt.title('training accuracy')\n","plt.plot(train_acc_vec)\n","plt.xlabel('epoch')\n","plt.subplot(2,2,4)\n","plt.title('testing accuracy')\n","plt.plot(test_acc_vec)\n","plt.xlabel('epoch')"],"execution_count":13,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Text(0.5, 0, 'epoch')"]},"metadata":{"tags":[]},"execution_count":13},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 720x720 with 4 Axes>"]},"metadata":{"tags":[],"needs_background":"light"}}]},{"cell_type":"code","metadata":{"id":"80pJz-ZeOSQQ","executionInfo":{"status":"aborted","timestamp":1628785393717,"user_tz":240,"elapsed":291,"user":{"displayName":"Avinash","photoUrl":"","userId":"03822310646586674578"}}},"source":[""],"execution_count":null,"outputs":[]}]}