{"cells":[{"metadata":{},"cell_type":"markdown","source":"In this notebook, we will perform transfer learning using EfficientNetB0 on MNIST. I will explain the architecture behind EfficientNetB0 as well as the concept behind transfer learning"},{"metadata":{"trusted":true},"cell_type":"code","source":"# Notebook parameters\n\nBATCH_SIZE = 64\nVALID_BATCH_SIZE = 100\nTEST_BATCH_SIZE = 100\nEPOCHS = 5\nNUM_CLASSES = 10\nSEED = 42\nEARLY_STOPPING = 25\nOUTPUT_DIR = '/kaggle/working/'\nMODEL_NAME = 'efficientnet-b0'","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!pip install efficientnet-pytorch","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Import necessary libraries"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport matplotlib.pyplot as plt\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nimport torch.optim as optim\nfrom torch.optim import lr_scheduler\nfrom torch.autograd import Variable\nfrom torch.utils.data import DataLoader, Dataset\n\nfrom sklearn.metrics import accuracy_score\nfrom PIL import Image, ImageOps, ImageEnhance\nfrom efficientnet_pytorch import EfficientNet","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Read in MNIST data:"},{"metadata":{"trusted":true},"cell_type":"code","source":"train = pd.read_csv('/kaggle/input/digit-recognizer/train.csv')\ntest = pd.read_csv('/kaggle/input/digit-recognizer/test.csv')\n\nprint('Shape of the training data: ', train.shape)\nprint('Shape of the test data: ', test.shape)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"sample_df = train.groupby('label').apply(lambda x: x.sample(n=1)).reset_index(drop = True)\nsample_df.drop(columns=['label'], inplace=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"nrows = 2\nncols = 5\nfig, axs = plt.subplots(nrows=nrows, ncols=ncols, gridspec_kw={'wspace': 0.01, 'hspace': 0.05},\n                       squeeze=True, figsize=(10,12))\n\nind_y = 0\nind_x = 0\nfor i, row in sample_df.iterrows():\n    if ind_y > ncols - 1:\n        ind_y = 0\n        ind_x += 1\n    sample_digit = sample_df.values[i, :].reshape((28, 28))\n    axs[ind_x, ind_y].axis('off')\n    axs[ind_x, ind_y].imshow(sample_digit, cmap='gray')\n    axs[ind_x, ind_y].set_title(\"Digit {}:\".format(i))\n    ind_y += 1\n\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.model_selection import train_test_split\n\n# Perform train, validation split\ntrain_df, valid_df = train_test_split(train, test_size = 0.2, random_state=SEED,stratify=train['label'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import cv2\n\n# Define custom data loader, \n# code adapted from https://www.kaggle.com/juiyangchang/cnn-with-pytorch-0-995-accuracy\n\nn_pixels = len(train_df.columns) - 1\n\n\nclass MNIST_Dataset(Dataset):\n    \"\"\"MNIST data set\"\"\"\n    \n    def __init__(self, df\n                ):\n        \n        if len(df.columns) == n_pixels:\n            # test data\n            self.X = df.values.reshape((-1,28,28)).astype(np.uint8)[:,:,:,None]\n            self.y = None\n            \n            self.X3 = np.full((self.X.shape[0], 3, 28, 28), 0.0)\n\n            for i, s in enumerate(self.X):\n                self.X3[i] = np.moveaxis(cv2.cvtColor(s, cv2.COLOR_GRAY2RGB), -1, 0)\n                \n        else:\n            # training/validation data\n            self.X = df.iloc[:,1:].values.reshape((-1,28,28)).astype(np.uint8)[:,:,:,None]\n            self.y = torch.from_numpy(df.iloc[:,0].values)\n            \n            self.X3 = np.full((self.X.shape[0], 3, 28, 28), 0.0)\n\n            for i, s in enumerate(self.X):\n                self.X3[i] = np.moveaxis(cv2.cvtColor(s, cv2.COLOR_GRAY2RGB), -1, 0)\n    \n    def __len__(self):\n        return len(self.X3)\n\n    def __getitem__(self, idx):\n        if self.y is not None:\n            return self.X3[idx] , self.y[idx]\n        else:\n            return self.X3[idx]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_dataset = MNIST_Dataset(train_df)\nvalid_dataset = MNIST_Dataset(valid_df)\ntest_dataset = MNIST_Dataset(test)\n\n\ntrain_loader = torch.utils.data.DataLoader(dataset=train_dataset,\n                                           batch_size=BATCH_SIZE,\n                                           shuffle=True)\nvalid_loader = torch.utils.data.DataLoader(dataset=valid_dataset,\n                                           batch_size=VALID_BATCH_SIZE, shuffle=False)\ntest_loader = torch.utils.data.DataLoader(dataset=test_dataset,\n                                           batch_size=TEST_BATCH_SIZE, shuffle=False)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","collapsed":true,"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":false},"cell_type":"markdown","source":"## Concept Behind Transfer Learning\n---\n\nFirst let's understand the concept behind transfer learning.\n\nAn analogy is the way humans learn. For example, we learn and gain knowledge in a certain field like math and statistics. We are able to take what we have learnt and apply to other fields.\n\nIn deep learning, there is a similar concept known as transfer learning. Transfer learning is the idea of overcoming the isolated learning paradigm and utilizing knowledge acquired for one task to solve related ones. \n\nFor Convolutional Neural Networks, transfer learning can be done by taking weights learnt by a model architecture on a specific task (for eg, Image classification of a predefined set of images) and using these weights as a starting point, with a modified architecture, letting the new model learn for more epochs.\n\n![image.png](attachment:image.png)\n\n(Image from: https://researchgate.net/figure/Illustration-of-transfer-learning-concept-where-the-rst-layers-in-network-A-and-network-B_fig2_316748306)\n","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"<ins>Great, but how does Transfer Learning help in training a CNN? </ins>\n\n* Generally, the rate of convergence of a CNN is faster when transfer learning is applied as compared to without."},{"metadata":{},"cell_type":"markdown","source":"## Transfer Learning using EfficientNet\n---\n\nYou may have seen solutions of winning solutions in kaggle using EfficientNet. B**ut what is EfficientNet?**\n\nEfficientNet is deep learning architecture designed by Google to tackle the problem of scaling Neural Networks (deciding how to best increase model size and increase accuracy). Given that there is a tradeoff between efficiency and accuracy in scaling CNNs, the idea by Google is to provide better accuracy and improve the efficiency of the models by reducing the parameters and FLOPS (Floating Point Operations Per Second) manifold. There is an excellent article on Medium that explains in detail, what this architecture does: https://towardsdatascience.com/efficientnet-scaling-of-convolutional-neural-networks-done-right-3fde32aef8ff"},{"metadata":{},"cell_type":"markdown","source":"We can take the pretrained weights and the architecture provided in EfficientNet, remove the \"head\" of the CNN network, and add our own custom head to form a new architecture. Then train it for some epochs for our model to learn specific to the task we set it to do "},{"metadata":{"trusted":true},"cell_type":"code","source":"## Load in pretrained effnet model and remove its head, replacing it with fully connected layer \n## that gives 10 outputs \ndef get_model(model_name='efficientnet-b0'):\n    model = EfficientNet.from_pretrained(model_name)\n    del model._fc\n    # # # use the same head as the baseline notebook.\n    model._fc = nn.Linear(1280, NUM_CLASSES)\n    return model","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"It's as simple as that! Just 6 lines of code to import the weights and architecture of pretrained efficientNet and replace its head."},{"metadata":{},"cell_type":"markdown","source":"Now, let's do some training with our efficientNet architecture"},{"metadata":{"trusted":true},"cell_type":"code","source":"import random\nimport os\n\ndef set_seed(seed: int = 42):\n    random.seed(seed)\n    np.random.seed(seed)\n    os.environ[\"PYTHONHASHSEED\"] = str(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)  # type: ignore","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"set_seed(SEED)\ndevice = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\noutput_dir = OUTPUT_DIR\n\n\nmodel = get_model(MODEL_NAME)\nmodel = model.to(device)\n\n    \n# # # get optimizer\noptimizer = optim.Adam(model.parameters(), lr=0.001)\n\n# # # get scheduler\nscheduler = lr_scheduler.CosineAnnealingLR(optimizer, T_max=10)\n\n# # # get loss\nloss_func = nn.CrossEntropyLoss()\n\nif torch.cuda.is_available():\n    model = model.cuda()\n    loss_func = loss_func.cuda()\n\nbest_val_accuracy = 0\nmin_val_loss = np.inf\nbest_epoch = 0\nbatches = 0\nepochs_no_improve = 0\nn_epochs_stop = EARLY_STOPPING\nfor epoch in range(EPOCHS):\n    running_loss = 0.0\n    targets = torch.empty(size=(BATCH_SIZE, )).to(device) \n    outputs = torch.empty(size=(BATCH_SIZE, )).to(device)\n    model.train()\n    for batch_idx, (data, target) in enumerate(train_loader):\n        batches += 1\n        data, target = Variable(data), Variable(target)\n        if torch.cuda.is_available():\n            data = data.type(torch.FloatTensor).cuda()\n            target = target.cuda()\n        targets = torch.cat((targets, target), 0)\n        optimizer.zero_grad()\n        output = model(data)\n        loss = loss_func(output, target)\n        output = torch.argmax(torch.softmax(output, dim=1), dim=1)\n        outputs = torch.cat((outputs, output), 0)\n        running_loss += loss.item()\n        loss.backward()\n        optimizer.step()\n    scheduler.step()\n    print('train/loss on EPOCH {}: {}'.format(epoch, running_loss/batches))\n    train_acc = accuracy_score(targets.cpu().detach().numpy().astype(int), \n                              outputs.cpu().detach().numpy().astype(int))\n    print('train/accuracy: {} for epoch {}'.format(train_acc, epoch))\n\n    model.eval()\n    # Validation loop\n    running_loss = 0.0\n    batches = 0\n    targets = torch.empty(size=(BATCH_SIZE, )).to(device) \n    outputs = torch.empty(size=(BATCH_SIZE, )).to(device) \n    for batch_idx, (data, target) in enumerate(valid_loader):\n        batches += 1\n        data, target = Variable(data, volatile=True), Variable(target)\n        if torch.cuda.is_available():\n            data = data.type(torch.FloatTensor).cuda()\n            target = target.cuda()\n        with torch.no_grad():\n            targets = torch.cat((targets, target), 0)\n            output = model(data)\n            loss = loss_func(output, target)\n            output = torch.argmax(torch.softmax(output, dim=1), dim=1)\n            outputs = torch.cat((outputs, output), 0)\n            running_loss += loss.item()\n\n    val_loss = running_loss/batches\n    print('val/loss: {}'.format(val_loss))\n    val_acc = accuracy_score(targets.cpu().detach().numpy().astype(int), \n                      outputs.cpu().detach().numpy().astype(int))\n    print('val/accuracy: {} for epoch {}'.format(val_acc, epoch))\n\n\n\n    # Model Checkpoint for best validation f1\n    if val_acc > best_val_accuracy:\n        best_val_accuracy = val_acc\n        min_val_loss = val_loss\n        print('Best val/acc: {} for epoch {}, saving model---->'.format(val_acc, epoch))\n        torch.save(model.state_dict(), \"{}/snapshot_epoch_{}.pth\".format(output_dir, epoch))\n        best_epoch = epoch\n        epochs_no_improve = 0\n    else:\n        epochs_no_improve += 1\n    if epochs_no_improve == n_epochs_stop:\n        print('Early stopping!')\n        break","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Prediction on Test Set"},{"metadata":{"trusted":true},"cell_type":"code","source":"def prediction(model, data_loader):\n    model.eval()\n    test_pred = torch.LongTensor()\n    \n    for i, data in enumerate(data_loader):\n        data = Variable(data, volatile=True)\n        if torch.cuda.is_available():\n            data = data.type(torch.FloatTensor).cuda()\n            \n        output = model(data)\n        pred = output.cpu().data.max(1, keepdim=True)[1]\n        test_pred = torch.cat((test_pred, pred), dim=0)\n        \n    return test_pred","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.load_state_dict(torch.load(\"snapshot_epoch_{}.pth\".format(best_epoch)))\ntest_pred = prediction(model, test_loader)\nsubmission = pd.DataFrame(np.c_[np.arange(1, len(test_dataset)+1)[:,None], test_pred.numpy()], \n                          columns=['ImageId', 'Label'])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission.to_csv('submission.csv', index=False)","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}