{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"sourceId":30892,"databundleVersionId":2715462,"sourceType":"competition"}],"dockerImageVersionId":30580,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"!pip install torchsummary\n!pip install torchgeometry","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-11-16T16:43:03.390448Z","iopub.execute_input":"2023-11-16T16:43:03.391239Z","iopub.status.idle":"2023-11-16T16:43:28.14757Z","shell.execute_reply.started":"2023-11-16T16:43:03.391201Z","shell.execute_reply":"2023-11-16T16:43:28.146369Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from torchsummary import summary\nfrom torchgeometry.losses import one_hot\nimport os\nimport pandas as pd\nimport numpy as np\nfrom PIL import Image\nimport cv2\nimport time\nimport imageio\nimport matplotlib.pyplot as plt\nimport time\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 import Tensor\nfrom torch.utils.data import Dataset, DataLoader, random_split\nfrom torchvision.transforms import Resize, PILToTensor, ToPILImage, Compose, InterpolationMode\nfrom collections import OrderedDict\nfrom PIL import ImageFilter\nimport torchvision.transforms as transforms\nimport random\nimport wandb","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:28.150515Z","iopub.execute_input":"2023-11-16T16:43:28.150912Z","iopub.status.idle":"2023-11-16T16:43:28.160318Z","shell.execute_reply.started":"2023-11-16T16:43:28.150873Z","shell.execute_reply":"2023-11-16T16:43:28.159109Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"!nvidia-smi -L","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:28.162301Z","iopub.execute_input":"2023-11-16T16:43:28.162646Z","iopub.status.idle":"2023-11-16T16:43:29.218821Z","shell.execute_reply.started":"2023-11-16T16:43:28.162618Z","shell.execute_reply":"2023-11-16T16:43:29.217662Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\ndevice","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.220589Z","iopub.execute_input":"2023-11-16T16:43:29.220914Z","iopub.status.idle":"2023-11-16T16:43:29.229398Z","shell.execute_reply.started":"2023-11-16T16:43:29.220885Z","shell.execute_reply":"2023-11-16T16:43:29.22838Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Number of class in the data set (3: neoplastic, non neoplastic, background)\nnum_classes = 3\n\n# Number of epoch\nepochs = 10\n\n# Hyperparameters for training \nlearning_rate = 2e-04\nbatch_size = 4\ndisplay_step = 50\n\n# Model path\ncheckpoint_path = '/kaggle/working/unet_model.pth'\npretrained_path = \"/kaggle/input/unet-checkpoint/unet_model.pth\"\n# Initialize lists to keep track of loss and accuracy\nloss_epoch_array = []\ntrain_accuracy = []\ntest_accuracy = []\nvalid_accuracy = []","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.233139Z","iopub.execute_input":"2023-11-16T16:43:29.233414Z","iopub.status.idle":"2023-11-16T16:43:29.241229Z","shell.execute_reply.started":"2023-11-16T16:43:29.233388Z","shell.execute_reply":"2023-11-16T16:43:29.240098Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transform = Compose([Resize((512, 512), interpolation=InterpolationMode.BILINEAR),\n                     PILToTensor()])","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.242667Z","iopub.execute_input":"2023-11-16T16:43:29.243075Z","iopub.status.idle":"2023-11-16T16:43:29.251778Z","shell.execute_reply.started":"2023-11-16T16:43:29.243022Z","shell.execute_reply":"2023-11-16T16:43:29.250756Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class UNetDataClass(Dataset):\n    def __init__(self, images_path, masks_path, transform):\n        super(UNetDataClass, self).__init__()\n        \n        images_list = os.listdir(images_path)\n        masks_list = os.listdir(masks_path)\n        \n        images_list = [images_path + image_name for image_name in images_list]\n        masks_list = [masks_path + mask_name for mask_name in masks_list]\n        \n        self.images_list = images_list\n        self.masks_list = masks_list\n        self.transform = transform\n        \n    def __getitem__(self, index):\n        img_path = self.images_list[index]\n        mask_path = self.masks_list[index]\n        \n        # Open image and mask\n        data = Image.open(img_path)\n        label = Image.open(mask_path)\n        \n        # Normalize\n        data = self.transform(data) / 255\n        label = self.transform(label) / 255\n        \n        label = torch.where(label>0.65, 1.0, 0.0)\n        \n        label[2, :, :] = 0.0001\n        label = torch.argmax(label, 0).type(torch.int64)\n        return data, label\n    \n    def __len__(self):\n        return len(self.images_list)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.253372Z","iopub.execute_input":"2023-11-16T16:43:29.25384Z","iopub.status.idle":"2023-11-16T16:43:29.267578Z","shell.execute_reply.started":"2023-11-16T16:43:29.253806Z","shell.execute_reply":"2023-11-16T16:43:29.266447Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from torch.utils.data import ConcatDataset","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.269129Z","iopub.execute_input":"2023-11-16T16:43:29.269449Z","iopub.status.idle":"2023-11-16T16:43:29.277191Z","shell.execute_reply.started":"2023-11-16T16:43:29.269421Z","shell.execute_reply":"2023-11-16T16:43:29.276111Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"images_path = \"/kaggle/input/bkai-igh-neopolyp/train/train/\"\nmasks_path =  \"/kaggle/input/bkai-igh-neopolyp/train_gt/train_gt/\"","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.278704Z","iopub.execute_input":"2023-11-16T16:43:29.279216Z","iopub.status.idle":"2023-11-16T16:43:29.28534Z","shell.execute_reply.started":"2023-11-16T16:43:29.279161Z","shell.execute_reply":"2023-11-16T16:43:29.284326Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_transform = Compose([\n    Resize((512, 512), interpolation=InterpolationMode.BILINEAR),\n    transforms.RandomVerticalFlip(p=0.5),\n    transforms.RandomHorizontalFlip(p=0.5),\n    transforms.RandomRotation(degrees = 45),\n    transforms.ColorJitter(brightness=0.2, contrast=0.2, saturation=0.2, hue=0.2),\n    transforms.RandomResizedCrop(size=(512, 512), scale=(0.8, 1.0)),\n    PILToTensor()\n])\n\nval_transform = Compose([\n    Resize((512, 512), interpolation=InterpolationMode.BILINEAR),\n    PILToTensor()\n    \n])\n\n\n\ntrain_dataset_not_aug = UNetDataClass(images_path= images_path,\n                             masks_path= masks_path,\n                             transform = val_transform)\n\ntrain_dataset_aug = UNetDataClass(images_path= images_path,\n                             masks_path= masks_path,\n                             transform = train_transform)\ntrain_dataset = ConcatDataset([train_dataset_not_aug, train_dataset_aug])\ntrain_dataset[0][0].shape\n\n#print(len(train_dataset))","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.286565Z","iopub.execute_input":"2023-11-16T16:43:29.287763Z","iopub.status.idle":"2023-11-16T16:43:29.357406Z","shell.execute_reply.started":"2023-11-16T16:43:29.287727Z","shell.execute_reply":"2023-11-16T16:43:29.355887Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_size = 0.8\nvalid_size = 0.2","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.359193Z","iopub.execute_input":"2023-11-16T16:43:29.359611Z","iopub.status.idle":"2023-11-16T16:43:29.367115Z","shell.execute_reply.started":"2023-11-16T16:43:29.359568Z","shell.execute_reply":"2023-11-16T16:43:29.36508Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_set, valid_set = random_split(train_dataset, \n                                    [int(train_size * len(train_dataset)) , \n                                     int(valid_size * len(train_dataset))])","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.368703Z","iopub.execute_input":"2023-11-16T16:43:29.369586Z","iopub.status.idle":"2023-11-16T16:43:29.376091Z","shell.execute_reply.started":"2023-11-16T16:43:29.369549Z","shell.execute_reply":"2023-11-16T16:43:29.374894Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"code","source":"train_dataloader = DataLoader(train_set, batch_size=8, shuffle=True)\nvalid_dataloader = DataLoader(valid_set, batch_size=8, shuffle=True)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.377456Z","iopub.execute_input":"2023-11-16T16:43:29.378255Z","iopub.status.idle":"2023-11-16T16:43:29.385475Z","shell.execute_reply.started":"2023-11-16T16:43:29.378216Z","shell.execute_reply":"2023-11-16T16:43:29.38438Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def double_conv(in_channels, out_channels):\n    return nn.Sequential(\n        nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1),\n        nn.BatchNorm2d(out_channels),\n        nn.ReLU(inplace=True),\n        nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1),\n        nn.BatchNorm2d(out_channels),\n        nn.ReLU(inplace=True)\n    )","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.393407Z","iopub.execute_input":"2023-11-16T16:43:29.393929Z","iopub.status.idle":"2023-11-16T16:43:29.401388Z","shell.execute_reply.started":"2023-11-16T16:43:29.393902Z","shell.execute_reply":"2023-11-16T16:43:29.400299Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def up_conv(in_channels, out_channels):\n    return nn.ConvTranspose2d(\n        in_channels, out_channels, kernel_size=2, stride=2\n    )","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.402936Z","iopub.execute_input":"2023-11-16T16:43:29.403681Z","iopub.status.idle":"2023-11-16T16:43:29.411351Z","shell.execute_reply.started":"2023-11-16T16:43:29.403624Z","shell.execute_reply":"2023-11-16T16:43:29.410382Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from torchvision.models import vgg16_bn\n\nclass VGGUnet(nn.Module):\n\n    def __init__(self, out_channels=2):\n        super().__init__()\n\n        self.encoder = vgg16_bn(pretrained=True).features\n        self.block1 = nn.Sequential(*self.encoder[:6])\n        self.block2 = nn.Sequential(*self.encoder[6:13])\n        self.block3 = nn.Sequential(*self.encoder[13:20])\n        self.block4 = nn.Sequential(*self.encoder[20:27])\n        self.block5 = nn.Sequential(*self.encoder[27:34])\n\n        self.bottleneck = nn.Sequential(*self.encoder[34:])\n        self.conv_bottleneck = double_conv(512, 1024)\n\n        self.up_conv6 = up_conv(1024, 512)\n        self.conv6 = double_conv(512 + 512, 512)\n        self.up_conv7 = up_conv(512, 256)\n        self.conv7 = double_conv(256 + 512, 256)\n        self.up_conv8 = up_conv(256, 128)\n        self.conv8 = double_conv(128 + 256, 128)\n        self.up_conv9 = up_conv(128, 64)\n        self.conv9 = double_conv(64 + 128, 64)\n        self.up_conv10 = up_conv(64, 32)\n        self.conv10 = double_conv(32 + 64, 32)\n        self.conv11 = nn.Conv2d(32, out_channels, kernel_size=1)\n\n    def forward(self, x):\n        block1 = self.block1(x)\n        block2 = self.block2(block1)\n        block3 = self.block3(block2)\n        block4 = self.block4(block3)\n        block5 = self.block5(block4)\n\n        bottleneck = self.bottleneck(block5)\n        x = self.conv_bottleneck(bottleneck)\n\n        x = self.up_conv6(x)\n        x = torch.cat([x, block5], dim=1)\n        x = self.conv6(x)\n\n        x = self.up_conv7(x)\n        x = torch.cat([x, block4], dim=1)\n        x = self.conv7(x)\n\n        x = self.up_conv8(x)\n        x = torch.cat([x, block3], dim=1)\n        x = self.conv8(x)\n\n        x = self.up_conv9(x)\n        x = torch.cat([x, block2], dim=1)\n        x = self.conv9(x)\n\n        x = self.up_conv10(x)\n        x = torch.cat([x, block1], dim=1)\n        x = self.conv10(x)\n\n        x = self.conv11(x)\n\n        return x\n","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.412552Z","iopub.execute_input":"2023-11-16T16:43:29.412905Z","iopub.status.idle":"2023-11-16T16:43:29.432378Z","shell.execute_reply.started":"2023-11-16T16:43:29.412877Z","shell.execute_reply":"2023-11-16T16:43:29.431087Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class CEDiceLoss(nn.Module):\n    def __init__(self, weights) -> None:\n        super(CEDiceLoss, self).__init__()\n        self.eps: float = 1e-6\n        self.weights: torch.Tensor = weights\n\n    def forward(\n            self,\n            input: torch.Tensor,\n            target: torch.Tensor) -> torch.Tensor:\n        if not torch.is_tensor(input):\n            raise TypeError(\"Input type is not a torch.Tensor. Got {}\"\n                            .format(type(input)))\n        if not len(input.shape) == 4:\n            raise ValueError(\"Invalid input shape, we expect BxNxHxW. Got: {}\"\n                             .format(input.shape))\n        if not input.shape[-2:] == target.shape[-2:]:\n            raise ValueError(\"input and target shapes must be the same. Got: {}\"\n                             .format(input.shape, input.shape))\n        if not input.device == target.device:\n            raise ValueError(\n                \"input and target must be in the same device. Got: {}\" .format(\n                    input.device, target.device))\n        if not self.weights.shape[1] == input.shape[1]:\n            raise ValueError(\"The number of weights must equal the number of classes\")\n        if not torch.sum(self.weights).item() == 1:\n            raise ValueError(\"The sum of all weights must equal 1\")\n            \n        # cross entropy loss\n        celoss = nn.CrossEntropyLoss(self.weights)(input, target)\n        \n        # compute softmax over the classes axis\n        input_soft = F.softmax(input, dim=1)\n\n        # create the labels one hot tensor\n        target_one_hot = one_hot(target, num_classes=input.shape[1],\n                                 device=input.device, dtype=input.dtype)\n\n        # compute the actual dice score\n        dims = (2, 3)\n        intersection = torch.sum(input_soft * target_one_hot, dims)\n        cardinality = torch.sum(input_soft + target_one_hot, dims)\n\n        dice_score = 2. * intersection / (cardinality + self.eps)\n        \n        dice_score = torch.sum(dice_score * self.weights, dim=1)\n        \n        return torch.mean(1. - dice_score) + celoss","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.433652Z","iopub.execute_input":"2023-11-16T16:43:29.43395Z","iopub.status.idle":"2023-11-16T16:43:29.450859Z","shell.execute_reply.started":"2023-11-16T16:43:29.433923Z","shell.execute_reply":"2023-11-16T16:43:29.44961Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def weights_init(model):\n    if isinstance(model, nn.Linear):\n        # Xavier Distribution\n        torch.nn.init.xavier_uniform_(model.weight)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.452981Z","iopub.execute_input":"2023-11-16T16:43:29.453849Z","iopub.status.idle":"2023-11-16T16:43:29.465919Z","shell.execute_reply.started":"2023-11-16T16:43:29.453806Z","shell.execute_reply":"2023-11-16T16:43:29.464871Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def save_model(model, optimizer, path):\n    checkpoint = {\n        \"model\": model.state_dict(),\n        \"optimizer\": optimizer.state_dict(),\n    }\n    torch.save(checkpoint, path)\n\ndef load_model(model, optimizer, path):\n    checkpoint = torch.load(path)\n    model.load_state_dict(checkpoint[\"model\"])\n    optimizer.load_state_dict(checkpoint['optimizer'])\n    return model, optimizer","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.467456Z","iopub.execute_input":"2023-11-16T16:43:29.468309Z","iopub.status.idle":"2023-11-16T16:43:29.477018Z","shell.execute_reply.started":"2023-11-16T16:43:29.468271Z","shell.execute_reply":"2023-11-16T16:43:29.47587Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Train function for each epoch\ndef train(train_dataloader, valid_dataloader,learing_rate_scheduler, epoch, display_step):\n    print(f\"Start epoch #{epoch+1}, learning rate for this epoch: {learing_rate_scheduler.get_last_lr()}\")\n    start_time = time.time()\n    train_loss_epoch = 0\n    test_loss_epoch = 0\n    last_loss = 999999999\n    model.train()\n    for i, (data,targets) in enumerate(train_dataloader):\n        \n        # Load data into GPU\n        data, targets = data.to(device), targets.to(device)\n\n        optimizer.zero_grad()\n        outputs = model(data)\n\n        # Backpropagation, compute gradients\n        loss = loss_function(outputs, targets.long())\n        loss.backward()\n\n        # Apply gradients\n        optimizer.step()\n        \n        # Save loss\n        train_loss_epoch += loss.item()\n        if (i+1) % display_step == 0:\n#             accuracy = float(test(test_loader))\n            print('Train Epoch: {} [{}/{} ({}%)]\\tLoss: {:.4f}'.format(\n                epoch + 1, (i+1) * len(data), len(train_dataloader.dataset), 100 * (i+1) * len(data) / len(train_dataloader.dataset), \n                loss.item()))\n                  \n    print(f\"Done epoch #{epoch+1}, time for this epoch: {time.time()-start_time}s\")\n    train_loss_epoch/= (i + 1)\n    \n    # Evaluate the validation set\n    model.eval()\n    with torch.no_grad():\n        for data, target in valid_dataloader:\n            data, target = data.to(device), target.to(device)\n            test_output = model(data)\n            test_loss = loss_function(test_output, target)\n            test_loss_epoch += test_loss.item()\n            \n    test_loss_epoch/= (i+1)\n    \n    return train_loss_epoch , test_loss_epoch","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.478542Z","iopub.execute_input":"2023-11-16T16:43:29.478974Z","iopub.status.idle":"2023-11-16T16:43:29.495142Z","shell.execute_reply.started":"2023-11-16T16:43:29.478933Z","shell.execute_reply":"2023-11-16T16:43:29.494023Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Test function\ndef test(dataloader):\n    test_loss = 0\n    correct = 0\n    with torch.no_grad():\n        for i, (data, targets) in enumerate(dataloader):\n            data, targets = data.to(device), targets.to(device)\n            outputs = model(data)\n            _, pred = torch.max(outputs, 1)\n            test_loss += targets.size(0)\n            correct += torch.sum(pred == targets).item()\n    return 100.0 * correct / test_loss","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.496871Z","iopub.execute_input":"2023-11-16T16:43:29.497282Z","iopub.status.idle":"2023-11-16T16:43:29.50755Z","shell.execute_reply.started":"2023-11-16T16:43:29.497243Z","shell.execute_reply":"2023-11-16T16:43:29.506276Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = VGGUnet(out_channels = 3)\n\n# model = nn.DataParallel(model)\n#checkpoint = torch.load(pretrained_path)\n\n#new_state_dict = OrderedDict()\n#for k, v in checkpoint['model'].items():\n#    name = k[7:] # remove `module.`\n#    new_state_dict[name] = v\n# load params\n#model.load_state_dict(new_state_dict)\nmodel.to(device)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:29.509112Z","iopub.execute_input":"2023-11-16T16:43:29.509999Z","iopub.status.idle":"2023-11-16T16:43:31.38893Z","shell.execute_reply.started":"2023-11-16T16:43:29.509957Z","shell.execute_reply":"2023-11-16T16:43:31.387761Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"summary(model, (3, 512, 512))","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:31.390381Z","iopub.execute_input":"2023-11-16T16:43:31.390741Z","iopub.status.idle":"2023-11-16T16:43:31.460171Z","shell.execute_reply.started":"2023-11-16T16:43:31.390707Z","shell.execute_reply":"2023-11-16T16:43:31.459025Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"weights = torch.Tensor([[0.4, 0.55, 0.05]]).cuda()\nloss_function = CEDiceLoss(weights)\n\n# Define the optimizer (Adam optimizer)\noptimizer = optim.Adam(params=model.parameters(), lr=learning_rate)\n#optimizer.load_state_dict(checkpoint['optimizer'])\n\n# Learning rate scheduler\nlearing_rate_scheduler = lr_scheduler.StepLR(optimizer, step_size=4, gamma=0.6)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:31.46142Z","iopub.execute_input":"2023-11-16T16:43:31.461712Z","iopub.status.idle":"2023-11-16T16:43:31.587895Z","shell.execute_reply.started":"2023-11-16T16:43:31.461685Z","shell.execute_reply":"2023-11-16T16:43:31.586737Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"save_model(model, optimizer, checkpoint_path)","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:31.589201Z","iopub.execute_input":"2023-11-16T16:43:31.589493Z","iopub.status.idle":"2023-11-16T16:43:32.249526Z","shell.execute_reply.started":"2023-11-16T16:43:31.589466Z","shell.execute_reply":"2023-11-16T16:43:32.248374Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"wandb.login(\n    # set the wandb project where this run will be logged\n#     project= \"PolypSegment\", \n    key = \"e646c0d3811286ec78074864658c743c1cdf8651\",\n)\nwandb.init(\n    project = \"bkao-neopolyp\"\n)\n# Training loop\ntrain_loss_array = []\ntest_loss_array = []\nlast_loss = 9999999999999\nfor epoch in range(epochs):\n    train_loss_epoch = 0\n    test_loss_epoch = 0\n    (train_loss_epoch, test_loss_epoch) = train(train_dataloader, \n                                              valid_dataloader, \n                                              learing_rate_scheduler, epoch, display_step)\n    \n    if test_loss_epoch < last_loss:\n        save_model(model, optimizer, checkpoint_path)\n        last_loss = test_loss_epoch\n        \n    learing_rate_scheduler.step()\n    train_loss_array.append(train_loss_epoch)\n    test_loss_array.append(test_loss_epoch)\n    wandb.log({\"Train loss\": train_loss_epoch, \"Valid loss\": test_loss_epoch})\n#     train_accuracy.append(test(train_loader))\n#     valid_accuracy.append(test(test_loader))\n#     print(\"Epoch {}: loss: {:.4f}, train accuracy: {:.4f}, valid accuracy:{:.4f}\".format(epoch + 1, \n#                                         train_loss_array[-1], train_accuracy[-1], valid_accuracy[-1]))","metadata":{"execution":{"iopub.status.busy":"2023-11-16T16:43:32.25134Z","iopub.execute_input":"2023-11-16T16:43:32.251749Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"torch.cuda.empty_cache()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# load_model(model, checkpoint)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.rcParams['figure.dpi'] = 90\nplt.rcParams['figure.figsize'] = (6, 4)\nepochs_array = range(epochs)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Plot Training and Test loss\nplt.plot(epochs_array, train_loss_array, 'g', label='Training loss')\n# plt.plot(epochs_array, test_loss_array, 'b', label='Test loss')\nplt.title('Training and Test loss')\nplt.xlabel('Epochs')\nplt.ylabel('Loss')\nplt.legend()\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# from torch.jit import load\n# model = UNet()\n# optimizer = optim.Adam(params=model.parameters(), lr=learning_rate)\n\n# checkpoint = torch.load(pretrained_path)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# optimizer.load_state_dict(checkpoint['optimizer'])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# from collections import OrderedDict\n# new_state_dict = OrderedDict()\n# for k, v in checkpoint['model'].items():\n#     name = k[7:] # remove `module.`\n#     new_state_dict[name] = v\n# # load params\n# model.load_state_dict(new_state_dict)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for i, (data, label) in enumerate(train_dataloader):\n    img = data\n    mask = label\n\n    print(type(img))\n    print(type(mask))\n    print(img.shape)\n\n\n    break","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, arr = plt.subplots(4, 3, figsize=(16, 12))\narr[0][0].set_title('Image')\narr[0][1].set_title('Segmentation')\narr[0][2].set_title('Predict')\n\nmodel.eval()\nwith torch.no_grad():\n    predict = model(img)\n    predict = predict\n    print(predict)\n\nfor i in range(4):\n    arr[i][0].imshow(img[i].permute(1, 2, 0));\n    \n    arr[i][1].imshow(F.one_hot(mask[i]).float())\n    \n    arr[i][2].imshow(F.one_hot(torch.argmax(predict[i], 0).cpu()).float())","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transform = Compose([Resize((224, 224), interpolation=InterpolationMode.BILINEAR),\n                     PILToTensor()])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class UNetTestDataClass(Dataset):\n    def __init__(self, images_path, transform):\n        super(UNetTestDataClass, self).__init__()\n        \n        images_list = os.listdir(images_path)\n        images_list = [images_path+i for i in images_list]\n        \n        self.images_list = images_list\n        self.transform = transform\n        \n    def __getitem__(self, index):\n        img_path = self.images_list[index]\n        data = Image.open(img_path)\n        h = data.size[1]\n        w = data.size[0]\n        data = self.transform(data) / 255        \n        return data, img_path, h, w\n    \n    def __len__(self):\n        return len(self.images_list)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"path = '/kaggle/input/bka0-neopolyp/test/test/'\nunet_test_dataset = UNetTestDataClass(path, transform)\ntest_dataloader = DataLoader(unet_test_dataset, batch_size=8, shuffle=True)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for i, (data, path, h, w) in enumerate(test_dataloader):\n    img = data\n    break","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, arr = plt.subplots(5, 2, figsize=(16, 12))\narr[0][0].set_title('Image');\narr[0][1].set_title('Predict');\n\nmodel.eval()\nwith torch.no_grad():\n    predict = model(img)\n\nfor i in range(5):\n    arr[i][0].imshow(img[i].permute(1, 2, 0));\n    arr[i][1].imshow(F.one_hot(torch.argmax(predict[i], 0).cpu()).float())","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.eval()\nif not os.path.isdir(\"/kaggle/working/predicted_masks\"):\n    os.mkdir(\"/kaggle/working/predicted_masks\")\nfor _, (img, path, H, W) in enumerate(test_dataloader):\n    a = path\n    b = img\n    h = H\n    w = W\n    \n    with torch.no_grad():\n        predicted_mask = model(b)\n    for i in range(len(a)):\n        image_id = a[i].split('/')[-1].split('.')[0]\n        filename = image_id + \".png\"\n        mask2img = Resize((h[i].item(), w[i].item()), interpolation=InterpolationMode.NEAREST)(ToPILImage()(F.one_hot(torch.argmax(predicted_mask[i], 0)).permute(2, 0, 1).float()))\n        mask2img.save(os.path.join(\"/kaggle/working/predicted_masks/\", filename))","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\n\n\n\ndef rle_to_string(runs):\n    return ' '.join(str(x) for x in runs)\n\ndef rle_encode_one_mask(mask):\n    pixels = mask.flatten()\n    pixels[pixels > 0] = 255\n    use_padding = False\n    if pixels[0] or pixels[-1]:\n        use_padding = True\n        pixel_padded = np.zeros([len(pixels) + 2], dtype=pixels.dtype)\n        pixel_padded[1:-1] = pixels\n        pixels = pixel_padded\n    \n    rle = np.where(pixels[1:] != pixels[:-1])[0] + 2\n    if use_padding:\n        rle = rle - 1\n    rle[1::2] = rle[1::2] - rle[:-1:2]\n    return rle_to_string(rle)\n\ndef mask2string(dir):\n    ## mask --> string\n    strings = []\n    ids = []\n    ws, hs = [[] for i in range(2)]\n    for image_id in os.listdir(dir):\n        id = image_id.split('.')[0]\n        path = os.path.join(dir, image_id)\n        print(path)\n        img = cv2.imread(path)[:,:,::-1]\n        h, w = img.shape[0], img.shape[1]\n        for channel in range(2):\n            ws.append(w)\n            hs.append(h)\n            ids.append(f'{id}_{channel}')\n            string = rle_encode_one_mask(img[:,:,channel])\n            strings.append(string)\n    r = {\n        'ids': ids,\n        'strings': strings,\n    }\n    return r\n\n\nMASK_DIR_PATH = '/kaggle/working/predicted_masks' # change this to the path to your output mask folder\ndir = MASK_DIR_PATH\nres = mask2string(dir)\ndf = pd.DataFrame(columns=['Id', 'Expected'])\ndf['Id'] = res['ids']\ndf['Expected'] = res['strings']\ndf.to_csv(r'output.csv', index=False)","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}