{"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_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"!pip install tez\n!pip install albumentation","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport albumentations\nimport matplotlib.pyplot as plt\nimport pandas as pd\n\nimport tez\nfrom tez.datasets import ImageDataset\nfrom tez.callbacks import EarlyStopping\n\nimport torch\nimport torch.nn as nn\nimport torchvision\n\nfrom sklearn import metrics, model_selection\n","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfx = pd.read_csv('../input/cassava-leaf-disease-classification/train.csv')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfx.head()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dfx.label.value_counts()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train, df_valid = model_selection.train_test_split(\n    dfx,\n    test_size=0.1,\n    random_state=42,\n    stratify=dfx.label.values\n)\n\ndf_train = df_train.reset_index(drop = True)\ndf_valid = df_valid.reset_index(drop = True)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.shape","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_valid.shape","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"image_path = \"../input/cassava-leaf-disease-classification/train_images/\"\n\n\ntrain_image_paths = [\n    os.path.join(image_path, x) for x in df_train.image_id.values\n]\n\nvalid_image_paths = [\n    os.path.join(image_path, x) for x in df_valid.image_id.values\n]","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_image_paths[:5]","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_targets = df_train.label.values\nvalid_targets = df_valid.label.values","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_dataset = ImageDataset(\n    image_paths= train_image_paths,\n    targets=train_targets,\n    \n    augmentations = None\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_dataset[0]","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def plot_img(image_dict):\n    image_tensor = image_dict[\"image\"]\n    target = image_dict[\"targets\"]\n    print(target)\n    plt.figure(figsize=(10,10))\n    image = image_tensor.permute(1, 2, 0) / 255\n    plt.imshow(image)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_img(train_dataset[0])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_aug = albumentations.Compose(\n    [\n        albumentations.RandomResizedCrop(256, 256),\n        albumentations.Transpose(p=0.5),\n        albumentations.HorizontalFlip(p=0.5),\n        albumentations.VerticalFlip(p=0.5),\n    ]\n\n)\n\n\nvalid_aug = albumentations.Compose(\n    [\n        albumentations.CenterCrop(256, 256, p=1.0),\n        albumentations.RandomResizedCrop(256, 256),\n        albumentations.Transpose(p=0.5),\n        albumentations.HorizontalFlip(p=0.5),\n        albumentations.VerticalFlip(p=0.5),\n    ]\n\n)\n\ntrain_dataset = ImageDataset(\n    image_paths= train_image_paths,\n    targets=train_targets,\n    \n    augmentations = train_aug\n)\n\n\nvalid_dataset = ImageDataset(\n    image_paths= valid_image_paths,\n    targets=valid_targets,\n    \n    augmentations = valid_aug\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_img(train_dataset[0])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class LeafModel(tez.Model):\n    def __init__(self, num_classes, pretrained=True):\n        super().__init__()\n        self.convnet = torchvision.models.resnet18(pretrained=pretrained)\n        self.convnet.fc = nn.Linear(512, num_classes)\n        self.step_scheduler_after = \"epoch\"\n    \n    def loss(self, outputs, targets):\n        if targets is None:\n            return None\n        return nn.CrossEntropyLoss()(outputs, targets)\n    \n    def monitor_matrics(self, outputs, targets):\n        outputs = torch.argmax(outputs, dim=1).cpu().detach().numpy()\n        targets = targets.cpu().detach().numpy()\n        acc = metrics.accuracy_score(targets, outputs)\n        return{\n            \"accuracy\":acc\n        }\n        \n    def fetch_optimizer(self):\n        opt = torch.optim.Adam(self.parameters(), lr = 1e-3)\n        return opt\n    \n    def fetch_sheduler(self):\n        sch = torch.optim.lr_scheduler.StepLR(self.optimizer, step_size=0.7)\n        return sch\n        \n    def forward(self, image, targets=None):\n        outputs = self.convnet(image)\n        if targets is not None:\n            loss = self.loss(outputs, targets)\n            mon_metrics = self.monitor_matrics(outputs, targets)\n            return outputs, loss, mon_metrics\n        return outputs, None, None","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = LeafModel(num_classes=dfx.label.nunique(),pretrained=True)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"img = train_dataset[10][\"image\"]\ny = train_dataset[10][\"targets\"]\n\n\nmodel(img.unsqueeze(0), y.unsqueeze(0))","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"es = EarlyStopping(\n    monitor=\"valid_accuracy\",\n    model_path=\"model.bin\",\n    patience=2,\n    mode = \"max\"\n)\n\nmodel.fit(\n    train_dataset,\n    valid_dataset=valid_dataset,\n    train_bs=32,\n    valid_bs=64,\n    device = 'cuda',\n    callbacks=[es],\n    fp16=True,\n    epochs = 10\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_dfx = pd.read_csv('../input/cassava-leaf-disease-classification/sample_submission.csv')\nimage_path = \"../input/cassava-leaf-disease-classification/test_images/\"\ntest_targets = test_dfx.label.values\n\ntest_image_paths = [\n    os.path.join(image_path, x) for x in test_dfx.image_id.values\n]\n\ntest_aug = albumentations.Compose(\n    [\n        albumentations.CenterCrop(256, 256, p=1.0),\n        albumentations.RandomResizedCrop(256, 256),\n        albumentations.Transpose(p=0.5),\n        albumentations.HorizontalFlip(p=0.5),\n        albumentations.VerticalFlip(p=0.5),\n    ]\n\n)\n\ntest_dataset = ImageDataset(\n    image_paths=test_image_paths,\n    targets=test_targets,\n    augmentations = test_aug\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"pred = model.predict(\n    test_dataset,\n    batch_size=64, \n    n_jobs = -1\n)\n\nfinal_preds = None\n\nfor p in pred:\n    if final_preds is None:\n        final_preds = p\n    else:\n        final_preds= np.vstack((final_preds, p))\n        \nfinal_preds = final_preds.argmax(axis=1)\ntest_dfx.label = final_preds\ntest_dfx.to_csv(\"submission.csv\", index=False)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}