{"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":"markdown","source":"# Imports","metadata":{}},{"cell_type":"code","source":"!pip install segmentation_models_pytorch --quiet","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:50:59.162410Z","iopub.execute_input":"2023-05-26T15:50:59.163132Z","iopub.status.idle":"2023-05-26T15:51:11.053029Z","shell.execute_reply.started":"2023-05-26T15:50:59.163086Z","shell.execute_reply":"2023-05-26T15:51:11.051839Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport numpy as np \nimport pandas as pd \nimport matplotlib.pyplot as plt\nfrom PIL import Image\nimport cv2\nfrom torch.utils.data import Dataset , DataLoader, random_split, SubsetRandomSampler\nfrom torchvision import transforms\nimport albumentations as A\nfrom albumentations.pytorch import ToTensorV2\nimport glob\nimport torch\nfrom tqdm.notebook import tqdm\nfrom skimage.io import imread\nimport segmentation_models_pytorch as smp\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import confusion_matrix\nfrom segmentation_models_pytorch.encoders import get_preprocessing_fn\nfrom segmentation_models_pytorch import utils\n%matplotlib inline","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:11.056839Z","iopub.execute_input":"2023-05-26T15:51:11.057425Z","iopub.status.idle":"2023-05-26T15:51:11.069889Z","shell.execute_reply.started":"2023-05-26T15:51:11.057393Z","shell.execute_reply":"2023-05-26T15:51:11.068931Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Data Directories**","metadata":{}},{"cell_type":"code","source":"ROOT_DIR = '/kaggle/input/airbus-ship-detection'\nTRAIN_DIR = os.path.join(ROOT_DIR, 'train_v2')\nTEST_DIR = os.path.join(ROOT_DIR, 'test_v2')\nTRAIN_SHIP_SEGMENTATIONS_DIR = os.path.join(ROOT_DIR, 'train_ship_segmentations_v2.csv')\nSAMPLE_SUBMISSION_DIR = os.path.join(ROOT_DIR, 'sample_submission_v2.csv')","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:11.071664Z","iopub.execute_input":"2023-05-26T15:51:11.072447Z","iopub.status.idle":"2023-05-26T15:51:11.080173Z","shell.execute_reply.started":"2023-05-26T15:51:11.072413Z","shell.execute_reply":"2023-05-26T15:51:11.079291Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Data Loading**","metadata":{}},{"cell_type":"code","source":"TRAIN_NAMES = os.listdir(TRAIN_DIR)\nTEST_NAMES = os.listdir(TEST_DIR)\nTRAIN_SHIP_SEGMENTATIONS_DF = pd.read_csv(TRAIN_SHIP_SEGMENTATIONS_DIR).set_index('ImageId')\nSAMPLE_SUBMISSION_DF = pd.read_csv(SAMPLE_SUBMISSION_DIR)\n\nprint(f\"TRAIN_NAMES length = {len(TRAIN_NAMES)}\\nTEST_NAMES length = {len(TEST_NAMES)}\")","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:11.083572Z","iopub.execute_input":"2023-05-26T15:51:11.084026Z","iopub.status.idle":"2023-05-26T15:51:12.231096Z","shell.execute_reply.started":"2023-05-26T15:51:11.083978Z","shell.execute_reply":"2023-05-26T15:51:12.229148Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Corrupted images list**","metadata":{}},{"cell_type":"code","source":"CORRUPTED_IMAGES =['6384c3e78.jpg','13703f040.jpg', '14715c06d.jpg', '33e0ff2d5.jpg',\n '4d4e09f2a.jpg', '877691df8.jpg', '8b909bb20.jpg', 'a8d99130e.jpg', \n 'ad55c3143.jpg', 'c8260c541.jpg', 'd6c7f17c7.jpg', 'dc3e7c901.jpg',\n 'e44dffe88.jpg', 'ef87bad36.jpg', 'f083256d8.jpg']","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:12.234237Z","iopub.execute_input":"2023-05-26T15:51:12.235546Z","iopub.status.idle":"2023-05-26T15:51:12.242675Z","shell.execute_reply.started":"2023-05-26T15:51:12.235466Z","shell.execute_reply":"2023-05-26T15:51:12.241776Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for names_set in [TRAIN_NAMES, TEST_NAMES]:\n for name in names_set:\n if name in CORRUPTED_IMAGES:\n names_set.remove(name)\n \nprint(f\"TRAIN_NAMES length = {len(TRAIN_NAMES)}\\nTEST_NAMES length = {len(TEST_NAMES)}\")","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:12.244589Z","iopub.execute_input":"2023-05-26T15:51:12.245128Z","iopub.status.idle":"2023-05-26T15:51:12.430001Z","shell.execute_reply.started":"2023-05-26T15:51:12.245098Z","shell.execute_reply":"2023-05-26T15:51:12.429068Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data View","metadata":{}},{"cell_type":"code","source":"def plot_images(num_of_images, title_image_dict, cmap_list=None, figsize=(12, 24), font_size=15, same_title=False, axis='off'):\n plt.rcParams.update({'font.size': font_size})\n plt.figure(figsize=figsize)\n\n if same_title:\n title = np.full(num_of_images, str(list(title_image_dict.keys())[0]))\n values = list(title_image_dict.values())[0]\n else:\n values = list(title_image_dict.values())\n title = list(title_image_dict.keys())\n\n rows = (num_of_images - 1) // 3 + 1 # Calculate the number of rows needed\n\n for i in range(num_of_images):\n plt.subplot(rows, 3, i + 1) # Use 3 as the number of columns\n plt.axis(axis)\n\n if cmap_list is not None:\n plt.imshow(values[i], cmap=cmap_list[i])\n else:\n plt.imshow(values[i])\n\n plt.title(f'{title[i]}')\n\n plt.tight_layout() # Adjust the layout to prevent overlapping","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:12.431616Z","iopub.execute_input":"2023-05-26T15:51:12.432272Z","iopub.status.idle":"2023-05-26T15:51:12.441841Z","shell.execute_reply.started":"2023-05-26T15:51:12.432236Z","shell.execute_reply":"2023-05-26T15:51:12.440901Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"first_train_image_path = (os.path.join(TRAIN_DIR,TRAIN_NAMES[0]))\nfirst_train_image = cv2.imread(first_train_image_path)\nfirst_train_image = cv2.cvtColor(first_train_image, cv2.COLOR_BGR2RGB)\nprint(f'{first_train_image.shape = }\\n')\n\nplot_images(num_of_images=1, title_image_dict={f\"{TRAIN_NAMES[0]}\":first_train_image}, figsize = (10,10), font_size=12, axis=True)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:12.443801Z","iopub.execute_input":"2023-05-26T15:51:12.444268Z","iopub.status.idle":"2023-05-26T15:51:12.886466Z","shell.execute_reply.started":"2023-05-26T15:51:12.444234Z","shell.execute_reply":"2023-05-26T15:51:12.884603Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display(TRAIN_SHIP_SEGMENTATIONS_DF)\nnum_of_unique_images = TRAIN_SHIP_SEGMENTATIONS_DF.index.nunique()\nnot_empty = pd.notna(TRAIN_SHIP_SEGMENTATIONS_DF.EncodedPixels)\nnum_of_empty_images = (~not_empty).sum()\nnum_of_masks = not_empty.sum()\nnum_of_non_empty_images = TRAIN_SHIP_SEGMENTATIONS_DF[not_empty].index.nunique()\n\nprint(f'{num_of_unique_images = } | {num_of_empty_images = } | {num_of_non_empty_images = } | {num_of_masks = }\\n')","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:12.888090Z","iopub.execute_input":"2023-05-26T15:51:12.888720Z","iopub.status.idle":"2023-05-26T15:51:13.022728Z","shell.execute_reply.started":"2023-05-26T15:51:12.888685Z","shell.execute_reply":"2023-05-26T15:51:13.021621Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**In conclusion the train_ship_segmentations_df contains:**\n\n* 192556 unique images\n* 150000 empty images\n* 42556 images with total of 81723 masks","metadata":{}},{"cell_type":"markdown","source":"# Deal With RLE (run-length encoding)","metadata":{}},{"cell_type":"code","source":"# create masks data frame\nmasks = TRAIN_SHIP_SEGMENTATIONS_DF.dropna(subset=['EncodedPixels'])\ndisplay(masks.head(10))","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.027885Z","iopub.execute_input":"2023-05-26T15:51:13.028235Z","iopub.status.idle":"2023-05-26T15:51:13.126868Z","shell.execute_reply.started":"2023-05-26T15:51:13.028209Z","shell.execute_reply":"2023-05-26T15:51:13.125055Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# we can see that the image '000155de5.jpg' has only one mask\nprint(masks['EncodedPixels']['000155de5.jpg'])","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.128361Z","iopub.execute_input":"2023-05-26T15:51:13.128718Z","iopub.status.idle":"2023-05-26T15:51:13.146258Z","shell.execute_reply.started":"2023-05-26T15:51:13.128686Z","shell.execute_reply":"2023-05-26T15:51:13.145140Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Explenation**\n\nThe first pair (264661, 17) indicates that there is a run of 17 pixels starting from the position 264,661 in the mask. Similarly, the next two pairs represent runs of 33 pixels starting from positions 265,429 and 266,197, respectively.","metadata":{}},{"cell_type":"code","source":"# turn rle example into a list of ints\nrle = [int(i) for i in masks['EncodedPixels']['000155de5.jpg'].split()]\n\n# turn list of ints into a list of (`start`, `length`) `pairs`\npairs = list(zip(rle[0:-1:2], rle[1::2])) \nprint(f'{pairs[:3] = }')","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.147691Z","iopub.execute_input":"2023-05-26T15:51:13.148428Z","iopub.status.idle":"2023-05-26T15:51:13.160216Z","shell.execute_reply.started":"2023-05-26T15:51:13.148390Z","shell.execute_reply":"2023-05-26T15:51:13.158836Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# convert into (x, y) coordinate where x is the column and y is the row\nfirst_pixel = pairs[0]\nfirst_pixel_coordinates = (first_pixel[0] % 768, first_pixel[0] // 768)\nprint(f\"first pixel: length = {first_pixel[1]} | position = {first_pixel[0]} | (x, y) coordinates = {first_pixel_coordinates}\")\n","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.162180Z","iopub.execute_input":"2023-05-26T15:51:13.162705Z","iopub.status.idle":"2023-05-26T15:51:13.169722Z","shell.execute_reply.started":"2023-05-26T15:51:13.162666Z","shell.execute_reply":"2023-05-26T15:51:13.168628Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def get_mask(image_name, df):\n shape = (768,768)\n image = np.zeros(shape[0]*shape[1], dtype=np.uint8)\n masks = df.loc[image_name]['EncodedPixels']\n if(type(masks) == float): return image.reshape(shape)\n if(type(masks) == str): masks = [masks]\n for mask in masks:\n s = mask.split()\n for i in range(len(s)//2):\n start = int(s[2*i]) - 1\n length = int(s[2*i+1])\n image[start:start+length] = 1\n return image.reshape(shape).T","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.171472Z","iopub.execute_input":"2023-05-26T15:51:13.171722Z","iopub.status.idle":"2023-05-26T15:51:13.181929Z","shell.execute_reply.started":"2023-05-26T15:51:13.171699Z","shell.execute_reply":"2023-05-26T15:51:13.180961Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def decode_mask(mask, shape=(768, 768)):\n pixels = mask.T.flatten()\n pixels = np.concatenate([[0], pixels, [0]])\n runs = np.where(pixels[1:] != pixels[:-1])[0] + 1\n runs[1::2] -= runs[::2]\n return ' '.join(str(x) for x in runs)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.183731Z","iopub.execute_input":"2023-05-26T15:51:13.184542Z","iopub.status.idle":"2023-05-26T15:51:13.191713Z","shell.execute_reply.started":"2023-05-26T15:51:13.184509Z","shell.execute_reply":"2023-05-26T15:51:13.190706Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Plot Mask Bounding Box\n\nI will use '000155de5.jpg' for my first example","metadata":{}},{"cell_type":"code","source":"# example image\nFIRST_EXAMPLE_IMAGE_NAME = '000155de5.jpg'\nexample_image_path = (os.path.join(TRAIN_DIR, FIRST_EXAMPLE_IMAGE_NAME))\nexample_image = cv2.imread(example_image_path)\nexample_image = cv2.cvtColor(example_image, cv2.COLOR_BGR2RGB)\n\nexample_mask = get_mask(FIRST_EXAMPLE_IMAGE_NAME, TRAIN_SHIP_SEGMENTATIONS_DF)\n\nfirst_example_dict = {\"Image\":example_image, \"Mask\":example_mask}\n\nplot_images(num_of_images=2, title_image_dict=first_example_dict, cmap_list=[None, 'gray'], figsize=(12,12))","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.193338Z","iopub.execute_input":"2023-05-26T15:51:13.194094Z","iopub.status.idle":"2023-05-26T15:51:13.735566Z","shell.execute_reply.started":"2023-05-26T15:51:13.194060Z","shell.execute_reply":"2023-05-26T15:51:13.734746Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Images can have multiple masks!**\n\nlets see image '000194a2d.jpg' for the second example ","metadata":{}},{"cell_type":"code","source":"# example image\nSECOND_EXAMPLE_IMAGE_NAME = '000194a2d.jpg'\nexample_image_path = (os.path.join(TRAIN_DIR, SECOND_EXAMPLE_IMAGE_NAME))\nexample_image = cv2.imread(example_image_path)\nexample_image = cv2.cvtColor(example_image, cv2.COLOR_BGR2RGB)\n\nexample_mask = get_mask(SECOND_EXAMPLE_IMAGE_NAME, TRAIN_SHIP_SEGMENTATIONS_DF)\nsecond_example_dict = {\"Image\":example_image, \"Mask\":example_mask}\n\n \nplot_images(num_of_images=2, title_image_dict=second_example_dict, cmap_list= [None, 'gray'], figsize = (12,12))","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:13.737208Z","iopub.execute_input":"2023-05-26T15:51:13.737843Z","iopub.status.idle":"2023-05-26T15:51:14.238850Z","shell.execute_reply.started":"2023-05-26T15:51:13.737809Z","shell.execute_reply":"2023-05-26T15:51:14.237974Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Prepare Data For Training","metadata":{}},{"cell_type":"code","source":"class ShipsDataSet(Dataset):\n def __init__(self, mode, transform=None):\n \n self.mode = mode\n self.transform = transform\n \n if mode == 'train':\n self.image_names = TRAIN_NAMES\n self.data_frame = TRAIN_SHIP_SEGMENTATIONS_DF\n self.directory = TRAIN_DIR\n if mode == 'test':\n self.image_names = TEST_NAMES\n self.data_frame = SAMPLE_SUBMISSION_DF\n self.directory = TEST_DIR\n \n \n\n def __getitem__(self, i):\n image = np.array(Image.open(os.path.join(self.directory,self.image_names[i])))\n \n if image.shape[:2] != (768,768):\n image = cv2.resize(image, (768,768))\n\n if self.mode == 'test':\n mask = np.zeros((768,768), dtype=np.uint8)\n else:\n mask = get_mask(self.image_names[i], self.data_frame)\n\n sample = {'image': image, 'mask': mask}\n \n if self.transform:\n sample = self.transform(image=image, mask=mask)\n\n \n return (sample['image'], sample['mask'])\n \n def __len__(self):\n return len(self.image_names)\n","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:14.240247Z","iopub.execute_input":"2023-05-26T15:51:14.240772Z","iopub.status.idle":"2023-05-26T15:51:14.254044Z","shell.execute_reply.started":"2023-05-26T15:51:14.240737Z","shell.execute_reply":"2023-05-26T15:51:14.252848Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"my_transform = A.Compose([\n A.Resize(224, 224),\n A.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),\n ToTensorV2()\n])","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:14.257467Z","iopub.execute_input":"2023-05-26T15:51:14.257756Z","iopub.status.idle":"2023-05-26T15:51:14.266831Z","shell.execute_reply.started":"2023-05-26T15:51:14.257731Z","shell.execute_reply":"2023-05-26T15:51:14.265959Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"BATCH_SIZE = 64\nNUM_SAMPLES = 1000\n\ntrain_set = ShipsDataSet(mode='train', transform = my_transform)\ntest_set = ShipsDataSet(mode='test', transform = my_transform)\n\n# Create validation datast\ntrain_size = int(0.8 * len(train_set))\nval_size = len(train_set) - train_size\ntrain_set, val_set = random_split(train_set, [train_size, val_size])\n\n\n# Create a random subset of indices\nsampled_indices_train = np.random.choice(len(train_set), NUM_SAMPLES, replace=False)\nsampled_indices_val = np.random.choice(len(val_set), min(NUM_SAMPLES//10, len(val_set)), replace=False)\n\n# Use the subset indices to create the SubsetRandomSampler\nsampler_train = SubsetRandomSampler(sampled_indices_train)\nsampler_val = SubsetRandomSampler(sampled_indices_val)\n\n# Create sub samples of the train and validation loaders\ntrain_loader_sample = DataLoader(dataset=train_set, batch_size=BATCH_SIZE, sampler=sampler_train)\nval_loader_sample = DataLoader(dataset=val_set, batch_size=BATCH_SIZE, sampler=sampler_val)\n\n# Create the data loaders from the dataset class\ntrain_loader = DataLoader(dataset=train_set, batch_size=BATCH_SIZE,shuffle=True)\nval_loader = DataLoader(dataset=val_set, batch_size=BATCH_SIZE)\ntest_loader = DataLoader(dataset=test_set, batch_size=BATCH_SIZE)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:14.268184Z","iopub.execute_input":"2023-05-26T15:51:14.268486Z","iopub.status.idle":"2023-05-26T15:51:14.302509Z","shell.execute_reply.started":"2023-05-26T15:51:14.268461Z","shell.execute_reply":"2023-05-26T15:51:14.301699Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model","metadata":{}},{"cell_type":"code","source":"ENCODER = 'resnet34'\nENCODER_WEIGHTS = 'imagenet'\nCLASSES = ['ships']\nACTIVATION = 'sigmoid' # could be None for logits or 'softmax2d' for multicalss segmentation\nDEVICE = 'cuda'\n\n# create segmentation model with pretrained encoder\n\nif os.path.exists('/kaggle/input/airbus-model-unet34/best_model.pth'):\n model = torch.load('/kaggle/input/airbus-model-unet34/best_model.pth', map_location=DEVICE)\n print(\"Model loaded successfully.\")\nelse:\n model = smp.Unet(\n encoder_name=ENCODER, \n encoder_weights=ENCODER_WEIGHTS, \n classes=len(CLASSES), \n activation=ACTIVATION)\n print(\"Saved model file not found.\")\n\n\npreprocessing_fn = smp.encoders.get_preprocessing_fn(ENCODER, ENCODER_WEIGHTS)\nloss = smp.utils.losses.DiceLoss()\nmetrics = [\n smp.utils.metrics.IoU(threshold=0.5),\n]\n\noptimizer = torch.optim.Adam([ \n dict(params=model.parameters(), lr=0.0001),\n])","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:14.304085Z","iopub.execute_input":"2023-05-26T15:51:14.304414Z","iopub.status.idle":"2023-05-26T15:51:14.935978Z","shell.execute_reply.started":"2023-05-26T15:51:14.304383Z","shell.execute_reply":"2023-05-26T15:51:14.934960Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Train & Evaluate Function","metadata":{}},{"cell_type":"code","source":"train_epoch = smp.utils.train.TrainEpoch(\n model, \n loss=loss, \n metrics=metrics, \n optimizer=optimizer,\n device=DEVICE,\n verbose=True,\n)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:14.972134Z","iopub.execute_input":"2023-05-26T15:51:14.972472Z","iopub.status.idle":"2023-05-26T15:51:15.013720Z","shell.execute_reply.started":"2023-05-26T15:51:14.972441Z","shell.execute_reply":"2023-05-26T15:51:15.012819Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"val_epoch = smp.utils.train.ValidEpoch(\n model,\n loss=loss,\n metrics=metrics,\n device=DEVICE,\n verbose=True,\n)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T15:51:15.015100Z","iopub.execute_input":"2023-05-26T15:51:15.015445Z","iopub.status.idle":"2023-05-26T15:51:15.023853Z","shell.execute_reply.started":"2023-05-26T15:51:15.015414Z","shell.execute_reply":"2023-05-26T15:51:15.022815Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"max_score = 0\n\nfor i in range(0, 1):\n \n print('\\nEpoch: {}'.format(i))\n train_logs = train_epoch.run(train_loader)\n valid_logs = val_epoch.run(val_loader)\n \n # do something (save model, change lr, etc.)\n if max_score < train_logs['iou_score']:\n max_score = train_logs['iou_score']\n torch.save(model, './best_model.pth')\n print('Model saved!')\n \n if i == 25:\n optimizer.param_groups[0]['lr'] = 1e-5\n print('Decrease decoder learning rate to 1e-5!')","metadata":{"execution":{"iopub.status.busy":"2023-05-26T16:13:17.655653Z","iopub.execute_input":"2023-05-26T16:13:17.656097Z","iopub.status.idle":"2023-05-26T17:13:47.778649Z","shell.execute_reply.started":"2023-05-26T16:13:17.656065Z","shell.execute_reply":"2023-05-26T17:13:47.777354Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Submission","metadata":{}},{"cell_type":"code","source":"test_df = SAMPLE_SUBMISSION_DF.copy()\ntest_df = test_df.set_index('ImageId')\ntest_df","metadata":{"execution":{"iopub.status.busy":"2023-05-26T17:31:47.274703Z","iopub.execute_input":"2023-05-26T17:31:47.275096Z","iopub.status.idle":"2023-05-26T17:31:47.290960Z","shell.execute_reply.started":"2023-05-26T17:31:47.275064Z","shell.execute_reply":"2023-05-26T17:31:47.289906Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def encode_minibatch(outputs):\n encoded_list = []\n for i in range(len(outputs)):\n predicted_mask = cv2.resize(outputs[i], (768, 768))\n encoded_list.append(decode_mask(predicted_mask))\n \n return encoded_list","metadata":{"execution":{"iopub.status.busy":"2023-05-26T17:59:24.020499Z","iopub.execute_input":"2023-05-26T17:59:24.020888Z","iopub.status.idle":"2023-05-26T17:59:24.028141Z","shell.execute_reply.started":"2023-05-26T17:59:24.020855Z","shell.execute_reply":"2023-05-26T17:59:24.025799Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"SUBMIT = True\n\nif SUBMIT:\n with torch.no_grad():\n size = test_loader.batch_size\n start = 0\n end = size\n for images, labels in tqdm(test_loader):\n images = images.to(DEVICE)\n labels = labels.to(DEVICE)\n outputs = model.to(DEVICE)(images)\n outputs = outputs.cpu().detach().numpy().squeeze()\n outputs = outputs.astype(np.uint8)\n\n if end > len(test_df):\n test_df['EncodedPixels'][start:] = encode_minibatch(outputs)\n else:\n test_df['EncodedPixels'][start:end] = encode_minibatch(outputs)\n\n start = end\n end += size\n \n test_df.to_csv('submission.csv', index=True)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T17:59:25.078536Z","iopub.execute_input":"2023-05-26T17:59:25.078905Z","iopub.status.idle":"2023-05-26T18:02:17.485373Z","shell.execute_reply.started":"2023-05-26T17:59:25.078874Z","shell.execute_reply":"2023-05-26T18:02:17.484433Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"![Screenshot 2023-05-26 at 22.33.36.png](attachment:4e3da49f-ab5f-4dc9-b9c9-c88f49c64ee5.png)","metadata":{},"attachments":{"4e3da49f-ab5f-4dc9-b9c9-c88f49c64ee5.png":{"image/png":"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"}}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}