{"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":"import argparse\nimport copy\nimport gc\nimport glob\nimport os\nimport random\nimport sys\nimport time\nfrom collections import defaultdict\nimport numpy as np\nimport pandas as pd\nimport torch\nfrom fastprogress import progress_bar\nfrom matplotlib import pyplot as plt\nfrom sklearn.model_selection import KFold\nfrom torch import optim\nimport warnings\nwarnings.filterwarnings(\"ignore\")","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-08-05T16:12:28.212446Z","iopub.execute_input":"2022-08-05T16:12:28.213070Z","iopub.status.idle":"2022-08-05T16:12:28.219253Z","shell.execute_reply.started":"2022-08-05T16:12:28.213034Z","shell.execute_reply":"2022-08-05T16:12:28.218276Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"!cp -r ../input/pytorch-segmentation-models-lib/ ./\n!pip config set global.disable-pip-version-check true\n!pip install -q ./pytorch-segmentation-models-lib/pretrainedmodels-0.7.4/pretrainedmodels-0.7.4\n!pip install -q ./pytorch-segmentation-models-lib/efficientnet_pytorch-0.6.3/efficientnet_pytorch-0.6.3\n!pip install -q ./pytorch-segmentation-models-lib/timm-0.4.12-py3-none-any.whl\n!pip install -q ./pytorch-segmentation-models-lib/segmentation_models_pytorch-0.2.0-py3-none-any.whl","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:12:28.251801Z","iopub.execute_input":"2022-08-05T16:12:28.252415Z","iopub.status.idle":"2022-08-05T16:13:12.526568Z","shell.execute_reply.started":"2022-08-05T16:12:28.252383Z","shell.execute_reply":"2022-08-05T16:13:12.525266Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class CFG:\n    def __init__(self, fold=0, seed=2020, train_bs=32, debug=False, pretrained_path=None):\n        self.n_fold = 5\n        self.fold = fold\n        self.seed = seed\n        self.train_bs = train_bs\n        self.valid_bs = train_bs  # may need to change this\n        self.debug = debug\n        self.img_size = [256, 256]\n        self.exp_name = 'Hubmap256-training'\n        self.epochs = 1000\n        self.lr = 2e-3\n        self.pretrained_path = pretrained_path\n        self.scheduler = \"CosineAnnealingLR\"\n        self.min_lr = 2e-4\n        self.T_max = int(30000 / self.train_bs * self.epochs) + 50\n        self.T_0 = 25\n        self.warmup_epochs = 10\n        self.wd = 1e-6\n        self.n_accumulate = max(1, 32 // self.train_bs)\n        self.num_classes = 1\n        self.device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n        self.backbone = \"swinunet\"\n        self.model_name = \"swinunet\"\n        self.model_urls = {\n            \"swinv2_tiny_window16_256\": \"../input/swinv2w/swinv2_tiny_patch4_window8_256.pth\",\n            \"swinv2_small_window8_256\": \"../input/swinv2w/swinv2_small_patch4_window8_256.pth\",\n            \"swinv2_small_window16_256\": \"../input/swinv2w/swinv2_small_patch4_window16_256.pth\",\n            \"swinv2_base_window16_256\": \"../input/swinv2w/swinv2_base_patch4_window16_256.pth\",\n        }\n\n        self.size = \"swinv2_base_window16_256\"\n\n        self.load_best_model = False\n\n        self.train_dataset = \"hap\"  # only all, hap, hubmap\n\n        self.dice_dataset = \"hap\"  # only all, hap, hubmap\n\n        self.only_dice = 0\n\n    def display(self):\n        print(f\"{self.exp_name}\")\n        print(f\"debug is {self.debug}\")\n        print(f\"seed is {self.seed}\")\n        print(f\"train_bs is {self.train_bs}\")\n        print(f\"img_size is {self.img_size}\")\n        print(f\"fold_no is {self.fold}\")\n        print(f\"backbone is {self.backbone}\")\n        print(f\"epochs is {self.epochs}\")\n        print(f\"Pretrained: {self.pretrained_path}\")","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:14.976226Z","iopub.execute_input":"2022-08-05T16:14:14.976590Z","iopub.status.idle":"2022-08-05T16:14:14.988689Z","shell.execute_reply.started":"2022-08-05T16:14:14.976556Z","shell.execute_reply":"2022-08-05T16:14:14.987666Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import glob\nimport os\n\nimport cv2\nimport numpy as np\nimport torch\nfrom skimage import io\nfrom torch.utils.data import Dataset, DataLoader\nimport albumentations as A\n\n\nclass TrainDataset(Dataset):\n    def __init__(self, graph_list=None, cfg=None, transforms=None, mode='train'):\n        self.graph_list = graph_list\n        # remove that one faulty image from train_csv\n        self.mode = mode\n        self.cfg = cfg\n        self.transforms = transforms\n        if cfg.train_dataset == \"hap\":\n            prefix = \"../input/hubmap-2022-256x256/\"\n        elif cfg.train_dataset == \"hubmap\":\n            prefix = \"../hubmap-256x256/\"\n        else:\n            prefix = \"../all_256/\"\n        self.image_paths = [prefix + \"train/\" + i for i in graph_list]\n        self.mask_paths = [prefix + \"masks/\" + i.replace(\"train\", \"mask\") for i in graph_list]\n\n    def __len__(self):\n        return len(self.image_paths)\n\n    def __getitem__(self, idx):\n\n        img = io.imread(self.image_paths[idx])\n        mask = cv2.imread(self.mask_paths[idx], cv2.IMREAD_GRAYSCALE)\n        if self.transforms:\n            data = self.transforms(image=img, mask=mask)\n            img = data['image']\n            mask = data['mask']\n        img = np.transpose(img, (2, 0, 1)) / 255.0\n        return torch.tensor(img), torch.tensor(mask)\n\n\nclass DiceDataset(Dataset):\n    def __init__(self, graph_list=None, cfg=None):\n        self.graph_list = graph_list\n        # remove that one faulty image from train_csv\n        self.cfg = cfg\n        if cfg.train_dataset == \"hap\":\n            prefix = \"../input/hubmap-2022-256x256/\"\n        elif cfg.train_dataset == \"hubmap\":\n            prefix = \"../hubmap-256x256/\"\n        else:\n            prefix = \"../all_256/\"\n        self.image_paths = [prefix + \"train/\" + i for i in graph_list]\n        self.mask_paths = [prefix + \"masks/\" + i.replace(\"train\", \"mask\") for i in graph_list]\n\n    def __len__(self):\n        return len(self.image_paths)\n\n    def __getitem__(self, idx):\n\n        img = io.imread(self.image_paths[idx])\n        mask = cv2.imread(self.mask_paths[idx], cv2.IMREAD_GRAYSCALE)\n        img = np.transpose(img, (2, 0, 1)) / 255.0\n        return torch.tensor(img), torch.tensor(mask)\n\n\ndef get_transforms(train=True, cfg=None):\n    data_transforms = {\n        \"train\": A.Compose([\n            #       A.Resize(*cfg.img_size, interpolation=cv2.INTER_NEAREST),\n            A.HorizontalFlip(p=0.5),\n            #         A.VerticalFlip(p=0.5),\n            A.ShiftScaleRotate(shift_limit=0.0625, scale_limit=0.05, rotate_limit=10, p=0.5),\n            A.OneOf([\n                A.GridDistortion(num_steps=5, distort_limit=0.05, p=1.0),\n                # #             A.OpticalDistortion(distort_limit=0.05, shift_limit=0.05, p=1.0),\n                A.ElasticTransform(alpha=1, sigma=50, alpha_affine=50, p=1.0)\n            ], p=0.25),\n            A.CoarseDropout(max_holes=8, max_height=cfg.img_size[0] // 20, max_width=cfg.img_size[1] // 20,\n                            min_holes=5, fill_value=0, mask_fill_value=0, p=0.5),\n        ], p=1.0),\n\n        \"valid\": A.Compose([\n            #             A.Resize(*cfg.img_size, interpolation=cv2.INTER_NEAREST),\n        ], p=1.0)\n    }\n    if train == True:\n        return data_transforms[\"train\"]\n    else:\n        return data_transforms['valid']\n\n\ndef prepare_train_loaders(fold, df, cfg, debug=False):\n    train_list = df.query(\"fold!=@fold\").reset_index(drop=True)[\"graph_name\"].values\n    valid_list = df.query(\"fold==@fold\").reset_index(drop=True)[\"graph_name\"].values\n\n    if debug:\n        train_list = train_list[:20]\n        valid_list = valid_list[:20]\n\n    train_dataset = TrainDataset(train_list, transforms=get_transforms(train=True, cfg=cfg), cfg=cfg, mode='train')\n    valid_dataset = TrainDataset(valid_list, transforms=get_transforms(train=False, cfg=cfg), cfg=cfg, mode='valid')\n\n    #     print(get_statistics(train_dataset))\n\n    train_loader = DataLoader(train_dataset, batch_size=cfg.train_bs if not cfg.debug else 20,\n                              num_workers=0, shuffle=True, pin_memory=True, drop_last=False)\n    valid_loader = DataLoader(valid_dataset, batch_size=cfg.valid_bs if not cfg.debug else 20,\n                              num_workers=0, shuffle=True, pin_memory=True)\n\n    return train_loader, valid_loader\n\ndef prepare_valid_loaders(cfg):\n    if cfg.train_dataset == \"hap\":\n        prefix = \"../input/hubmap-2022-256x256/\"\n    elif cfg.train_dataset == \"hubmap\":\n        prefix = \"../hubmap-256x256/\"\n    else:\n        prefix = \"../all_256/\"\n    dice_graph_path_list = glob.glob(prefix + \"train/*\")\n    dice_graph_name_list = [i[i.rindex(\"/\") + 1:] for i in dice_graph_path_list]\n    dice_dataset = DiceDataset(dice_graph_name_list, cfg=cfg)\n    dice_loader = DataLoader(dice_dataset,  num_workers=0, shuffle=True, batch_size=1, pin_memory=True)\n    return dice_loader","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.030793Z","iopub.execute_input":"2022-08-05T16:14:15.031133Z","iopub.status.idle":"2022-08-05T16:14:15.057234Z","shell.execute_reply.started":"2022-08-05T16:14:15.031103Z","shell.execute_reply":"2022-08-05T16:14:15.056230Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import numpy as np\nimport torch\nfrom fastai.learner import Metric\nfrom fastai.torch_core import flatten_check\nfrom torch import nn\nimport torch.nn.functional as F\n\nclass DiceScore(nn.Module):\n    def __init__(self, weight=None, size_average=True):\n        super(DiceScore, self).__init__()\n\n    def forward(self, inputs, targets, smooth=1):\n        # comment out if your model contains a sigmoid or equivalent activation layer\n        inputs = F.sigmoid(inputs)\n\n        # flatten label and prediction tensors\n        inputs = inputs.view(-1)\n        targets = targets.view(-1)\n\n        intersection = (inputs * targets).sum()\n        dice = (2. * intersection + smooth) / (inputs.sum() + targets.sum() + smooth)\n\n        return dice\n\n\nclass DiceBCELoss(nn.Module):\n    # Formula Given above.\n    def __init__(self, weight=None, size_average=True):\n        super(DiceBCELoss, self).__init__()\n\n    def forward(self, inputs, targets, smooth=1):\n        # comment out if your model contains a sigmoid or equivalent activation layer\n        #         inputs = nnF.sigmoid(inputs)\n\n        # flatten label and prediction tensors\n\n        inputs = inputs.view(-1)\n        targets = targets.view(-1)\n\n        BCE = F.binary_cross_entropy_with_logits(inputs, targets, reduction='mean')\n\n        inputs = F.sigmoid(inputs)\n        intersection = (inputs * targets).sum()\n        dice_loss = 1 - (2. * intersection + smooth) / (inputs.sum() + targets.sum() + smooth)\n\n        Dice_BCE = BCE + dice_loss\n\n        return Dice_BCE\n\n\nclass Dice_th_pred(Metric):\n    def __init__(self, ths=np.arange(0.1, 0.9, 0.01), axis=1):\n        self.axis = axis\n        self.ths = ths\n        self.reset()\n\n    def reset(self):\n        self.inter = torch.zeros(len(self.ths))\n        self.union = torch.zeros(len(self.ths))\n\n    def accumulate(self, p, t):\n        pred, targ = flatten_check(p, t)\n        for i, th in enumerate(self.ths):\n            p = (pred > th).float()\n            self.inter[i] += (p * targ).float().sum().item()\n            self.union[i] += (p + targ).float().sum().item()\n\n    @property\n    def value(self):\n        dices = torch.where(self.union > 0.0, 2.0 * self.inter / self.union,\n                            torch.zeros_like(self.union))\n        return dices","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.095885Z","iopub.execute_input":"2022-08-05T16:14:15.096173Z","iopub.status.idle":"2022-08-05T16:14:15.112246Z","shell.execute_reply.started":"2022-08-05T16:14:15.096146Z","shell.execute_reply":"2022-08-05T16:14:15.111040Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from itertools import chain\n\nimport numpy as np\nimport torch\nfrom timm.models.layers import to_2tuple, DropPath, trunc_normal_\nfrom torch import nn\nimport torch.utils.checkpoint as checkpoint\nimport torch.nn.functional as F\nfrom torchvision import models\n\nclass Mlp(nn.Module):\n    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):\n        super().__init__()\n        out_features = out_features or in_features\n        hidden_features = hidden_features or in_features\n        self.fc1 = nn.Linear(in_features, hidden_features)\n        self.act = act_layer()\n        self.fc2 = nn.Linear(hidden_features, out_features)\n        self.drop = nn.Dropout(drop)\n\n    def forward(self, x):\n        x = self.fc1(x)\n        x = self.act(x)\n        x = self.drop(x)\n        x = self.fc2(x)\n        x = self.drop(x)\n        return x\n\n\ndef window_partition(x, window_size):\n    \"\"\"\n    Args:\n        x: (B, H, W, C)\n        window_size (int): window size\n    Returns:\n        windows: (num_windows*B, window_size, window_size, C)\n    \"\"\"\n\n    B, H, W, C = x.shape\n    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)\n    windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C)\n    return windows\n\n\ndef window_reverse(windows, window_size, H, W):\n    \"\"\"\n    Args:\n        windows: (num_windows*B, window_size, window_size, C)\n        window_size (int): Window size\n        H (int): Height of image\n        W (int): Width of image\n    Returns:\n        x: (B, H, W, C)\n    \"\"\"\n    B = int(windows.shape[0] / (H * W / window_size / window_size))\n    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)\n    x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)\n    return x\n\n\nclass WindowAttention(nn.Module):\n    r\"\"\" Window based multi-head self attention (W-MSA) module with relative position bias.\n    It supports both of shifted and non-shifted window.\n    Args:\n        dim (int): Number of input channels.\n        window_size (tuple[int]): The height and width of the window.\n        num_heads (int): Number of attention heads.\n        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True\n        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0\n        proj_drop (float, optional): Dropout ratio of output. Default: 0.0\n        pretrained_window_size (tuple[int]): The height and width of the window in pre-training.\n    \"\"\"\n\n    def __init__(self, dim, window_size, num_heads, qkv_bias=True, attn_drop=0., proj_drop=0.,\n                 pretrained_window_size=[0, 0]):\n\n        super().__init__()\n        self.dim = dim\n        self.window_size = window_size  # Wh, Ww\n        self.pretrained_window_size = pretrained_window_size\n        self.num_heads = num_heads\n\n        self.logit_scale = nn.Parameter(torch.log(10 * torch.ones((num_heads, 1, 1))), requires_grad=True)\n\n        # mlp to generate continuous relative position bias\n        self.cpb_mlp = nn.Sequential(nn.Linear(2, 512, bias=True),\n                                     nn.ReLU(inplace=True),\n                                     nn.Linear(512, num_heads, bias=False))\n\n        # get relative_coords_table\n        relative_coords_h = torch.arange(-(self.window_size[0] - 1), self.window_size[0], dtype=torch.float32)\n        relative_coords_w = torch.arange(-(self.window_size[1] - 1), self.window_size[1], dtype=torch.float32)\n        relative_coords_table = torch.stack(\n            torch.meshgrid([relative_coords_h,\n                            relative_coords_w])).permute(1, 2, 0).contiguous().unsqueeze(0)  # 1, 2*Wh-1, 2*Ww-1, 2\n        if pretrained_window_size[0] > 0:\n            relative_coords_table[:, :, :, 0] /= (pretrained_window_size[0] - 1)\n            relative_coords_table[:, :, :, 1] /= (pretrained_window_size[1] - 1)\n        else:\n            relative_coords_table[:, :, :, 0] /= (self.window_size[0] - 1)\n            relative_coords_table[:, :, :, 1] /= (self.window_size[1] - 1)\n        relative_coords_table *= 8  # normalize to -8, 8\n        relative_coords_table = torch.sign(relative_coords_table) * torch.log2(\n            torch.abs(relative_coords_table) + 1.0) / np.log2(8)\n\n        self.register_buffer(\"relative_coords_table\", relative_coords_table)\n\n        # get pair-wise relative position index for each token inside the window\n        coords_h = torch.arange(self.window_size[0])\n        coords_w = torch.arange(self.window_size[1])\n        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww\n        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww\n        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww\n        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2\n        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0\n        relative_coords[:, :, 1] += self.window_size[1] - 1\n        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1\n        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww\n        self.register_buffer(\"relative_position_index\", relative_position_index)\n\n        self.qkv = nn.Linear(dim, dim * 3, bias=False)\n        if qkv_bias:\n            self.q_bias = nn.Parameter(torch.zeros(dim))\n            self.v_bias = nn.Parameter(torch.zeros(dim))\n        else:\n            self.q_bias = None\n            self.v_bias = None\n        self.attn_drop = nn.Dropout(attn_drop)\n        self.proj = nn.Linear(dim, dim)\n        self.proj_drop = nn.Dropout(proj_drop)\n        self.softmax = nn.Softmax(dim=-1)\n\n    def forward(self, x, mask=None):\n        \"\"\"\n        Args:\n            x: input features with shape of (num_windows*B, N, C)\n            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None\n        \"\"\"\n        B_, N, C = x.shape\n        qkv_bias = None\n        if self.q_bias is not None:\n            qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias, requires_grad=False), self.v_bias))\n        qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)\n        qkv = qkv.reshape(B_, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)\n        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)\n\n        # cosine attention\n        attn = (F.normalize(q, dim=-1) @ F.normalize(k, dim=-1).transpose(-2, -1))\n        logit_scale = torch.clamp(self.logit_scale, max=torch.log(torch.tensor(1. / 0.01))).exp()\n        attn = attn * logit_scale\n\n        relative_position_bias_table = self.cpb_mlp(self.relative_coords_table).view(-1, self.num_heads)\n        relative_position_bias = relative_position_bias_table[self.relative_position_index.view(-1)].view(\n            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1)  # Wh*Ww,Wh*Ww,nH\n        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww\n        relative_position_bias = 16 * torch.sigmoid(relative_position_bias)\n        attn = attn + relative_position_bias.unsqueeze(0)\n\n        if mask is not None:\n            nW = mask.shape[0]\n            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)\n            attn = attn.view(-1, self.num_heads, N, N)\n            attn = self.softmax(attn)\n        else:\n            attn = self.softmax(attn)\n\n        attn = self.attn_drop(attn)\n\n        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)\n        x = self.proj(x)\n        x = self.proj_drop(x)\n        return x\n\n    def extra_repr(self) -> str:\n        return f'dim={self.dim}, window_size={self.window_size}, ' \\\n               f'pretrained_window_size={self.pretrained_window_size}, num_heads={self.num_heads}'\n\n    def flops(self, N):\n        # calculate flops for 1 window with token length of N\n        flops = 0\n        # qkv = self.qkv(x)\n        flops += N * self.dim * 3 * self.dim\n        # attn = (q @ k.transpose(-2, -1))\n        flops += self.num_heads * N * (self.dim // self.num_heads) * N\n        #  x = (attn @ v)\n        flops += self.num_heads * N * N * (self.dim // self.num_heads)\n        # x = self.proj(x)\n        flops += N * self.dim * self.dim\n        return flops\n\n\nclass SwinTransformerBlock(nn.Module):\n    r\"\"\" Swin Transformer Block.\n    Args:\n        dim (int): Number of input channels.\n        input_resolution (tuple[int]): Input resulotion.\n        num_heads (int): Number of attention heads.\n        window_size (int): Window size.\n        shift_size (int): Shift size for SW-MSA.\n        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.\n        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True\n        drop (float, optional): Dropout rate. Default: 0.0\n        attn_drop (float, optional): Attention dropout rate. Default: 0.0\n        drop_path (float, optional): Stochastic depth rate. Default: 0.0\n        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU\n        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm\n        pretrained_window_size (int): Window size in pre-training.\n    \"\"\"\n\n    def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0,\n                 mlp_ratio=4., qkv_bias=True, drop=0., attn_drop=0., drop_path=0.,\n                 act_layer=nn.GELU, norm_layer=nn.LayerNorm, pretrained_window_size=0):\n        super().__init__()\n        self.dim = dim\n        self.input_resolution = input_resolution\n        self.num_heads = num_heads\n        self.window_size = window_size\n        self.shift_size = shift_size\n        self.mlp_ratio = mlp_ratio\n        if min(self.input_resolution) <= self.window_size:\n            # if window size is larger than input resolution, we don't partition windows\n            self.shift_size = 0\n            self.window_size = min(self.input_resolution)\n        assert 0 <= self.shift_size < self.window_size, \"shift_size must in 0-window_size\"\n\n        self.norm1 = norm_layer(dim)\n        self.attn = WindowAttention(\n            dim, window_size=to_2tuple(self.window_size), num_heads=num_heads,\n            qkv_bias=qkv_bias, attn_drop=attn_drop, proj_drop=drop,\n            pretrained_window_size=to_2tuple(pretrained_window_size))\n\n        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()\n        self.norm2 = norm_layer(dim)\n        mlp_hidden_dim = int(dim * mlp_ratio)\n        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)\n\n        if self.shift_size > 0:\n            # calculate attention mask for SW-MSA\n            H, W = self.input_resolution\n            img_mask = torch.zeros((1, H, W, 1))  # 1 H W 1\n            h_slices = (slice(0, -self.window_size),\n                        slice(-self.window_size, -self.shift_size),\n                        slice(-self.shift_size, None))\n            w_slices = (slice(0, -self.window_size),\n                        slice(-self.window_size, -self.shift_size),\n                        slice(-self.shift_size, None))\n            cnt = 0\n            for h in h_slices:\n                for w in w_slices:\n                    img_mask[:, h, w, :] = cnt\n                    cnt += 1\n\n            mask_windows = window_partition(img_mask, self.window_size)  # nW, window_size, window_size, 1\n            mask_windows = mask_windows.view(-1, self.window_size * self.window_size)\n            attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)\n            attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))\n        else:\n            attn_mask = None\n\n        self.register_buffer(\"attn_mask\", attn_mask)\n\n    def forward(self, x):\n        H, W = self.input_resolution\n        B, L, C = x.shape\n        assert L == H * W, \"input feature has wrong size\"\n\n        shortcut = x\n        x = x.view(B, H, W, C)\n\n        # cyclic shift\n        if self.shift_size > 0:\n            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))\n        else:\n            shifted_x = x\n\n        # partition windows\n        x_windows = window_partition(shifted_x, self.window_size)  # nW*B, window_size, window_size, C\n        x_windows = x_windows.view(-1, self.window_size * self.window_size, C)  # nW*B, window_size*window_size, C\n\n        # W-MSA/SW-MSA\n        attn_windows = self.attn(x_windows, mask=self.attn_mask)  # nW*B, window_size*window_size, C\n\n        # merge windows\n        attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C)\n        shifted_x = window_reverse(attn_windows, self.window_size, H, W)  # B H' W' C\n\n        # reverse cyclic shift\n        if self.shift_size > 0:\n            x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))\n        else:\n            x = shifted_x\n        x = x.view(B, H * W, C)\n        x = shortcut + self.drop_path(self.norm1(x))\n\n        # FFN\n        x = x + self.drop_path(self.norm2(self.mlp(x)))\n\n        return x\n\n    def extra_repr(self) -> str:\n        return f\"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, \" \\\n               f\"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}\"\n\n    def flops(self):\n        flops = 0\n        H, W = self.input_resolution\n        # norm1\n        flops += self.dim * H * W\n        # W-MSA/SW-MSA\n        nW = H * W / self.window_size / self.window_size\n        flops += nW * self.attn.flops(self.window_size * self.window_size)\n        # mlp\n        flops += 2 * H * W * self.dim * self.dim * self.mlp_ratio\n        # norm2\n        flops += self.dim * H * W\n        return flops\n\n\nclass PatchMerging(nn.Module):\n    r\"\"\" Patch Merging Layer.\n    Args:\n        input_resolution (tuple[int]): Resolution of input feature.\n        dim (int): Number of input channels.\n        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm\n    \"\"\"\n\n    def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm):\n        super().__init__()\n        self.input_resolution = input_resolution\n        self.dim = dim\n        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)\n        self.norm = norm_layer(2 * dim)\n\n    def forward(self, x):\n        \"\"\"\n        x: B, H*W, C\n        \"\"\"\n        H, W = self.input_resolution\n        B, L, C = x.shape\n        assert L == H * W, \"input feature has wrong size\"\n        assert H % 2 == 0 and W % 2 == 0, f\"x size ({H}*{W}) are not even.\"\n\n        x = x.view(B, H, W, C)\n\n        x0 = x[:, 0::2, 0::2, :]  # B H/2 W/2 C\n        x1 = x[:, 1::2, 0::2, :]  # B H/2 W/2 C\n        x2 = x[:, 0::2, 1::2, :]  # B H/2 W/2 C\n        x3 = x[:, 1::2, 1::2, :]  # B H/2 W/2 C\n        x = torch.cat([x0, x1, x2, x3], -1)  # B H/2 W/2 4*C\n        x = x.view(B, -1, 4 * C)  # B H/2*W/2 4*C\n\n        x = self.reduction(x)\n        x = self.norm(x)\n\n        return x\n\n    def extra_repr(self) -> str:\n        return f\"input_resolution={self.input_resolution}, dim={self.dim}\"\n\n    def flops(self):\n        H, W = self.input_resolution\n        flops = (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim\n        flops += H * W * self.dim // 2\n        return flops\n\n\nclass BasicLayer(nn.Module):\n    \"\"\" A basic Swin Transformer layer for one stage.\n    Args:\n        dim (int): Number of input channels.\n        input_resolution (tuple[int]): Input resolution.\n        depth (int): Number of blocks.\n        num_heads (int): Number of attention heads.\n        window_size (int): Local window size.\n        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.\n        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True\n        drop (float, optional): Dropout rate. Default: 0.0\n        attn_drop (float, optional): Attention dropout rate. Default: 0.0\n        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0\n        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm\n        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None\n        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.\n        pretrained_window_size (int): Local window size in pre-training.\n    \"\"\"\n\n    def __init__(self, dim, input_resolution, depth, num_heads, window_size,\n                 mlp_ratio=4., qkv_bias=True, drop=0., attn_drop=0.,\n                 drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False,\n                 pretrained_window_size=0):\n\n        super().__init__()\n        self.dim = dim\n        self.input_resolution = input_resolution\n        self.depth = depth\n        self.use_checkpoint = use_checkpoint\n\n        # build blocks\n        self.blocks = nn.ModuleList([\n            SwinTransformerBlock(dim=dim, input_resolution=input_resolution,\n                                 num_heads=num_heads, window_size=window_size,\n                                 shift_size=0 if (i % 2 == 0) else window_size // 2,\n                                 mlp_ratio=mlp_ratio,\n                                 qkv_bias=qkv_bias,\n                                 drop=drop, attn_drop=attn_drop,\n                                 drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,\n                                 norm_layer=norm_layer,\n                                 pretrained_window_size=pretrained_window_size)\n            for i in range(depth)])\n\n        # patch merging layer\n        if downsample is not None:\n            self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer)\n        else:\n            self.downsample = None\n\n    def forward(self, x):\n        for blk in self.blocks:\n            if self.use_checkpoint:\n                x = checkpoint.checkpoint(blk, x)\n            else:\n                x = blk(x)\n        if self.downsample is not None:\n            x = self.downsample(x)\n        return x\n\n    def extra_repr(self) -> str:\n        return f\"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}\"\n\n    def flops(self):\n        flops = 0\n        for blk in self.blocks:\n            flops += blk.flops()\n        if self.downsample is not None:\n            flops += self.downsample.flops()\n        return flops\n\n    def _init_respostnorm(self):\n        for blk in self.blocks:\n            nn.init.constant_(blk.norm1.bias, 0)\n            nn.init.constant_(blk.norm1.weight, 0)\n            nn.init.constant_(blk.norm2.bias, 0)\n            nn.init.constant_(blk.norm2.weight, 0)\n\n\nclass PatchEmbed(nn.Module):\n    r\"\"\" Image to Patch Embedding\n    Args:\n        img_size (int): Image size.  Default: 224.\n        patch_size (int): Patch token size. Default: 4.\n        in_chans (int): Number of input image channels. Default: 3.\n        embed_dim (int): Number of linear projection output channels. Default: 96.\n        norm_layer (nn.Module, optional): Normalization layer. Default: None\n    \"\"\"\n\n    def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):\n        super().__init__()\n        img_size = to_2tuple(img_size)\n        patch_size = to_2tuple(patch_size)\n        patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]]\n        self.img_size = img_size\n        self.patch_size = patch_size\n        self.patches_resolution = patches_resolution\n        self.num_patches = patches_resolution[0] * patches_resolution[1]\n\n        self.in_chans = in_chans\n        self.embed_dim = embed_dim\n\n        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)\n        if norm_layer is not None:\n            self.norm = norm_layer(embed_dim)\n        else:\n            self.norm = None\n\n    def forward(self, x):\n        B, C, H, W = x.shape\n        # FIXME look at relaxing size constraints\n        assert H == self.img_size[0] and W == self.img_size[1], \\\n            f\"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]}).\"\n        x = self.proj(x).flatten(2).transpose(1, 2)  # B Ph*Pw C\n        if self.norm is not None:\n            x = self.norm(x)\n        return x\n\n    def flops(self):\n        Ho, Wo = self.patches_resolution\n        flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1])\n        if self.norm is not None:\n            flops += Ho * Wo * self.embed_dim\n        return flops\n\n\nclass SwinTransformerV2(nn.Module):\n    r\"\"\" Swin Transformer\n        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -\n          https://arxiv.org/pdf/2103.14030\n    Args:\n        img_size (int | tuple(int)): Input image size. Default 224\n        patch_size (int | tuple(int)): Patch size. Default: 4\n        in_chans (int): Number of input image channels. Default: 3\n        num_classes (int): Number of classes for classification head. Default: 1000\n        embed_dim (int): Patch embedding dimension. Default: 96\n        depths (tuple(int)): Depth of each Swin Transformer layer.\n        num_heads (tuple(int)): Number of attention heads in different layers.\n        window_size (int): Window size. Default: 7\n        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4\n        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True\n        drop_rate (float): Dropout rate. Default: 0\n        attn_drop_rate (float): Attention dropout rate. Default: 0\n        drop_path_rate (float): Stochastic depth rate. Default: 0.1\n        norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm.\n        ape (bool): If True, add absolute position embedding to the patch embedding. Default: False\n        patch_norm (bool): If True, add normalization after patch embedding. Default: True\n        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False\n        pretrained_window_sizes (tuple(int)): Pretrained window sizes of each layer.\n    \"\"\"\n\n    def __init__(self, img_size=224, patch_size=4, in_chans=3, num_classes=1000,\n                 embed_dim=96, depths=[2, 2, 6, 2], num_heads=[3, 6, 12, 24],\n                 window_size=7, mlp_ratio=4., qkv_bias=True,\n                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1,\n                 norm_layer=nn.LayerNorm, ape=False, patch_norm=True,\n                 use_checkpoint=False, pretrained_window_sizes=[0, 0, 0, 0], **kwargs):\n        super().__init__()\n\n        self.num_classes = num_classes\n        self.num_layers = len(depths)\n        self.embed_dim = embed_dim\n        self.ape = ape\n        self.patch_norm = patch_norm\n        self.num_features = int(embed_dim * 2 ** (self.num_layers - 1))\n        self.mlp_ratio = mlp_ratio\n\n        # split image into non-overlapping patches\n        self.patch_embed = PatchEmbed(\n            img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim,\n            norm_layer=norm_layer if self.patch_norm else None)\n        num_patches = self.patch_embed.num_patches\n        patches_resolution = self.patch_embed.patches_resolution\n        self.patches_resolution = patches_resolution\n\n        # absolute position embedding\n        if self.ape:\n            self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim))\n            trunc_normal_(self.absolute_pos_embed, std=.02)\n\n        self.pos_drop = nn.Dropout(p=drop_rate)\n\n        # stochastic depth\n        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule\n\n        # build layers\n        self.layers = nn.ModuleList()\n        for i_layer in range(self.num_layers):\n            layer = BasicLayer(dim=int(embed_dim * 2 ** i_layer),\n                               input_resolution=(patches_resolution[0] // (2 ** i_layer),\n                                                 patches_resolution[1] // (2 ** i_layer)),\n                               depth=depths[i_layer],\n                               num_heads=num_heads[i_layer],\n                               window_size=window_size,\n                               mlp_ratio=self.mlp_ratio,\n                               qkv_bias=qkv_bias,\n                               drop=drop_rate, attn_drop=attn_drop_rate,\n                               drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],\n                               norm_layer=norm_layer,\n                               downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,\n                               use_checkpoint=use_checkpoint,\n                               pretrained_window_size=pretrained_window_sizes[i_layer])\n            self.layers.append(layer)\n\n        self.norm = norm_layer(self.num_features)\n        self.avgpool = nn.AdaptiveAvgPool1d(1)\n        self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()\n\n        self.apply(self._init_weights)\n        for bly in self.layers:\n            bly._init_respostnorm()\n\n    def _init_weights(self, m):\n        if isinstance(m, nn.Linear):\n            trunc_normal_(m.weight, std=.02)\n            if isinstance(m, nn.Linear) and m.bias is not None:\n                nn.init.constant_(m.bias, 0)\n        elif isinstance(m, nn.LayerNorm):\n            nn.init.constant_(m.bias, 0)\n            nn.init.constant_(m.weight, 1.0)\n\n    @torch.jit.ignore\n    def no_weight_decay(self):\n        return {'absolute_pos_embed'}\n\n    @torch.jit.ignore\n    def no_weight_decay_keywords(self):\n        return {\"cpb_mlp\", \"logit_scale\", 'relative_position_bias_table'}\n\n    def forward_features(self, x):\n        x = self.patch_embed(x)\n        if self.ape:\n            x = x + self.absolute_pos_embed\n        x = self.pos_drop(x)\n\n        for layer in self.layers:\n            x = layer(x)\n\n        x = self.norm(x)  # B L C\n        x = self.avgpool(x.transpose(1, 2))  # B C 1\n        x = torch.flatten(x, 1)\n        return x\n\n    def extra_features(self, x):\n        x = self.patch_embed(x)\n        if self.ape:\n            x = x + self.absolute_pos_embed\n        x = self.pos_drop(x)\n        feature = []\n\n        for layer in self.layers:\n            x = layer(x)\n            bs, n, f = x.shape\n            h = int(n ** 0.5)\n\n            feature.append(x.view(-1, h, h, f).permute(0, 3, 1, 2).contiguous())\n        return feature\n\n    def get_unet_feature(self, x):\n        x = self.patch_embed(x)\n        if self.ape:\n            x = x + self.absolute_pos_embed\n        x = self.pos_drop(x)\n        bs, n, f = x.shape\n        h = int(n ** 0.5)\n        feature = [x.view(-1, h, h, f).permute(0, 3, 1, 2).contiguous()]\n\n        for layer in self.layers:\n            x = layer(x)\n            bs, n, f = x.shape\n            h = int(n ** 0.5)\n\n            feature.append(x.view(-1, h, h, f).permute(0, 3, 1, 2).contiguous())\n        return feature\n\n    def forward(self, x):\n        x = self.forward_features(x)\n        x = self.head(x)\n        return x\n\n    def flops(self):\n        flops = 0\n        flops += self.patch_embed.flops()\n        for i, layer in enumerate(self.layers):\n            flops += layer.flops()\n        flops += self.num_features * self.patches_resolution[0] * self.patches_resolution[1] // (2 ** self.num_layers)\n        flops += self.num_features * self.num_classes\n        return flops\n\n\ndef swin_v2(size, img_size=256, in_22k=False, config=None,**kwargs):\n    if size == \"swinv2_tiny_window16_256\":\n        model = SwinTransformerV2(img_size=img_size, window_size=16, embed_dim=96, depths=[2, 2, 6, 2],\n                                  num_heads=[3, 6, 12, 24], **kwargs)\n        checkpoint = torch.load(config.model_urls[size])[\"model\"]\n        if img_size != 256:\n            del checkpoint[\"layers.0.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.0.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.0.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.0.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.0.blocks.1.attn.relative_position_index\"]\n            del checkpoint[\"layers.1.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.1.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.1.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.1.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.1.blocks.1.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.1.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.2.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.2.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.3.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.3.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.3.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.4.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.4.attn.relative_position_index\"]\n            del checkpoint[\"layers.2.blocks.5.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.5.attn.relative_coords_table\"]\n            del checkpoint[\"layers.2.blocks.5.attn.relative_position_index\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_position_index\"]\n        model.load_state_dict(checkpoint, strict=False)\n    elif size == \"swinv2_small_window8_256\":\n        model = SwinTransformerV2(img_size=img_size, window_size=8, embed_dim=96, depths=[2, 2, 18, 2],\n                                  num_heads=[3, 6, 12, 24], **kwargs)\n        checkpoint = torch.load(config.model_urls[size])[\"model\"]\n        if img_size != 256:\n            del checkpoint[\"layers.0.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.1.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.3.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.5.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.7.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.9.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.11.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.13.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.15.attn_mask\"]\n            del checkpoint[\"layers.2.blocks.17.attn_mask\"]\n\n        model.load_state_dict(checkpoint, strict=False)\n    elif size == \"swinv2_small_window16_256\":\n        model = SwinTransformerV2(img_size=img_size, window_size=16, embed_dim=96, depths=[2, 2, 18, 2],\n                                  num_heads=[3, 6, 12, 24], **kwargs)\n        checkpoint = torch.load(config.model_urls[size])[\"model\"]\n        if img_size != 256:\n            del checkpoint[\"layers.0.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.1.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_position_index\"]\n        model.load_state_dict(checkpoint, strict=False)\n    elif size == \"swinv2_base_window16_256\":\n        model = SwinTransformerV2(img_size=img_size, window_size=16, embed_dim=128, depths=[2, 2, 18, 2],\n                                  num_heads=[4, 8, 16, 32], **kwargs)\n        checkpoint = torch.load(config.model_urls[size])[\"model\"]\n        if img_size != 256:\n            del checkpoint[\"layers.0.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.1.blocks.1.attn_mask\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.0.attn.relative_position_index\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_coords_table\"]\n            del checkpoint[\"layers.3.blocks.1.attn.relative_position_index\"]\n        model.load_state_dict(checkpoint, strict=False)\n\n    return model\n\n\nclass PSPModule(nn.Module):\n    # In the original inmplementation they use precise RoI pooling\n    # Instead of using adaptative average pooling\n    def __init__(self, in_channels, bin_sizes=[1, 2, 4, 6]):\n        super(PSPModule, self).__init__()\n        out_channels = in_channels // len(bin_sizes)\n        self.stages = nn.ModuleList([self._make_stages(in_channels, out_channels, b_s)\n                                     for b_s in bin_sizes])\n        self.bottleneck = nn.Sequential(\n            nn.Conv2d(in_channels + (out_channels * len(bin_sizes)), in_channels,\n                      kernel_size=3, padding=1, bias=False),\n            nn.BatchNorm2d(in_channels),\n            nn.ReLU(inplace=True),\n            nn.Dropout2d(0.1)\n        )\n\n    def _make_stages(self, in_channels, out_channels, bin_sz):\n        prior = nn.AdaptiveAvgPool2d(output_size=bin_sz)\n        conv = nn.Conv2d(in_channels, out_channels, kernel_size=1, bias=False)\n        bn = nn.BatchNorm2d(out_channels)\n        relu = nn.ReLU(inplace=True)\n        return nn.Sequential(prior, conv, bn, relu)\n\n    def forward(self, features):\n        h, w = features.size()[2], features.size()[3]\n        pyramids = [features]\n        pyramids.extend([F.interpolate(stage(features), size=(h, w), mode='bilinear',\n                                       align_corners=True) for stage in self.stages])\n        output = self.bottleneck(torch.cat(pyramids, dim=1))\n        return output\n\n\nclass ResNet(nn.Module):\n    def __init__(self, in_channels=3, output_stride=16, backbone='resnet101', pretrained=True):\n        super(ResNet, self).__init__()\n        model = getattr(models, backbone)(pretrained)\n        if not pretrained or in_channels != 3:\n            self.initial = nn.Sequential(\n                nn.Conv2d(in_channels, 64, 7, stride=2, padding=3, bias=False),\n                nn.BatchNorm2d(64),\n                nn.ReLU(inplace=True),\n                nn.MaxPool2d(kernel_size=3, stride=2, padding=1)\n            )\n            # initialize_weights(self.initial)\n        else:\n            self.initial = nn.Sequential(*list(model.children())[:4])\n\n        self.layer1 = model.layer1\n        self.layer2 = model.layer2\n        self.layer3 = model.layer3\n        self.layer4 = model.layer4\n\n        if output_stride == 16:\n            s3, s4, d3, d4 = (2, 1, 1, 2)\n        elif output_stride == 8:\n            s3, s4, d3, d4 = (1, 1, 2, 4)\n\n        if output_stride == 8:\n            for n, m in self.layer3.named_modules():\n                if 'conv1' in n and (backbone == 'resnet34' or backbone == 'resnet18'):\n                    m.dilation, m.padding, m.stride = (d3, d3), (d3, d3), (s3, s3)\n                elif 'conv2' in n:\n                    m.dilation, m.padding, m.stride = (d3, d3), (d3, d3), (s3, s3)\n                elif 'downsample.0' in n:\n                    m.stride = (s3, s3)\n\n        for n, m in self.layer4.named_modules():\n            if 'conv1' in n and (backbone == 'resnet34' or backbone == 'resnet18'):\n                m.dilation, m.padding, m.stride = (d4, d4), (d4, d4), (s4, s4)\n            elif 'conv2' in n:\n                m.dilation, m.padding, m.stride = (d4, d4), (d4, d4), (s4, s4)\n            elif 'downsample.0' in n:\n                m.stride = (s4, s4)\n\n    def forward(self, x):\n        x = self.initial(x)\n        x1 = self.layer1(x)\n        print(\"x1\", x1.shape)\n        x2 = self.layer2(x1)\n        print(\"x2\", x2.shape)\n        x3 = self.layer3(x2)\n        print(\"x3\", x3.shape)\n        x4 = self.layer4(x3)\n        print(\"x4\", x4.shape)\n\n        return [x1, x2, x3, x4]\n\n\ndef up_and_add(x, y):\n    return F.interpolate(x, size=(y.size(2), y.size(3)), mode='bilinear', align_corners=True) + y\n\n\nclass FPN_fuse(nn.Module):\n    def __init__(self, feature_channels=[256, 512, 1024, 2048], fpn_out=256):\n        super(FPN_fuse, self).__init__()\n        assert feature_channels[0] == fpn_out\n        self.conv1x1 = nn.ModuleList([nn.Conv2d(ft_size, fpn_out, kernel_size=1)\n                                      for ft_size in feature_channels[1:]])\n        self.smooth_conv = nn.ModuleList([nn.Conv2d(fpn_out, fpn_out, kernel_size=3, padding=1)]\n                                         * (len(feature_channels) - 1))\n        self.conv_fusion = nn.Sequential(\n            nn.Conv2d(len(feature_channels) * fpn_out, fpn_out, kernel_size=3, padding=1, bias=False),\n            nn.BatchNorm2d(fpn_out),\n            nn.ReLU(inplace=True)\n        )\n\n    def forward(self, features):\n        features[1:] = [conv1x1(feature) for feature, conv1x1 in zip(features[1:], self.conv1x1)]  ##\n        P = [up_and_add(features[i], features[i - 1]) for i in reversed(range(1, len(features)))]\n        P = [smooth_conv(x) for smooth_conv, x in zip(self.smooth_conv, P)]\n        P = list(reversed(P))\n        P.append(features[-1])  # P = [P1, P2, P3, P4]\n        H, W = P[0].size(2), P[0].size(3)\n        P[1:] = [F.interpolate(feature, size=(H, W), mode='bilinear', align_corners=True) for feature in P[1:]]\n\n        x = self.conv_fusion(torch.cat((P), dim=1))\n        return x\n\n\nclass UperNet_swin(nn.Module):\n    # Implementing only the object path\n    def __init__(self, size=\"swinv2_small_window16_256\", config=None, img_size=256, num_classes=1, in_channels=3, pretrained=True):\n        super(UperNet_swin, self).__init__()\n\n        self.backbone = swin_v2(size=size, img_size=img_size, config=config)\n        if size.split(\"_\")[1] in [\"small\", \"tiny\"]:\n            feature_channels = [192, 384, 768, 768]\n        elif size.split(\"_\")[1] in [\"base\"]:\n            feature_channels = [256, 512, 1024, 1024]\n        self.PPN = PSPModule(feature_channels[-1])\n        self.FPN = FPN_fuse(feature_channels, fpn_out=feature_channels[0])\n        self.head = nn.Conv2d(feature_channels[0], num_classes, kernel_size=3, padding=1)\n\n    def forward(self, x):\n        input_size = (x.size()[2], x.size()[3])\n\n        features = self.backbone.extra_features(x)\n        features[-1] = self.PPN(features[-1])\n        x = self.head(self.FPN(features))\n\n        x = F.interpolate(x, size=input_size, mode='bilinear')\n        return x\n\n    def get_backbone_params(self):\n        return self.backbone.parameters()\n\n    def get_decoder_params(self):\n        return chain(self.PPN.parameters(), self.FPN.parameters(), self.head.parameters())\n\n    def freeze_bn(self):\n        for module in self.modules():\n            if isinstance(module, nn.BatchNorm2d): module.eval()\n\n\nfrom segmentation_models_pytorch.base import modules as md\n\nfrom timm.models.layers.cbam import *\n\n\nclass DecoderBlock(nn.Module):\n    def __init__(\n            self,\n            in_channels,\n            skip_channels,\n            out_channels,\n            use_batchnorm=True,\n            attention_type=None,\n    ):\n        super().__init__()\n        self.conv1 = md.Conv2dReLU(\n            in_channels + skip_channels,\n            out_channels,\n            kernel_size=3,\n            padding=1,\n            use_batchnorm=use_batchnorm,\n        )\n        if attention_type == \"cbam\":\n            self.attention1 = CbamModule(channels=in_channels + skip_channels)\n        else:\n            self.attention1 = md.Attention(attention_type, in_channels=in_channels + skip_channels)\n        self.conv2 = md.Conv2dReLU(\n            out_channels,\n            out_channels,\n            kernel_size=3,\n            padding=1,\n            use_batchnorm=use_batchnorm,\n        )\n        if attention_type == \"cbam\":\n            self.attention2 = CbamModule(channels=out_channels)\n        else:\n            self.attention2 = md.Attention(attention_type, in_channels=out_channels)\n        self.in_channels = in_channels\n        self.out_channels = out_channels\n        self.skip_channels = skip_channels\n\n    def forward(self, x, skip=None):\n        if skip is None:\n            x = F.interpolate(x, scale_factor=2, mode=\"nearest\")\n        else:\n            if x.shape[-1] != skip.shape[-1]:\n                x = F.interpolate(x, scale_factor=2, mode=\"nearest\")\n        if skip is not None:\n            # print(x.shape,skip.shape)\n            x = torch.cat([x, skip], dim=1)\n            x = self.attention1(x)\n        x = self.conv1(x)\n        x = self.conv2(x)\n        x = self.attention2(x)\n        return x\n\n\nclass CenterBlock(nn.Sequential):\n    def __init__(self, in_channels, out_channels, use_batchnorm=True):\n        conv1 = md.Conv2dReLU(\n            in_channels,\n            out_channels,\n            kernel_size=3,\n            padding=1,\n            use_batchnorm=use_batchnorm,\n        )\n        conv2 = md.Conv2dReLU(\n            out_channels,\n            out_channels,\n            kernel_size=3,\n            padding=1,\n            use_batchnorm=use_batchnorm,\n        )\n        super().__init__(conv1, conv2)\n\n\nclass UnetDecoder(nn.Module):\n    def __init__(\n            self,\n            encoder_channels,\n            decoder_channels,\n            n_blocks=5,\n            use_batchnorm=True,\n            attention_type=None,\n            center=False,\n    ):\n        super().__init__()\n\n        if n_blocks != len(decoder_channels):\n            raise ValueError(\n                \"Model depth is {}, but you provide `decoder_channels` for {} blocks.\".format(\n                    n_blocks, len(decoder_channels)\n                )\n            )\n\n        # remove first skip with same spatial resolution\n        encoder_channels = encoder_channels[1:]\n        # reverse channels to start from head of encoder\n        encoder_channels = encoder_channels[::-1]\n\n        # computing blocks input and output channels\n        head_channels = encoder_channels[0]\n        in_channels = [head_channels] + list(decoder_channels[:-1])\n        skip_channels = list(encoder_channels[1:]) + [0]\n\n        out_channels = decoder_channels\n\n        if center:\n            self.center = CenterBlock(head_channels, head_channels, use_batchnorm=use_batchnorm)\n        else:\n            self.center = nn.Identity()\n\n        # combine decoder keyword arguments\n        kwargs = dict(use_batchnorm=use_batchnorm, attention_type=attention_type)\n        blocks = [\n            DecoderBlock(in_ch, skip_ch, out_ch, **kwargs)\n            for in_ch, skip_ch, out_ch in zip(in_channels, skip_channels, out_channels)\n        ]\n        self.blocks = nn.ModuleList(blocks)\n\n    def forward(self, *features):\n\n        features = features[1:]  # remove first skip with same spatial resolution\n        features = features[::-1]  # reverse channels to start from head of encoder\n\n        head = features[0]\n        skips = features[1:]\n\n        x = self.center(head)\n        for i, decoder_block in enumerate(self.blocks):\n            skip = skips[i] if i < len(skips) else None\n            x = decoder_block(x, skip)\n            # y_i = self.upsample1(y_i)\n        # hypercol = torch.cat([y0,y1,y2,y3,y4], dim=1)\n\n        return x\n\n\nclass SegmentationHead(nn.Sequential):\n    def __init__(self, in_channels, out_channels, kernel_size=3, upsampling=1):\n        conv2d = nn.Conv2d(in_channels, out_channels, kernel_size=kernel_size, padding=kernel_size // 2)\n        upsampling = nn.UpsamplingBilinear2d(scale_factor=upsampling) if upsampling > 1 else nn.Identity()\n        super().__init__(conv2d, upsampling)\n\n\nclass unet_swin(nn.Module):\n\n    def __init__(\n            self, config, size=\"small\", img_size=256  # \"base\" \"large\"\n    ):\n        super().__init__()\n\n        self.encoder = swin_v2(size=size, img_size=img_size, config=config)\n\n        if size.split(\"_\")[1] in [\"small\", \"tiny\"]:\n            feature_channels = (3, 192, 384, 768, 768)\n        elif size.split(\"_\")[1] in [\"base\"]:\n            feature_channels = (3, 256, 512, 1024, 1024)\n        self.decoder = UnetDecoder(encoder_channels=feature_channels, n_blocks=4, decoder_channels=(512, 256, 128, 64),\n                                   attention_type=None)\n\n        self.segmentation_head = SegmentationHead(in_channels=64, out_channels=1, kernel_size=3, upsampling=4\n                                                  )\n\n    def forward(self, input):\n        encoder_featrue = self.encoder.get_unet_feature(input)\n        decoder_output = self.decoder(*encoder_featrue)\n        masks = self.segmentation_head(decoder_output)\n\n        return masks","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.275966Z","iopub.execute_input":"2022-08-05T16:14:15.277007Z","iopub.status.idle":"2022-08-05T16:14:15.628900Z","shell.execute_reply.started":"2022-08-05T16:14:15.276964Z","shell.execute_reply":"2022-08-05T16:14:15.627679Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import gc\nimport time\n\nimport cv2\nfrom torch.cuda import amp\nimport torch\nfrom torch.optim import lr_scheduler\nfrom tqdm import tqdm\n\n@torch.no_grad()\ndef valid_one_epoch(model, dataloader, device, epoch, optimizer, cfg):\n    model.eval()\n\n    total_val_loss = 0\n    total_val_score = 0\n\n    pbar = tqdm(enumerate(dataloader), total=len(dataloader), desc='Valid ')\n\n    for step, (images, masks) in pbar:\n        images = images.to(device, dtype=torch.float)\n        masks = masks.to(device, dtype=torch.float)\n\n        y_pred = model(images)\n        criterion = DiceBCELoss()\n        loss = criterion(y_pred, masks)\n        dice_score = DiceScore()(y_pred, masks).detach().item()\n\n        loss = loss.detach().item()\n        total_val_loss += loss\n        total_val_score += dice_score\n\n    print(f'\\nTesting epoch {epoch} ')\n    print(f'Total DiceBCE loss: {total_val_loss / len(dataloader):.4f}')\n    print(f'Total average Dice Score: {total_val_score / len(dataloader):.4f}')\n\n    torch.cuda.empty_cache()\n    gc.collect()\n\n    return total_val_loss / len(dataloader), total_val_score / len(dataloader)\n\n\ndef train_one_epoch(model, optimizer, scheduler, dataloader, device, epoch, cfg):\n    model.train()\n    scaler = amp.GradScaler()\n    total_loss = 0\n    pbar = tqdm(enumerate(dataloader), total=len(dataloader), desc='Train ', leave=False)\n    data_size = 0\n    total_dice_score = 0\n    for step, (images, masks) in pbar:\n        images = images.to(device, dtype=torch.float)\n        masks = masks.to(device, dtype=torch.float)\n\n        batch_size = images.size(0)\n        data_size += batch_size\n\n        with amp.autocast(enabled=True):\n            y_pred = model(images)\n            criterion = DiceBCELoss()\n            loss = criterion(y_pred, masks)\n            dice_score = DiceScore()(y_pred, masks).detach().item()\n\n        scaler.scale(loss / cfg.n_accumulate).backward()\n\n        if ((step + 1) % cfg.n_accumulate == 0 or (step + 1) == len(dataloader)):\n\n            scaler.step(optimizer)\n            scaler.update()\n            # zero the parameter gradients\n            optimizer.zero_grad()\n\n            if scheduler is not None:\n                scheduler.step()\n\n        loss = loss.detach().item()\n        total_loss += loss\n        total_dice_score += dice_score\n\n        pbar.set_postfix(desc=f'Loss={loss:.4f} DiceScore= {dice_score:.4f}  Batch_id={step}')\n\n    print(f'\\nTraining epoch {epoch} ')\n    print(f'Total DiceBCE loss: {total_loss / len(dataloader):.4f}')\n    print(f'Total average Dice Score: {total_dice_score / len(dataloader):.4f}')\n\n    torch.cuda.empty_cache()\n    gc.collect()\n\n    return (total_loss / len(dataloader), total_dice_score / len(dataloader))\n\n\ndef build_model(cfg):\n    model = unet_swin(img_size=256, size=cfg.size, config=cfg)\n    model.to(cfg.device)\n    return model\n\n\ndef load_model(path, cfg=None):\n    model = build_model(cfg)\n    model.load_state_dict(torch.load(path))\n    model.eval()\n    return model\n\n\ndef save_img(data, name, out):\n    data = data.float().cpu().numpy()\n    img = cv2.imencode('.png', (data * 255).astype(np.uint8))[1]\n    out.writestr(name, img)\n\n\ndef fetch_scheduler(optimizer, cfg):\n    if cfg.scheduler == 'CosineAnnealingLR':\n        scheduler = lr_scheduler.CosineAnnealingLR(optimizer, T_max=cfg.T_max,\n                                                   eta_min=cfg.min_lr)\n    elif cfg.scheduler == 'CosineAnnealingWarmRestarts':\n        scheduler = lr_scheduler.CosineAnnealingWarmRestarts(optimizer, T_0=cfg.T_0,\n                                                             eta_min=cfg.min_lr)\n    elif cfg.scheduler == 'ReduceLROnPlateau':\n        scheduler = lr_scheduler.ReduceLROnPlateau(optimizer,\n                                                   mode='min',\n                                                   factor=0.1,\n                                                   patience=7,\n                                                   threshold=0.0001,\n                                                   min_lr=cfg.min_lr, )\n    elif cfg.scheduer == 'ExponentialLR':\n        scheduler = lr_scheduler.ExponentialLR(optimizer, gamma=0.85)\n\n    else:\n        scheduler = None\n\n    return scheduler\n\n\nclass Model_pred:\n    def __init__(self, model, dl, tta: bool = True, half: bool = False, config=None):\n        self.model = model\n        self.dl = dl\n        self.tta = tta\n        self.half = half\n        self.config = config\n\n    def __iter__(self):\n        self.model.eval()\n        name_list = self.dl.dataset.graph_list\n        count = 0\n        with torch.no_grad():\n            for x, y in iter(self.dl):\n                if self.config.device != \"cpu\":\n                    x = x.to(self.config.device)\n                if self.half:\n                    x = x.half()\n                x = x.type(torch.float)\n                p = self.model(x)\n                py = torch.sigmoid(p).detach()\n                if self.tta:\n                    # x,y,xy flips as TTA\n                    flips = [[-1], [-2], [-2, -1]]\n                    for f in flips:\n                        p = self.model(torch.flip(x, f))\n                        p = torch.flip(p, f)\n                        py += torch.sigmoid(p).detach()\n                    py /= (1 + len(flips))\n                if y is not None and len(y.shape) == 4 and py.shape != y.shape:\n                    py = F.upsample(py, size=(y.shape[-2], y.shape[-1]), mode=\"bilinear\")\n                py = py.permute(0, 2, 3, 1).float().cpu()\n                batch_size = len(py)\n                for i in range(batch_size):\n                    taget = y[i].detach().cpu() if y is not None else None\n                    yield py[i], taget, name_list[count]\n                    count += 1\n\n    def __len__(self):\n        return len(self.dl.dataset)\n","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.631383Z","iopub.execute_input":"2022-08-05T16:14:15.632195Z","iopub.status.idle":"2022-08-05T16:14:15.661035Z","shell.execute_reply.started":"2022-08-05T16:14:15.632158Z","shell.execute_reply":"2022-08-05T16:14:15.660014Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def score(weight_path):\n    score_lindex = weight_path.rindex(\"_\") + 1\n    score_rindex = weight_path.rindex(\".\")\n    return float(weight_path[score_lindex:score_rindex])\n\ndef set_seed(seed=42):\n    '''Sets the seed of the entire notebook so results are the same every time we run.\n    This is for REPRODUCIBILITY.'''\n    print(f\"Setting seed as {seed}\")\n    np.random.seed(seed)\n    random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)\n    # When running on the CuDNN backend, two further options must be set\n    torch.backends.cudnn.deterministic = True\n    torch.backends.cudnn.benchmark = False\n    # Set a fixed value for the hash seed\n    os.environ['PYTHONHASHSEED'] = str(seed)\n    print('> SEEDING DONE')\n\n\ndef initialise_config(debug=False, train_bs=4, fold=0, pretrained_path=None):\n    cfg = CFG(fold=fold, train_bs=train_bs, debug=debug, pretrained_path=pretrained_path)\n    set_seed(cfg.seed)\n    return cfg\n\n\ndef create_folds(cfg=None):\n    if cfg.train_dataset == \"hap\":\n        image_name_list = [ i[i.rindex(\"/\"):]for i in glob.glob(\"../input/hubmap-2022-256x256/train/*.png\")]\n    elif cfg.train_dataset == \"hubmap\":\n        image_name_list = [i[i.rindex(\"/\"):] for i in glob.glob(\"../hubmap-256x256/train/*.png\")]\n    else:\n        image_name_list = [i[i.rindex(\"/\"):] for i in glob.glob(\"../all_256/train/*.png\")]\n\n    df = pd.DataFrame({\"graph_name\":image_name_list})\n    skf = KFold(n_splits=cfg.n_fold, shuffle=True, random_state=cfg.seed)\n    for fold, idxes in enumerate(skf.split(range(len(df)))):\n        df.loc[idxes[1], 'fold'] = fold\n    return df\n\n\ndef run_training(model, optimizer, scheduler, device, num_epochs, fold, train_loader, valid_loader, cfg):\n    if device != \"cpu\":\n        print(\"cuda: {}\\n\".format(torch.cuda.get_device_name()))\n\n    start = time.time()\n    best_model_wts = copy.deepcopy(model.state_dict())\n    best_dice = 0\n    history = defaultdict(list)\n\n    if cfg.only_dice != 1:\n        for epoch in range(1, num_epochs + 1):\n\n            print(f'Epoch {epoch}/{num_epochs}', end='')\n            train_loss, train_score = train_one_epoch(model, optimizer, scheduler,\n                                                      dataloader=train_loader,\n                                                      device=cfg.device, epoch=epoch, cfg=cfg)\n\n            val_loss, val_score = valid_one_epoch(model, valid_loader,\n                                                  device=cfg.device,\n                                                  epoch=epoch,\n                                                  optimizer=optimizer, cfg=cfg)\n\n            history['epoch'].append(epoch)\n            history['Train Loss'].append(train_loss)\n            history['Valid Loss'].append(val_loss)\n            history['Valid Scores'].append(val_score)\n\n            print(f'Train Loss: {train_loss} | Valid Loss: {val_loss}')\n            print(f'Train Score: {train_score} | Valid Dice Score: {val_score}')\n\n            # deep copy the model\n            if val_score >= best_dice:\n                os.system(f\"rm models/fold_{fold}/{cfg.size}_*\")\n\n                print(f\"Valid Score Improved ({best_dice:0.4f} ---> {val_score:0.4f})\")\n                best_dice = val_score\n                best_model_wts = copy.deepcopy(model.state_dict())\n                PATH = f\"models/fold_{fold}/{cfg.size}_{val_score:0.4f}.pth\"\n                torch.save(model.state_dict(), PATH)\n\n                print(f\"Model Saved\")\n\n            print()\n            print()\n\n        end = time.time()\n        time_elapsed = end - start\n        print('Training complete in {:.0f}h {:.0f}m {:.0f}s'.format(\n            time_elapsed // 3600, (time_elapsed % 3600) // 60, (time_elapsed % 3600) % 60))\n\n        model.load_state_dict(best_model_wts)\n\n        plt.subplot(1, 2, 1, frameon=False)\n        plt.title(f'fold_{fold}_train_loss')\n        plt.xlabel('Epoch')\n        plt.plot(history['epoch'], history['Train Loss'], \"r\")\n\n        plt.subplot(1, 2, 2, frameon=False)\n        plt.title(f'fold_{fold}_test_dice')\n        plt.xlabel('Epoch')\n        plt.plot(history['epoch'], history['Valid Scores'], \"b\")\n\n        plt.savefig(f\"models/fold_{fold}/metric_fold_{fold}.jpg\")\n        plt.close()\n\n    dice_loader = prepare_valid_loaders(cfg)\n    mp = Model_pred(model, dice_loader, config=cfg)\n    dice = Dice_th_pred(np.arange(0.2, 0.7, 0.01))\n    for p in progress_bar(mp):\n        dice.accumulate(p[0], p[1])\n    # save_img(p[0], p[2], out)\n    gc.collect()\n    dices = dice.value\n    noise_ths = dice.ths\n    best_dice = dices.max()\n    best_thr = noise_ths[dices.argmax()]\n    plt.figure(figsize=(8, 4))\n    plt.plot(noise_ths, dices, color='blue')\n    plt.vlines(x=best_thr, ymin=dices.min(), ymax=dices.max(), colors='black')\n    d = dices.max() - dices.min()\n    plt.text(noise_ths[-1] - 0.1, best_dice - 0.1 * d, f'DICE = {best_dice:.3f}', fontsize=12)\n    plt.text(noise_ths[-1] - 0.1, best_dice - 0.2 * d, f'TH = {best_thr:.3f}', fontsize=12)\n    plt.savefig(f'models/fold_{fold}/save.jpg')\n    plt.close()\n\n    weight_path = glob.glob(f\"models/fold_{fold}/{cfg.size}*.pth\")[0]\n    down_index = weight_path.rindex(\"_\")\n    new_weight_path = weight_path[:down_index] + f\"_{best_thr:.3f}\" + weight_path[down_index:]\n    os.rename(weight_path, new_weight_path)\n\n    return model, history","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.662901Z","iopub.execute_input":"2022-08-05T16:14:15.663507Z","iopub.status.idle":"2022-08-05T16:14:15.689684Z","shell.execute_reply.started":"2022-08-05T16:14:15.663471Z","shell.execute_reply":"2022-08-05T16:14:15.688577Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def main(cfg):\n    cfg.display()\n    print(f'#' * 30)\n    print(f'### Fold: {cfg.fold}')\n    print(f'#' * 30)\n\n    train_loader, valid_loader = prepare_train_loaders(fold=cfg.fold,\n                                                     df=create_folds(cfg),\n                                                     debug=cfg.debug,\n                                                     cfg=cfg)\n    if cfg.load_best_model:\n        models = glob.glob(f\"models/fold_{cfg.fold}/{cfg.size}_*.pth\")\n        models = sorted(models, key=lambda i: score(i), reverse=True)\n        model = load_model(models[0], cfg=cfg).to(cfg.device)\n        print(\"Load Pretrained Model: \" + models[0])\n    elif cfg.pretrained_path is None:\n        model = unet_swin(img_size=256, size=cfg.size, config=cfg).to(cfg.device)\n    else:\n        model = load_model(cfg.pretrained_path, cfg=cfg).to(cfg.device)\n        print(\"Load pretrained Model: \" + cfg.pretrained_path)\n\n    optimizer = optim.Adam(model.parameters(), lr=cfg.lr, weight_decay=cfg.wd)\n    scheduler = fetch_scheduler(optimizer, cfg=cfg)\n    run_training(model, optimizer, scheduler,\n                     device=cfg.device,\n                     num_epochs=cfg.epochs, fold=cfg.fold,\n                     train_loader=train_loader,\n                     valid_loader=valid_loader,\n                     cfg=cfg)","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.691368Z","iopub.execute_input":"2022-08-05T16:14:15.691945Z","iopub.status.idle":"2022-08-05T16:14:15.705612Z","shell.execute_reply.started":"2022-08-05T16:14:15.691858Z","shell.execute_reply":"2022-08-05T16:14:15.704632Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"! mkdir models","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:15.708174Z","iopub.execute_input":"2022-08-05T16:14:15.708822Z","iopub.status.idle":"2022-08-05T16:14:16.759884Z","shell.execute_reply.started":"2022-08-05T16:14:15.708787Z","shell.execute_reply":"2022-08-05T16:14:16.758547Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cfg = initialise_config(train_bs=8, fold=0)\nif not os.path.exists(f\"models/fold_{cfg.fold}\"):\n    os.mkdir(f\"models/fold_{cfg.fold}\")\n\nmain(cfg)","metadata":{"execution":{"iopub.status.busy":"2022-08-05T16:14:16.762064Z","iopub.execute_input":"2022-08-05T16:14:16.762436Z","iopub.status.idle":"2022-08-05T16:26:24.355494Z","shell.execute_reply.started":"2022-08-05T16:14:16.762395Z","shell.execute_reply":"2022-08-05T16:26:24.354100Z"},"trusted":true},"execution_count":null,"outputs":[]}]}