{"cells":[{"metadata":{},"cell_type":"markdown","source":"

If you fork it, please give an upvote!

"},{"metadata":{},"cell_type":"markdown","source":"# About\n* In this notebook, I use ArcFace instead of Linear. .\n* I will made datasets public in the inference section .\n\n# Source Kernels\n* This notebook was written by refering these great kernels below, so please don't forget to **check and upvote** them.\n* [Pytorch Efficientnet Baseline [Train] AMP+Aug](https://www.kaggle.com/khyeh0719/pytorch-efficientnet-baseline-train-amp-aug)\n* [Pytorch Efficientnet Baseline [Inference] TTA](https://www.kaggle.com/khyeh0719/pytorch-efficientnet-baseline-inference-tta)\n\n# Inference Link\n\n* Inference part is here: https://www.kaggle.com/wenlia/arcface-inference-tta"},{"metadata":{"trusted":true},"cell_type":"code","source":"package_paths = [\n '../input/pytorch-image-models/pytorch-image-models-master', #'../input/efficientnet-pytorch-07/efficientnet_pytorch-0.7.0'\n '../input/image-fmix/FMix-master'\n]\nimport sys; \n\nfor pth in package_paths:\n sys.path.append(pth)\n \nfrom fmix import sample_mask, make_low_freq_image, binarise_mask","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"from glob import glob\nfrom sklearn.model_selection import GroupKFold, StratifiedKFold\nimport cv2\nfrom skimage import io\nimport torch\nfrom torch import nn\nimport os\nfrom datetime import datetime\nimport time\nimport random\nimport cv2\nimport math\nimport torchvision\nfrom torchvision import transforms\nimport pandas as pd\nimport numpy as np\nfrom tqdm import tqdm\nfrom torch.nn import Parameter\nimport matplotlib.pyplot as plt\nfrom torch.utils.data import Dataset,DataLoader\nfrom torch.utils.data.sampler import SequentialSampler, RandomSampler\nfrom torch.cuda.amp import autocast, GradScaler\nfrom torch.nn.modules.loss import _WeightedLoss\nimport torch.nn.functional as F\n\nimport timm\n\nimport sklearn\nimport warnings\nimport joblib\nfrom sklearn.metrics import roc_auc_score, log_loss\nfrom sklearn import metrics\nimport warnings\nimport cv2\nimport pydicom\n#from efficientnet_pytorch import EfficientNet\nfrom scipy.ndimage.interpolation import zoom","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"CFG = {\n 'fold_num': 5,\n 'seed': 719,\n 'model_arch': 'tf_efficientnet_b4_ns',\n 'img_size': 512,\n 'epochs': 10,\n 'train_bs': 24,\n 'valid_bs': 24,\n 'T_0': 10,\n 'lr': 1e-4,\n 'min_lr': 1e-6,\n 'weight_decay':1e-6,\n 'num_workers': 4,\n 'accum_iter': 2, # suppoprt to do batch accumulation for backprop with effectively larger batch size\n 'verbose_step': 1,\n 'device': 'cuda:0',\n 'classifier_type': 'arcface', # arcface,linear\n 'cosine_scale': 15,\n 'cosine_margin': 0.4\n}","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"train = pd.read_csv('../input/cassava-leaf-disease-classification/train.csv')\ntrain.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train.label.value_counts()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"> We could do stratified validation split in each fold to make each fold's train and validation set looks like the whole train set in target distributions."},{"metadata":{"trusted":true},"cell_type":"code","source":"submission = pd.read_csv('../input/cassava-leaf-disease-classification/sample_submission.csv')\nsubmission.head()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Helper Functions"},{"metadata":{"trusted":true},"cell_type":"code","source":"def seed_everything(seed):\n random.seed(seed)\n os.environ['PYTHONHASHSEED'] = str(seed)\n np.random.seed(seed)\n torch.manual_seed(seed)\n torch.cuda.manual_seed(seed)\n torch.backends.cudnn.deterministic = True\n torch.backends.cudnn.benchmark = True\n \ndef get_img(path):\n im_bgr = cv2.imread(path)\n im_rgb = im_bgr[:, :, ::-1]\n #print(im_rgb)\n return im_rgb\n\nimg = get_img('../input/cassava-leaf-disease-classification/train_images/1000015157.jpg')\nplt.imshow(img)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Dataset"},{"metadata":{"trusted":true},"cell_type":"code","source":"def rand_bbox(size, lam):\n W = size[0]\n H = size[1]\n cut_rat = np.sqrt(1. - lam)\n cut_w = np.int(W * cut_rat)\n cut_h = np.int(H * cut_rat)\n\n # uniform\n cx = np.random.randint(W)\n cy = np.random.randint(H)\n\n bbx1 = np.clip(cx - cut_w // 2, 0, W)\n bby1 = np.clip(cy - cut_h // 2, 0, H)\n bbx2 = np.clip(cx + cut_w // 2, 0, W)\n bby2 = np.clip(cy + cut_h // 2, 0, H)\n return bbx1, bby1, bbx2, bby2\n\n\nclass CassavaDataset(Dataset):\n def __init__(self, df, data_root, \n transforms=None, \n output_label=True, \n one_hot_label=False,\n do_fmix=False, \n fmix_params={\n 'alpha': 1., \n 'decay_power': 3., \n 'shape': (CFG['img_size'], CFG['img_size']),\n 'max_soft': True, \n 'reformulate': False\n },\n do_cutmix=False,\n cutmix_params={\n 'alpha': 1,\n }\n ):\n \n super().__init__()\n self.df = df.reset_index(drop=True).copy()\n self.transforms = transforms\n self.data_root = data_root\n self.do_fmix = do_fmix\n self.fmix_params = fmix_params\n self.do_cutmix = do_cutmix\n self.cutmix_params = cutmix_params\n \n self.output_label = output_label\n self.one_hot_label = one_hot_label\n \n if output_label == True:\n self.labels = self.df['label'].values\n #print(self.labels)\n \n if one_hot_label is True:\n self.labels = np.eye(self.df['label'].max()+1)[self.labels]\n #print(self.labels)\n \n def __len__(self):\n return self.df.shape[0]\n \n def __getitem__(self, index: int):\n \n # get labels\n if self.output_label:\n target = self.labels[index]\n \n img = get_img(\"{}/{}\".format(self.data_root, self.df.loc[index]['image_id']))\n\n if self.transforms:\n img = self.transforms(image=img)['image']\n \n if self.do_fmix and np.random.uniform(0., 1., size=1)[0] > 0.5:\n with torch.no_grad():\n #lam, mask = sample_mask(**self.fmix_params)\n \n lam = np.clip(np.random.beta(self.fmix_params['alpha'], self.fmix_params['alpha']),0.6,0.7)\n \n # Make mask, get mean / std\n mask = make_low_freq_image(self.fmix_params['decay_power'], self.fmix_params['shape'])\n mask = binarise_mask(mask, lam, self.fmix_params['shape'], self.fmix_params['max_soft'])\n \n fmix_ix = np.random.choice(self.df.index, size=1)[0]\n fmix_img = get_img(\"{}/{}\".format(self.data_root, self.df.iloc[fmix_ix]['image_id']))\n\n if self.transforms:\n fmix_img = self.transforms(image=fmix_img)['image']\n\n mask_torch = torch.from_numpy(mask)\n \n # mix image\n img = mask_torch*img+(1.-mask_torch)*fmix_img\n\n #print(mask.shape)\n\n #assert self.output_label==True and self.one_hot_label==True\n\n # mix target\n rate = mask.sum()/CFG['img_size']/CFG['img_size']\n target = rate*target + (1.-rate)*self.labels[fmix_ix]\n #print(target, mask, img)\n #assert False\n \n if self.do_cutmix and np.random.uniform(0., 1., size=1)[0] > 0.5:\n #print(img.sum(), img.shape)\n with torch.no_grad():\n cmix_ix = np.random.choice(self.df.index, size=1)[0]\n cmix_img = get_img(\"{}/{}\".format(self.data_root, self.df.iloc[cmix_ix]['image_id']))\n if self.transforms:\n cmix_img = self.transforms(image=cmix_img)['image']\n \n lam = np.clip(np.random.beta(self.cutmix_params['alpha'], self.cutmix_params['alpha']),0.3,0.4)\n bbx1, bby1, bbx2, bby2 = rand_bbox((CFG['img_size'], CFG['img_size']), lam)\n\n img[:, bbx1:bbx2, bby1:bby2] = cmix_img[:, bbx1:bbx2, bby1:bby2]\n\n rate = 1 - ((bbx2 - bbx1) * (bby2 - bby1) / (CFG['img_size'] * CFG['img_size']))\n target = rate*target + (1.-rate)*self.labels[cmix_ix]\n \n #print('-', img.sum())\n #print(target)\n #assert False\n \n # do label smoothing\n #print(type(img), type(target))\n if self.output_label == True:\n return img, target\n else:\n return img","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Define Train\\Validation Image Augmentations"},{"metadata":{"trusted":true},"cell_type":"code","source":"from albumentations import (\n HorizontalFlip, VerticalFlip, IAAPerspective, ShiftScaleRotate, CLAHE, RandomRotate90,\n Transpose, ShiftScaleRotate, Blur, OpticalDistortion, GridDistortion, HueSaturationValue,\n IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, MedianBlur, IAAPiecewiseAffine, RandomResizedCrop,\n IAASharpen, IAAEmboss, RandomBrightnessContrast, Flip, OneOf, Compose, Normalize, Cutout, CoarseDropout, ShiftScaleRotate, CenterCrop, Resize\n)\n\nfrom albumentations.pytorch import ToTensorV2\n\ndef get_train_transforms():\n return Compose([\n RandomResizedCrop(CFG['img_size'], CFG['img_size']),\n Transpose(p=0.5),\n HorizontalFlip(p=0.5),\n VerticalFlip(p=0.5),\n ShiftScaleRotate(p=0.5),\n HueSaturationValue(hue_shift_limit=0.2, sat_shift_limit=0.2, val_shift_limit=0.2, p=0.5),\n RandomBrightnessContrast(brightness_limit=(-0.1,0.1), contrast_limit=(-0.1, 0.1), p=0.5),\n Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225], max_pixel_value=255.0, p=1.0),\n CoarseDropout(p=0.5),\n Cutout(p=0.5),\n ToTensorV2(p=1.0),\n ], p=1.)\n \n \ndef get_valid_transforms():\n return Compose([\n CenterCrop(CFG['img_size'], CFG['img_size'], p=1.),\n Resize(CFG['img_size'], CFG['img_size']),\n Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225], max_pixel_value=255.0, p=1.0),\n ToTensorV2(p=1.0),\n ], p=1.)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Classifier module"},{"metadata":{"trusted":true},"cell_type":"code","source":"# ====================================================\n# stronger classifier module\n# default is linear \n# other include in {CircleLoss,Arcface,Cosface,AMSoftmax}\n# ====================================================\nclass ContrastiveLoss(nn.Module):\n def __init__(self, margin=0.3, **kwargs):\n super(ContrastiveLoss, self).__init__()\n self.margin = margin\n\n def forward(self, inputs, targets):\n n = inputs.size(0)\n # Compute similarity matrix\n sim_mat = torch.matmul(inputs, inputs.t())\n targets = targets\n loss = list()\n c = 0\n\n for i in range(n):\n pos_pair_ = torch.masked_select(sim_mat[i], targets == targets[i])\n\n # move itself\n pos_pair_ = torch.masked_select(pos_pair_, pos_pair_ < 1)\n neg_pair_ = torch.masked_select(sim_mat[i], targets != targets[i])\n\n pos_pair_ = torch.sort(pos_pair_)[0]\n neg_pair_ = torch.sort(neg_pair_)[0]\n\n neg_pair = torch.masked_select(neg_pair_, neg_pair_ > self.margin)\n\n neg_loss = 0\n\n pos_loss = torch.sum(-pos_pair_ + 1)\n if len(neg_pair) > 0:\n neg_loss = torch.sum(neg_pair)\n loss.append(pos_loss + neg_loss)\n\n loss = sum(loss) / n\n return loss\n\n\nclass CircleLoss(nn.Module):\n def __init__(self, in_features, out_features, s=256, m=0.25):\n super(CircleLoss, self).__init__()\n self.weight = Parameter(torch.Tensor(out_features, in_features))\n self._s = s\n self._m = m\n self.reset_parameters()\n\n def reset_parameters(self):\n nn.init.kaiming_uniform_(self.weight, a=math.sqrt(5))\n\n def __call__(self, bn_feat, targets):\n\n sim_mat = F.linear(F.normalize(bn_feat), F.normalize(self.weight))\n alpha_p = F.relu(-sim_mat.detach() + 1 + self._m)\n alpha_n = F.relu(sim_mat.detach() + self._m)\n delta_p = 1 - self._m\n delta_n = self._m\n\n s_p = self._s * alpha_p * (sim_mat - delta_p)\n s_n = self._s * alpha_n * (sim_mat - delta_n)\n\n one_hot = torch.zeros(sim_mat.size(), device=targets.device)\n one_hot.scatter_(1, targets.view(-1, 1).long(), 1)\n\n pred_class_logits = one_hot * s_p + (1.0 - one_hot) * s_n\n\n return pred_class_logits\n\n\nclass Arcface(nn.Module):\n r\"\"\"Implement of large margin arc distance: :\n Args:\n in_features: size of each input sample\n out_features: size of each output sample\n s: norm of input feature\n m: margin\n cos(theta + m)\n \"\"\"\n def __init__(self, in_features, out_features, s=30.0, m=0.30, easy_margin=False, ls_eps=0.0):\n super(Arcface, self).__init__()\n self.in_features = in_features\n self.out_features = out_features\n self.s = s\n self.m = m\n self.ls_eps = ls_eps # label smoothing\n self.weight = Parameter(torch.FloatTensor(out_features, in_features))\n nn.init.xavier_uniform_(self.weight)\n\n self.easy_margin = easy_margin\n self.cos_m = math.cos(m)\n self.sin_m = math.sin(m)\n self.th = math.cos(math.pi - m)\n self.mm = math.sin(math.pi - m) * m\n\n def forward(self, input, label):\n #print(\"input.shape\",input.shape,label.shape,self.in_features,self.weight.shape)\n #print(\"label arcface time device\",label.device)\n # --------------------------- cos(theta) & phi(theta) ---------------------------\n cosine = F.linear(F.normalize(input), F.normalize(self.weight))\n if not self.training:\n return cosine\n \n sine = torch.sqrt(1.0 - torch.pow(cosine, 2))\n phi = cosine * self.cos_m - sine * self.sin_m\n phi = phi.type_as(cosine)\n if self.easy_margin:\n phi = torch.where(cosine > 0, phi, cosine)\n else:\n phi = torch.where(cosine > self.th, phi, cosine - self.mm)\n # --------------------------- convert label to one-hot ---------------------------\n # one_hot = torch.zeros(cosine.size(), requires_grad=True, device='cuda')\n one_hot = torch.zeros(cosine.size(), device='cuda')\n \n #print(\"label device is \",label.device)\n #print(\"one_hot device is \",one_hot.device)\n one_hot.scatter_(1, label.view(-1, 1).long(), 1)\n if self.ls_eps > 0:\n one_hot = (1 - self.ls_eps) * one_hot + self.ls_eps / self.out_features\n # -------------torch.where(out_i = {x_i if condition_i else y_i) -------------\n output = (one_hot * phi) + ((1.0 - one_hot) * cosine)\n output *= self.s\n\n return output\n\n\nclass Cosface(nn.Module):\n r\"\"\"Implement of large margin cosine distance: :\n Args:\n in_features: size of each input sample\n out_features: size of each output sample\n s: norm of input feature\n m: margin\n cos(theta) - m\n \"\"\"\n\n def __init__(self, in_features, out_features, s=30.0, m=0.30):\n super(Cosface, self).__init__()\n self.in_features = in_features\n self.out_features = out_features\n self.s = s\n self.m = m\n self.weight = Parameter(torch.FloatTensor(out_features, in_features))\n nn.init.xavier_uniform_(self.weight)\n\n def forward(self, input, label):\n # --------------------------- cos(theta) & phi(theta) ---------------------------\n cosine = F.linear(F.normalize(input), F.normalize(self.weight))\n phi = cosine - self.m\n # --------------------------- convert label to one-hot ---------------------------\n one_hot = torch.zeros(cosine.size(), device='cuda')\n # one_hot = one_hot.cuda() if cosine.is_cuda else one_hot\n one_hot.scatter_(1, label.view(-1, 1).long(), 1)\n # -------------torch.where(out_i = {x_i if condition_i else y_i) -------------\n output = (one_hot * phi) + ((1.0 - one_hot) * cosine) # you can use torch.where if your torch.__version__ is 0.4\n output *= self.s\n # print(output)\n\n return output\n\n def __repr__(self):\n return self.__class__.__name__ + '(' \\\n + 'in_features=' + str(self.in_features) \\\n + ', out_features=' + str(self.out_features) \\\n + ', s=' + str(self.s) \\\n + ', m=' + str(self.m) + ')'\n\n\nclass AMSoftmax(nn.Module):\n def __init__(self, in_features, out_features, s=30.0, m=0.30):\n super(AMSoftmax, self).__init__()\n self.m = m\n self.s = s\n self.in_feats = in_features\n self.W = torch.nn.Parameter(torch.randn(in_features, out_features), requires_grad=True)\n self.ce = nn.CrossEntropyLoss()\n nn.init.xavier_normal_(self.W, gain=1)\n\n def forward(self, x, lb):\n assert x.size()[0] == lb.size()[0]\n assert x.size()[1] == self.in_feats\n x_norm = torch.norm(x, p=2, dim=1, keepdim=True).clamp(min=1e-12)\n x_norm = torch.div(x, x_norm)\n w_norm = torch.norm(self.W, p=2, dim=0, keepdim=True).clamp(min=1e-12)\n w_norm = torch.div(self.W, w_norm)\n costh = torch.mm(x_norm, w_norm)\n # print(x_norm.shape, w_norm.shape, costh.shape)\n lb_view = lb.view(-1, 1)\n delt_costh = torch.zeros(costh.size(), device='cuda').scatter_(1, lb_view, self.m)\n costh_m = costh - delt_costh\n costh_m_s = self.s * costh_m\n return costh_m_s","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Model"},{"metadata":{"trusted":true},"cell_type":"code","source":"class CassvaImgClassifier(nn.Module):\n def __init__(self, model_arch, n_class, pretrained=False):\n super().__init__()\n \n backbone = timm.create_model(model_arch, pretrained=True)\n n_features = backbone.classifier.in_features\n self.backbone = nn.Sequential(*backbone.children())[:-1]\n# for key, value in self.backbone.named_parameters():\n# print(key,value.shape)\n \n \n# for i in self.backbone.children():\n# print(i)\n \n \"\"\"\n self.backbone = nn.Sequential(*backbone.children())[:-2]\n if CFG.POOLING_METHOD == \"GeM\":\n print(\"using GeM pooling\")\n self.gap = GeM()\n else:\n print(\"using AdaptiveAvgPool2D\")\n self.gap = nn.AdaptiveAvgPool2d((1, 1))\n \"\"\"\n #self.model = timm.create_model(model_arch, pretrained=pretrained)\n #n_features = self.model.classifier.in_features\n #self.model.classifier = nn.Linear(n_features, n_class)\n if CFG['classifier_type'] == \"arcface\":\n print('using {} with s:{}, m: {}'.format(CFG['classifier_type'],CFG['cosine_scale'],CFG['cosine_margin']))\n self.classifier = Arcface(n_features, n_class,s=CFG['cosine_scale'], m=CFG['cosine_margin'])\n elif CFG['classifier_type'] == \"cosface\":\n print('using {} with s:{}, m: {}'.format(CFG['classifier_type'],CFG['cosine_scale'],CFG['cosine_margin']))\n self.classifier = Cosface(n_features, n_class,s=CFG['cosine_scale'], m=CFG['cosine_margin'])\n elif CFG['classifier_type'] == \"amsoftmax\":\n print('using {} with s:{}, m: {}'.format(CFG['classifier_type'],CFG['cosine_scale'],CFG['cosine_margin']))\n self.classifier = AMSoftmax(n_features, n_class,s=CFG['cosine_scale'], m=CFG['cosine_margin'])\n elif CFG['classifier_type'] == \"circle\":\n print('using {} with s:{}, m: {}'.format(CFG['classifier_type'],CFG['cosine_scale'],CFG['cosine_margin']))\n self.classifier = CircleLoss(n_features, n_class,s=CFG['cosine_scale'], m=CFG['cosine_margin'])\n else:\n self.classifier = nn.Linear(n_features, n_class) # origin\n '''\n self.model.classifier = nn.Sequential(\n nn.Dropout(0.3),\n #nn.Linear(n_features, hidden_size,bias=True), nn.ELU(),\n nn.Linear(n_features, n_class, bias=True)\n )\n '''\n def forward_features(self, x):\n x = self.backbone(x)\n return x\n def forward(self, x,label=None):\n \n x = self.forward_features(x)\n # x = self.pool(feats).view(x.size(0), -1) # orgin\n #x = self.gap(feats)\n #x = x.view(x.size(0), -1)\n \n if CFG['classifier_type'] in ('arcface','cosface','amsoftmax','circle'):\n #print(\"label classifer input time device\",label.device)\n x = self.classifier(x,label)\n else:\n x = self.classifier(x)\n return x #, feats\n# x = self.model(x)\n# return x","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Training APIs"},{"metadata":{"trusted":true},"cell_type":"code","source":"def prepare_dataloader(df, trn_idx, val_idx, data_root='../input/cassava-leaf-disease-classification/train_images/'):\n \n from catalyst.data.sampler import BalanceClassSampler\n \n train_ = df.loc[trn_idx,:].reset_index(drop=True)\n valid_ = df.loc[val_idx,:].reset_index(drop=True)\n \n train_ds = CassavaDataset(train_, data_root, transforms=get_train_transforms(), output_label=True, one_hot_label=False, do_fmix=False, do_cutmix=False)\n valid_ds = CassavaDataset(valid_, data_root, transforms=get_valid_transforms(), output_label=True)\n \n train_loader = torch.utils.data.DataLoader(\n train_ds,\n batch_size=CFG['train_bs'],\n pin_memory=False,\n drop_last=False,\n shuffle=True, \n num_workers=CFG['num_workers'],\n #sampler=BalanceClassSampler(labels=train_['label'].values, mode=\"downsampling\")\n )\n val_loader = torch.utils.data.DataLoader(\n valid_ds, \n batch_size=CFG['valid_bs'],\n num_workers=CFG['num_workers'],\n shuffle=False,\n pin_memory=False,\n )\n return train_loader, val_loader\n\ndef train_one_epoch(epoch, model, loss_fn, optimizer, train_loader, device, scheduler=None, schd_batch_update=False):\n model.train()\n\n t = time.time()\n running_loss = None\n\n pbar = tqdm(enumerate(train_loader), total=len(train_loader))\n for step, (imgs, image_labels) in pbar:\n imgs = imgs.to(device).float()\n image_labels = image_labels.to(device).long()\n\n #print(image_labels.shape, exam_label.shape)\n with autocast():\n image_preds = model(imgs,image_labels) #output = model(input)\n #print(image_preds.shape, exam_pred.shape)\n\n loss = loss_fn(image_preds, image_labels)\n \n scaler.scale(loss).backward()\n\n if running_loss is None:\n running_loss = loss.item()\n else:\n running_loss = running_loss * .99 + loss.item() * .01\n\n if ((step + 1) % CFG['accum_iter'] == 0) or ((step + 1) == len(train_loader)):\n # may unscale_ here if desired (e.g., to allow clipping unscaled gradients)\n\n scaler.step(optimizer)\n scaler.update()\n optimizer.zero_grad() \n \n if scheduler is not None and schd_batch_update:\n scheduler.step()\n\n if ((step + 1) % CFG['verbose_step'] == 0) or ((step + 1) == len(train_loader)):\n description = f'epoch {epoch} loss: {running_loss:.4f}'\n \n pbar.set_description(description)\n \n if scheduler is not None and not schd_batch_update:\n scheduler.step()\n \ndef valid_one_epoch(epoch, model, loss_fn, val_loader, device, scheduler=None, schd_loss_update=False):\n model.eval()\n\n t = time.time()\n loss_sum = 0\n sample_num = 0\n image_preds_all = []\n image_targets_all = []\n \n pbar = tqdm(enumerate(val_loader), total=len(val_loader))\n for step, (imgs, image_labels) in pbar:\n imgs = imgs.to(device).float()\n image_labels = image_labels.to(device).long()\n \n image_preds = model(imgs) #output = model(input)\n #print(image_preds.shape, exam_pred.shape)\n image_preds_all += [torch.argmax(image_preds, 1).detach().cpu().numpy()]\n image_targets_all += [image_labels.detach().cpu().numpy()]\n \n loss = loss_fn(image_preds, image_labels)\n \n loss_sum += loss.item()*image_labels.shape[0]\n sample_num += image_labels.shape[0] \n\n if ((step + 1) % CFG['verbose_step'] == 0) or ((step + 1) == len(val_loader)):\n description = f'epoch {epoch} loss: {loss_sum/sample_num:.4f}'\n pbar.set_description(description)\n \n image_preds_all = np.concatenate(image_preds_all)\n image_targets_all = np.concatenate(image_targets_all)\n print('validation multi-class accuracy = {:.4f}'.format((image_preds_all==image_targets_all).mean()))\n \n if scheduler is not None:\n if schd_loss_update:\n scheduler.step(loss_sum/sample_num)\n else:\n scheduler.step()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# reference: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/173733\nclass MyCrossEntropyLoss(_WeightedLoss):\n def __init__(self, weight=None, reduction='mean'):\n super().__init__(weight=weight, reduction=reduction)\n self.weight = weight\n self.reduction = reduction\n\n def forward(self, inputs, targets):\n lsm = F.log_softmax(inputs, -1)\n\n if self.weight is not None:\n lsm = lsm * self.weight.unsqueeze(0)\n print(\"targets.shape\",targets.shape,\"lsm.shape\",lsm.shape)\n loss = -(targets * lsm).sum(-1)\n\n if self.reduction == 'sum':\n loss = loss.sum()\n elif self.reduction == 'mean':\n loss = loss.mean()\n\n return loss","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\n\nclass SmoothCrossEntropyLoss(_WeightedLoss):\n def __init__(self, weight=None, reduction='mean', smoothing=0.1):\n super().__init__(weight=weight, reduction=reduction)\n self.smoothing = smoothing\n self.weight = weight\n self.reduction = reduction\n\n @staticmethod\n def _smooth_one_hot(targets:torch.Tensor, n_classes:int, smoothing=0.0):\n assert 0 <= smoothing < 1\n with torch.no_grad():\n targets = torch.empty(size=(targets.size(0), n_classes),\n device=targets.device) \\\n .fill_(smoothing /(n_classes-1)) \\\n .scatter_(1, targets.data.unsqueeze(1), 1.-smoothing)\n return targets\n\n def forward(self, inputs, targets):\n targets = SmoothCrossEntropyLoss._smooth_one_hot(targets, inputs.size(-1),\n self.smoothing)\n lsm = F.log_softmax(inputs, -1)\n\n if self.weight is not None:\n lsm = lsm * self.weight.unsqueeze(0)\n\n loss = -(targets * lsm).sum(-1)\n\n if self.reduction == 'sum':\n loss = loss.sum()\n elif self.reduction == 'mean':\n loss = loss.mean()\n\n return loss","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Main Loop"},{"metadata":{"trusted":true},"cell_type":"code","source":"if __name__ == '__main__':\n # for training only, need nightly build pytorch\n\n seed_everything(CFG['seed'])\n \n folds = StratifiedKFold(n_splits=CFG['fold_num'], shuffle=True, random_state=CFG['seed']).split(np.arange(train.shape[0]), train.label.values)\n \n for fold, (trn_idx, val_idx) in enumerate(folds):\n # we'll train fold 0 first\n# if fold > 0:\n# break \n\n print('Training with {} started'.format(fold))\n\n print(len(trn_idx), len(val_idx))\n train_loader, val_loader = prepare_dataloader(train, trn_idx, val_idx, data_root='../input/cassava-leaf-disease-classification/train_images/')\n\n device = torch.device(CFG['device'])\n \n model = CassvaImgClassifier(CFG['model_arch'], train.label.nunique(), pretrained=True).to(device)\n scaler = GradScaler() \n optimizer = torch.optim.Adam(model.parameters(), lr=CFG['lr'], weight_decay=CFG['weight_decay'])\n #scheduler = torch.optim.lr_scheduler.StepLR(optimizer, gamma=0.1, step_size=CFG['epochs']-1)\n scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, T_0=CFG['T_0'], T_mult=1, eta_min=CFG['min_lr'], last_epoch=-1)\n #scheduler = torch.optim.lr_scheduler.OneCycleLR(optimizer=optimizer, pct_start=0.1, div_factor=25, \n # max_lr=CFG['lr'], epochs=CFG['epochs'], steps_per_epoch=len(train_loader))\n \n loss_tr = SmoothCrossEntropyLoss().to(device) # nn.CrossEntropyLoss().to(device)\n loss_fn = SmoothCrossEntropyLoss().to(device)\n \n for epoch in range(CFG['epochs']):\n train_one_epoch(epoch, model, loss_tr, optimizer, train_loader, device, scheduler=scheduler, schd_batch_update=False)\n\n with torch.no_grad():\n valid_one_epoch(epoch, model, loss_fn, val_loader, device, scheduler=None, schd_loss_update=False)\n \n# if not os.path.exists(\"use_mycross\"):\n# os.makedirs(\"use_mycross\")\n torch.save(model.state_dict(),'{}_fold_{}_{}'.format(CFG['model_arch'], fold, epoch))\n \n #torch.save(model.cnn_model.state_dict(),'{}/cnn_model_fold_{}_{}'.format(CFG['model_path'], fold, CFG['tag']))\n# del model, optimizer, train_loader, val_loader, scaler, scheduler\n# torch.cuda.empty_cache()","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}