kernel_id
int64
24.2k
23.3M
prompt
stringlengths
8
1.85M
completetion
stringlengths
1
182k
comp_name
stringlengths
5
57
14,240,018
class enet_v2(nn.Module): def __init__(self, backbone, out_dim, pretrained=False): super(enet_v2, self ).__init__() self.enet = timm.create_model(backbone, pretrained=pretrained) in_ch = self.enet.classifier.in_features self.myfc = nn.Linear(in_ch, out_dim) self.enet.classifier = nn.Identity() def forward(self, x): x...
from glob import glob from sklearn.model_selection import GroupKFold, StratifiedKFold import cv2 from skimage import io import torch from torch import nn import os from datetime import datetime import time import random import cv2 import torchvision from torchvision import transforms import pandas as pd import numpy as...
Cassava Leaf Disease Classification
14,240,018
def load_state(model_path): model = CustomResNext(CFG.model_name, pretrained=False) try: model.load_state_dict(torch.load(model_path)['model'], strict=True) state_dict = torch.load(model_path)['model'] except: state_dict = torch.load(model_path)['model'] state_dict = {k[7:] if k.startswith('module.')else k: state_dic...
CFG = { 'fold_num': 5, 'seed': 719, 'model_arch': 'tf_efficientnet_b4_ns', 'img_size': 512, 'epochs': 10, 'train_bs': 32, 'valid_bs': 32, 'lr': 1e-4, 'num_workers': 4, 'accum_iter': 1, 'verbose_step': 1, 'device': 'cuda:0', 'tta': 3, 'used_epochs': [4,5,6,7], 'weights': [1,1,1,1] }
Cassava Leaf Disease Classification
14,240,018
model = CustomResNext(CFG.model_name, pretrained=False) states = [load_state(MODEL_DIR+f'{CFG.model_name}_fold{fold}.pth')for fold in CFG.trn_fold] test_dataset = TestDataset(test, transform=get_transforms(data='valid')) test_loader = DataLoader(test_dataset, batch_size=CFG.batch_size, shuffle=False, num_workers=CFG.n...
train = pd.read_csv('.. /input/cassava-leaf-disease-classification/train.csv') train.head()
Cassava Leaf Disease Classification
14,240,018
plt.style.use('seaborn-white') sns.set_style('white') plt.rcParams['figure.figsize'] = [16, 10] plt.rcParams['font.size'] = 14 %matplotlib inline warnings.filterwarnings("ignore") BatchNormalization, Activation, GlobalAveragePooling2D, MaxPooling2D, concatenate, Reshape, Add, multiply) t_start = time.time()<define_...
train.label.value_counts()
Cassava Leaf Disease Classification
14,240,018
basic_name = f'Unet_resnet' save_model_name = basic_name + '.model' submission_file = basic_name + '.csv' TRAIN_IMAGE_DIR = '.. /input/train/images/' TRAIN_MASK_DIR = '.. /input/train/masks/' TEST_IMAGE_DIR = '.. /input/test/images/' img_size = 101 seed=1994 batch_size = 128 epochs = 120<compute_train_metric>
submission = pd.read_csv('.. /input/cassava-leaf-disease-classification/sample_submission.csv') submission.head()
Cassava Leaf Disease Classification
14,240,018
def get_iou_vector(A, B): batch_size = A.shape[0] metric = [] for batch in range(batch_size): t, p = A[batch]>0, B[batch]>0 intersection = np.logical_and(t, p) union = np.logical_or(t, p) iou =(np.sum(intersection > 0)+ 1e-10)/(np.sum(union > 0)+ 1e-10) thresholds = np.arange(0.5, 1, 0.05) s = [] for thresh in thre...
def seed_everything(seed): random.seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = True def get_img(path): im_bgr = cv2.imread(path) im_rgb = im_bgr[:, :, ::-1] r...
Cassava Leaf Disease Classification
14,240,018
train_df = pd.read_csv(".. /input/train.csv", index_col="id", usecols=[0]) train_df["images"] = [np.array(load_img(TRAIN_IMAGE_DIR + "{}.png".format(idx), color_mode = "grayscale")) / 255 for idx in tqdm_notebook(train_df.index)] train_df["masks"] = [np.array(load_img(TRAIN_MASK_DIR + "{}.png".format(idx), color_mode ...
class CassavaDataset(Dataset): def __init__( self, df, data_root, transforms=None, output_label=True ): super().__init__() self.df = df.reset_index(drop=True ).copy() self.transforms = transforms self.data_root = data_root self.output_label = output_label def __len__(self): return self.df.shape[0] def __getitem__(sel...
Cassava Leaf Disease Classification
14,240,018
X = np.array(train_df.images.tolist() ).reshape(-1, img_size, img_size, 1) y = np.array(train_df.masks.tolist() ).reshape(-1, img_size, img_size, 1) x_train, x_valid, y_train, y_valid = \ train_test_split( X, y, test_size=0.2, stratify=train_df.coverage_class, random_state=seed )<define_search_model>
HorizontalFlip, VerticalFlip, IAAPerspective, ShiftScaleRotate, CLAHE, RandomRotate90, Transpose, ShiftScaleRotate, Blur, OpticalDistortion, GridDistortion, HueSaturationValue, IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, MedianBlur, IAAPiecewiseAffine, RandomResizedCrop, IAASharpen, IAAEmboss, RandomBrightnessCon...
Cassava Leaf Disease Classification
14,240,018
def BatchActivate(x): x = BatchNormalization()(x) x = Activation('elu' )(x) return x def convolution_block(x, filters, size, strides=(1,1), padding='same', activation=True): x = Conv2D(filters, size, strides=strides, padding=padding )(x) if activation==True: x = BatchActivate(x) return x def residual_block(blockInp...
class CassvaImgClassifier(nn.Module): def __init__(self, model_arch, n_class, pretrained=False): super().__init__() self.model = timm.create_model(model_arch, pretrained=pretrained) n_features = self.model.classifier.in_features self.model.classifier = nn.Linear(n_features, n_class) def forward(self, x): x = self.mod...
Cassava Leaf Disease Classification
14,240,018
def build_model(input_layer, start_neurons, DropoutRatio=0.5): conv1 = unet_layer(input_layer,start_neurons * 1,use_csSE_ratio=2) pool1 = MaxPooling2D(( 2,2))(conv1) pool1 = Dropout(DropoutRatio/3 )(pool1) conv2 = unet_layer(pool1, start_neurons * 2,use_csSE_ratio=2) pool2 = MaxPooling2D(( 2,2))(conv2) pool2 = Dro...
if __name__ == '__main__': seed_everything(CFG['seed']) folds = StratifiedKFold(n_splits=CFG['fold_num'] ).split(np.arange(train.shape[0]), train.label.values) for fold,(trn_idx, val_idx)in enumerate(folds): if fold > 0: break print('Inference fold {} started'.format(fold)) valid_ = train.loc[val_idx,:].reset_index(d...
Cassava Leaf Disease Classification
14,240,018
def do_augmentation(seqs, seq2_train, X_train, y_train): seq_det = seqs.to_deterministic() X_train_aug = seq_det.augment_image(X_train) X_train_aug = seq2_train.augment_image(X_train_aug) y_train_aug = seq_det.augment_image(y_train) if y_train_aug.shape !=(101, 101): X_train_aug = ia.imresize_single_image(X_train_au...
OUTPUT_DIR = './' if not os.path.exists(OUTPUT_DIR): os.makedirs(OUTPUT_DIR) test['label'] = np.argmax(tst_preds, axis=1) test.head()
Cassava Leaf Disease Classification
14,240,018
<train_model><EOS>
test.to_csv(OUTPUT_DIR+'submission.csv', index=False )
Cassava Leaf Disease Classification
14,476,793
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<choose_model_class>
sys.path.append('.. /input/timm-pytorch-image-models/pytorch-image-models-master') warnings.filterwarnings("ignore" )
Cassava Leaf Disease Classification
14,476,793
model1 = load_model(save_model_name, custom_objects={'my_iou_metric':my_iou_metric}) input_x = model1.layers[0].input output_layer = model1.layers[-1].input model2 = Model(input_x, output_layer) model2.compile(loss=symmetric_lovasz, optimizer=Adam(lr=0.01), metrics=[my_iou_metric_2] )<train_on_grid>
DATA_PATH = '.. /input/cassava-leaf-disease-classification/' TRAIN_DIR = DATA_PATH + 'train_images/' TEST_DIR = DATA_PATH + 'test_images/' MODEL_PATH = '.. /input/cassavanet-baseline-models/' N_TTA = 8 HEIGHT = 512 WIDTH = 512 CHANNELS = 3 N_CLASSES = 5 MODEL_LIST = [0,1,2,3,4,5] IMG_MEAN = [0.485, 0.456, 0.406] IMG_ST...
Cassava Leaf Disease Classification
14,476,793
early_stopping = EarlyStopping(monitor='val_my_iou_metric_2', mode = 'max',patience=30, verbose=1) model_checkpoint = ModelCheckpoint(save_model_name,monitor='val_my_iou_metric_2', mode = 'max', save_best_only=True, verbose=1) reduce_lr = ReduceLROnPlateau(monitor='val_my_iou_metric_2', mode = 'max',factor=0.5, patie...
def seed_everything(seed): random.seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = True SEED = 1234 seed_everything(SEED) DEVICE = torch.device("cuda:0" if torch....
Cassava Leaf Disease Classification
14,476,793
model = load_model(save_model_name,custom_objects={'my_iou_metric_2': my_iou_metric_2, 'symmetric_lovasz': symmetric_lovasz} )<predict_on_test>
class CassavaNet(nn.Module): def __init__(self, model_name=None, pretrained=False): super().__init__() self.model_name = model_name if model_name == 'deit_base_patch16_224' or model_name == 'deit_base_patch16_384': self.model = torch.hub.load('facebookresearch/deit:main', model_name, pretrained=pretrained) else: self....
Cassava Leaf Disease Classification
14,476,793
def predict_result(model,x_test,img_size): x_test_reflect = np.array([np.fliplr(x)for x in x_test]) preds_test = model.predict(x_test ).reshape(-1, img_size, img_size) preds_test2_refect = model.predict(x_test_reflect ).reshape(-1, img_size, img_size) preds_test += np.array([ np.fliplr(x)for x in preds_test2_refect]...
class GetData(Dataset): def __init__(self, Dir, FNames, labels,Type): self.dir = Dir self.fnames = FNames self.lbs = labels self.type = Type def __len__(self): return len(self.fnames) def __getitem__(self, index): x = imread(os.path.join(self.dir, self.fnames[index])) if "train" in self.type: aug_data = train_transfor...
Cassava Leaf Disease Classification
14,476,793
thresholds_ori = np.linspace(0.3, 0.7, 31) thresholds = np.log(thresholds_ori/(1-thresholds_ori)) ious = np.array([iou_metric_batch(y_valid, preds_valid > threshold)for threshold in tqdm_notebook(thresholds)]) print(ious);<categorify>
Aug_Norm = A.Normalize(mean=IMG_MEAN, std=IMG_STD, max_pixel_value=255.0, p=1.0) test_aug = Compose([ A.HorizontalFlip(p=0.5), A.VerticalFlip(p=0.5), A.ShiftScaleRotate(p = 1.0), A.ColorJitter(brightness=0.1, contrast=0.2, saturation=0.2, hue=0.00, always_apply=False, p=1.0), A.RandomCrop(height= HEIGHT, width = WIDTH...
Cassava Leaf Disease Classification
14,476,793
def rle_encode(im): pixels = im.flatten(order='F') pixels = np.concatenate([[0], pixels, [0]]) runs = np.where(pixels[1:] != pixels[:-1])[0] + 1 runs[1::2] -= runs[::2] return ' '.join(str(x)for x in runs )<predict_on_test>
models = [] count = 0 for model_fpath in os.listdir(MODEL_PATH): if count in MODEL_LIST: print("Model Loaded:",model_fpath) model_name_split = model_fpath.split('_f')[0] model = CassavaNet(model_name_split,pretrained = False) info = torch.load(MODEL_PATH + model_fpath,map_location = torch.device(DEVICE)) model.load_s...
Cassava Leaf Disease Classification
14,476,793
test_images = os.listdir(TEST_IMAGE_DIR) x_test = np.array([(np.array(load_img(TEST_IMAGE_DIR + "{}".format(idx), color_mode = "grayscale")))/ 255 for idx in tqdm_notebook(test_images)] ).reshape(-1, img_size, img_size, 1) preds_test = predict_result(model,x_test,img_size) pred_dict = {idx[:10]: rle_encode(np.round(...
submission = pd.DataFrame() list_files = os.listdir(TEST_DIR) submission['image_id'] = pd.Series(list_files) submission['label'] = 0 submission.head()
Cassava Leaf Disease Classification
14,476,793
sub = pd.DataFrame.from_dict(pred_dict,orient='index') sub.index.names = ['id'] sub.columns = ['rle_mask'] sub.to_csv(submission_file )<train_model>
start_time = time.time() BATCH_SIZE = 1 test_set = GetData(TEST_DIR,submission['image_id'], submission['label'], Type = 'test') test_loader = DataLoader(test_set, batch_size=BATCH_SIZE, shuffle=False, num_workers=8,pin_memory = True) with torch.no_grad() : for i,(images,labels)in enumerate(test_loader): voting = np.z...
Cassava Leaf Disease Classification
14,476,793
<set_options><EOS>
submission.to_csv('submission.csv',index=False) submission.head()
Cassava Leaf Disease Classification
14,527,205
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<init_hyperparams>
') ') ')
Cassava Leaf Disease Classification
14,527,205
def set_learning_rate(optimizer, lr): for param_group in optimizer.param_groups: param_group['lr'] = lr def get_learning_rate(optimizer): return optimizer.param_groups[0]['lr'] class LearningRate() : def __init__(self, initial_lr, iteration_type): self.initial_lr = initial_lr self.iteration_type = iteration_type def ge...
!pip install.. /input/cassava-models/Keras_Applications-1.0.8-py3-none-any.whl !pip install.. /input/cassava-models/efficientnet-1.1.0-py3-none-any.whl
Cassava Leaf Disease Classification
14,527,205
def save_checkpoint(checkpoint_path, model, optimizer): state = {'state_dict': model.state_dict() , 'optimizer' : optimizer.state_dict() } torch.save(state, checkpoint_path) print('model saved to %s' % checkpoint_path) def load_checkpoint(checkpoint_path, model, optimizer): state = torch.load(checkpoint_path) mode...
AUTO = tf.data.experimental.AUTOTUNE EPOCHS = 20 BATCH_SIZE = 32 * strategy.num_replicas_in_sync IMAGE_SIZE = [512, 512] SEED = 123 LR = 0.0001 TTA = 10 VERBOSE = 2 N_CLASSES = 5 TEST_FILENAMES = '.. /input/cassava-leaf-disease-classification/test_images/*.jpg'
Cassava Leaf Disease Classification
14,527,205
def train_model(model,epochs, learning_rate,loss_function, optimizer, dataset, dataset_val, batch_size ): snapshot = SnapshotLR( initial_lr=0.000001,max_lr=0.0001, total_iters=epochs,n_cycles=30, iteration_type="epochs" ) scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode="max", factor=0.97, m...
def data_augment(image, image_name): p_spatial = tf.random.uniform([], 0, 1.0, dtype = tf.float32) p_rotate = tf.random.uniform([], 0, 1.0, dtype = tf.float32) p_pixel_1 = tf.random.uniform([], 0, 1.0, dtype = tf.float32) p_pixel_2 = tf.random.uniform([], 0, 1.0, dtype = tf.float32) p_pixel_3 = tf.random.uniform([]...
Cassava Leaf Disease Classification
14,527,205
<categorify><EOS>
def get_model() : with strategy.scope() : inp = tf.keras.layers.Input(shape =(*IMAGE_SIZE, 3)) x = efn.EfficientNetB5(weights = None, include_top = False )(inp) x = tf.keras.layers.GlobalAveragePooling2D()(x) x = tf.keras.layers.Dropout(0.2 )(x) output = tf.keras.layers.Dense(N_CLASSES, activation = 'softmax' )(x) ...
Cassava Leaf Disease Classification
14,520,819
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<normalization>
package_path = '.. /input/pytorch-image-models/pytorch-image-models-master'
Cassava Leaf Disease Classification
14,520,819
def load_image(path,pad=True, mask = False): flip = False if "_gael" in path: path = path.replace("_gael","") flip = True img = cv2.imread(str(path)) if flip: img = cv2.flip(img, 0) img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) height, width, _ = img.shape if pad: if height % 32 == 0: y_min_pad = 0 y_max_pad = 0 else...
from glob import glob from sklearn.model_selection import GroupKFold, StratifiedKFold import cv2 from skimage import io import torch from torch import nn import os from datetime import datetime import time import random import cv2 import torchvision from torchvision import transforms import pandas as pd import numpy as...
Cassava Leaf Disease Classification
14,520,819
def pred_resize(predictions,height,width): if height % 32 == 0: y_min_pad = 0 y_max_pad = 0 else: y_pad = 32 - height % 32 y_min_pad = int(y_pad / 2) y_max_pad = y_pad - y_min_pad if width % 32 == 0: x_min_pad = 0 x_max_pad = 0 else: x_pad = 32 - width % 32 x_min_pad = int(x_pad / 2) x_max_pad = x_pad - x_min_pad p...
CFG = { 'fold_num': 5, 'seed': 719, 'model_arch': 'vit_base_patch16_384', 'img_size': 384, 'epochs': 10, 'train_bs': 32, 'valid_bs': 32, 'lr': 1e-4, 'num_workers': 4, 'accum_iter': 1, 'verbose_step': 1, 'device': 'cuda:0', 'used_folds':[0,2,3], 'used_epochs': [7,8,9], 'tta': 3 }
Cassava Leaf Disease Classification
14,520,819
def stack_and_resize(predictions,height,width): preds_stacked = np.vstack(predictions)[:, 0, :, :] return pred_resize(preds_stacked,height,width )<set_options>
train = pd.read_csv('.. /input/cassava-leaf-disease-classification/train.csv') train.head()
Cassava Leaf Disease Classification
14,520,819
warnings.filterwarnings("ignore" )<load_from_csv>
train.label.value_counts()
Cassava Leaf Disease Classification
14,520,819
directory = '.. /input/tgs-salt-identification-challenge' depths_df = pd.read_csv(os.path.join(directory, 'train.csv')) train_path = os.path.join(directory, 'train') test_path = os.path.join(directory, 'test') ids_val = _pickle.load(open(".. /input/intermediatetgs/val_index.obj","rb")) ids_train = _pickle.load(open...
submission = pd.read_csv('.. /input/cassava-leaf-disease-classification/sample_submission.csv') submission.head()
Cassava Leaf Disease Classification
14,520,819
epoch =100 learning_rate = 0.0001 loss_fn = torch.nn.BCELoss() optimizer = torch.optim.Adam(model.parameters() , lr=learning_rate, eps=1e-7) load_checkpoint(".. /input/80pytorchgamma/high_val_iou_tgs.pth", model, optimizer) model = train_model(model,epoch,learning_rate, loss_fn,optimizer, dataset, dataset_val,32 )<pr...
class CassavaDataset(Dataset): def __init__( self, df, data_root, transforms=None, output_label=True ): super().__init__() self.df = df.reset_index(drop=True ).copy() self.transforms = transforms self.data_root = data_root self.output_label = output_label def __len__(self): return self.df.shape[0] def __getitem__(sel...
Cassava Leaf Disease Classification
14,520,819
val_predictions = [] val_masks = [] for image, mask in tqdm_notebook(data.DataLoader(dataset_val, batch_size = 32)) : image = Variable(image.type(torch.FloatTensor ).cuda()) y_pred = model(image ).cpu().data.numpy() val_predictions.append(y_pred) val_masks.append(mask) val_predictions_stacked = stack_and_resize(val_...
HorizontalFlip, VerticalFlip, IAAPerspective, ShiftScaleRotate, CLAHE, RandomRotate90, Transpose, ShiftScaleRotate, Blur, OpticalDistortion, GridDistortion, HueSaturationValue, IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, MedianBlur, IAAPiecewiseAffine, RandomResizedCrop, IAASharpen, IAAEmboss, RandomBrightnessCon...
Cassava Leaf Disease Classification
14,520,819
metric_by_threshold = [] for threshold in np.linspace(0.3, 0.7, 50): val_binary_prediction =(val_predictions_stacked > threshold ).astype(int) iou_values = [] for y_mask, p_mask in zip(val_masks_stacked, val_binary_prediction): iou = jaccard_similarity_score(y_mask.flatten() , p_mask.flatten()) iou_values.append(iou)...
class ViTBase16(nn.Module): def __init__(self, n_classes, pretrained=False): super(ViTBase16, self ).__init__() self.model = timm.create_model(CFG['model_arch'], pretrained=pretrained) self.model.head = nn.Linear(self.model.head.in_features, n_classes) def forward(self, x): x = self.model(x) return x
Cassava Leaf Disease Classification
14,520,819
threshold = best_threshold binary_prediction =(all_predictions_stacked > threshold ).astype(int) all_masks = [] for p_mask in list(binary_prediction): p_mask = rle_encoding(p_mask) all_masks.append(' '.join(map(str, p_mask)) )<save_to_csv>
if __name__ == '__main__': tst_preds_avg = [] seed_everything(CFG['seed']) for fold in CFG['used_folds']: print('Inference fold {} started'.format(fold)) test = pd.DataFrame() test['image_id'] = list(os.listdir('.. /input/cassava-leaf-disease-classification/test_images/')) test_ds = CassavaDataset(test, '.. /input/cas...
Cassava Leaf Disease Classification
14,520,819
submit = pd.DataFrame([test_file_list, all_masks] ).T submit.columns = ['id', 'rle_mask'] submit.to_csv('submit_baseline_torch.csv', index = False )<load_from_csv>
test['label'] = np.argmax(np.mean(tst_preds_avg, axis=0), axis=-1) test.head()
Cassava Leaf Disease Classification
14,520,819
<count_missing_values><EOS>
test.to_csv('submission.csv', index=False )
Cassava Leaf Disease Classification
14,462,070
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<data_type_conversions>
package_paths = ['.. /input/imgclsmob/imgclsmob-master', '.. /input/efficientnetpytorch/EfficientNet-PyTorch-master', '.. /input/timm-pytorch-image-models/pytorch-image-models-master'] for package_path in package_paths: sys.path.append(package_path )
Cassava Leaf Disease Classification
14,462,070
tgs = ".. /input" tgs1 = ".. /input/tgs-salt-identification-challenge" def getImage(imgId): path = Path(tgs+"/train/images/")/ '{}'.format(imgId) img = imread(path) return img.astype(np.uint8) def getGrayImage(imgId): path = Path(tgs+"/train/images/")/'{}'.format(imgId) img = imread(path ).astype(np.uint8) img = c...
import numpy as np import pandas as pd import os import torch import random from albumentations import * from albumentations.pytorch import ToTensorV2 from torch.utils.data import Dataset, DataLoader import timm import torch.nn as nn import torch.nn.functional as F from efficientnet_pytorch import EfficientNet from pyt...
Cassava Leaf Disease Classification
14,462,070
thresholds = [0.5, 0.55, 0.60, 0.65, 0.70, 0.75, 0.80, 0.85, 0.90, 0.95] n_thresholds = len(thresholds) def IoUhelper(TrueMask, predictedMask): intersection = cv2.bitwise_and(TrueMask, predictedMask) union = cv2.bitwise_or(TrueMask, predictedMask) intersectionCnt = cv2.countNonZero(intersection) unionCnt = cv2.coun...
def seed_everything(seed): random.seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = True
Cassava Leaf Disease Classification
14,462,070
def rle_encoding(mask): mask = mask.ravel() encoding = "" i = 0 while(i<len(mask)) : currCnt = 0 start = i if(mask[i] == 255): while(i < len(mask)and mask[i] == 255): currCnt+=1 i+=1 encoding+=(" "+ str(start+1)+ " " + str(currCnt)) else: i+=1 return encoding.strip()<categorify>
class CasDataset(Dataset): def __init__(self, df, path, transforms): self.df = df self.path = path self.transforms = transforms def __len__(self): return self.df.shape[0] def __getitem__(self, idx): image = cv2.imread(self.path+self.df.loc[idx, 'image_id']) image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) if self.trans...
Cassava Leaf Disease Classification
14,462,070
imgId = train[haveMask].index[18] rle_orig = train[haveMask].rle_mask[18] img1 = getImage(imgId+".png") img2 = getMask(imgId+".png") rle1 = rle_encoding(np.transpose(img2)) print("Original Encoding ", rle_orig) print("Len1 = ", len(rle_orig.split())) print("-------") print("Encoding ", rle1) print("Len2 = ", len(r...
class CasModel(nn.Module): def __init__(self, num_classes=5, model="resnet34"): super().__init__() if model == "resnet16": self.backbone = ptcv_get_model("resnet16", pretrained=False) self.backbone.features.final_pool = nn.AdaptiveAvgPool2d(1) self.backbone.output = nn.Sequential(nn.Linear(512, num_classes)) elif mod...
Cassava Leaf Disease Classification
14,462,070
TRAIN_IMAGE_DIR = tgs+'/train/images' TRAIN_MASK_DIR = tgs+'/train/masks' TEST_MASK_DIR = tgs+'/test/images' im_height = 128 im_width = 128 train_image_list = os.listdir(TRAIN_IMAGE_DIR) train_mask_list = os.listdir(TRAIN_MASK_DIR) test_image_list = os.listdir(TEST_MASK_DIR )<prepare_x_and_y>
def inference_one_epoch(model, data_loader, device): model.eval() image_preds_all = [] for step,(imgs)in enumerate(data_loader): imgs = imgs.to(device ).float() image_preds = model(imgs) image_preds_all += [torch.softmax(image_preds, 1 ).detach().cpu().numpy() ] image_preds_all = np.concatenate(image_preds_all, axis=0...
Cassava Leaf Disease Classification
14,462,070
X_train_image = np.zeros(( len(train_image_list), im_height, im_width, 1), dtype = np.uint8) Y_train_mask = np.zeros(( len(train_mask_list), im_height, im_width, 1), dtype = np.uint8) for i in tqdm(range(len(train_image_list))): imgId = train_image_list[i] img = getGrayImage(imgId) img = cv2.resize(img,(im_height, i...
seed_everything(960630) test_df = pd.DataFrame() test_df['image_id'] = list(os.listdir('.. /input/cassava-leaf-disease-classification/test_images/')) test_data = CasDataset(test_df, '.. /input/cassava-leaf-disease-classification/test_images/', transforms_test )
Cassava Leaf Disease Classification
14,462,070
<count_values>
test_loader = DataLoader( test_data, batch_size=32, num_workers=8, shuffle=False, pin_memory=False, )
Cassava Leaf Disease Classification
14,462,070
print(np.count_nonzero(Y_train_mask[2])) print(cntOne[2] )<count_unique_values>
test_preds = [] device = torch.device('cuda:0') model = CasModel(model="resnet34" ).to(device) for fold in range(5): model.load_state_dict(torch.load(f'.. /input/casresnet34/resnet34_f_{fold}.pth')) with torch.no_grad() : for t in range(tta): test_preds += [inference_one_epoch(model, test_loader, device)]
Cassava Leaf Disease Classification
14,462,070
x = np.array([0, 1.0, 3.0, 1.6]) bins = np.array([0, 1.0, 2.5, 4.0, 10.0]) inds = np.digitize(x, bins) np.unique(inds )<count_unique_values>
test_preds = np.mean(test_preds, axis=0) test_df['label'] = np.argmax(test_preds, axis=1) test_df.head()
Cassava Leaf Disease Classification
14,462,070
<split><EOS>
test_df.to_csv('submission.csv', index=False )
Cassava Leaf Disease Classification
13,863,118
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<normalization>
!pip install --quiet /kaggle/input/kerasapplications !pip install --quiet /kaggle/input/efficientnet-git
Cassava Leaf Disease Classification
13,863,118
def augment(X, Y): print("FlipLR") X1 = np.append(X, [np.fliplr(x)for x in X], axis = 0) Y1 = np.append(Y, [np.fliplr(y)for y in Y], axis = 0) print("Roll") X = np.append(X1, [np.roll(x, 40, axis = 1)for x in X1], axis = 0) Y = np.append(Y1, [np.roll(y, 40, axis = 1)for y in Y1], axis = 0) m = X.shape[0] np.rando...
def seed_everything(seed=0): random.seed(seed) np.random.seed(seed) tf.random.set_seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) os.environ['TF_DETERMINISTIC_OPS'] = '1' seed = 0 seed_everything(seed) warnings.filterwarnings('ignore' )
Cassava Leaf Disease Classification
13,863,118
print(train_Y[70, :, :, 0]) train_X =(train_X.astype(np.float32)/255.0) train_Y =(train_Y/255.0 ).astype(np.bool ).astype(np.uint8) val_X =(val_X.astype(np.float32)/255.0) val_Y =(val_Y/255.0 ).astype(np.bool ).astype(np.uint8) print(train_X.dtype, train_Y.dtype) print(train_Y[70, :, :, 0] )<import_modules>
strategy = tf.distribute.get_strategy() AUTO = tf.data.experimental.AUTOTUNE REPLICAS = strategy.num_replicas_in_sync print(f'REPLICAS: {REPLICAS}' )
Cassava Leaf Disease Classification
13,863,118
import keras <import_modules>
BATCH_SIZE = 16 * REPLICAS HEIGHT = 512 WIDTH = 512 CHANNELS = 3 N_CLASSES = 5 TTA_STEPS = 8
Cassava Leaf Disease Classification
13,863,118
from keras.models import * from keras.layers import * from keras.optimizers import * from keras.callbacks import *<define_search_model>
def data_augment(image, label): p_spatial = tf.random.uniform([], 0, 1.0, dtype=tf.float32) p_rotate = tf.random.uniform([], 0, 1.0, dtype=tf.float32) p_pixel_1 = tf.random.uniform([], 0, 1.0, dtype=tf.float32) p_pixel_2 = tf.random.uniform([], 0, 1.0, dtype=tf.float32) p_pixel_3 = tf.random.uniform([], 0, 1.0, dty...
Cassava Leaf Disease Classification
13,863,118
def batchActivate(x): x = BatchNormalization()(x) x = Activation('relu' )(x) return x def residualBlock(blockInput, numChannel, matchChannel = False): x = batchActivate(blockInput) x = Conv2D(numChannel,(3, 3), activation= None, padding = "same", use_bias = False )(x) x = batchActivate(x) x = Conv2D(numChannel,(3,...
def get_name(file_path): parts = tf.strings.split(file_path, os.path.sep) name = parts[-1] return name def decode_image(image_data): image = tf.image.decode_jpeg(image_data, channels=3) image = tf.cast(image, tf.float32)/ 255.0 return image def center_crop(image): image = tf.reshape(image, [600, 800, CHANNELS]) ...
Cassava Leaf Disease Classification
13,863,118
inputs = Input(shape =(im_height, im_width, 1)) init = Conv2D(16,(7, 7), activation="relu", padding = "same" )(inputs) conv1 = residualBlock(init, 16) conv1 = residualBlock(conv1, 16) c1 = residualBlock(conv1, 16) c1 = BatchNormalization()(c1) p1 = MaxPooling2D(pool_size=(2, 2))(c1) conv2 = residualBlock(p1, 32, ...
model_path_list = glob.glob('/kaggle/input/cassava-leaf-disease-training-with-tpu-v2-pods/*.h5') model_path_list.sort() print('Models to predict:') print(*model_path_list, sep=' ' )
Cassava Leaf Disease Classification
13,863,118
earlyStopping = EarlyStopping(patience = 20, verbose = 1) checkpointer = ModelCheckpoint('model-tgs-salt-1.h5',monitor = 'val_my_iou_metric', verbose=1, save_best_only=True) reducelr=ReduceLROnPlateau(monitor='val_my_iou_metric',patience=5, min_lr=0.00001, verbose=1,factor=0.5) epochs = 50 batch_size = 64 history = ...
def model_fn(input_shape, N_CLASSES): inputs = L.Input(shape=input_shape, name='input_image') base_model = efn.EfficientNetB4(input_tensor=inputs, include_top=False, weights=None, pooling='avg') x = L.Dropout (.5 )(base_model.output) output = L.Dense(N_CLASSES, activation='softmax', name='output' )(x) model = Model...
Cassava Leaf Disease Classification
13,863,118
model1 = load_model("./model-tgs-salt-1.h5", custom_objects = {'my_iou_metric':my_iou_metric} )<set_options>
files_path = f'{database_base_path}test_images/' test_size = len(os.listdir(files_path)) test_preds = np.zeros(( test_size, N_CLASSES)) for model_path in model_path_list: print(model_path) K.clear_session() model.load_weights(model_path) if TTA_STEPS > 0: test_ds = get_dataset(files_path, tta=True ).repeat() ct_steps...
Cassava Leaf Disease Classification
13,863,118
<define_variables><EOS>
submission = pd.DataFrame({'image_id': image_names, 'label': test_preds}) submission.to_csv('submission.csv', index=False) display(submission.head() )
Cassava Leaf Disease Classification
13,643,882
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<predict_on_test>
import numpy as np import os import pandas as pd from fastai.vision.all import *
Cassava Leaf Disease Classification
13,643,882
totIoU = 0.0 m = val_X.shape[0] for i in tqdm(range(m)) : img = val_X[i:i+1, :, :] origMask = val_Y[i:i+1, :, :] predMask = model1.predict(img) origMask = cv2.inRange(origMask[0, :, :, 0], 0.55, 255) predMask = cv2.inRange(predMask[0, :, :, 0], 0.55, 255) totIoU += meanHit(origMask, predMask) print("mean IoU = ", t...
set_seed(42 )
Cassava Leaf Disease Classification
13,643,882
X_test_image = X_test_image/255.0<set_options>
train_df = pd.read_csv(dataset_path/'train.csv' )
Cassava Leaf Disease Classification
13,643,882
gc.collect()<categorify>
train_df['path'] = train_df['image_id'].map(lambda x:dataset_path/'train_images'/x) train_df = train_df.drop(columns=['image_id']) train_df = train_df.sample(frac=1 ).reset_index(drop=True) train_df.head(10 )
Cassava Leaf Disease Classification
13,643,882
origSize =(101, 101) submit_names = [] submit_rleMasks = [] for i in tqdm(range(len(test_image_list))): img = X_test_image[i:i+1,:,:] predMask = model1.predict(img) predMask = cv2.inRange(predMask[0, :, :, 0], 0.55, 255) predMask = cv2.resize(predMask, origSize) submit_names.append(test_image_list[i][:-4]) submit_...
dls = ImageDataLoaders.from_df(train_df, splitter=RandomSplitter(0.2, seed=42), label_col=0, fn_col=1, bs=bs, item_tfms=item_tfms, batch_tfms=batch_tfms )
Cassava Leaf Disease Classification
13,643,882
sub = pd.DataFrame({'id':submit_names, 'rle_mask': submit_rleMasks}) print(sub.shape) sub.head()<feature_engineering>
learn = cnn_learner(dls, resnet50, metrics=[error_rate, accuracy] ).to_native_fp16()
Cassava Leaf Disease Classification
13,643,882
for i in tqdm(range(len(test_image_list))): if(len(sub.iloc[i, 1])== 0): sub.iloc[i, 1] = np.nan<save_to_csv>
learn.freeze() learn.fine_tune(1, cbs=[MixUp(0.5)] )
Cassava Leaf Disease Classification
13,643,882
sub.to_csv("tgsModel1.csv", index = False )<install_modules>
learn.save('estagio-1' )
Cassava Leaf Disease Classification
13,643,882
!pip3 install pycocotools<set_options>
learn = learn.load('estagio-1' )
Cassava Leaf Disease Classification
13,643,882
%matplotlib inline %reload_ext autoreload %autoreload 2 print(torch.__version__) torch.cuda.is_available() torch.backends.cudnn.benchmark=True<load_from_csv>
learn = learn.to_native_fp32()
Cassava Leaf Disease Classification
13,643,882
MASKS_FN = 'train.csv' TRAIN_DN = Path('train/images/') MASKS_DN = Path('train/masks/') TEST = Path('test/images/') PATH = Path('/kaggle/input/tgs-salt-identification-challenge/') PATH128 = Path('/tmp/128/') TMP = Path('/tmp/') MODEL = Path('/tmp/model/') PRETRAINED = Path('/kaggle/input/is-there-salt-resnet34/m...
learn.save('estagio-2' )
Cassava Leaf Disease Classification
13,643,882
train_names_png = [TRAIN_DN/f for f in os.listdir(PATH/TRAIN_DN)] train_names = list(seg.index.values) masks_names_png = [MASKS_DN/f for f in os.listdir(PATH/MASKS_DN)] test_names_png = [TEST/f for f in os.listdir(PATH/TEST)]<categorify>
sample_df = pd.read_csv(dataset_path/'sample_submission.csv') sample_df.head()
Cassava Leaf Disease Classification
13,643,882
with ThreadPoolExecutor(4)as e: e.map(resize_mask, train_names_png )<categorify>
_sample_df = sample_df.copy() _sample_df['path'] = _sample_df['image_id'].map(lambda x:dataset_path/'test_images'/x) _sample_df = _sample_df.drop(columns=['image_id']) test_dl = dls.test_dl(_sample_df )
Cassava Leaf Disease Classification
13,643,882
with ThreadPoolExecutor(4)as e: e.map(resize_mask, masks_names_png )<categorify>
test_dl.show_batch()
Cassava Leaf Disease Classification
13,643,882
with ThreadPoolExecutor(4)as e: e.map(resize_mask, test_names_png )<define_variables>
preds, _ = learn.tta(dl=test_dl, n=8, beta=0 )
Cassava Leaf Disease Classification
13,643,882
PATH = PATH128<import_modules>
sample_df['label'] = preds.argmax(dim=-1 ).numpy()
Cassava Leaf Disease Classification
13,643,882
<compute_test_metric><EOS>
sample_df.to_csv('submission.csv',index=False )
Cassava Leaf Disease Classification
13,055,692
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<load_pretrained>
Path.ls = lambda x: list(x.iterdir()) device = torch.device("cuda" if torch.cuda.is_available() else "cpu") effnet_path = '.. /input/efficientnet-pytorch/' sys.path.append(effnet_path) test_path = Path(".. /input/cassava-leaf-disease-classification/test_images") test_fnames = test_path.ls()
Cassava Leaf Disease Classification
13,055,692
def get_base() : layers = cut_model(f(True), cut) return nn.Sequential(*layers) def load_pretrained(model, path): weights = torch.load(PRETRAINED, map_location=lambda storage, loc: storage) model.load_state_dict(weights, strict=False) return model<categorify>
test_df = pd.read_csv(".. /input/cassava-leaf-disease-classification/sample_submission.csv") train_df = pd.read_csv(".. /input/cassava-leaf-disease-classification/train.csv") num_classes = train_df['label'].nunique()
Cassava Leaf Disease Classification
13,055,692
class SaveFeatures() : features=None def __init__(self, m): self.hook = m.register_forward_hook(self.hook_fn) def hook_fn(self, module, input, output): self.features = output def remove(self): self.hook.remove()<concatenate>
mean = [0.4589, 0.5314, 0.3236] std = [0.2272, 0.2297, 0.2200] test_tfms = albumentations.Compose([ albumentations.RandomResizedCrop(256, 256), albumentations.HorizontalFlip(p=0.5), albumentations.HueSaturationValue( hue_shift_limit=0.2, sat_shift_limit=0.2, val_shift_limit=0.2, p=0.5 ), albumentations.RandomBrightne...
Cassava Leaf Disease Classification
13,055,692
class UnetBlock(nn.Module): def __init__(self, up_in, x_in, n_out): super().__init__() up_out = x_out = n_out//2 self.x_conv = nn.Conv2d(x_in, x_out, 1) self.tr_conv = nn.ConvTranspose2d(up_in, up_out, 2, stride=2) self.bn = nn.BatchNorm2d(n_out) def forward(self, up_p, x_p): up_p = self.tr_conv(up_p) x_p = self.x_...
test_dl = make_dataloaders(df=test_df, split="test" )
Cassava Leaf Disease Classification
13,055,692
class UnetModel() : def __init__(self,model,name='unet'): self.model,self.name = model,name def get_layer_groups(self, precompute): lgs = list(split_by_idxs(children(self.model.rn), [lr_cut])) return lgs + [children(self.model)[1:]]<prepare_x_and_y>
class EfficientNetModel(nn.Module): def __init__(self, arch="b4", dropout=0.2, n_out=5, pretrained=True, freeze=True): super().__init__() if pretrained: self.model = EfficientNet.from_pretrained(f"efficientnet-{arch}") if freeze: for p in self.model.parameters() : p.requires_grad = False else: self.model = EfficientNe...
Cassava Leaf Disease Classification
13,055,692
x_names = [f'{x}.png' for x in train_names] x_names_path = np.array([str(TRAIN_DN/x)for x in x_names]) y_names = [x for x in x_names] y_names_path = np.array([str(MASKS_DN/x)for x in x_names] )<set_options>
model = model = EfficientNetModel(pretrained=False, freeze=False ).to(device) model.load_state_dict(torch.load(".. /input/pytorch-better-normalization-onecycle-lr-train/effnet.pt", map_location=device)) model.eval() ;
Cassava Leaf Disease Classification
13,055,692
aug_tfms = [RandomRotate(4, tfm_y=TfmType.CLASS), RandomFlip(tfm_y=TfmType.CLASS), RandomLighting(0.05, 0.05, tfm_y=TfmType.CLASS)] <split>
def inference_one_pass(model, test_dl): model.eval() all_preds = [] with torch.no_grad() : for batch in tqdm(test_dl): preds = model(batch.to(device)) all_preds.append(preds) return torch.cat(all_preds, dim=0 )
Cassava Leaf Disease Classification
13,055,692
lr=3e-3 wd=1e-7 lrs = np.array([lr/100,lr/10,lr]) n_folds = 8 out=np.zeros(( 18000,sz,sz)) alpha = 0 for i in range(n_folds): val_size = 4000//n_folds val_idxs=list(range(i*val_size,(i+1)*val_size)) (( val_x,trn_x),(val_y,trn_y)) = split_by_idx(val_idxs, x_names_path, y_names_path) test_x = np.array(test_names_png) ...
num_passes = 5 tta = None for _ in range(num_passes): all_preds = inference_one_pass(model, test_dl) if tta is None: tta = all_preds else: tta += all_preds tta /= float(num_passes) label_preds = tta.argmax(dim=1 )
Cassava Leaf Disease Classification
13,055,692
out = out/n_folds alpha = alpha/n_folds<categorify>
test_df['label'] = label_preds.cpu().numpy()
Cassava Leaf Disease Classification
13,055,692
<create_dataframe><EOS>
test_df.to_csv("submission.csv", index=False )
Cassava Leaf Disease Classification
13,021,276
<SOS> metric: CategorizationAccuracy Kaggle data source: cassava-leaf-disease-classification<set_options>
tez_path = '.. /input/tez-lib/' effnet_path = '.. /input/efficientnet-pytorch/' sys.path.append(tez_path) sys.path.append(effnet_path )
Cassava Leaf Disease Classification
13,021,276
plt.style.use('seaborn-white') sns.set_style("white") <train_model>
import os import albumentations import pandas as pd import numpy as np import tez from tez.datasets import ImageDataset import torch import torch.nn as nn from torch.nn import functional as F from efficientnet_pytorch import EfficientNet
Cassava Leaf Disease Classification
13,021,276
img_size_ori = 101 img_size_target = 128 def upsample(img): if img_size_ori == img_size_target: return img return resize(img,(img_size_target, img_size_target), mode='constant', preserve_range=True) def downsample(img): if img_size_ori == img_size_target: return img return resize(img,(img_size_ori, img_size_ori), mode...
class LeafModel(tez.Model): def __init__(self, num_classes): super().__init__() self.effnet = EfficientNet.from_name("efficientnet-b4") self.dropout = nn.Dropout(0.1) self.out = nn.Linear(1792, num_classes) self.step_scheduler_after = "epoch" def forward(self, image, targets=None): batch_size, _, _, _ = image.shape ...
Cassava Leaf Disease Classification
13,021,276
debug=False train_df = pd.read_csv(".. /input/tgs-salt-identification-challenge/train.csv", index_col="id", usecols=[0]) depths_df = pd.read_csv(".. /input/tgs-salt-identification-challenge/depths.csv", index_col="id") train_df = train_df.join(depths_df) test_df = depths_df[~depths_df.index.isin(train_df.index)] if ...
test_aug = albumentations.Compose([ albumentations.RandomResizedCrop(256, 256), albumentations.Transpose(p=0.5), albumentations.HorizontalFlip(p=0.5), albumentations.VerticalFlip(p=0.5), albumentations.HueSaturationValue( hue_shift_limit=0.2, sat_shift_limit=0.2, val_shift_limit=0.2, p=0.5 ), albumentations.RandomBri...
Cassava Leaf Disease Classification
13,021,276
def dice_coef(y_true, y_pred): y_true_f = K.flatten(y_true) y_pred = K.cast(y_pred, 'float32') y_pred_f = K.cast(K.greater(K.flatten(y_pred), 0.5), 'float32') intersection = y_true_f * y_pred_f score = 2.* K.sum(intersection)/(K.sum(y_true_f)+ K.sum(y_pred_f)) return score def dice_loss(y_true, y_pred): smooth = 1. ...
dfx = pd.read_csv(".. /input/cassava-leaf-disease-classification/sample_submission.csv") image_path = ".. /input/cassava-leaf-disease-classification/test_images/" test_image_paths = [os.path.join(image_path, x)for x in dfx.image_id.values] test_targets = dfx.label.values test_dataset = ImageDataset( image_paths=test_...
Cassava Leaf Disease Classification
13,021,276
train_df["images"] = [np.array(load_img(".. /input/tgs-salt-identification-challenge/train/images/{}.png".format(idx), grayscale=True)) / 255 for idx in tqdm_notebook(train_df.index)]<feature_engineering>
train_dfx = pd.read_csv(".. /input/cassava-leaf-disease-classification/train.csv") model = LeafModel(num_classes=train_dfx.label.nunique()) model.load(".. /input/leafmodel/model.bin" )
Cassava Leaf Disease Classification
13,021,276
train_df["masks"] = [np.array(load_img(".. /input/tgs-salt-identification-challenge/train/masks/{}.png".format(idx), grayscale=True)) / 255 for idx in tqdm_notebook(train_df.index)]<feature_engineering>
final_preds = None for j in range(5): preds = model.predict(test_dataset, batch_size=32, n_jobs=-1, device="cuda") temp_preds = None for p in preds: if temp_preds is None: temp_preds = p else: temp_preds = np.vstack(( temp_preds, p)) if final_preds is None: final_preds = temp_preds else: final_preds += temp_preds fina...
Cassava Leaf Disease Classification
13,021,276
train_df["coverage"] = train_df.masks.map(np.sum)/ pow(img_size_ori, 2 )<categorify>
final_preds = final_preds.argmax(axis=1 )
Cassava Leaf Disease Classification
13,021,276
def cov_to_class(val): for i in range(0, 11): if val * 10 <= i : return i train_df["coverage_class"] = train_df.coverage.map(cov_to_class )<define_search_model>
dfx.label = final_preds
Cassava Leaf Disease Classification
13,021,276
def conv_block(m, dim, acti, bn, res, do=0): n = Conv2D(dim, 3, activation=acti, padding='same' )(m) n = BatchNormalization()(n)if bn else n n = Dropout(do )(n)if do else n n = Conv2D(dim, 3, activation=acti, padding='same' )(n) n = BatchNormalization()(n)if bn else n return Concatenate()([m, n])if res else n d...
dfx.to_csv("submission.csv", index=False )
Cassava Leaf Disease Classification
13,048,268
n_split=6 skf=StratifiedKFold(n_splits=n_split) models=[] historys=[] epochs = 200 batch_size = 32 if debug: epochs=3 sub_model_list=[1,0,2,3,5,2] for i in range(n_split): print('reading '+str(i)+' model') if i==1: data_root='baseline-0-760-0-143-6fold-split-1st/' elif i==2: data_root='fork-of-baseline-0-760-0-143-6f...
%matplotlib inline pd.set_option('display.max_rows', None) pd.set_option('display.max_columns', None )
Cassava Leaf Disease Classification
13,048,268
threshes=[] for i,(trdex,valdex)in enumerate(skf.split(X=train_df.index.values,y=train_df.coverage_class.values)) : ids_valid=train_df.index.values[valdex] y_valid=np.array(train_df.loc[ids_valid].masks.map(upsample ).tolist() ).reshape(-1, img_size_target, img_size_target, 1) x_valid=np.array(train_df.loc[ids_valid]....
data_dir = '/kaggle/input/cassava-leaf-disease-classification' train = pd.read_csv(os.path.join(data_dir, 'train.csv')) sub = pd.read_csv(os.path.join(data_dir, 'sample_submission.csv'))
Cassava Leaf Disease Classification
13,048,268
def RLenc(img, order='F', format=True): bytes = img.reshape(img.shape[0] * img.shape[1], order=order) runs = [] r = 0 pos = 1 for c in bytes: if(c == 0): if r != 0: runs.append(( pos, r)) pos += r r = 0 pos += 1 else: r += 1 if r != 0: runs.append(( pos, r)) pos += r r = 0 if format: z = '' for rr in runs: z += '{} ...
def seed_everything(seed): random.seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = True
Cassava Leaf Disease Classification