def main(args):
test_ds = MnistDataset(
args.test_image_file,
args.test_label_file,
transform=transforms.Compose([ToTensor()]),
)
test_loader = torch.utils.data.DataLoader(
test_ds,
batch_size=args.batch_size,
collate_fn=collate_fn,
shuffle=False,
)
model = Net().to(device)
model.load_state_dict(torch.load(args.checkpoint))
model.eval()
predicts = []
truths = []
with torch.no_grad():
for i, sample in enumerate(test_loader):
X, Y_true = sample["X"].to(device), sample["Y"].to(device)
output = model(X)
predicts.append(torch.argmax(output, dim=1))
truths.append(Y_true)
predicts = torch.cat(predicts, dim=0)
truths = torch.cat(truths, dim=0)
acc = torch.sum(torch.eq(predicts, truths))
print("Acc: {:.4f}".format(acc / len(predicts)))
def train(epochs, train_loader, dev_loader, lr, seed, log_interval,
output_dir):
"""Train the model. Store snapshot models in the output_dir alongside
evaluations on the dev set after each epoch
"""
model = Net()
optimizer = optim.Adam(model.parameters(), lr=lr)
measure_size(model)
use_cuda = torch.cuda.is_available()
device = torch.device("cuda:0" if use_cuda else "cpu")
print("Using device: ", device)
if use_cuda:
torch.cuda.manual_seed(seed)
else:
torch.manual_seed(seed)
#torch.backends.cudnn.benchmark = False
#torch.backends.cudnn.deterministic = True
model.to(device)
for epoch in range(1, epochs):
model.train()
total_loss = 0.0
for batch_idx, (data, target) in enumerate(train_loader):
if use_cuda:
data, target = data.to(device), target.to(device)
data = data.unsqueeze_(1)
optimizer.zero_grad()
output = model(data)
loss = F.nll_loss(output, target)
total_loss += loss.item()
loss.backward()
optimizer.step()
if batch_idx % log_interval == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
print("Total loss = %.6f" % (total_loss / len(train_loader.dataset)))
test(model, dev_loader,
os.path.join(output_dir, 'dev-eer-' + str(epoch)))
torch.save(model, os.path.join(output_dir,
'iter' + str(epoch) + '.mdl'))
def load_model(model_path, TEXT=None, LABEL=None):
#for saved model (.pt)
if '.pt' in model_path:
if torch.typename(torch.load(model_path)) == 'OrderedDict':
if 'tut' in model_path:
INPUT_DIM = len(TEXT.vocab)
EMBEDDING_DIM = 100
N_FILTERS = 100
FILTER_SIZES = [3, 4, 5]
OUTPUT_DIM = 1
DROPOUT = 0.5
PAD_IDX = TEXT.vocab.stoi[TEXT.pad_token]
model = CNN(INPUT_DIM, EMBEDDING_DIM, N_FILTERS, FILTER_SIZES,
OUTPUT_DIM, DROPOUT, PAD_IDX)
elif 'mnist' in model_path:
model = Net()
elif 'HELOC' or 'heloc' in model_path:
input_size = 22
model = MLP(input_size)
model.load_state_dict(torch.load(model_path))
else:
model = torch.load(model_path)
#for pretrained model
elif model_path == 'VGG19':
model = models.vgg19(pretrained=True)
elif model_path == 'ResNet50':
model = models.resnet50(pretrained=True)
elif model_path == 'DenseNet161':
model = models.densenet161(pretrained=True)
model.eval()
if cuda_available():
model.cuda()
return model
cpu = False
if cpu:
device = torch.device('cpu')
else:
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# Image transformations
transform = transforms.Compose([
transforms.Resize(size=(32, 32)),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])
model = Net()
# 加载模型
print(model)
if cpu:
checkpoint = torch.load('./weights/chen_liveness_mobilenetv2.pth.tar',
map_location=lambda storage, loc: storage)
else:
checkpoint = torch.load('./weights/chen_liveness_mobilenetv2.pth.tar')
print('gpu')
new_state_dict = OrderedDict()
# 用了nn.DataParallel的模型需要处理才能在cpu上使用
'''for k, v in checkpoint.items():
print('===> Loading datasets')
train_set = get_training_set(opt.data_dir, opt.lr_train_dataset,
opt.hr_train_dataset, opt.upscale_factor,
opt.patch_size, opt.data_augmentation)
training_data_loader = DataLoader(dataset=train_set,
num_workers=opt.threads,
batch_size=opt.batchSize,
shuffle=True)
print('===> Building model ', opt.model_type)
#model = SRCNN()
#model = VDSR()
#model = LGCNet()
#model = EEGAN(4)
model = Mymodel(3, 64, opt.scale)
#model = DDBPN(opt)
#model = RDN(opt.scale)
#model = RCAN(opt)
#model = SAN(opt)
#model = torch.nn.DataParallel(model, device_ids=gpus_list)
criterion = nn.L1Loss() #nn.MSELoss()#nn.L1Loss()
#################################
#################################
# pre_net = Mymodel_pre(3,64,4)
# pre_path = 'checkpoints_my_x4/model_my_x4_epoch_929.pth'
# checkpoint = torch.load(pre_path)
# pre_net.load_state_dict(checkpoint["model"].state_dict())
# author: frendy
# site: http://frendy.vip/
# time: 28/06/2017
import os
import torch
import torch.optim as optim
import torch.nn as nn
from torch.autograd import Variable
from config import MODEL_PATH, DATA_PATH
from models.model import transform, Net
from data import loadTrainData
trainset, trainloader = loadTrainData()
net = Net()
if os.path.exists(MODEL_PATH):
net.load_state_dict(torch.load(MODEL_PATH))
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)
for epoch in range(2): # loop over the dataset multiple times
running_loss = 0.0
for i, data in enumerate(trainloader, 0):
# get the inputs
inputs, labels = data
# wrap them in Variable
inputs, labels = Variable(inputs), Variable(labels)
# author: frendy
# site: http://frendy.vip/
# time: 28/06/2017
import torch
import torchvision
from torch.autograd import Variable
from config import MODEL_PATH
from models.model import Net, classes
from data import loadTestData
from utils.image import imshow
testset, testloader = loadTestData()
net = Net()
net.load_state_dict(torch.load(MODEL_PATH))
dataiter = iter(testloader)
images, labels = dataiter.next()
# print images
imshow(torchvision.utils.make_grid(images))
print('GroundTruth: ', ' '.join('%5s' % classes[labels[j]] for j in range(4)))
outputs = net(Variable(images))
_, predicted = torch.max(outputs.data, 1)
print('Predicted: ',
' '.join('%5s' % classes[predicted[j][0]] for j in range(4)))
import torch
import torch.optim as optim
import utils.train_repro_util as repro_util
from models.model import NetVanilla as Net
import utils.utils_global as utils_global
from utils.utils_global import ProgressBar
from properties.properties_repro import PropertiesRepro
from properties.properties_global import PropertiesGlobal
properties_global = PropertiesGlobal()
properties_repro = PropertiesRepro()
training_generator, validation_generator = properties_repro.get_dataloaders()
# declare net and loss
net = Net()
# load model
net.load_state_dict(
torch.load(properties_repro.path_model_to_load,
map_location=properties_repro.device))
# send net to cuda
net.to(properties_repro.device)
# define optimizer
optimizer = optim.Adam(net.parameters(), lr=properties_repro.lr)
# define learning rate scheduling
milestones = np.array([25, 37, 50, 62])
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer,
milestones=milestones,
gamma=0.5)
'train':
datasets.ImageFolder(root = os.path.join(data_path, 'train'),
transform=image_transforms['train']),
'valid':
datasets.ImageFolder(root = os.path.join(data_path, 'valid'),
transform=image_transforms['valid'])
}
# Dataloader iterators, make sure to shuffle
dataloaders = {
'train': DataLoader(data['train'], batch_size = 64, shuffle = True, num_workers = 8, pin_memory = True),
'valid': DataLoader(data['valid'], batch_size =64, shuffle = True, num_workers = 8, pin_memory = True),
}
net = Net()
print(net)
net.to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.01, momentum=0.9, weight_decay=5e-5)
epoches=300
eval_acc_list = []
is_best = False
###训练网络
for epoch in range(epoches):
train_loss = 0.
train_acc = 0.
for inputs, targets in dataloaders['train']:
sys.path.append("../")
from utils.loading import load_data
from models.model import Net
from torch.utils.data import DataLoader
with open("../data/configuration.json", "r") as file:
data = json.load(file)
data = json.loads(json.dumps(data))
path = data['data']['paths']["train"]
type_ = "train/audio"
BATCH_SIZE = 32
N_EPOCHS = 5
N_classes = 30
lr = 0.001
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
net = Net(BATCH_SIZE, 1, 128, 1).to(device)
MAX_NUM_WAVS_PER_CLASS = 2**27 - 1 # ~134M
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(net.parameters(), lr=lr, momentum=0.9)
def train(epoch, data):
net.train().to(device)
# zero the parameter gradients
optimizer.zero_grad()
inputs, labels = data
# print(type(inputs))
inputs = torch.from_numpy(np.asarray(inputs).astype(np.float32))
permutation = torch.randperm(inputs.size()[0])
running_loss = 0
class Trainer(object):
def __init__(self, args):
self.args = args
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
self.prepare_data()
self.setup_train()
def prepare_data(self):
train_val = MnistDataset(
self.args.train_image_file,
self.args.train_label_file,
transform=transforms.Compose([ToTensor()]),
)
train_len = int(0.8 * len(train_val))
train_ds, val_ds = torch.utils.data.random_split(
train_val, [train_len, len(train_val) - train_len]
)
print("Train {}, val {}".format(len(train_ds), len(val_ds)))
self.train_loader = torch.utils.data.DataLoader(
train_ds,
batch_size=self.args.batch_size,
collate_fn=collate_fn,
shuffle=True,
)
self.val_loader = torch.utils.data.DataLoader(
val_ds,
batch_size=self.args.batch_size,
collate_fn=collate_fn,
shuffle=False,
)
def setup_train(self):
self.model = Net().to(self.device)
self.optimizer = torch.optim.SGD(self.model.parameters(), lr=self.args.lr)
self.criterion = nn.CrossEntropyLoss().to(self.device)
if not os.path.isdir(self.args.ckpt):
os.mkdir(self.args.ckpt)
def train_one_epoch(self):
train_loss = 0.0
self.model.train()
for i, sample in enumerate(self.train_loader):
X, Y_true = sample["X"].to(self.device), sample["Y"].to(self.device)
self.optimizer.zero_grad()
output = self.model(X)
loss = self.criterion(output, Y_true)
loss.backward()
self.optimizer.step()
train_loss += loss.item()
return train_loss / len(self.train_loader)
def evaluate(self):
val_loss = 0.0
self.model.eval()
predicts = []
truths = []
with torch.no_grad():
for i, sample in enumerate(self.val_loader):
X, Y_true = sample["X"].to(self.device), sample["Y"].to(self.device)
output = self.model(X)
loss = self.criterion(output, Y_true)
val_loss += loss.item()
predicts.append(torch.argmax(output, dim=1))
truths.append(Y_true)
predicts = torch.cat(predicts, dim=0)
truths = torch.cat(truths, dim=0)
acc = torch.sum(torch.eq(predicts, truths))
return acc / len(predicts), val_loss / (len(self.val_loader))
def run(self):
min_loss = 10e4
max_acc = 0
for epoch in range(self.args.epochs):
train_loss = self.train_one_epoch()
val_acc, val_loss = self.evaluate()
if val_acc > max_acc:
max_acc = val_acc
torch.save(
self.model.state_dict(),
os.path.join(
self.args.ckpt,
"{}_{}_{:.4f}.pth".format(self.args.name, epoch, max_acc),
),
)
print(
"Epoch {}, loss {:.4f}, val_acc {:.4f}".format(
epoch, train_loss, val_acc
)
)
import torch.optim as optim
import utils.utils_global as utils_global
from models.model import NetVanilla as Net
import utils.train_dsac_util as ransac_util
import utils.train_dsac_sub as sub
from utils.utils_global import ProgressBar
from properties.properties_dsac import PropertiesDsac
from properties.properties_global import PropertiesGlobal
properties_global = PropertiesGlobal()
properties_dsac = PropertiesDsac()
# Loading pre-trained model
model = Net()
model.load_state_dict(torch.load(properties_dsac.model_path, map_location=properties_global.device))
model.to(properties_global.device)
training_generator, validation_generator = properties_dsac.get_dataloaders()
optimizer = optim.Adam(model.parameters(), lr=properties_dsac.learning_rate)
print('INFO TRAINING: ')
print('repro error threshold: {} pixels'.format(properties_dsac.inlier_threshold))
print('number of refinement iters: {}'.format(properties_dsac.number_refinement_iterations))
if properties_dsac.starting_epoch == 0:
losses_log = np.zeros((properties_dsac.max_epochs, 3))
else:
losses_old = np.loadtxt(os.path.join(properties_dsac.save_in_folder, 'log ransac.txt'))
use_cuda = torch.cuda.is_available()
state_dict = torch.load(config["model"])
df = pd.read_csv('kaggle_template.csv')
features_test = np.load(
os.path.join(config["load_dir"] + '/_feature_test_assigment.npy'))
features_test_crop = np.load(
os.path.join(config["load_dir"] + '/_feature_test_assigment_crop.npy'))
if config["concatenate"]:
features_test = np.concatenate((features_test, features_test_crop), axis=1)
#Model
model = Net(features_test.shape[1])
model.load_state_dict(state_dict['model'])
model.eval()
if use_cuda:
print('Using GPU')
model.cuda()
else:
print('Using CPU')
for dossier, sous_dossiers, fichiers in os.walk(config["test_dir"]):
for num, fichier in enumerate(fichiers):
num_photo = df.loc[df['Id'] == fichier.split('.')[0]].index[0]
data = torch.tensor(features_test[num])
if use_cuda:
data = data.cuda()
dataloders = {
x: torch.utils.data.DataLoader(features_datasets[x],
batch_size=config["batchsize"],
shuffle=True,
num_workers=4)
for x in ['train_images', 'val_images']
}
dataset_sizes = {
x: len(features_datasets[x])
for x in ['train_images', 'val_images']
}
# Neural network and optimizer
# We define neural net in model.py so that it can be reused by the evaluate.py script
model = Net(features_train.shape[1])
print(model)
if use_cuda:
print('Using GPU')
model.cuda()
else:
print('Using CPU')
optimizer = optim.SGD(model.parameters(),
lr=config["lr"],
momentum=config["momentum"])
criterion = nn.CrossEntropyLoss()
# Run the functions and save the best model in the function model_ft.
def setup_train(self):
self.model = Net().to(self.device)
self.optimizer = torch.optim.SGD(self.model.parameters(), lr=self.args.lr)
self.criterion = nn.CrossEntropyLoss().to(self.device)
if not os.path.isdir(self.args.ckpt):
os.mkdir(self.args.ckpt)
cpu = False
if cpu:
device = torch.device('cpu')
else:
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# Image transformations
transform = transforms.Compose([
transforms.Resize(size=(32, 32)),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.226, 0.226, 0.226])
])
model = Net()
# 加载模型
print(model)
if cpu:
checkpoint = torch.load('./weights/best_model-20200904.pth.tar',
map_location=lambda storage, loc: storage)
else:
checkpoint = torch.load('./weights/best_model-20200904.pth.tar')
#new_state_dict = OrderedDict()
# 用了nn.DataParallel的模型需要处理才能在cpu上使用
'''for k, v in checkpoint.items():
name = k[7:] # remove module.
dataloaders = {
'train':
DataLoader(data['train'],
batch_size=args.batch_size,
shuffle=True,
num_workers=8,
pin_memory=True),
'valid':
DataLoader(data['valid'],
batch_size=args.batch_size,
shuffle=True,
num_workers=8,
pin_memory=True),
}
model = Net()
#model = MobileNet(inputs)
MOMENTUM = 0.9
WEIGHT_DECAY = 5e-5
print(model)
'''if args.resume_weights:
mask_checkpoint = torch.load(args.resume_weights)
new_mask_state_dict = OrderedDict()
for k, v in mask_checkpoint.items():
name = k[7:] # remove module.
new_mask_state_dict[name] = v