def __init__(self, args):
self.lr = args.lr
self.wd = args.wd
self.epochs = 200
self.dataset = args.dataset
self.test_uncertainty = args.test_uncertainty
self.vote = args.vote
self.n_models = args.n_models
self.device = torch.device('cuda')
torch.manual_seed(8734)
self.ensemble = [
models.LeNet().to(self.device) for _ in range(self.n_models)
]
self.attach_optimizers()
if self.dataset == 'mnist':
self.data_train, self.data_test = datagen.load_mnist()
elif self.dataset == 'cifar':
self.data_train, self.data_test = datagen.load_cifar()
self.best_test_acc = 0.
self.best_test_loss = np.inf
print(self.ensemble[0], ' X {}'.format(self.n_models))
print('step1:load the datasets...')
data, label = read_img(path) # 调用read_img()函数,读取图片数据和对应的标签
x_train, y_train, x_val, y_val = shuffle_data(
data, label) # 调用shuffle_data()函数，打乱数据集，并划分数据集
# step2: 构建模型并开始训练、测试
print('step2: build the model and training...')
# 占位符
x = tf.placeholder(tf.float32, shape=[None, w, h, c], name='x')
y_ = tf.placeholder(tf.int32, shape=[
None,
], name='y_')
logits, pred = lenet.LeNet(x) # pred 是经过softmax处理过的
loss = tf.losses.sparse_softmax_cross_entropy(labels=y_, logits=logits)
#train_op = tf.train.AdadeltaOptimizer(learning_rate=0.001).minimize(loss)
train_op = tf.train.AdamOptimizer(learning_rate=0.001).minimize(
loss) # 和上面的优化器不同
correct_prediction = tf.equal(tf.cast(tf.argmax(logits, 1), tf.int32), y_)
acc = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
num_epoches = 10
batch_size = 64
train_losses = []
train_acces = []
val_losses = []
val_acces = []
sess = tf.InteractiveSession()
download=False) # 为True需要先下载，然后处理；为False使用下载完成的
test_dataset = datasets.MNIST(root='./MNIST_data',
train=False,
transform=data_tf)
train_loader = DataLoader(train_dataset, batch_size=batch_size,
shuffle=True) #训练集，shuffle为True时打乱数据，让数据更有选择随机性
test_loader = DataLoader(test_dataset, batch_size=batch_size,
shuffle=False) #对测试集进行迭代器编号
# 60000张训练集，batch_size为50 => 60000/50=1200,即每次模型输入50个数据，要进行1200次
# print(len(train_loader))
# 选择模型
model = lenet.LeNet()
#print(model) # 查看模型架构
criterion = nn.CrossEntropyLoss() # 定义损失函数
optimizer = optim.SGD(model.parameters(), lr=learning_rate) # 定义优化器
#'''训练之前定义一些空列表来添加训练得到的一些值 '''
train_losses = []
train_acces = []
eval_losses = []
eval_acces = []
for epoch in range(num_epoches):
# 每次epoch训练
train_loss = 0
train_acc = 0
model.train()
from models import preact_resnet_model
from models import alexnet
from models import lenet
_MODEL_DICT = {
'resnet18': resnet_model.ResNet18(),
'resnet34': resnet_model.ResNet34(),
'resnet50': resnet_model.ResNet50(),
'resnet101': resnet_model.ResNet101(),
'resnet152': resnet_model.ResNet152(),
'preact_resnet18': preact_resnet_model.PreActResNet18(),
'preact_resnet34': preact_resnet_model.PreActResNet34(),
'preact_resnet50': preact_resnet_model.PreActResNet50(),
'preact_resnet101': preact_resnet_model.PreActResNet101(),
'preact_resnet152': preact_resnet_model.PreActResNet152(),
'alexnet': alexnet.AlexNet(),
'lenet': lenet.LeNet()
}
def _invalid_model_name():
raise ValueError("Not a valid model name")
def fetch_teacher(model_name, model_dicts=_MODEL_DICT):
return model_dicts[model_name]
def fetch_student(model_name, model_dicts=_MODEL_DICT):
return model_dicts[model_name]
num_workers=2)
testset = torchvision.datasets.CIFAR100(root='./data',
train=False,
download=True,
transform=transform_test)
testloader = torch.utils.data.DataLoader(testset,
batch_size=100,
shuffle=False,
num_workers=2)
num_classes = 100
else:
raise "only support dataset CIFAR10 or CIFAR100"
if args.model == "lenet":
net = lenet.LeNet(num_classes=num_classes)
elif args.model == "vgg16":
net = vgg.vgg16(num_classes=num_classes, pretrained=args.pretrain)
elif args.model == "vgg16_bn":
net = vgg.vgg16_bn(num_classes=num_classes, pretrained=args.pretrain)
elif args.model == "resnet18":
net = resnet.resnet18(num_classes=num_classes, pretrained=args.pretrain)
elif args.model == "resnet34":
net = resnet.resnet18(num_classes=num_classes, pretrained=args.pretrain)
elif args.model == "resnet50":
net = resnet.resnet50(num_classes=num_classes, pretrained=args.pretrain)
elif args.model == "resnetv2_18":
net = resnet_v2.resnet18(num_classes=num_classes, pretrained=args.pretrain)