def __init__(self):
self.name = 'densenet'
self.model_filename = 'networks/models/densenet.h5'
self.growth_rate = 12
self.depth = 100
self.compression = 0.5
self.num_classes = 10
self.img_rows, self.img_cols = 32, 32
self.img_channels = 3
self.batch_size = 64 # 64 or 32 or other
self.epochs = 250
self.iterations = 782
self.weight_decay = 0.0001
self.log_filepath = r'networks/models/densenet/'
self.acc = 0.9467 # Precalculated result for cifar10
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self):
self.name = 'wide_resnet'
self.model_filename = 'networks/models/wide_resnet.h5'
self.depth = 16
self.wide = 8
self.num_classes = 10
self.img_rows, self.img_cols = 32, 32
self.img_channels = 3
self.batch_size = 128
self.epochs = 200
self.iterations = 391
self.weight_decay = 0.0005
self.log_filepath = r'networks/models/wide_resnet/'
self.acc = 0.9534 # Precalculated result for cifar10
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self):
self.name = 'lecun_net'
self.model_filename = 'networks/models/lecun_net.h5'
self.num_classes = 10
self.input_shape = 32, 32, 3
self.batch_size = 128
self.epochs = 200
self.iterations = 391
self.weight_decay = 0.0001
self.log_filepath = r'networks/models/lecun_net/'
self.acc = 0.7488 # Precalculated result for cifar10
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self, epochs=350, batch_size=128, load_weights=True):
self.name = 'pure_cnn'
self.model_filename = 'networks/models/pure_cnn.h5'
self.num_classes = 10
self.input_shape = 32, 32, 3
self.batch_size = batch_size
self.epochs = epochs
self.learn_rate = 1.0e-4
self.log_filepath = r'networks/models/pure_cnn/'
if load_weights:
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self, epochs=200, batch_size=128, load_weights=True):
self.name = 'lecun_net'
self.model_filename = 'networks/models/lecun_net.h5'
self.num_classes = 10
self.input_shape = 32, 32, 3
self.batch_size = batch_size
self.epochs = epochs
self.iterations = 391
self.weight_decay = 0.0001
self.log_filepath = r'networks/models/lecun_net/'
if load_weights:
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self):
self.name = 'capsnet'
self.model_filename = 'networks/models/capsnet.h5'
self.num_classes = 10
self.input_shape = 32, 32, 3
self.num_routes = 3
self.batch_size = 128
self._model = CapsNetv1(input_shape=self.input_shape,
n_class=self.num_classes,
n_route=self.num_routes)
try:
self._model.load_weights(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model.load_weights(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
def __init__(self, epochs=250, batch_size=64, load_weights=True):
self.name = 'densenet'
self.model_filename = 'networks/models/densenet.h5'
self.growth_rate = 12
self.depth = 100
self.compression = 0.5
self.num_classes = 10
self.img_rows, self.img_cols = 32, 32
self.img_channels = 3
self.batch_size = batch_size
self.epochs = epochs
self.iterations = 782
self.weight_decay = 0.0001
self.log_filepath = r'networks/models/densenet/'
if load_weights:
try:
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to load', self.name)
print('Downloading model')
try:
download_model(self.name)
self._model = load_model(self.model_filename)
self.param_count = self._model.count_params()
print('Successfully loaded', self.name)
except (ImportError, ValueError, OSError):
print('Failed to download model')
