wrn_c.py - production (composable)
Example: Instantiate a stock WRN model
from wrn_c import WRN
wrn = WRN(depth=28, k=10)
Example: Compose and Train a WRN model
example():
''' Example for constructing/training a WRN model on CIFAR-10
'''
# Example of constructing a mini WRN
wrn = WRN(depth=14, k=2, input_shape=(32, 32, 3), n_classes=10)
wrn.model.compile(loss='sparse_categorical_crossentropy', optimizer='adam', metrics=['acc'])
wrn.model.summary()
wrn.cifar10()