Open sonfire186 opened 4 years ago
I see different examples in README.md and archs.py
x = Dense(args.num_features, kernel_initializer='he_normal', kernel_regularizer=regularizers.l2(weight_decay))(x) x = BatchNormalization()(x) output = ArcFace(10, regularizer=regularizers.l2(weight_decay))([x, y])
x = Dense(512, kernel_initializer='he_normal')(x) x = BatchNormalization()(x) output = ArcFace(num_classes=10)([x, y])
Which examples is correct?
The first one seems better as it includes weight decay regularization, but both are correct.
I see different examples in README.md and archs.py
Which examples is correct?