Open robinsonkwame opened 4 years ago
Needs a new block for
# Kente dataset parameters kente_val_frac = 1./5 kente_bias = True kente_rep_dim = 128 kente_architecture = 1 # choose one of the implemented architectures kente_normal = 0 kente_outlier = -1
I assume that the kente_val_frac is training dependent and not a population assumption.
kente_val_frac
Needs a new block for
I assume that the
kente_val_frac
is training dependent and not a population assumption.