Open hugddygff opened 6 years ago
The options of decoder does not have the usage of model ensemble .
Does someone implement it ?
And can you share the code or the experience to implement it?
Thanks!
W_0 = utils.weight_variable([FLAGS.z_dim, 64 * GEN_DIMENSION / 2 * IMAGE_SIZE / 16 * IMAGE_SIZE / 16],
NameError: name 'utils' is not defined
It would be great if someone can add this ensembling feature :)
The options of decoder does not have the usage of model ensemble .
Does someone implement it ?
And can you share the code or the experience to implement it?
Thanks!