Open jeelsheth1010 opened 5 years ago
In paper, Table 1 states that including shared embeddings, hierarchical latent space and Orthogonal Regularization in model gives the best results. So does this implementation include all the above three things?
It is a simple implementation without tags, and just the truncated latent is used now.
@ewrfcas @lkj8389 Thanks for the response. Have u conditioned ur model with labels?
Regards
In paper, Table 1 states that including shared embeddings, hierarchical latent space and Orthogonal Regularization in model gives the best results. So does this implementation include all the above three things?