Prior attentions in Aligner Model #3209
PranjalyaDS
started this conversation in
General
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
In the unsupervised alignment model, which is used in MLP Mixer TTS model as well, in the dataloader a
attn_priorcomponent is used. I wanted to know what is its role in the Alignment module and its training?Also, how resourceful and necessary is it for better training of the model?
Thank you.
Beta Was this translation helpful? Give feedback.
All reactions