Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Correct a typo :) #11

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Correct a typo :) #11

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion workspace/uncond/Experiments.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -102,4 +102,4 @@ For the CP+Linear setting, it only takes less than half minutes to generate a so
We find that the generated pieces of REMI-XL tend to stick to some patterns and occasionally fall to loop them for a quite long time, or even the entire song. The quality within the loop is suprisingly organized (clearly arpeggio in left hand, melody line in right hand), but also a bit of tedious. The samples from CP-linear have a rather different texture, the "structure" diversity is richer but it's also more aggressive in selecting pitches.

### About EOS
It turns out that REMI-XL failed in generating EOS sequence, which implies the sequence length might exceed the "effective length" becuase of it's rnn nature. In fact, we also tried remi+linear and cp+linear, and both of them success in this criterion.
It turns out that REMI-XL failed in generating EOS sequence, which implies the sequence length might exceed the "effective length" becuase of it's rnn nature. In fact, we also tried remi+linear and cp+linear, and both of them succeed in this criterion.