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

Jump to bottom

Add composer prefix tokens to TokenizerLazy #23

Merged
merged 3 commits into from
Aug 12, 2023
Merged

Add composer prefix tokens to TokenizerLazy #23

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
049e0ed
added composer prefix tokens
loubbrad Aug 12, 2023
3e69b34
fix composer name config usage
loubbrad Aug 12, 2023
8c22f07
fix roadmap
loubbrad Aug 12, 2023
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
Prev Previous commit
fix roadmap
  • Loading branch information
@loubbrad
loubbrad committed Aug 12, 2023
commit 8c22f0727a777cbf5ac6619f785ed0fa02c0a5dd
2 changes: 1 addition & 1 deletion ROADMAP.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@ As it stands, the basic functionality of the repository is implemented and teste
* [x] **~~Add further pre-processing tests~~**

Add further MidiDict pre-processing tests to improve dataset quality. Some ideas are checking for the frequency of note messages (an average of > 15 p/s or < 2 p/s is a bad sign). I'm open to any suggestions for MidiDict preprocessing tests. Properly cleaning pre-training datasets has a huge effect on model quality and robustness.
* [] **~~Add meta-token prefix support for LazyTokenizer~~**
* [x] **~~Add meta-token prefix support for LazyTokenizer~~**

Investigate the possibility of adding meta-tokens to the prefix in LazyTokenizer. Some examples could be genre, composer, or data source tags. This might require a rewrite of how sequence prefixes are handled.
* [x] **~~Add 'ending soon' token to lazy tokenizer~~**
Expand Down