KimberleyJSN

/

melbandroformer

like 30

Hi Ryan

I would love to see my model used in Audacity and you have permission to use it any way you like, sorry for not putting a license I don't have any experience with them and they seem quite legally complicated, is me saying you have permission ok? I also gave you permission on my github repo for the model https://github.com/KimberleyJensen/Mel-Band-Roformer-Vocal-Model/issues/7 .

I hope this ok please let me know if it isn't.

Hi Kimberley,
Thanks for the response! No worries, let me ask if having your permission alone is enough to satisfy the requirements we'd need to distribute the model, etc.

Anyway, really appreciate your work!

Hello! Just to follow up on this... it seems that explicit permission isn't really enough for us to distribute this model with our Audacity plugins. To do so, I think we need it to be tagged with an open source license.
For example, we will be releasing this dereverb model with our next release (https://huggingface.co/anvuew/dereverb_mel_band_roformer) . We're able to do so as this dereverb model repo is tagged as gpl-3.0.

Thanks!
Ryan

Hey Ryan I have now added a gpl-3.0 license, sorry about not having one before, the licenses are actually a lot less complicated than I thought.

Fantastic! And no worries at all.

Thanks so much! We'll include this model for our next big update to our plugins -- super excited to give users the ability to extract vocals with this model.. it's sooo good!

Hi @KimberlyJSN, I left you a note on your github. I was wondering if you would consider changing the License to MIT or Apache? It would make it simpler for inclusion with other projects as GPL 3 can be a little restrictive and including it in upstream work might not be feasible. I'd specifically like to create a MLX set of weights based on your work to provide an option for MacOS.