squirrelon

Hi, It is actually not using transformers, those would be too slow. It is using a combination of CNN's and linear layers. Correct, it uses embedings, not waveforms or spectrograms. The inputs are midis, some of which I made myself in FL Studio. The model creates a "latent representation" from each midi, I can then sample randomly from this latent space to get an original piece. The most important part is the preprocessing in my opinion.

Feel free to comment any enhancement suggestions or points you have.

Had a colleague submit a paper with literal AI slop left in the text, got hit with a nasty revision request. Check your drafts before you submit, people. The reviewers will find it.

Solo founders typically spend so much time building an app completely invested in. They start neglecting the UX (marketing as well) and then they wonder why the app is not successful.

I would definitely suggest sending the beta beta version to friends and family before investing 90% more time on it.

I think this is actually valid, if you think about it. Some days I go by with constantly thinking be it about work, in messages, or simply on social media. Taking the time to stare at walls actually provides ability to step back, calm down, and actually random thoughts will start appearing out of which some can be insightful.

40k people are not under thread, I am getting AI contractor job offers every month on UpWork, I am glad I haven't accepted more than one as it is just not worth to do.