Very good content, I felt like you missed some points here and there but I really like how you expand every keyword encountered and provide some short context about it. It was definitely useful to understand the paper better.
@camdenparsons5114
Жыл бұрын
So glad they open-source their audio tokenizer. Hope Google will follow suit in next releases
@user-kn2vv3rv3i
9 ай бұрын
I am wondering is it possible to train a model like MusicGen on a local PC with one RTX 4090 or a pair of them? How long it may take for it? I am also interested to know the estimated time for fine-tuning the model?
@eamram
Жыл бұрын
Could you do a 10-15 minute summary, rather than 2:30 hours?
@thesahil301290
Жыл бұрын
Please let him do the good work rather than making a summary so that you could have a gist about it in less time. These long videos would be helpful for people who really want to build something, and not just plug and play.
Пікірлер: 5