Spaces:

marcop

/

musika

like 333

Build error

If we compare audio-generation to image-generation, I can say that this AI is on the same level as DALL-E 1 is in terms of image-generation. Music is almost coherent. The fact that this model is released for free for anyone to use is mindblowing, big kudos to the creator of the space, Marco Pasini and Jan Schlüter. Although it would be cool to have more control, to change seed and other parameters