vykintas
@vykintas
I think given current reliability of generative models (LLMs, stable diffusion etc.) the best use is to generate music. Cant wait until it will be released since its more fault tolerant than both visual art and text generation
1 reply
0 recast
0 reaction
vykintas
@vykintas
And it looks to be done: https://google-research.github.io/seanet/musiclm/examples/
0 reply
0 recast
0 reaction