1 comment

[ 3.8 ms ] story [ 15.1 ms ] thread
Awesome work. The quality of output of stable diffusion is already amazing, but the speed could really be improved. This probably can help a great deal. Can this be easily integrated in available generation tools?