main
[Versatile Diffusion] Add versatile diffusion model (#1283)
Clean up flaky behaviour on Slow CUDA Pytorch Push Tests (#4759)
move tests to nightly (#4451)