Closed JackKelly closed 2 years ago
create_u_net
torch.optim.lr_scheduler.CosineAnnealingLR
ReduceLROnPlateau
fit_flat_cos
Ranger
create_u_net
not to skip the coord conv channels?torch.optim.lr_scheduler.CosineAnnealingLR
. Their code usesReduceLROnPlateau
withfit_flat_cos
)Ranger
)