I am using pytorch lightning with Adam optmiser to train a BYOL model, the model and pipeline works fine while training, when I stop and resume from checkpoint, it raise this error. I didn't do anything just resume from previous checkpoint, have no idea why there is an error like this, I upgrade my bolts and pytorch lightning to the mast (I think its the latest version)

my opmizer is like this, and Adam is the original Adam provided by pytorch