Should I save optimizer ?

[mudimingquedeyinmoujia]

It seems the script main.py and main_dit.py only save model using accelerator.save_model(model, save_dir) and do not save the state of optimizer, so is this a negligence or the character of accelerator ? Will this influence resume training ?