How to perform breakpoint training

[zhenzi0322]

No description provided.

[williamyang1991]

You need to save all these files and recover the training by loading all these models and optimizers.

GP-UNIT/train.py

Lines 200 to 208 in d59695e

	torch.save(
	{
	#"g": netG.state_dict(),
	#"d": netD.state_dict(),
	"g_ema": netG_ema.state_dict(),
	"sampler": sampler.icp.netT.state_dict(),
	#"g_optim": optimizer_G.state_dict(),
	#"d_optim": optimizer_D.state_dict(),
	#"args": args,