Comment by bwfan123

thanks !, I am curious how this relates to the recent "monarch" announcement - which has similar goals of facilitating large scale fault tolerant training [1].

[1] https://github.com/pytorch-labs/monarch/issues/175#issuecomm...