Re:GShard: Scaling giant models with conditional computation...
by ddorian43 on Thursday July 02, 2020 @06:14AM
(#23711703)
What is the cost when getting all data from network compared to local disks ? Is this connected to the cloud not offering local persistent disks ? Will this still work in colocation ?
[ Parent
| on HN ]
There are no comments on this comment yet.