This repository was archived by the owner on Nov 1, 2021. It is now read-only.

Description
Hey guys,
I'm looking to implement this in a multi-node environment, managed by SLURM. As far as I can tell this code has not been developed for such an environment.
I have been able to get the client-server model from https://github.com/twitter/torch-ipc running, but I thought that before I went ahead and began development on my own multi-node implementation of allReduce following that as an example, I would just ask to see if I was doing anything incorrectly or if there other existing multi-node torch implementations of sgd.
Thanks