icon-lab / SynDiff

Official PyTorch implementation of SynDiff described in the paper (https://arxiv.org/abs/2207.08208).
Other
244 stars 37 forks source link

The server socket cannot be initialized #21

Open xychen2022 opened 1 year ago

xychen2022 commented 1 year ago

Hi, I got the following distributed learning related problem when running the code using example data. How to solve it? Thanks

[W socket.cpp:401] [c10d] The server socket cannot be initialized on [::]:6021 (errno: 97 - Address family not supported by protocol). [W socket.cpp:401] [c10d] The server socket has failed to bind to 0.0.0.0:6021 (errno: 98 - Address already in use). [E socket.cpp:435] [c10d] The server socket has failed to listen on any local network address.

onat-dalmaz commented 1 year ago

It seems that cpp packages related to distributed learning are not properly installed. Please check torch.multiprocessing torch.distributed

and their required packages too.