openai/glow

Open MPI

andvalenzuela opened this issue · 0 comments

Hi,

When training with CUDA_VISIBLE_DEVICES="0,1,2" mpiexec -n 3 python3 train.py --data_dir celeba-tfr --problem 'celeba' on 3GPUs I get the following warning:

**[[51353,1],1]: A high-performance Open MPI point-to-point messaging module
was unable to find any relevant network interfaces:

Module: OpenFabrics (openib)
Host: dev-vi-mltrain-01

Another transport will be used instead, although this may result in
lower performance.

NOTE: You can disable this warning by setting the MCA parameter
btl_base_warn_component_unused to 0.**

Does anyone know what it means and how to fix it?

Thank you in advance.