From fa99171e47a3e1690c2ed5071857d4edd9c23ef0 Mon Sep 17 00:00:00 2001 From: janEbert <janpublicebert@posteo.net> Date: Thu, 10 Oct 2024 17:25:47 +0200 Subject: [PATCH] Fix Gloo name --- pytorch-ddp-example/run.sbatch | 2 +- pytorch-fsdp-example/run.sbatch | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/pytorch-ddp-example/run.sbatch b/pytorch-ddp-example/run.sbatch index 19c49e3..7997600 100644 --- a/pytorch-ddp-example/run.sbatch +++ b/pytorch-ddp-example/run.sbatch @@ -30,7 +30,7 @@ export MASTER_PORT=54123 # Prevent NCCL not figuring out how to initialize. export NCCL_SOCKET_IFNAME=ib0 -# Prevent GLOO not being able to communicate. +# Prevent Gloo not being able to communicate. export GLOO_SOCKET_IFNAME=ib0 srun env -u CUDA_VISIBLE_DEVICES python -u -m torchrun_jsc \ diff --git a/pytorch-fsdp-example/run.sbatch b/pytorch-fsdp-example/run.sbatch index 2e31024..1f5941f 100644 --- a/pytorch-fsdp-example/run.sbatch +++ b/pytorch-fsdp-example/run.sbatch @@ -30,7 +30,7 @@ export MASTER_PORT=54123 # Prevent NCCL not figuring out how to initialize. export NCCL_SOCKET_IFNAME=ib0 -# Prevent GLOO not being able to communicate. +# Prevent Gloo not being able to communicate. export GLOO_SOCKET_IFNAME=ib0 # For the example, we get a bunch of information from the -- GitLab