Merge pull request #207 from aws-samples/KeitaW-patch-1

Bump to AWS_OFI_NCCL_VERSION=1.8.1-aws
aws-samples · Mar 11, 2024 · 1a02173 · 1a02173
2 parents f5e0b7b + 33ad5e0
commit 1a02173
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 3 deletions.
diff --git a/2.ami_and_containers/containers/pytorch/0.nvcr-pytorch-aws.dockerfile b/2.ami_and_containers/containers/pytorch/0.nvcr-pytorch-aws.dockerfile
@@ -25,7 +25,7 @@ ENV DEBIAN_FRONTEND=noninteractive
 # The three must-be-built packages.
 # Efa-installer>=1.29.0 required for nccl>=2.19.0 to avoid libfabric NCCL error.
 ENV EFA_INSTALLER_VERSION=1.30.0
-ENV AWS_OFI_NCCL_VERSION=1.7.4-aws
+ENV AWS_OFI_NCCL_VERSION=1.8.1-aws
 ENV NCCL_TESTS_VERSION=master
 
 RUN apt-get update -y

diff --git a/3.test_cases/3.MPT/0.llm-foundry.Dockerfile b/3.test_cases/3.MPT/0.llm-foundry.Dockerfile
@@ -1,7 +1,7 @@
 FROM mosaicml/llm-foundry:2.2.0_cu121_flash2_aws-latest
 
 ARG EFA_INSTALLER_VERSION=1.30.0
-ARG AWS_OFI_NCCL_VERSION=v1.7.4-aws
+ARG AWS_OFI_NCCL_VERSION=1.8.1-aws
 ARG NCCL_TESTS_VERSION=master
 ARG NCCL_VERSION=v2.18.6-1
 ARG LLM_FOUNDRY_VERSION=v0.4.0

diff --git a/3.test_cases/7.tensorflow-distributed/0.tensorflow.Dockerfile b/3.test_cases/7.tensorflow-distributed/0.tensorflow.Dockerfile
@@ -1,7 +1,7 @@
 FROM nvcr.io/nvidia/tensorflow:23.10-tf2-py3
 
 ARG EFA_INSTALLER_VERSION=1.30.0
-ARG AWS_OFI_NCCL_VERSION=v1.7.4-aws
+ARG AWS_OFI_NCCL_VERSION=1.8.1-aws
 ARG NCCL_TESTS_VERSION=master
 ARG NCCL_VERSION=v2.18.6-1
 ARG OPEN_MPI_PATH=/opt/amazon/openmpi