Skip to content

Fix PyTorch 2.8 SageMaker Package Regressions #5520

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
DevakiBolleneni merged 19 commits into from
Dec 2, 2025
Merged

Fix PyTorch 2.8 SageMaker Package Regressions #5520

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
8756884
run tests
Nov 28, 2025
937054a
rebuild and run tests
Nov 28, 2025
4769818
update sagemaker pin and rebuild image
Nov 29, 2025
80adedf
rebuild image with sm sdk
Dec 1, 2025
b107c51
pin versions and rebuild image
Dec 1, 2025
826a943
Fix version pins and rebuild image
Dec 1, 2025
c65ba08
pins s3fs to prevent the downgrade and rebuild
Dec 1, 2025
57a767b
resolve incompatibilty issue
Dec 1, 2025
9c87c39
retry tests
Dec 1, 2025
0dc80eb
install newer version of s3fs
Dec 1, 2025
0b0f708
revert config file
Dec 1, 2025
797fc9a
revert version pin changes and pin sm version
Dec 1, 2025
1a979a8
build image
Dec 1, 2025
6128b85
put back awscli and boto3 pins
Dec 1, 2025
55ee780
Empty commit to trigger rebuild
Dec 1, 2025
1c78f58
fix typo & rerun
Dec 1, 2025
55b15e3
fix SM version pin & rerun
Dec 2, 2025
7d8e24e
add sniffio explicitly & rerun
Dec 2, 2025
c4034a7
revert toml file
Dec 2, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 6 additions & 4 deletions pytorch/training/docker/2.8/py3/Dockerfile.cpu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -188,7 +188,7 @@ RUN pip install --no-cache-dir \
"requests>=2.32.0" \
"setuptools>=70.0.0" \
"urllib3>=2.5.0" \
"awscli" \
awscli \
opencv-python==4.11.0.86 \
mpi4py \
jinja2>=3.1.6 \
Expand Down Expand Up @@ -282,10 +282,12 @@ WORKDIR /
# Install SM packages
RUN pip install --no-cache-dir -U \
smclarify \
"sagemaker>=2" \
"sagemaker==2.254.1" \
sagemaker-experiments \
sagemaker-pytorch-training \
sagemaker-training
sagemaker-training \
# Add sniffio explicitly as it's not included in sagemaker==2.254.1 dependencies
sniffio

# Install extra packages
RUN pip install --no-cache-dir -U \
Expand All @@ -297,7 +299,7 @@ RUN pip install --no-cache-dir -U \
scikit-learn \
seaborn \
shap \
cloudpickle
cloudpickle

# Copy workaround script for incorrect hostname
COPY changehostname.c /
Expand Down
8 changes: 5 additions & 3 deletions pytorch/training/docker/2.8/py3/cu129/Dockerfile.gpu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -250,10 +250,12 @@ RUN pip install --no-cache-dir -U \
"awscli<1.42.50" \
"boto3<1.40.50" \
smclarify \
"sagemaker>=2" \
"sagemaker==2.254.1" \
sagemaker-experiments \
sagemaker-pytorch-training \
sagemaker-training
sagemaker-training \
# Add sniffio explicitly as it's not included in sagemaker==2.254.1 dependencies
sniffio

# Install extra packages
RUN pip install --no-cache-dir -U \
Expand All @@ -265,7 +267,7 @@ RUN pip install --no-cache-dir -U \
shap \
scikit-learn \
seaborn \
cloudpickle
cloudpickle

COPY setup_oss_compliance.sh setup_oss_compliance.sh
RUN bash setup_oss_compliance.sh ${PYTHON} && rm setup_oss_compliance.sh
Expand Down