Skip to content

Addressing package regression in Pytorch-training-2.7 image #5478

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
bhanutejagk merged 17 commits into from
Nov 18, 2025
Merged

Addressing package regression in Pytorch-training-2.7 image #5478

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
17 commits
Select commit Hold shift + click to select a range
06fc4e3
Building the image and basic testing
Nov 14, 2025
d7b38fc
Addressed package regression caused by smclarify dependency s3fs
Nov 14, 2025
ceda39b
building only training image
Nov 14, 2025
f0222bf
explicitly declared typer, langcodes, languade_data to fix package re…
Nov 17, 2025
6477875
Removed AML2_CPU_ARM64_US_EAST_1 since ami not available
Nov 17, 2025
c6e4908
removed version pins for awscli and boto3
Nov 17, 2025
70ab7e6
added boto3 package
Nov 17, 2025
225cb82
Merge remote-tracking branch 'upstream/master'
Nov 17, 2025
47390e4
removed boto3 declaration for cpu image
Nov 17, 2025
db4b83f
added boto3 without any specific pin
Nov 17, 2025
da859ec
Merge remote-tracking branch 'upstream/master'
Nov 17, 2025
6244749
pinned awscli and boto3 to the versions in prod image
Nov 17, 2025
dc36343
added typer langcodes and language_data in samageker image recipe aswell
Nov 18, 2025
dd65cec
reduced the version of awscli and boto3
Nov 18, 2025
0620e08
revereted awscli and boto3 to 1.42.61
Nov 18, 2025
8f16238
updated awscli and bot3 versions in gpu to 1.42.61 to match those in cpu
Nov 18, 2025
13a70af
reveted back toml file
Nov 18, 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
13 changes: 11 additions & 2 deletions pytorch/training/docker/2.7/py3/Dockerfile.cpu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,11 +241,14 @@ RUN pip install --no-cache-dir -U torch==${PYTORCH_VERSION} \
fastai==2.8.2 \
accelerate \
# pin numpy requirement for fastai dependency
# requires explicit declaration of spacy, thic, blis
# requires explicit declaration of spacy, thic, blis, typer, langcodes, language_data
spacy \
#thinc 8.3.6 is not compatible with numpy 1.26.4 (sagemaker doesn't support latest numpy)
thinc==8.3.4 \
blis \
typer \
langcodes \
language_data \
numpy \
&& pip uninstall -y dataclasses

Expand Down Expand Up @@ -312,15 +315,21 @@ RUN pip install --no-cache-dir -U torch==${PYTORCH_VERSION} \
fastai==2.8.2 \
accelerate \
# pin numpy requirement for fastai dependency
# requires explicit declaration of spacy, thic, blis
# requires explicit declaration of spacy, thic, blis, typer, langcodes, language_data
spacy \
thinc==8.3.4 \
blis \
numpy \
typer \
langcodes \
language_data \
&& pip uninstall -y dataclasses

# Install SM packages
RUN pip install --no-cache-dir -U \
# address package regression caused by smclarify depedency s3fs"
"awscli<=1.42.61" \
"boto3<=1.40.61" \
smclarify \
"sagemaker>=2.9.0,<3" \
"sagemaker-experiments<1" \
Expand Down
8 changes: 7 additions & 1 deletion pytorch/training/docker/2.7/py3/cu128/Dockerfile.gpu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -143,11 +143,14 @@ RUN pip install --no-cache-dir \
"tornado>=6.5.1" \
"fastai==2.8.2" \
# pin numpy requirement for fastai dependency
# requires explicit declaration of spacy, thic, blis
# requires explicit declaration of spacy, thic, blis, typer, langcodes, language_data
spacy \
#thinc 8.3.6 is not compatible with numpy 1.26.4 (sagemaker doesn't support latest numpy)
"thinc==8.3.4" \
blis \
typer \
langcodes \
language_data \
"jinja2>=3.1.6"\
"typing-extensions>=4.14.1" \
&& pip uninstall -y dataclasses
Expand Down Expand Up @@ -195,6 +198,9 @@ RUN chmod +x /usr/local/bin/start_with_right_hostname.sh

# Install SM packages
RUN pip install --no-cache-dir -U \
# address package regression caused by smclarify depedency s3fs"
"awscli<=1.42.61" \
"boto3<=1.40.61" \
smclarify \
"sagemaker>=2.9.0,<3" \
"sagemaker-experiments<1" \
Expand Down
2 changes: 2 additions & 0 deletions test/test_utils/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,10 +140,12 @@ def get_ami_id_ssm(region_name, parameter_path):
ami_name_pattern="Deep Learning ARM64 AMI OSS Nvidia Driver GPU PyTorch 2.2.? (Ubuntu 20.04) ????????",
IncludeDeprecated=True,
)

AML2_CPU_ARM64_US_EAST_1 = get_ami_id_boto3(
region_name="us-east-1",
ami_name_pattern="Deep Learning ARM64 Base OSS Nvidia Driver GPU AMI (Amazon Linux 2) ????????",
)

PT_GPU_PY3_BENCHMARK_IMAGENET_AMI_US_EAST_1 = "ami-0673bb31cc62485dd"
PT_GPU_PY3_BENCHMARK_IMAGENET_AMI_US_WEST_2 = "ami-02d9a47bc61a31d43"

Expand Down