Merge branch 'master' into master

aws · Nov 12, 2024 · b8be6f5 · b8be6f5
2 parents 7a549ca + cff4e48
commit b8be6f5
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 2 deletions.
diff --git a/huggingface/pytorch/inference/docker/2.1/py3/Dockerfile.cpu b/huggingface/pytorch/inference/docker/2.1/py3/Dockerfile.cpu
@@ -86,6 +86,8 @@ RUN apt-get update \
  && update-alternatives --install /usr/bin/c++ c++ /usr/bin/g++ 100 \
  && apt-get clean
 
+RUN apt-get update && apt-get install -y cmake
+
 # Install OpenMPI
 RUN wget https://download.open-mpi.org/release/open-mpi/v4.1/openmpi-${OPEN_MPI_VERSION}.tar.gz \
  && gunzip -c openmpi-$OPEN_MPI_VERSION.tar.gz | tar xf - \
@@ -184,12 +186,14 @@ RUN curl -o /license.txt  https://aws-dlc-licenses.s3.amazonaws.com/pytorch-2.1/
 
 # install Hugging Face libraries and its dependencies
 RUN pip install --no-cache-dir \
+    kenlm==0.1 \
     transformers[sentencepiece,audio,vision]==${TRANSFORMERS_VERSION} \
     diffusers==${DIFFUSERS_VERSION} \
     peft==${PEFT_VERSION} \
     "accelerate>=0.25.0" \
     "protobuf>=3.19.5,<=3.20.2" \
-    "sagemaker-huggingface-inference-toolkit==2.3"
+    "sagemaker-huggingface-inference-toolkit==2.3" \
+    "huggingface_hub==0.23.2"
 
 # Install IPEx and its dependencies
 # from source is mandatory for cutomized AWS PyTorch binaries: https://github.com/intel/intel-extension-for-pytorch/issues/317

diff --git a/huggingface/pytorch/inference/docker/2.1/py3/cu118/Dockerfile.gpu b/huggingface/pytorch/inference/docker/2.1/py3/cu118/Dockerfile.gpu
@@ -102,6 +102,8 @@ RUN apt-get update \
  && rm -rf /var/lib/apt/lists/* \
  && apt-get clean
 
+ RUN apt-get update && apt-get install -y cmake
+
 # Install NCCL
 RUN cd /tmp \
  && git clone https://github.com/NVIDIA/nccl.git -b v${NCCL_VERSION}-1 \
@@ -217,13 +219,15 @@ RUN curl -o /license.txt  https://aws-dlc-licenses.s3.amazonaws.com/pytorch-2.0/
 
 # install Hugging Face libraries and its dependencies
 RUN pip install --no-cache-dir \
+    kenlm==0.1 \
 	transformers[sentencepiece,audio,vision]==${TRANSFORMERS_VERSION} \
     diffusers==${DIFFUSERS_VERSION} \
     peft==${PEFT_VERSION} \
     "accelerate>=0.25.0" \
 	"protobuf>=3.19.5,<=3.20.2" \
 	"numpy>=1.22.2,<1.24" \
-	"sagemaker-huggingface-inference-toolkit==2.3"
+	"sagemaker-huggingface-inference-toolkit==2.3" \
+    "huggingface_hub==0.23.2"
 
 RUN HOME_DIR=/root \
  && curl -o ${HOME_DIR}/oss_compliance.zip https://aws-dlinfra-utilities.s3.amazonaws.com/oss_compliance.zip \

diff --git a/release_images_training.yml b/release_images_training.yml
@@ -64,3 +64,16 @@ release_images:
       example: False
       disable_sm_tag: True
       force_release: False
+  6:
+    framework: "huggingface_pytorch"
+    version: "2.3.0"
+    hf_transformers: "4.46.1"
+    arch_type: "x86"
+    training:
+      device_types: ["gpu"]
+      python_versions: [ "py311" ]
+      os_version: "ubuntu20.04"
+      cuda_version: "cu121"
+      example: False
+      disable_sm_tag: False
+      force_release: False
diff --git a/test/sagemaker_tests/huggingface/inference/requirements.txt b/test/sagemaker_tests/huggingface/inference/requirements.txt
@@ -27,4 +27,5 @@ fabric
 invoke
 gitpython
 toml
+huggingface_hub==0.23.2
 transformers==4.28.1