Available Amazon SageMaker Images - Amazon SageMaker

Available Amazon SageMaker Images

The following SageMaker images are available in Amazon SageMaker Studio. SageMaker images contain the latest Amazon SageMaker Python SDK and the latest version of the kernel. The name in brackets ([ ]) is the resource identifier of the SageMaker image as specified in the Amazon Resource Name (ARN) for the SageMaker image. For more information, see Deep Learning Containers Images.

  • Base Python [python-3.6]

    Official Python 3.6 image from DockerHub with boto3 and AWS CLI included.

  • Base Python 2.0 [sagemaker-base-python-38]

    Official Python 3.8 image from DockerHub with boto3 and AWS CLI included.

  • Data Science [datascience-1.0]

    Data Science is a Python 3.7 conda image with the most commonly used Python packages and libraries, such as NumPy and SciKit Learn.

  • Data Science 2.0 [sagemaker-datascience-38]

    Data Science 2.0 is a Python 3.8 conda image with the most commonly used Python packages and libraries, such as NumPy and SciKit Learn.

  • SparkMagic [sagemaker-sparkmagic]

    Anaconda Individual Edition with PySpark and Spark kernels. For more information, see sparkmagic.

  • MXNet 1.6 Python 3.6 (optimized for CPU) [mxnet-1.6-cpu-py36]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.6 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for MXNet 1.6.0 .

  • MXNet 1.6 Python 3.6 (optimized for GPU) [mxnet-1.6-gpu-py36]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.6 with CUDA 10.1 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for MXNet 1.6.0 .

  • MXNet 1.8 Python 3.7 (optimized for CPU) [mxnet-1.8-cpu-py37-ubuntu16.04-v1]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.8 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for AWS MX 1.8.0 .

  • MXNet 1.8 Python 3.7 (optimized for GPU) [mxnet-1.8-gpu-py37-cu110-ubuntu16.04-v1]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.8 with CUDA 11.0 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for AWS MX 1.8.0 .

  • MXNet 1.9 Python 3.8 (optimized for CPU) [mxnet-1.9-cpu-py38-ubuntu20.04-sagemaker-v1.0]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.9 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for MX 1.9.0 on SageMaker .

  • MXNet 1.9 Python 3.8 (optimized for GPU) [mxnet-1.9-gpu-py38-cu112-ubuntu20.04-sagemaker-v1.0]

    The AWS Deep Learning Containers for AWS MX powered by Apache MXNet 1.9 with CUDA 11.2 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for MX 1.9.0 on SageMaker .

  • PyTorch 1.10 Python 3.8 (optimized for CPU) [pytorch-1.10-cpu-py38]

    The AWS Deep Learning Containers for PyTorch 1.10 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.10.2 on SageMaker .

  • PyTorch 1.10 Python 3.8 (optimized for GPU) [pytorch-1.10-gpu-py38]

    The AWS Deep Learning Containers for PyTorch 1.10 with CUDA 11.3 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.10.2 on SageMaker .

  • PyTorch 1.4 Python 3.6 (optimized for CPU) [pytorch-1.4-cpu-py36]

    The AWS Deep Learning Containers for PyTorch 1.4 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v3.2 for PyTorch .

  • PyTorch 1.4 Python 3.6 (optimized for GPU) [pytorch-1.4-gpu-py36]

    The AWS Deep Learning Containers for PyTorch 1.4 with CUDA 10.1 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v3.2 for PyTorch .

  • PyTorch 1.6 Python 3.6 (optimized for CPU) [pytorch-1.6-cpu-py36-ubuntu16.04-v1]

    The AWS Deep Learning Containers for PyTorch 1.6 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.6.0 .

  • PyTorch 1.6 Python 3.6 (optimized for GPU) [pytorch-1.6-gpu-py36-cu110-ubuntu18.04-v3]

    The AWS Deep Learning Containers for PyTorch 1.6 with CUDA 11.0 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.6.0 with CUDA 11.0 .

  • PyTorch 1.8 Python 3.6 (optimized for CPU) [1.8.1-cpu-py36]

    The AWS Deep Learning Containers for PyTorch 1.8 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.8.0 .

  • PyTorch 1.8 Python 3.6 (optimized for GPU) [pytorch-1.8-gpu-py36]

    The AWS Deep Learning Containers for PyTorch 1.8 with CUDA 11.1 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for PyTorch 1.8.0 .

  • TensorFlow 1.15 Python 3.6 (optimized for CPU) [tensorflow-1.15-cpu-py36]

    The AWS Deep Learning Containers for TensorFlow 1.15 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers with TensorFlow 1.15.3 .

  • TensorFlow 1.15 Python 3.6 (optimized for GPU) [tensorflow-1.15-gpu-py36]

    The AWS Deep Learning Containers for TensorFlow 1.15 with CUDA 10.0 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers with TensorFlow 1.15.3 .

  • TensorFlow 1.15 Python 3.7 (optimized for CPU) [tensorflow-1.15-cpu-py37-ubuntu18.04-v7]

    The AWS Deep Learning Containers for TensorFlow 1.15 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v7.0 for TensorFlow .

  • TensorFlow 1.15 Python 3.7 (optimized for GPU) [tensorflow-1.15-gpu-py37-cu110-ubuntu18.04-v8]

    The AWS Deep Learning Containers for TensorFlow 1.15 with CUDA 11.0 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v7.0 for TensorFlow .

  • TensorFlow 2.1 Python 3.6 (optimized for CPU) [tensorflow-2.1-cpu-py36]

    The AWS Deep Learning Containers for TensorFlow 2.1 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v6.2 for Tensorflow .

  • TensorFlow 2.1 Python 3.6 (optimized for GPU) [tensorflow-2.1-gpu-py36]

    The AWS Deep Learning Containers for TensorFlow 2.1 with CUDA 10.1 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers v6.2 for Tensorflow .

  • TensorFlow 2.3 Python 3.7 (optimized for CPU) [tensorflow-2.3-cpu-py37-ubuntu18.04-v1]

    The AWS Deep Learning Containers for TensorFlow 2.3 include containers for training on CPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers with TensorFlow 2.3.0 .

  • TensorFlow 2.3 Python 3.7 (optimized for GPU) [tensorflow-2.3-gpu-py37-cu110-ubuntu18.04-v3]

    The AWS Deep Learning Containers for TensorFlow 2.3 with CUDA 11.0 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for TensorFlow 2.3.1 with CUDA 11.0 .

  • TensorFlow 2.6 Python 3.8 (optimized for CPU) [tensorflow-2.6-cpu-py38-ubuntu20.04-v1]

    The AWS Deep Learning Containers for TensorFlow 2.6 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for TensorFlow 2.6 .

  • TensorFlow 2.6 Python 3.8 (optimized for GPU) [tensorflow-2.6-gpu-py38-cu112-ubuntu20.04-v1]

    The AWS Deep Learning Containers for TensorFlow 2.6 with CUDA 11.2 include containers for training on GPU, optimized for performance and scale on AWS. For more information, see AWS Deep Learning Containers for TensorFlow 2.6 .