View a markdown version of this page

Deep Learning AMIs Release Notes - AWS Deep Learning AMIs
Base DLAMIsSingle-Framework DLAMIsMulti-Framework DLAMISOCI Supported DLAMIs

Deep Learning AMIs Release Notes

Here you can find detailed release notes for all currently supported AWS Deep Learning AMIs (DLAMI) options.

For release notes for DLAMI frameworks that we no longer support, see the Unsupported Framework Release Notes Archive section of the DLAMI Framework Support Policy page.

Release Notes

Release Notes for Base DLAMIs

X86 Base DLAMI Release Notes

Below are the release notes for X86 Base DLAMI:

GPU

ARM64 Base DLAMI Release Notes

Below are the release notes for ARM64 Base DLAMI:

GPU

AWS Neuron

Release Notes for Single Framework DLAMIs

PyTorch DLAMIs

X86 PyTorch DLAMI Release Notes

Below are the Release notes for X86 PyTorch DLAMIs:

GPU
AWS Neuron

    ARM64 PyTorch DLAMI Release Notes

    Below are the Release notes for ARM64 PyTorch DLAMIs:

    GPU

    AWS Neuron

    Release Notes for Multi-Framework DLAMIs

    Tip

    If you use only one machine learning framework, then we recommend a single-framework DLAMI.

    Multi-Framework DLAMI Release Notes

    Below are the release notes for Multi-Framework X86 DLAMI:

    GPU

    AWS Neuron

    SOCI Supported DLAMIs

    The following DLAMIs include pre-installed SOCI (Seekable OCI) snapshotter for container image lazy loading and parallel pull. Use nerdctl --snapshotter soci to pull Deep Learning Container images with the -soci tag suffix for faster container startup.

    X86 SOCI Supported DLAMIs

    Below are the X86 DLAMIs with SOCI support:

    ARM64 SOCI Supported DLAMIs

    Below are the ARM64 DLAMIs with SOCI support: