release_images_training.yml

---
release_images:
  1:
    framework: "pytorch"
    version: "2.0.0"
    customer_type: "ec2"
    arch_type: "x86"
    training:
      device_types: ["cpu", "gpu"]
      python_versions: ["py310"]
      os_version: "ubuntu20.04"
      cuda_version: "cu118"
      example: False
      disable_sm_tag: False  # [Default: False] This option is not used by Example images
      force_release: False
  2:
    framework: "pytorch"
    version: "2.0.0"
    customer_type: "ec2"
    arch_type: "x86"
    training:
      device_types: ["gpu"]
      python_versions: ["py310"]
      os_version: "ubuntu20.04"
      cuda_version: "cu118"
      example: True
      disable_sm_tag: False  # [Default: False] This option is not used by Example images
      force_release: False
  3:
    framework: "pytorch"
    version: "2.0.0"
    arch_type: "x86"
    customer_type: "sagemaker"
    training:
      device_types: ["cpu", "gpu"]
      python_versions: ["py310"]
      os_version: "ubuntu20.04"
      cuda_version: "cu118"
      example: False
      disable_sm_tag: False
      force_release: False