Check flavour of cuda getting installed in nvcr.io/nvidia/pytorch:22.09-py3

1

Hi,

We are trying to see which flavour of cuda is installed on image nvcr.io/nvidia/pytorch:22.09-py3

flavors of Cuda -

Three flavors of images are provided: Docker

  • base: Includes the CUDA runtime (cudart)
  • runtime: Builds on the base and includes the [CUDA math libraries], and [NCCL]. A runtime image that also includes [cuDNN] is available.
  • devel: Builds on the runtime and includes headers, development tools for building CUDA images. These images are particularly useful for multi-stage builds.

Found DockerFile for pytorch/Dockerfile at main · pytorch/pytorch · GitHub which points to direct install of cuda, does not mention flavor gets installed. Any directions to check the installation will be helpful.

2

The NVIDIA NGC containers are devel containers and allow you to build your application in them or to use them as a base for your custom container. The linked Dockerfile is used to build the PyTorch container on docker.hub and ja unrelated.