[feature-request] Update pytorch neuron inference container for current neuron SDK

[AdventLee]

Checklist

• [X] I've prepended issue tag with type of change: [feature]
• [X] (If applicable) I've documented below the DLC image/dockerfile this relates to
• [ ] (If applicable) I've documented the tests I've run on the DLC image
• [X] I'm using an existing DLC image listed here: https://docs.aws.amazon.com/deep-learning-containers/latest/devguide/deep-learning-containers-images.html
• [ ] I've built my own container based off DLC (and I've attached the code used to build my own image)

Concise Description: Neuron containers for inference use old SDK (1.19), stale pytorch versions (1.10.2), py3.7 and ubuntu 18.04. Out of sync with current neuron version and neuron training containers.

DLC image/dockerfile: Pytorch neuron inference container

Is your feature request related to a problem? Please describe. A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

There is a version mismatch between current neuron SDK, pytorch, python and available container between training and inference.

Describe the solution you'd like A clear and concise description of what you want to happen.

Update inference container to versions consistent with training container, and current neuron SDK. This is SDK 2.4.0, pytorch 1.11.0, python 3.8 and ubuntu 20.04.

Eg. 763104351884.dkr.ecr.us-west-2.amazonaws.com/pytorch-inference-neuron:1.11.0-neuron-py38-sdk2.4.0-ubuntu20.04

Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered.

N/A

Additional context Add any other context or screenshots about the feature request here.

N/A

[j-adamczyk]

+1, those outdated versions everywhere, and different versions for different frameworks/libraries/dependencies, is inexcusable

[Limess]

We've just started looking into using neuron and immediately found it very off putting that the state of the deep learning containers is a bit of a mess.

We'd also like it to be clearer on the E3/Sagemaker split for these images - what platform are the current ones intended for?