Deprecate Trainer.root_gpu

[ananthsub]

Proposed refactor

Deprecate this property:
https://github.com/PyTorchLightning/pytorch-lightning/blob/cf64f3443474a93d23b5afb0417e4a60298006e6/pytorch_lightning/trainer/trainer.py#L1982-L1984

Motivation

This property is redundant with the Strategy's root_device property, which generalizes this check.
Following #11758, this property is not used anywhere else in the framework. This is only used in tests, which can be updated to use trainer.strategy.root_device instead

Pitch

Deprecate this property, and recommend using trainer.strategy.root_device instead
Regarding how to deprecate the property, this issue provides an example: #10615

Additional context

---

If you enjoy Lightning, check out our other projects! ⚡

• Metrics: Machine learning metrics for distributed, scalable PyTorch applications.

• Lite: enables pure PyTorch users to scale their existing code on any kind of device while retaining full control over their own loops and optimization logic.

• Flash: The fastest way to get a Lightning baseline! A collection of tasks for fast prototyping, baselining, fine-tuning, and solving problems with deep learning.

• Bolts: Pretrained SOTA Deep Learning models, callbacks, and more for research and production with PyTorch Lightning and PyTorch.

• Lightning Transformers: Flexible interface for high-performance research using SOTA Transformers leveraging Pytorch Lightning, Transformers, and Hydra.

cc @Borda @justusschock @awaelchli @rohitgr7 @tchaton @akihironitta @kaushikb11 @ananthsub @ninginthecloud @jjenniferdai

[vpavankalyan]

I would like to work on this issue

[rohitgr7]

hey @vpavankalyan!
I have assigned this issue to you. Feel free to send a PR :)