Support `channels_last` with training

Nobody has claimed this yet.

## 🚀 Feature

I'd like to try out some `channels_last` training to see if it improves performance (https://pytorch.org/tutorials/intermediate/memory_format_tutorial.html)

I'm not entirely sure what the best way to do this with lightning is but I also think it should probably be one of those features that you set on the trainer and it just magically works.

### Motivation

Using the `channels_last` memory format can improve performance in some cases

### Pitch

Add a trainer flag that does whatever is needed for `channels_last` so

```
trainer = pl.Trainer( ..., memory_format='channels_last')
```

or something like that and then before anything happens with training/testing you need to convert the module

```
if self.memory_format == 'channels_last':
lightning_module = lightning_module.to(memory_format=torch.channels_last)
```

and each batch in the train/test/val loops

```
if self.memory_format == 'channels_last':
batch = batch.to(memory_format=torch.channels_last)
lightning_module(batch)
```

### Alternatives

I have no idea but I assume I could do this without changing lightning although I'm not sure how yet

### Additional context

I am not sure I will be able to PR this one but I'm not opposed to trying

cc @borda @carmocca @justusschock @awaelchli

Start by reading the Trainer configuration and the module setup, then trace how batches move through the train, validation, and test loops. Review PyTorch's channels_last memory-format guidance linked in the issue. Done means a Trainer option consistently applies the requested memory format to the module and relevant batches during training, validation, and testing.

Written by the indexing model from the issue text.