Additionally check number of devices of Trainer for evaluation before warning when using DistributedSampler

[function2-llx]

Hi everyone,

When using DistributedSampler (e.g., when using DDPStrategy) during validation/test, a warning will suggest setting Trainer(devices=1) even if it already is. How about additionally checking if the number of devices of Trainer is 1 here?

https://github.com/Lightning-AI/lightning/blob/d8e5e7f889646e2ae3f480941b4c9e18434e994d/src/pytorch_lightning/trainer/connectors/data_connector.py#L301