Race condition between `AsyncCheckpointIO` and `WandbLogger`

Nobody has claimed this yet.

### Bug description

When `AsyncCheckpointIO` is used and `WandbLogger` is configured to log model artifacts, there is a race condition that can occur. The notion of race conditions with `AsyncCheckpointIO` is already documented in #14035, but I feel that the topic of race conditions specifically involving `WandbLogger` may be distinct enough to warrant its own issue.

In short, the race condition happens when `WandbLogger` tries to log a checkpoint artifact that doesn't exist yet because `AsyncCheckpointIO` hasn't finished writing it.

I wouldn't know how to implement a thread lock or future callback in `WandbLogger`, but it would be nice if there's an easy solution here.

### What version are you seeing the problem on?

v2.0

### How to reproduce the bug

_No response_

### Error messages and logs

```
# Error messages and logs here please
```

### Environment

Current environment

```
#- Lightning Component (e.g. Trainer, LightningModule, LightningApp, LightningWork, LightningFlow):
#- PyTorch Lightning Version (e.g., 1.5.0):
#- Lightning App Version (e.g., 0.5.2):
#- PyTorch Version (e.g., 2.0):
#- Python version (e.g., 3.9):
#- OS (e.g., Linux):
#- CUDA/cuDNN version:
#- GPU models and configuration:
#- How you installed Lightning(`conda`, `pip`, source):
#- Running environment of LightningApp (e.g. local, cloud):
```

### More info

_No response_

cc @lantiga @morganmcg1 @borisdayma @scottire @parambharat

Start by tracing the interaction between AsyncCheckpointIO and WandbLogger, focusing on when checkpoint writing completes and when artifact logging begins. Reproduce the reported race if possible; done means WandbLogger does not attempt to log a checkpoint artifact before AsyncCheckpointIO has finished writing it.

Written by the indexing model from the issue text.