SaveConfigCallback saves config to the current working directory instead of the logging directory.

Nobody has claimed this yet.

### Bug description

According to the doc, `SaveConfigCallback` should save config files to dynamic logging directories like `lightning_logs/version_7/config.yaml` by default, but it seems like it just saves to the current working directory, which is the default `trainer.log_dir`.

### How to reproduce the bug

```python
# main.py
from pytorch_lightning.cli import LightningCLI
from pytorch_lightning.demos.boring_classes import BoringModel, BoringDataModule

def main():
LightningCLI(BoringModel, BoringDataModule)

if __name__ == "__main__":
main()
```

```bash
python main.py fit
```

You'll notice that a `config.yaml` is created in the current working directory rather than the log directory.

### Error messages and logs

_No response_

### Environment

Current environment

```
* CUDA:
- GPU:
- NVIDIA GeForce GTX 1060 6GB
- available: True
- version: 11.7
* Lightning:
- lightning-utilities: 0.8.0
- pytorch-lightning: 2.0.0
- torch: 2.0.0
- torchmetrics: 0.11.4
* Packages:
- aiohttp: 3.8.4
- aiosignal: 1.3.1
- async-timeout: 4.0.2
- attrs: 22.2.0
- certifi: 2022.12.7
- charset-normalizer: 3.1.0
- cmake: 3.26.0
- docstring-parser: 0.15
- filelock: 3.10.0
- frozenlist: 1.3.3
- fsspec: 2023.3.0
- idna: 3.4
- importlib-resources: 5.12.0
- jinja2: 3.1.2
- jsonargparse: 4.20.0
- lightning-utilities: 0.8.0
- lit: 16.0.0
- markupsafe: 2.1.2
- mpmath: 1.3.0
- multidict: 6.0.4
- networkx: 3.0
- numpy: 1.24.2
- nvidia-cublas-cu11: 11.10.3.66
- nvidia-cuda-cupti-cu11: 11.7.101
- nvidia-cuda-nvrtc-cu11: 11.7.99
- nvidia-cuda-runtime-cu11: 11.7.99
- nvidia-cudnn-cu11: 8.5.0.96
- nvidia-cufft-cu11: 10.9.0.58
- nvidia-curand-cu11: 10.2.10.91
- nvidia-cusolver-cu11: 11.4.0.1
- nvidia-cusparse-cu11: 11.7.4.91
- nvidia-nccl-cu11: 2.14.3
- nvidia-nvtx-cu11: 11.7.91
- packaging: 23.0
- pip: 23.0.1
- pytorch-lightning: 2.0.0
- pyyaml: 6.0
- requests: 2.28.2
- setuptools: 67.4.0
- sympy: 1.11.1
- tdgu: 0.1.0
- torch: 2.0.0
- torchmetrics: 0.11.4
- tqdm: 4.65.0
- triton: 2.0.0
- typeshed-client: 2.2.0
- typing-extensions: 4.5.0
- urllib3: 1.26.15
- wheel: 0.38.4
- yarl: 1.8.2
* System:
- OS: Linux
- architecture:
- 64bit
- ELF
- processor: x86_64
- python: 3.10.10
- version: #137-Ubuntu SMP Wed Jun 15 13:33:07 UTC 2022
```

### More info

_No response_

cc @carmocca @mauvilsa @borda

Start by tracing LightningCLI and SaveConfigCallback using the provided main.py reproduction and `python main.py fit`. Check how the callback chooses its output path relative to trainer.log_dir; done means config.yaml is written under the dynamic logging directory, such as lightning_logs/version_7, rather than the current working directory.

Written by the indexing model from the issue text.