Wandb logger conflict using LightningCLI

[robmarkcole]

Description

When using the Wandb logger with Lightning CLI a config.yaml will be written to the working dir and cause subsequent runs to error and fail until this file is deleted. This is due to Lightning-AI/pytorch-lightning#17168

The workaround is to pass LightningCLI(save_config_kwargs={"overwrite": True}) in torchgeo.main.py

Steps to reproduce

In lightning config:

trainer:
  min_epochs: 10
  max_epochs: 50
  logger:
    class_path: lightning.pytorch.loggers.wandb.WandbLogger
    init_args:
      project: my-proj

Will result in

RuntimeError: SaveConfigCallback expected ./config.yaml to NOT exist. Aborting to avoid overwriting results of a previous run. You can delete the previous config file, set `LightningCLI(save_config_callback=None)` to disable config saving, or set `LightningCLI(save_config_kwargs={"overwrite": True})` to overwrite the config file.

Version

0.5.0

[adamjstewart]

I'm fine with this workaround. Want to submit a PR?