_ExperimentWriter overwrites CSV log file after loading checkpoint of training model

[f48r1]

CSV logger provided by lightning writes metrics while training a model.
If you resume training via a checkpoint and reuse the CSV logger, unfortunately, the previous logger file is first removed and recreated.
The reference code is here : https://github.com/Lightning-AI/pytorch-lightning/blob/master/src/lightning/fabric/loggers/csv_logs.py#L266
To avoid this inconvenience I have commented on the use of the function here : https://github.com/Lightning-AI/pytorch-lightning/blob/master/src/lightning/fabric/loggers/csv_logs.py#L210
Suggested modifications ?