Lightgbm keeps crashing my R session

[lg1000]

I am trying to use lightgbm to predict a dataset with about 10.000 rows on a windows 10 vm. However, like 50% of the time the model gets ether stuck or my R session crashes with the following warning message of lgbm:

This should have nothing to do with my dataset, as every other tree based model from parsnip and modeltime works fine all the time.

I did update windows and parallels desktop as well and still having the same issue.