Hey Leona,
Could you send a screenshot of your pipeline, please? Specifically how you configure the File sink plugin.
It seems to me that the parts are just there because it's the natural output format from Spark's parallel processing. If you look at the data inside the 'part-r-0000' file, is it CSV data?
At LiveRamp (the company I work at), we created and use a GCS merge file plugin, that lets us combine all of those part files into one file. If there is some interest in it, we could make sure we open-source it.
Best,
Baptiste Benet