OOMKilled With A 256Mi Limit, But The Python Data Job 'only Uses' Half Of It???

I raised the limit for the Python data job to 256Mi after the first OOMKill. Then it happened again:

I'm afraid to just keep doubling it. My team lead asked 'is it a leak?' and I honestly don't know how to tell.

I don't want to just throw more memory at the problem every time, especially if it's a real leak. How do people tell teh difference in practice? Which metrics should I actually be looking at for the Python data job?