publications - How much informal writing should there be in a research article?

I am starting to write a research article in computer science area. While writing, I am using very formal terminologies and sentences which occasionally get quite boring. Sometimes, funny side of my personality pops up, and I end up writing some informal sentences which I have to delete later, picturing a very serious looking reviewer reading my paper draft. I always have this idea that a research article should be narrated like a story (keeping the facts and assumptions valid) to make it interested for the readers and stories often get quite informal. Don't they?

So the question is: How informal can I go in writing a research article even if my assumptions are correct? Do reviewers mind if the language used is not quite formal?