Context_recall returns 1 when context is empty

[lalehsg]

[ ] I have checked the documentation and related resources and couldn't resolve my bug.

Describe the bug all instances in my evaluation set where context is [""], are getting context_recall=1. I am using GPT3.5 as the judge.

Code to Reproduce just feed an example with context = [""] and compute the metric.

Expected behavior I expect the metric value to be 0.

[repetitioestmaterstudiorum]

I wonder the same! I evaluated 69 questions without context (just ['']) on context_recall and context_precision. Of those 138 test cases, I get 88 with a non-zero score! Scores range anywhere from 0.2 to 1. There seems to be a bug regarding the calculation of these metrics.