microsoft/LLaVA-Med

How were conversation metrics calculated in the new Benchmark tests with LLaVA 1.5?

NicoZenith opened this issue · 0 comments

How were conversation metrics calculated in the new Benchmark tests with LLaVA 1.5?