Comment by MeetingsBrowser

Are there any concrete benchmarks for comparing models for different types of programming tasks?