OPT-tier according to benchmarks
#2
by
ChuckMcSneed
- opened
Is it really THAT bad?
There was a broken model in this merge unfortunately. The merge between codellama python and codellama chat was fucked. I'll be releasing a better one in the coming days