Human eval results?
#2
by
rombodawg
- opened
Are we able to get this model evaluated on humaneval? So we can compare the scores to the models that were used in the merge
My anecdote is that it's way better than wizardcoder.
My anecdote is that it's way better than wizardcoder.
the little bit of testing i have done so far, agrees with that. ( for python at least. my only use case for AI )
I've been trying to code with C# and Forms/.Net, and it does a better job with it than pretty much every other model I've tried.