The benchmarks you have on the model card for this version are for the base model, can you add some that show the instruct model performance?
Also can we get some coding benchmarks like humaneval for both base and instruct models?
· Sign up or log in to comment
