Request for evaluate code

[Marcowky]

Hi, thank you for your great work and the open-source release!

How do I evaluate the model's performance on each dataset?
Could you please provide some scripts or commands for the evaluation?
Or could you share the code repo you use for evaluation?

Thanks!

[2561681244]

Hello, the author claims that pass@1 is equivalent to mean@16 (or avg@16).

#28