What is the difference between "finetune" and "final-finetune" in `/example`.

[eric8607242]

Hello,

Thanks for the amazing repo!

I'm wondering what is the difference between "finetune" and "final-finetune" in /example.
Do we train the model and the mask score in the finetune stage, and only train the optimized model in the final-finetune stage?

Is there a way to directly save the optimized model and load the optimized model instead of loading the patched model and optimizing to get the pruned model?

Big thanks again for the great work!

[xhqi0109]

Hello,

Thanks for the amazing repo!
I also want to know the difference between "finetune" and "final-finetune" in /example.
Big thanks again for the great work!