Adding new datasets and new LLMs

[jihyukkim-nlp]

Hello, thank you for publishing awesome work and sharing codes!

I would like to to apply MoP to other datasets and other LLMs.

• Could you give us some guidance on adding new models or new datasets?

In addition, for having experiments on shared datasets,

• I am also wondering what each directory/file means in this link.

For example, bbh directory contains three inner-directories such as raw, raw_original, and source. And, each of the three contains execute and induce.
How is each of them used for experiments?

Best regards,
Jihyuk Kim