Which dataset is used for fine-tuning CodeGen25-7B-multi resulting in CodeGen25-7B-mono?

[iCSawyer]

I have checked the CodeGen2.5 technical report and the CodeGen2 paper, but I cannot find a clue. Can you kindly provide some information about the details? Thank you ~