How to prepare csv data for Multi-label task with XGBoost

[a-sharma123]

Hi,

I wanted to use the Sagemaker XGboost for my multi-label text classification task, however i do not know the format the csv data needs to be in for multi-label
Given that i have n features of the text already through vectorizing (f1,f2,f3....) and each text can be tagged with multiple labels (l1,l2,...)
i tried formatting the data as
[l1,l2..],f1,f2,f3....fn
[l4,l8..],f1,f2,f3....fn
But it did not work. How should i format the csv file in this case?