Non-English lyrics in the dataset

[xhevahir]

I see the dataset has a fair number of lyrics in other languages; I noticed some in at least French, Italian, Spanish, and Portuguese. Would the model benefit from removing all of the non-English lyrics?

[yy1lab]

Hi, thanks for your comments. We only take English lyrics to train lyrics encoder model. xhevahir <[email protected]> 于2019年8月30日周五 上午4:54写道：

…

I see the dataset has a fair number of lyrics in other languages; I noticed some in at least French, Italian, Spanish, and Portuguese. Would the model benefit from removing all of the non-English lyrics? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#2?email_source=notifications&email_token=AMQGHTCZWKWWLB5A4RBP7UTQHASNVA5CNFSM4ISGHLS2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HIJIUWA>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AMQGHTENIE4ZLFLBDWDSY2DQHASNVANCNFSM4ISGHLSQ> .