search

shansongliu / MU-LLaMA

MU-LLaMA: Music Understanding Large Language Model
GNU General Public License v3.0
221 stars 16 forks source link

Distinguishing QA and captioning #28

Closed joemzhao closed 1 month ago

joemzhao commented 1 month ago

Hello!

Thanks for the nice work! I'd like to ask how did you separate the QA and the captioning subtasks? In EvalMusicQA.json, there is no information on this, but in the paper, Tab2 and Tab3 report QA and captioning results separately.

Thank you!

shansongliu commented 1 month ago

You can distinguish them by the questions. The captioning task's questions have been described in the paper. Please take a look.