huizhangdb / pmemo Goto Github PK

Hello, and thank you very much for sharing this amazing dataset with the community. I was wondering if it would be possible to have the independent dynamic annotations from each listener?

Thank you!

Hi HuiZhangDB,
greatwork to you and the team. I am looking the dataset for my school study purpose. However the dataset is not avaiable to download via the link in your blog post. Could you please provide me a different way to access the dataset. Great thanks and have a nice weekend.
Isa
[email protected]

Can you pls explain in your code what is.
Adding some comments will surely help.

• distdir
• dynamic_distdir
• all_dynamic_distfile

Hi Hui Zhang,

Thank you for creating this wonderful project. I found some duplication in the current dataset(2019 ver.) (e.g. music Id 996 & 997, 90 & 92) and their static annotations are different. I wonder is the duplication intentional?

Thanks in advance.
Kexin

Metadata of tracks, e.g. title, displayed artist, are usually ambiguous. Could your team add "ISRC" of tracks as one attribute of the metadata? If so, the dataset will be more useful for users with access to many full length audios.

btw: some songs have comma symbol in their artist attribute. It causes trouble to csv parser. Could your team fix it?

Thanks very much.