[KongchangAI]
Model

MERT2

A multi-view music representation encoder that provides semantic tokens for music generation models. It performs strongly on the MARBLE music understanding benchmark.

Timeline (last 90 days)

Oct 14

MERT2 is a multi-view music representation encoder that achieved the best results on 14 of the 15 tasks in the MARBLE music understanding benchmark.

Unverified50%
Oct 14

YuE2's model weights, code, and the two auxiliary models MERT2 and SheetSage2 have all been released on GitHub and Hugging Face, and the project page is map-yue2.github.io.

Unverified50%

All Facts (2)

Source Articles