Issues related to voice datasets

[Re-dot-art]

When I was using the freematch algorithm to train on the esc50 dataset, I encountered the following problem:

I downloaded the relevant data of the model to a folder under the USB framework and added the folder to the code, but the problem still occurred.
How can I solve it?

[github-actions]

Stale issue message