Semantic Representation

[Uneasy-Z]

How can I get semantic representations from wavtokenizer?
As the paper reads:
We extract embeddings corresponding to the discrete codebooks of an acoustic codec model as its respective representations and evaluate the classification accuracy of the codec model on ARCH datasets using its representations.

wish you all the best！

[jishengpeng]

How can I get semantic representations from wavtokenizer? As the paper reads: We extract embeddings corresponding to the discrete codebooks of an acoustic codec model as its respective representations and evaluate the classification accuracy of the codec model on ARCH datasets using its representations.

wish you all the best！

README, wavtokenizer.encode_infer(). Thank you for your attention.

[Uneasy-Z]

ok~ Thank you for your prompt reply~