VevoCheckpoints

Models & Types · AiDotNet.TextToSpeech.CodecBased

The files Vevo’s release reads (models/vc/vevo/infer_*.py).

Constructors

Constructor Summary
VevoCheckpoints()

Properties

Property Summary
string Directory { get; set; } The amphion/Vevo directory: tokenizer/vq8192, tokenizer/vq32, contentstyle_modeling/PhoneToVq8192 and Vq32ToVq8192, acoustic_modeling/Vq8192ToMels and Vocoder.
string Hubert { get; set; } HuBERT-Large’s weights (facebook/hubert-large-ll60k’s pytorch_model.bin or model.safetensors; torchaudio’s HUBERT_LARGE is the same fairseq model).