from transformers import PretrainedConfig from typing import Sequence class VocoderConfig(PretrainedConfig): model_type = "archinetai/vocoder-v1"