configs/JSUT.py

_base_ = [
    "./_base_/archs/hifi_svc.py",
]

speaker_mapping = {'jsut': 0,}

model = dict(
    type="HiFiSVC",
    speaker_encoder=dict(
        input_size=len(speaker_mapping),
    ),
)

preprocessing = dict(
    text_features_extractor=dict(
        type="ContentVec",
    ),
    pitch_extractor=dict(
        type="ParselMouthPitchExtractor",
        keep_zeros=False,
        f0_min=40.0,
        f0_max=1600.0,
    ),
    energy_extractor=dict(
        type="RMSEnergyExtractor",
    ),
    augmentations=[
        dict(
            type="RandomPitchShifting",
            key_shifts=[-5., 5.],
            probability=1.5,
        ),
        dict(
            type="RandomTimeStretching",
            factors=[0.8, 1.2],
            probability=0.75,
        )
    ],
)