OpenVoiceV2Options
Options & Configuration · AiDotNet.TextToSpeech.VoiceCloning
Options for OpenVoiceV2 TTS model.
Inheritance: ModelOptions → TtsModelOptions → CodecTtsOptions → VoiceCloningOptions → OpenVoiceV2Options
For Beginners
These options configure the OpenVoiceV2 model. Default values follow the original paper settings.
Constructors
| Constructor | Summary |
|---|---|
OpenVoiceV2Options() |
Initializes a new instance with default values. |
OpenVoiceV2Options(OpenVoiceV2Options other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
OpenVoiceTrainingOptions Training { get; set; } |
Auxiliary IPA prior, waveform reconstruction and adversarial training configuration. |
double Beta1 { get; set; } |
First AdamW moment coefficient used by the released recipe. |
double Beta2 { get; set; } |
Second AdamW moment coefficient used by the released recipe. |
double DropoutRate { get; set; } |
Dropout probability in the auxiliary IPA training encoder. |
double Epsilon { get; set; } |
AdamW numerical-stability epsilon used by the released recipe. |
int DecoderDim { get; set; } |
Gets or sets the hidden width of the decoder. |
int EncoderDim { get; set; } |
Gets or sets the encoder width of the reference configuration. |
int MelChannels { get; set; } |
Number of mel bands used by the waveform reconstruction objective. |
int NumEncoderLayers { get; set; } |
Number of layers in the auxiliary IPA training encoder. |
int NumHeads { get; set; } |
Attention heads in the auxiliary IPA training encoder. |