FishSpeechOptions

Options & Configuration · AiDotNet.TextToSpeech.CodecBased

Options for FishSpeech (Fish Speech 1.4).

Inheritance: ModelOptions → TtsModelOptions → FishSpeechFamilyOptions → FishSpeechOptions

For Beginners

These settings size Fish Speech 1.4 and control how it speaks and trains; the defaults reproduce the released model.

How It Works

The defaults are Fish Speech 1.4’s (fishaudio/fish-speech-1.4): the same Dual-AR sizes as 1.5 over a 32,000-token byte-level BPE vocabulary with one <|semantic|> token, a 4,096-token context and dropout 0.1; the same Firefly-GAN. Text goes through the Chinese number normalizer as the 1.4 API does by default. Training follows the release’s fine-tuning recipe (ReleaseFineTune).

Constructors

Constructor Summary
FishSpeechOptions() Creates the released Fish Speech 1.4 configuration.