Struct WasmVoice 

pub struct WasmVoice { /* private fields */ }

WASM-compatible voice synthesizer. Loads model from bytes rather than file paths.

Implementations

impl WasmVoice

pub fn load_from_bytes( model_bytes: &[u8], config_json: &str, ) -> Result<Self, PiperError>

Load from in-memory model and config data.

Arguments

• model_bytes - ONNX model file contents
• config_json - config.json file contents as string

pub fn synthesize_ids( &mut self, phoneme_ids: &[i64], speaker_id: Option<i64>, language_id: Option<i64>, noise_scale: f32, length_scale: f32, noise_w: f32, ) -> Result<WasmSynthesisResult, PiperError>

Synthesize from pre-computed phoneme IDs (no G2P needed). This is the primary API for WASM since G2P may not be available.

pub fn config(&self) -> &VoiceConfig

Get the loaded config

pub fn has_speaker_id(&self) -> bool

Whether the model accepts a speaker ID input

pub fn has_language_id(&self) -> bool

Whether the model accepts a language ID input

pub fn has_prosody(&self) -> bool

Whether the model accepts prosody features input

pub fn capabilities(&self) -> &WasmModelCapabilities

Get model capabilities

Trait Implementations

impl Debug for WasmVoice

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.