voxr 0.7.0

Audio and speech: codecs, DSP, G2P, voices, and TTS/ASR pipelines over boostr models
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

//! Port of misaki 0.9.4's English G2P (`misaki.en`, Apache-2.0,
//! <https://github.com/hexgrad/misaki>), the G2P Kokoro-82M was trained on.
//!
//! Input is raw text or tagged tokens, as misaki builds them from spaCy; the
//! out-of-vocabulary fallback is a separate component. Output matches
//! `misaki.en.G2P(trf=False, british=..., fallback=None, unk=...)`. The US
//! and GB gold and silver lexicons are embedded zlib-compressed; see
//! `assets/misaki/build_lexicon.py`.

mod chars;
mod data;
pub mod front;
pub mod g2p;
mod lexicon;
mod num2words;
mod number;
mod preprocess;
mod resolve;
mod retokenize;
mod special;
mod stem;
mod stress;
mod subtokenize;
pub mod token;

pub use front::TextFrontEnd;
pub use g2p::{DEFAULT_UNK, G2pOutput, KOKORO_UNK, MisakiG2p, OovFallback};
pub use token::MToken;