feat: add expressive offline Russian book TTS

2026-07-21 21:28:48 +03:00
parent 28312e3fbc
commit cf82ab61eb
47 changed files with 127843 additions and 124 deletions
@@ -38,6 +38,10 @@ Before accepting the model, synthesize a held-out Russian validation list at mul
 `noise_scale`, and `noise_scale_w` settings, measure real-time factor on the target phone, and listen for skipped
 words, unstable stress, clicks, and repeated phonemes. A successful export alone is not a quality gate.

+The epoch-500 book-reading profile selected by listening comparison is stored in
+`aletheia_book_profile.json`. Keep its duration, noise, and sentence-pause values together when evaluating the ONNX
+model or integrating it into the reader.
+
 On native Windows, first run the one-batch CUDA check:

 ```powershell