Word-wise intonation model for cross-language TTS systems

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.