F5-TTS

Esta página de Soniqo documenta F5-TTS tal como está implementado en speech-swift / speech-core. Los enlaces a Hugging Face aparecen debajo de las notas de integración.

Primero página interna

Las tarjetas y menús apuntan primero aquí; los enlaces al modelo fuente y a los bundles siguen disponibles en esta página.

Resumen

Modelo	F5-TTS
Rol	Zero-shot voice cloning from a short reference plus transcript
Backend	MLX fp16
Salida	24 kHz mono waveform
Idiomas	English and Mandarin, mixed EN/ZH supported
Licencia	CC-BY-NC-4.0 weights; non-commercial bundle
Estado	Ready through speech speak --engine f5 and the F5TTS Swift product
Fuente	SWivid F5-TTS
Producto Swift	`F5TTS`
CLI / runtime	`speech speak --engine f5`

Uso

El fragmento siguiente refleja la API o el comando actual expuesto por speech-swift.

# Clone a voice from a ~10 s reference clip plus its transcript.
.build/release/speech speak "Hello from my cloned voice." \
  --engine f5 \
  --voice-sample reference.wav \
  --f5-reference-text "Text spoken in the reference clip." \
  -o cloned.wav

Enlaces del modelo

Notas de implementación

16 flow steps are the default (RTF 0.57 on an M5 Pro, lower is faster); --f5-steps 32 for maximum fidelity, 12 runs at RTF 0.43 — ASR roundtrips are word-identical across all three.
--f5-reference-text is required: cloning conditions on the reference transcript, and quality tracks how cleanly the reference clip ends on a sentence boundary.
Mandarin runs through a bundled pinyin lexicon with tone sandhi baked in at export time (99.2% token-exact against the upstream rjieba/pypinyin frontend), so Swift needs no external tokenizer.
Deterministic per --f5-seed; --f5-speed scales pacing, CFG strength and sway sampling are exposed as flags; about 0.8 GB peak RSS makes it the smallest cloning engine in the package.