Higgs TTS 3
توثق هذه الصفحة من Soniqo نموذج Higgs TTS 3 كما هو منفذ في speech-swift / speech-core. روابط Hugging Face موجودة أدناه بعد ملاحظات الدمج.
الصفحة الداخلية أولا
بطاقات الصفحة الرئيسية وقوائم الوثائق تشير إلى هذه الصفحة أولا؛ وتبقى روابط النموذج والحزم داخلها.
لمحة سريعة
| النموذج | Higgs TTS 3 |
|---|---|
| الدور | Conversational TTS and zero-shot cloning with inline emotion, style, and SFX tags |
| Backend | MLX bf16 backbone with float32 codec |
| الإخراج | 24 kHz mono waveform, 25 frames per second |
| اللغات | 100+ languages |
| الرخصة | Boson Higgs TTS 3 Research and Non-Commercial License |
| الحالة | Ready through speech speak --engine higgs and the HiggsTTS Swift product |
| المصدر | Boson AI higgs-tts-3-4b |
| منتج Swift | HiggsTTS |
| CLI / runtime | speech speak --engine higgs |
الاستخدام
المقتطف أدناه يطابق API أو الأمر الحالي في speech-swift.
# Expressive cloned speech with an inline emotion tag.
.build/release/speech speak "<|emotion:elation|> We did it — the demo works!" \
--engine higgs \
--voice-sample reference.wav \
--higgs-ref-text "Text spoken in the reference clip." \
-o higgs.wavروابط النموذج
ملاحظات التنفيذ
- RTF 0.78 on an M5 Pro for cloned and plain synthesis alike: sampling stays on the GPU with the delay ramp masked on-device, and asyncEval overlaps decode with the end-of-codes state machine.
- Cloning gates: English lexical overlap 1.00 and Mandarin CER 0.000 on ASR roundtrips; the language model and codec are parity-validated against the upstream implementations in both directions.
- Control tags are ordinary tokens inside the target text: <|emotion:...|>, <|style:whispering|>, <|sfx:laughter|>, <|prosody:pause|> and friends.
- Encoding a 12 s reference costs about 0.6 s once and the codes are reusable across generations; omit --voice-sample for reference-free synthesis in a stock voice.
- The CLI defaults to temperature 0.8 because the upstream 1.0 wobbles on single sentences; expect about 8.6 GB peak RSS for the 4B backbone.