F5-TTS

此 Soniqo 页面记录本地 speech-swift / speech-core 实现中的 F5-TTS。Hugging Face 包链接放在集成说明之后。

先进入站内页面

首页卡片和文档菜单先指向这里；源模型和权重包链接仍在本页提供。

概览

模型	F5-TTS
用途	Zero-shot voice cloning from a short reference plus transcript
后端	MLX fp16
输出	24 kHz mono waveform
语言	English and Mandarin, mixed EN/ZH supported
许可证	CC-BY-NC-4.0 weights; non-commercial bundle
状态	Ready through speech speak --engine f5 and the F5TTS Swift product
来源	SWivid F5-TTS
Swift 产品	`F5TTS`
CLI / 运行时	`speech speak --engine f5`

使用

下面的片段对应当前 speech-swift 仓库暴露的 API 或命令。

# 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

模型链接

实现说明

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.