FishAudio S1
FishAudio S1 is a leading text-to-speech (TTS) model trained on more than 2 million hours of audio data in multiple languages.
Supported languages:
- English (en)
- Chinese (zh)
- Japanese (ja)
- German (de)
- French (fr)
- Spanish (es)
- Korean (ko)
- Arabic (ar)
- Russian (ru)
- Dutch (nl)
- Italian (it)
- Polish (pl)
- Portuguese (pt)
Please refer to
Fish Speech Github for more info.
Demo available at
Fish Audio Playground.
Visit the
Fish Audio website for blog & tech report.
Emotion and Tone Support
FishAudio S1 supports a variety of emotional, tone, and special markers to enhance speech synthesis:
1. Emotional markers:
(angry) (sad) (disdainful) (excited) (surprised) (satisfied) (unhappy) (anxious) (hysterical) (delighted) (scared) (worried) (indifferent) (upset) (impatient) (nervous) (guilty) (scornful) (frustrated) (depressed) (panicked) (furious) (empathetic) (embarrassed) (reluctant) (disgusted) (keen) (moved) (proud) (relaxed) (grateful) (confident) (interested) (curious) (confused) (joyful) (disapproving) (negative) (denying) (astonished) (serious) (sarcastic) (conciliative) (comforting) (sincere) (sneering) (hesitating) (yielding) (painful) (awkward) (amused)
2. Tone markers:
(in a hurry tone) (shouting) (screaming) (whispering) (soft tone)
3. Special markers:
(laughing) (chuckling) (sobbing) (crying loudly) (sighing) (panting) (groaning) (crowd laughing) (background laughter) (audience laughing)
Special markers with corresponding onomatopoeia:
- Laughing: Ha,ha,ha
- Chuckling: Hmm,hmm
Model Variants and Performance
FishAudio S1 includes the following models:
- S1 (4B, proprietary): The full-sized model.
- S1-mini (0.5B): A distilled version of S1.
Both S1 and S1-mini incorporate online Reinforcement Learning from Human Feedback (RLHF).
Seed TTS Eval Metrics (English, auto eval, based on OpenAI gpt-4o-transcribe, speaker distance using Revai/pyannote-wespeaker-voxceleb-resnet34-LM):
- S1:
- WER (Word Error Rate): 0.008
- CER (Character Error Rate): 0.004
- Distance: 0.332
- S1-mini:
- WER (Word Error Rate): 0.011
- CER (Character Error Rate): 0.005
- Distance: 0.380
License
This model is permissively licensed under the CC-BY-NC-SA-4.0 license.
