Text-to-Speech
Vui
English
tts
speech-synthesis
voice-cloning
streaming
conversational
qwen3-tts
conversational-speech
context-aware
cpu-inference
on-device
small-model
Instructions to use fluxions/vui with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Vui
How to use fluxions/vui with Vui:
# !pip install git+https://github.com/fluxions-ai/vui import torchaudio from vui.inference import render from vui.model import Vui, model = Vui.from_pretrained().cuda() waveform = render( model, "Hey, here is some random stuff, usually something quite long as the shorter the text the less likely the model can cope!", ) print(waveform.shape) torchaudio.save("out.opus", waveform[0], 22050) - Notebooks
- Google Colab
- Kaggle
Add babble probe gate for vui-190k (linear hidden-state probe, AUC 0.92)
Browse files- babble_probe-190k.pt +3 -0
babble_probe-190k.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb47f73592248c4129f2fa2f1851f1e61965a3d6af1d1597c5a55e0be96f2e0a
|
| 3 |
+
size 4931
|