|
import sys |
|
import os |
|
|
|
os.environ["COQUI_TOS_AGREED"] = "1" |
|
|
|
import gradio as gr |
|
from TTS.api import TTS |
|
|
|
model_names = TTS().list_models() |
|
m = model_names[0] |
|
print(model_names) |
|
tts = TTS(m, gpu=False) |
|
tts.to("cpu") |
|
|
|
br_ = """ |
|
<script> |
|
var par = document.createElement("p"); |
|
var text = document.createTextNode("fhsgdjrs hgrtsfya"); |
|
par.appendChild(text); |
|
document.body.appendChild(par); |
|
</script> |
|
""" |
|
|
|
def predict(prompt, language, audio_file_pth, mic_file_path, use_mic, agree): |
|
if agree == True: |
|
if use_mic == True: |
|
if mic_file_path is not None: |
|
speaker_wav=mic_file_path |
|
else: |
|
gr.Warning("Please record your voice with Microphone, or uncheck Use Microphone to use reference audios") |
|
return ( |
|
None, |
|
None, |
|
) |
|
|
|
else: |
|
speaker_wav=audio_file_pth |
|
|
|
if len(prompt)<2: |
|
gr.Warning("Please give a longer prompt text") |
|
return ( |
|
None, |
|
None, |
|
) |
|
if len(prompt)>10000: |
|
gr.Warning("Text length limited to 10000 characters for this demo, please try shorter text") |
|
return ( |
|
None, |
|
None, |
|
) |
|
try: |
|
if language == "fr": |
|
if m.find("your") != -1: |
|
language = "fr-fr" |
|
if m.find("/fr/") != -1: |
|
language = None |
|
tts.tts_to_file( |
|
text=prompt, |
|
file_path="output.wav", |
|
speaker_wav=speaker_wav, |
|
language=language |
|
) |
|
except RuntimeError as e : |
|
if "device-assert" in str(e): |
|
|
|
gr.Warning("Unhandled Exception encounter, please retry in a minute") |
|
print("Cuda device-assert Runtime encountered need restart") |
|
sys.exit("Exit due to cuda device-assert") |
|
else: |
|
raise e |
|
|
|
return ( |
|
gr.make_waveform( |
|
audio="output.wav", |
|
), |
|
"output.wav", |
|
) |
|
else: |
|
gr.Warning("Please accept the Terms & Condition!") |
|
return ( |
|
None, |
|
None, |
|
) |
|
|
|
|
|
title = "XTTS Glz's remake (Fonctional Text-2-Speech)" |
|
|
|
description = f""" |
|
<a href="https://huggingface.co/coqui/XTTS-v1">XTTS</a> is a Voice generation model that lets you clone voices into different languages by using just a quick 3-second audio clip. |
|
<br/> |
|
XTTS is built on previous research, like Tortoise, with additional architectural innovations and training to make cross-language voice cloning and multilingual speech generation possible. |
|
<br/> |
|
This is the same model that powers our creator application <a href="https://coqui.ai">Coqui Studio</a> as well as the <a href="https://docs.coqui.ai">Coqui API</a>. In production we apply modifications to make low-latency streaming possible. |
|
<br/> |
|
Leave a star on the Github <a href="https://github.com/coqui-ai/TTS">TTS</a>, where our open-source inference and training code lives. |
|
<br/>{br_} |
|
<p>For faster inference without waiting in the queue, you should duplicate this space and upgrade to GPU via the settings. |
|
<br/> |
|
<a href="https://huggingface.co/spaces/coqui/xtts?duplicate=true"> |
|
<img style="margin-top: 0em; margin-bottom: 0em" src="https://bit.ly/3gLdBN6" alt="Duplicate Space"></a> |
|
</p> |
|
""" |
|
|
|
article = """ |
|
<div style='margin:20px auto;'> |
|
<p>By using this demo you agree to the terms of the Coqui Public Model License at https://coqui.ai/cpml</p> |
|
</div> |
|
""" |
|
examples = [ |
|
[ |
|
"Hello, World !, here is an example of light voice cloning. Try to upload your best audio samples quality", |
|
"en", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Je suis un lycéen français de 17 ans, passioner par la Cyber-Sécuritée et les models d'IA.", |
|
"fr", |
|
"examples/male.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Als ich sechs war, sah ich einmal ein wunderbares Bild", |
|
"de", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Cuando tenía seis años, vi una vez una imagen magnífica", |
|
"es", |
|
"examples/male.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Quando eu tinha seis anos eu vi, uma vez, uma imagem magnífica", |
|
"pt", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Kiedy miałem sześć lat, zobaczyłem pewnego razu wspaniały obrazek", |
|
"pl", |
|
"examples/male.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Un tempo lontano, quando avevo sei anni, vidi un magnifico disegno", |
|
"it", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Bir zamanlar, altı yaşındayken, muhteşem bir resim gördüm", |
|
"tr", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Когда мне было шесть лет, я увидел однажды удивительную картинку", |
|
"ru", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Toen ik een jaar of zes was, zag ik op een keer een prachtige plaat", |
|
"nl", |
|
"examples/male.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"Když mi bylo šest let, viděl jsem jednou nádherný obrázek", |
|
"cs", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
[ |
|
"当我还只有六岁的时候， 看到了一副精彩的插画", |
|
"zh-cn", |
|
"examples/female.wav", |
|
None, |
|
False, |
|
True, |
|
], |
|
] |
|
|
|
|
|
|
|
gr.Interface( |
|
fn=predict, |
|
inputs=[ |
|
gr.Textbox( |
|
label="Text Prompt", |
|
info="One or two sentences at a time is better", |
|
value="Hello, World !, here is an example of light voice cloning. Try to upload your best audio samples quality", |
|
), |
|
gr.Dropdown( |
|
label="Language", |
|
info="Select an output language for the synthesised speech", |
|
choices=[ |
|
"en", |
|
"es", |
|
"fr", |
|
"de", |
|
"it", |
|
"pt", |
|
"pl", |
|
"tr", |
|
"ru", |
|
"nl", |
|
"cs", |
|
"ar", |
|
"zh-cn", |
|
], |
|
max_choices=1, |
|
value="en", |
|
), |
|
gr.Audio( |
|
label="Reference Audio", |
|
info="Click on the ✎ button to upload your own target speaker audio", |
|
type="filepath", |
|
value="examples/female.wav", |
|
), |
|
gr.Audio(source="microphone", |
|
type="filepath", |
|
info="Use your microphone to record audio", |
|
label="Use Microphone for Reference"), |
|
gr.Checkbox(label="Check to use Microphone as Reference", |
|
value=False, |
|
info="Notice: Microphone input may not work properly under traffic",), |
|
gr.Checkbox( |
|
label="Agree", |
|
value=True, |
|
info="I agree to the terms of the Coqui Public Model License at https://coqui.ai/cpml", |
|
), |
|
], |
|
outputs=[ |
|
gr.Video(label="Waveform Visual"), |
|
gr.Audio(label="Synthesised Audio"), |
|
], |
|
title=title, |
|
description=description, |
|
article=article, |
|
examples=examples, |
|
).queue().launch(debug=True) |
