Kit-Lemonfoot commited on
Commit
dcb0521
1 Parent(s): 9bf15a9

Added more Holostars, more Phase, Kizuna, updated Pekora and Kaela, and fixed many bugs.

Browse files
Files changed (2) hide show
  1. app.py +78 -22
  2. edgetts_db.py +230 -0
app.py CHANGED
@@ -27,20 +27,23 @@ from lib.infer_pack.models import (
27
  )
28
  from vc_infer_pipeline import VC
29
  from config import Config
 
 
30
  config = Config()
31
  logging.getLogger("numba").setLevel(logging.WARNING)
32
  limitation = os.getenv("SYSTEM") == "spaces"
33
  #limitation=True
 
34
 
35
  audio_mode = []
36
  f0method_mode = []
37
  if limitation is True:
38
  f0method_info = "PM is better for testing, RMVPE is better for finalized generations. (Default: RMVPE)"
39
- audio_mode = ["TTS Audio", "Upload audio"]
40
  f0method_mode = ["pm", "rmvpe"]
41
  else:
42
  f0method_info = "PM is fast but low quality, crepe and harvest are slow but good quality, RMVPE is the best of both worlds. (Default: RMVPE)"
43
- audio_mode = ["TTS Audio", "Youtube", "Upload audio"]
44
  f0method_mode = ["pm", "crepe", "harvest", "rmvpe"]
45
 
46
  #if os.path.isfile("rmvpe.pt"):
@@ -54,7 +57,7 @@ vcArr.append(VC(32000, config))
54
  vcArr.append(VC(40000, config))
55
  vcArr.append(VC(48000, config))
56
 
57
- def infer(name, path, index, vc_audio_mode, vc_input, vc_upload, tts_text, tts_voice, f0_up_key, f0_method, index_rate, filter_radius, resample_sr, rms_mix_rate, protect):
58
  try:
59
  #Setup audio
60
  if vc_audio_mode == "Input path" or "Youtube" and vc_input != "":
@@ -71,17 +74,35 @@ def infer(name, path, index, vc_audio_mode, vc_input, vc_upload, tts_text, tts_v
71
  audio = librosa.to_mono(audio.transpose(1, 0))
72
  if sampling_rate != 16000:
73
  audio = librosa.resample(audio, orig_sr=sampling_rate, target_sr=16000)
74
- elif vc_audio_mode == "TTS Audio":
 
75
  if len(tts_text) > 250 and limitation:
76
  return "Text is too long.", None
77
- if tts_text is None or tts_voice is None:
78
  return "You need to enter text and select a voice.", None
79
- asyncio.run(edge_tts.Communicate(tts_text, "-".join(tts_voice.split('-')[:-1])).save("tts.mp3"))
80
- audio, sr = librosa.load("tts.mp3", sr=16000, mono=True)
 
 
 
 
81
  duration = audio.shape[0] / sr
82
  if duration > 30 and limitation:
83
  return "Your text generated an audio that was too long.", None
84
  vc_input = "tts.mp3"
 
 
 
 
 
 
 
 
 
 
 
 
 
85
  times = [0, 0, 0]
86
  f0_up_key = int(f0_up_key)
87
 
@@ -264,9 +285,11 @@ def change_audio_mode(vc_audio_mode):
264
  gr.Slider.update(visible=False),
265
  gr.Audio.update(visible=False),
266
  gr.Button.update(visible=False),
267
- # TTS
268
  gr.Textbox.update(visible=False),
269
- gr.Dropdown.update(visible=False)
 
 
270
  )
271
  elif vc_audio_mode == "Upload audio":
272
  return (
@@ -284,9 +307,11 @@ def change_audio_mode(vc_audio_mode):
284
  gr.Slider.update(visible=False),
285
  gr.Audio.update(visible=False),
286
  gr.Button.update(visible=False),
287
- # TTS
288
  gr.Textbox.update(visible=False),
289
- gr.Dropdown.update(visible=False)
 
 
290
  )
291
  elif vc_audio_mode == "Youtube":
292
  return (
@@ -306,9 +331,11 @@ def change_audio_mode(vc_audio_mode):
306
  gr.Button.update(visible=True),
307
  # TTS
308
  gr.Textbox.update(visible=False),
309
- gr.Dropdown.update(visible=False)
 
 
310
  )
311
- elif vc_audio_mode == "TTS Audio":
312
  return (
313
  # Input & Upload
314
  gr.Textbox.update(visible=False),
@@ -326,7 +353,31 @@ def change_audio_mode(vc_audio_mode):
326
  gr.Button.update(visible=False),
327
  # TTS
328
  gr.Textbox.update(visible=True),
329
- gr.Dropdown.update(visible=True)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
330
  )
331
  else:
332
  return (
@@ -346,14 +397,15 @@ def change_audio_mode(vc_audio_mode):
346
  gr.Button.update(visible=False),
347
  # TTS
348
  gr.Textbox.update(visible=False, interactive=True),
349
- gr.Dropdown.update(visible=False, interactive=True)
 
 
350
  )
351
 
352
  if __name__ == '__main__':
353
  load_hubert()
354
  categories = load_model()
355
- tts_voice_list = asyncio.get_event_loop().run_until_complete(edge_tts.list_voices())
356
- voices = [f"{v['ShortName']}-{v['Gender']}" for v in tts_voice_list]
357
  with gr.Blocks(theme=gr.themes.Base()) as app:
358
  gr.Markdown(
359
  "# <center> VTuber RVC Models\n"
@@ -383,7 +435,7 @@ if __name__ == '__main__':
383
  )
384
  with gr.Row():
385
  with gr.Column():
386
- vc_audio_mode = gr.Dropdown(label="Input voice", choices=audio_mode, allow_custom_value=False, value="TTS Audio")
387
  # Input and Upload
388
  vc_input = gr.Textbox(label="Input audio path", visible=False)
389
  vc_upload = gr.Audio(label="Upload audio file", visible=False, interactive=True)
@@ -397,7 +449,9 @@ if __name__ == '__main__':
397
  vc_audio_preview = gr.Audio(label="Audio Preview", visible=False)
398
  # TTS
399
  tts_text = gr.Textbox(visible=True, label="TTS text", info="Text to speech input (There is a limit of 250 characters)", interactive=True)
400
- tts_voice = gr.Dropdown(label="Edge-tts speaker", choices=voices, visible=True, allow_custom_value=False, value="en-US-AnaNeural-Female", interactive=True)
 
 
401
  with gr.Column():
402
  vc_transform0 = gr.Number(label="Transpose", value=0, info='Type "12" to change from male to female voice. Type "-12" to change female to male voice')
403
  f0method0 = gr.Radio(
@@ -489,6 +543,7 @@ if __name__ == '__main__':
489
  resample_sr0,
490
  rms_mix_rate0,
491
  protect0,
 
492
  ],
493
  outputs=[vc_log, vc_output]
494
  )
@@ -519,15 +574,16 @@ if __name__ == '__main__':
519
  vc_combined_output,
520
  vc_combine,
521
  tts_text,
522
- tts_voice
 
523
  ]
524
  )
525
  gr.Markdown(
526
  "## <center>Credit to:\n"
527
  "#### <center>Original devs:\n"
528
- "<center>the RVC Project, lj1995, zomehwh \n\n"
529
  "#### <center>Model creators:\n"
530
- "<center>dacoolkid44, Hijack, Maki Ligon, megaaziib, KitLemonfoot, yeey5, Sui, MahdeenSky, Itaxhix, Acato, Kyuubical, Listra92, IshimaIshimsky, ZomballTH, Jotape91, RigidSpinner, RandomAssBettel, Mimizukari, Oida, Shu-Kun, Nhat Minh, Ardha27, Legitdark, TempoHawk, 0x3e9, Kaiaya, Skeetawn, Sonphantrung, Pianissimo, Gloomwastragic, Sunesu, Aimbo, Act8113, Blyxeen\n"
531
  )
532
  if limitation is True:
533
  app.queue(concurrency_count=1, max_size=20, api_open=config.api).launch(share=config.colab)
 
27
  )
28
  from vc_infer_pipeline import VC
29
  from config import Config
30
+ from edgetts_db import tts_order_voice
31
+
32
  config = Config()
33
  logging.getLogger("numba").setLevel(logging.WARNING)
34
  limitation = os.getenv("SYSTEM") == "spaces"
35
  #limitation=True
36
+ language_dict = tts_order_voice
37
 
38
  audio_mode = []
39
  f0method_mode = []
40
  if limitation is True:
41
  f0method_info = "PM is better for testing, RMVPE is better for finalized generations. (Default: RMVPE)"
42
+ audio_mode = ["Edge-TTS", "Upload audio", "Record Audio"]
43
  f0method_mode = ["pm", "rmvpe"]
44
  else:
45
  f0method_info = "PM is fast but low quality, crepe and harvest are slow but good quality, RMVPE is the best of both worlds. (Default: RMVPE)"
46
+ audio_mode = ["Edge-TTS", "Youtube", "Upload audio", "Record Audio"]
47
  f0method_mode = ["pm", "crepe", "harvest", "rmvpe"]
48
 
49
  #if os.path.isfile("rmvpe.pt"):
 
57
  vcArr.append(VC(40000, config))
58
  vcArr.append(VC(48000, config))
59
 
60
+ def infer(name, path, index, vc_audio_mode, vc_input, vc_upload, tts_text, tts_voice, f0_up_key, f0_method, index_rate, filter_radius, resample_sr, rms_mix_rate, protect, record_button):
61
  try:
62
  #Setup audio
63
  if vc_audio_mode == "Input path" or "Youtube" and vc_input != "":
 
74
  audio = librosa.to_mono(audio.transpose(1, 0))
75
  if sampling_rate != 16000:
76
  audio = librosa.resample(audio, orig_sr=sampling_rate, target_sr=16000)
77
+ tts_text = "Uploaded Audio"
78
+ elif vc_audio_mode == "Edge-TTS":
79
  if len(tts_text) > 250 and limitation:
80
  return "Text is too long.", None
81
+ if tts_text is None or tts_voice is None or tts_text=="":
82
  return "You need to enter text and select a voice.", None
83
+ voice = language_dict[tts_voice]
84
+ asyncio.run(edge_tts.Communicate(tts_text, voice).save("tts.mp3"))
85
+ try:
86
+ audio, sr = librosa.load("tts.mp3", sr=16000, mono=True)
87
+ except:
88
+ return "ERROR: Invalid characters for the chosen TTS speaker. (Change your TTS speaker to one that supports your language!)", None
89
  duration = audio.shape[0] / sr
90
  if duration > 30 and limitation:
91
  return "Your text generated an audio that was too long.", None
92
  vc_input = "tts.mp3"
93
+ elif vc_audio_mode == "Record Audio":
94
+ if record_button is None:
95
+ return "Please record some audio.", None
96
+ sampling_rate, audio = record_button
97
+ duration = audio.shape[0] / sampling_rate
98
+ if duration > 60 and limitation:
99
+ return "Too long! Please record an audio file that is less than 1 minute.", None
100
+ audio = (audio / np.iinfo(audio.dtype).max).astype(np.float32)
101
+ if len(audio.shape) > 1:
102
+ audio = librosa.to_mono(audio.transpose(1, 0))
103
+ if sampling_rate != 16000:
104
+ audio = librosa.resample(audio, orig_sr=sampling_rate, target_sr=16000)
105
+ tts_text = "Recorded Audio"
106
  times = [0, 0, 0]
107
  f0_up_key = int(f0_up_key)
108
 
 
285
  gr.Slider.update(visible=False),
286
  gr.Audio.update(visible=False),
287
  gr.Button.update(visible=False),
288
+ # EdgeTTS
289
  gr.Textbox.update(visible=False),
290
+ gr.Dropdown.update(visible=False),
291
+ # Record Own
292
+ gr.Audio.update(visible=False)
293
  )
294
  elif vc_audio_mode == "Upload audio":
295
  return (
 
307
  gr.Slider.update(visible=False),
308
  gr.Audio.update(visible=False),
309
  gr.Button.update(visible=False),
310
+ # EdgeTTS
311
  gr.Textbox.update(visible=False),
312
+ gr.Dropdown.update(visible=False),
313
+ # Record Own
314
+ gr.Audio.update(visible=False)
315
  )
316
  elif vc_audio_mode == "Youtube":
317
  return (
 
331
  gr.Button.update(visible=True),
332
  # TTS
333
  gr.Textbox.update(visible=False),
334
+ gr.Dropdown.update(visible=False),
335
+ # Record Own
336
+ gr.Audio.update(visible=False)
337
  )
338
+ elif vc_audio_mode == "Edge-TTS":
339
  return (
340
  # Input & Upload
341
  gr.Textbox.update(visible=False),
 
353
  gr.Button.update(visible=False),
354
  # TTS
355
  gr.Textbox.update(visible=True),
356
+ gr.Dropdown.update(visible=True),
357
+ # Record Own
358
+ gr.Audio.update(visible=False)
359
+ )
360
+ elif vc_audio_mode == "Record Audio":
361
+ return (
362
+ # Input & Upload
363
+ gr.Textbox.update(visible=False),
364
+ gr.Audio.update(visible=False),
365
+ # Youtube
366
+ gr.Dropdown.update(visible=False),
367
+ gr.Textbox.update(visible=False),
368
+ gr.Dropdown.update(visible=False),
369
+ gr.Button.update(visible=False),
370
+ gr.Audio.update(visible=False),
371
+ gr.Audio.update(visible=False),
372
+ gr.Audio.update(visible=False),
373
+ gr.Slider.update(visible=False),
374
+ gr.Audio.update(visible=False),
375
+ gr.Button.update(visible=False),
376
+ # TTS
377
+ gr.Textbox.update(visible=False),
378
+ gr.Dropdown.update(visible=False),
379
+ # Record Own
380
+ gr.Audio.update(visible=True)
381
  )
382
  else:
383
  return (
 
397
  gr.Button.update(visible=False),
398
  # TTS
399
  gr.Textbox.update(visible=False, interactive=True),
400
+ gr.Dropdown.update(visible=False, interactive=True),
401
+ # Record Own
402
+ gr.Audio.update(visible=False)
403
  )
404
 
405
  if __name__ == '__main__':
406
  load_hubert()
407
  categories = load_model()
408
+ voices = list(language_dict.keys())
 
409
  with gr.Blocks(theme=gr.themes.Base()) as app:
410
  gr.Markdown(
411
  "# <center> VTuber RVC Models\n"
 
435
  )
436
  with gr.Row():
437
  with gr.Column():
438
+ vc_audio_mode = gr.Dropdown(label="Input voice", choices=audio_mode, allow_custom_value=False, value="Edge-TTS")
439
  # Input and Upload
440
  vc_input = gr.Textbox(label="Input audio path", visible=False)
441
  vc_upload = gr.Audio(label="Upload audio file", visible=False, interactive=True)
 
449
  vc_audio_preview = gr.Audio(label="Audio Preview", visible=False)
450
  # TTS
451
  tts_text = gr.Textbox(visible=True, label="TTS text", info="Text to speech input (There is a limit of 250 characters)", interactive=True)
452
+ tts_voice = gr.Dropdown(label="Edge-tts speaker", choices=voices, visible=True, allow_custom_value=False, value="English-Ana (Female)", interactive=True)
453
+ # Record Own
454
+ record_button = gr.Audio(source="microphone", label="Record your own audio", visible=False, interactive=True)
455
  with gr.Column():
456
  vc_transform0 = gr.Number(label="Transpose", value=0, info='Type "12" to change from male to female voice. Type "-12" to change female to male voice')
457
  f0method0 = gr.Radio(
 
543
  resample_sr0,
544
  rms_mix_rate0,
545
  protect0,
546
+ record_button
547
  ],
548
  outputs=[vc_log, vc_output]
549
  )
 
574
  vc_combined_output,
575
  vc_combine,
576
  tts_text,
577
+ tts_voice,
578
+ record_button
579
  ]
580
  )
581
  gr.Markdown(
582
  "## <center>Credit to:\n"
583
  "#### <center>Original devs:\n"
584
+ "<center>the RVC Project, lj1995, zomehwh, sysf\n\n"
585
  "#### <center>Model creators:\n"
586
+ "<center>dacoolkid44, Hijack, Maki Ligon, megaaziib, KitLemonfoot, yeey5, Sui, MahdeenSky, Itaxhix, Acato, Kyuubical, Listra92, IshimaIshimsky, ZomballTH, Jotape91, RigidSpinner, RandomAssBettel, Mimizukari, Oida, Shu-Kun, Nhat Minh, Ardha27, Legitdark, TempoHawk, 0x3e9, Kaiaya, Skeetawn, Sonphantrung, Pianissimo, RavenCutie21, Gloomwastragic, Sunesu, Aimbo, Act8113, Blyxeen\n"
587
  )
588
  if limitation is True:
589
  app.queue(concurrency_count=1, max_size=20, api_open=config.api).launch(share=config.colab)
edgetts_db.py ADDED
@@ -0,0 +1,230 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ tts_order_voice = {'English-Jenny (Female)': 'en-US-JennyNeural',
2
+ 'English-Guy (Male)': 'en-US-GuyNeural',
3
+ 'English-Ana (Female)': 'en-US-AnaNeural',
4
+ 'English-Aria (Female)': 'en-US-AriaNeural',
5
+ 'English-Christopher (Male)': 'en-US-ChristopherNeural',
6
+ 'English-Eric (Male)': 'en-US-EricNeural',
7
+ 'English-Michelle (Female)': 'en-US-MichelleNeural',
8
+ 'English-Roger (Male)': 'en-US-RogerNeural',
9
+ 'Spanish (Mexican)-Dalia (Female)': 'es-MX-DaliaNeural',
10
+ 'Spanish (Mexican)-Jorge- (Male)': 'es-MX-JorgeNeural',
11
+ 'Korean-Sun-Hi- (Female)': 'ko-KR-SunHiNeural',
12
+ 'Korean-InJoon- (Male)': 'ko-KR-InJoonNeural',
13
+ 'Thai-Premwadee- (Female)': 'th-TH-PremwadeeNeural',
14
+ 'Thai-Niwat- (Male)': 'th-TH-NiwatNeural',
15
+ 'Vietnamese-HoaiMy- (Female)': 'vi-VN-HoaiMyNeural',
16
+ 'Vietnamese-NamMinh- (Male)': 'vi-VN-NamMinhNeural',
17
+ 'Japanese-Nanami- (Female)': 'ja-JP-NanamiNeural',
18
+ 'Japanese-Keita- (Male)': 'ja-JP-KeitaNeural',
19
+ 'French-Denise- (Female)': 'fr-FR-DeniseNeural',
20
+ 'French-Eloise- (Female)': 'fr-FR-EloiseNeural',
21
+ 'French-Henri- (Male)': 'fr-FR-HenriNeural',
22
+ 'Brazilian-Francisca- (Female)': 'pt-BR-FranciscaNeural',
23
+ 'Brazilian-Antonio- (Male)': 'pt-BR-AntonioNeural',
24
+ 'Indonesian-Ardi- (Male)': 'id-ID-ArdiNeural',
25
+ 'Indonesian-Gadis- (Female)': 'id-ID-GadisNeural',
26
+ 'Hebrew-Avri- (Male)': 'he-IL-AvriNeural',
27
+ 'Hebrew-Hila- (Female)': 'he-IL-HilaNeural',
28
+ 'Italian-Isabella- (Female)': 'it-IT-IsabellaNeural',
29
+ 'Italian-Diego- (Male)': 'it-IT-DiegoNeural',
30
+ 'Italian-Elsa- (Female)': 'it-IT-ElsaNeural',
31
+ 'Dutch-Colette- (Female)': 'nl-NL-ColetteNeural',
32
+ 'Dutch-Fenna- (Female)': 'nl-NL-FennaNeural',
33
+ 'Dutch-Maarten- (Male)': 'nl-NL-MaartenNeural',
34
+ 'Malese-Osman- (Male)': 'ms-MY-OsmanNeural',
35
+ 'Malese-Yasmin- (Female)': 'ms-MY-YasminNeural',
36
+ 'Norwegian-Pernille- (Female)': 'nb-NO-PernilleNeural',
37
+ 'Norwegian-Finn- (Male)': 'nb-NO-FinnNeural',
38
+ 'Swedish-Sofie- (Female)': 'sv-SE-SofieNeural',
39
+ 'ArabicSwedish-Mattias- (Male)': 'sv-SE-MattiasNeural',
40
+ 'Arabic-Hamed- (Male)': 'ar-SA-HamedNeural',
41
+ 'Arabic-Zariyah- (Female)': 'ar-SA-ZariyahNeural',
42
+ 'Greek-Athina- (Female)': 'el-GR-AthinaNeural',
43
+ 'Greek-Nestoras- (Male)': 'el-GR-NestorasNeural',
44
+ 'German-Katja- (Female)': 'de-DE-KatjaNeural',
45
+ 'German-Amala- (Female)': 'de-DE-AmalaNeural',
46
+ 'German-Conrad- (Male)': 'de-DE-ConradNeural',
47
+ 'German-Killian- (Male)': 'de-DE-KillianNeural',
48
+ 'Afrikaans-Adri- (Female)': 'af-ZA-AdriNeural',
49
+ 'Afrikaans-Willem- (Male)': 'af-ZA-WillemNeural',
50
+ 'Ethiopian-Ameha- (Male)': 'am-ET-AmehaNeural',
51
+ 'Ethiopian-Mekdes- (Female)': 'am-ET-MekdesNeural',
52
+ 'Arabic (UAD)-Fatima- (Female)': 'ar-AE-FatimaNeural',
53
+ 'Arabic (UAD)-Hamdan- (Male)': 'ar-AE-HamdanNeural',
54
+ 'Arabic (Bahrain)-Ali- (Male)': 'ar-BH-AliNeural',
55
+ 'Arabic (Bahrain)-Laila- (Female)': 'ar-BH-LailaNeural',
56
+ 'Arabic (Algeria)-Ismael- (Male)': 'ar-DZ-IsmaelNeural',
57
+ 'Arabic (Egypt)-Salma- (Female)': 'ar-EG-SalmaNeural',
58
+ 'Arabic (Egypt)-Shakir- (Male)': 'ar-EG-ShakirNeural',
59
+ 'Arabic (Iraq)-Bassel- (Male)': 'ar-IQ-BasselNeural',
60
+ 'Arabic (Iraq)-Rana- (Female)': 'ar-IQ-RanaNeural',
61
+ 'Arabic (Jordan)-Sana- (Female)': 'ar-JO-SanaNeural',
62
+ 'Arabic (Jordan)-Taim- (Male)': 'ar-JO-TaimNeural',
63
+ 'Arabic (Kuwait)-Fahed- (Male)': 'ar-KW-FahedNeural',
64
+ 'Arabic (Kuwait)-Noura- (Female)': 'ar-KW-NouraNeural',
65
+ 'Arabic (Lebanon)-Layla- (Female)': 'ar-LB-LaylaNeural',
66
+ 'Arabic (Lebanon)-Rami- (Male)': 'ar-LB-RamiNeural',
67
+ 'Arabic (Libya)-Iman- (Female)': 'ar-LY-ImanNeural',
68
+ 'Arabic (Libya)-Omar- (Male)': 'ar-LY-OmarNeural',
69
+ 'Arabic (Morocco)-Jamal- (Male)': 'ar-MA-JamalNeural',
70
+ 'Arabic (Morocco)-Mouna- (Female)': 'ar-MA-MounaNeural',
71
+ 'Arabic (Oman)-Abdullah- (Male)': 'ar-OM-AbdullahNeural',
72
+ 'Arabic (Oman)-Aysha- (Female)': 'ar-OM-AyshaNeural',
73
+ 'Arabic (Qatar)-Amal- (Female)': 'ar-QA-AmalNeural',
74
+ 'Arabic (Qatar)-Moaz- (Male)': 'ar-QA-MoazNeural',
75
+ 'Arabic (Syrian Arab Republic)-Amany- (Female)': 'ar-SY-AmanyNeural',
76
+ 'Arabic (Syrian Arab Republic)-Laith- (Male)': 'ar-SY-LaithNeural',
77
+ 'Arabic (Tunisia)-Hedi- (Male)': 'ar-TN-HediNeural',
78
+ 'Arabic (Tunisia)-Reem- (Female)': 'ar-TN-ReemNeural',
79
+ 'Arabic (Yemen )-Maryam- (Female)': 'ar-YE-MaryamNeural',
80
+ 'Arabic (Yemen )-Saleh- (Male)': 'ar-YE-SalehNeural',
81
+ 'Azerbaijani-Babek- (Male)': 'az-AZ-BabekNeural',
82
+ 'Azerbaijani-Banu- (Female)': 'az-AZ-BanuNeural',
83
+ 'Bulgarian-Borislav- (Male)': 'bg-BG-BorislavNeural',
84
+ 'Bulgarian-Kalina- (Female)': 'bg-BG-KalinaNeural',
85
+ 'Bengali (Bangladesh)-Nabanita- (Female)': 'bn-BD-NabanitaNeural',
86
+ 'Bengali (Bangladesh)-Pradeep- (Male)': 'bn-BD-PradeepNeural',
87
+ 'Bengali (India)-Bashkar- (Male)': 'bn-IN-BashkarNeural',
88
+ 'Bengali (India)-Tanishaa- (Female)': 'bn-IN-TanishaaNeural',
89
+ 'Bosniak (Bosnia and Herzegovina)-Goran- (Male)': 'bs-BA-GoranNeural',
90
+ 'Bosniak (Bosnia and Herzegovina)-Vesna- (Female)': 'bs-BA-VesnaNeural',
91
+ 'Catalan (Spain)-Joana- (Female)': 'ca-ES-JoanaNeural',
92
+ 'Catalan (Spain)-Enric- (Male)': 'ca-ES-EnricNeural',
93
+ 'Czech (Czech Republic)-Antonin- (Male)': 'cs-CZ-AntoninNeural',
94
+ 'Czech (Czech Republic)-Vlasta- (Female)': 'cs-CZ-VlastaNeural',
95
+ 'Welsh (UK)-Aled- (Male)': 'cy-GB-AledNeural',
96
+ 'Welsh (UK)-Nia- (Female)': 'cy-GB-NiaNeural',
97
+ 'Danish (Denmark)-Christel- (Female)': 'da-DK-ChristelNeural',
98
+ 'Danish (Denmark)-Jeppe- (Male)': 'da-DK-JeppeNeural',
99
+ 'German (Austria)-Ingrid- (Female)': 'de-AT-IngridNeural',
100
+ 'German (Austria)-Jonas- (Male)': 'de-AT-JonasNeural',
101
+ 'German (Switzerland)-Jan- (Male)': 'de-CH-JanNeural',
102
+ 'German (Switzerland)-Leni- (Female)': 'de-CH-LeniNeural',
103
+ 'English (Australia)-Natasha- (Female)': 'en-AU-NatashaNeural',
104
+ 'English (Australia)-William- (Male)': 'en-AU-WilliamNeural',
105
+ 'English (Canada)-Clara- (Female)': 'en-CA-ClaraNeural',
106
+ 'English (Canada)-Liam- (Male)': 'en-CA-LiamNeural',
107
+ 'English (UK)-Libby- (Female)': 'en-GB-LibbyNeural',
108
+ 'English (UK)-Maisie- (Female)': 'en-GB-MaisieNeural',
109
+ 'English (UK)-Ryan- (Male)': 'en-GB-RyanNeural',
110
+ 'English (UK)-Sonia- (Female)': 'en-GB-SoniaNeural',
111
+ 'English (UK)-Thomas- (Male)': 'en-GB-ThomasNeural',
112
+ 'English (Hong Kong)-Sam- (Male)': 'en-HK-SamNeural',
113
+ 'English (Hong Kong)-Yan- (Female)': 'en-HK-YanNeural',
114
+ 'English (Ireland)-Connor- (Male)': 'en-IE-ConnorNeural',
115
+ 'English (Ireland)-Emily- (Female)': 'en-IE-EmilyNeural',
116
+ 'English (India)-Neerja- (Female)': 'en-IN-NeerjaNeural',
117
+ 'English (India)-Prabhat- (Male)': 'en-IN-PrabhatNeural',
118
+ 'English (Kenya)-Asilia- (Female)': 'en-KE-AsiliaNeural',
119
+ 'English (Kenya)-Chilemba- (Male)': 'en-KE-ChilembaNeural',
120
+ 'English (Nigeria)-Abeo- (Male)': 'en-NG-AbeoNeural',
121
+ 'English (Nigeria)-Ezinne- (Female)': 'en-NG-EzinneNeural',
122
+ 'English (New Zealand)-Mitchell- (Male)': 'en-NZ-MitchellNeural',
123
+ 'English (Philippines)-James- (Male)': 'en-PH-JamesNeural',
124
+ 'English (Philippines)-Rosa- (Female)': 'en-PH-RosaNeural',
125
+ 'English (Singapore)-Luna- (Female)': 'en-SG-LunaNeural',
126
+ 'English (Singapore)-Wayne- (Male)': 'en-SG-WayneNeural',
127
+ 'English (Tanzania)-Elimu- (Male)': 'en-TZ-ElimuNeural',
128
+ 'English (Tanzania)-Imani- (Female)': 'en-TZ-ImaniNeural',
129
+ 'English (South Africa)-Leah- (Female)': 'en-ZA-LeahNeural',
130
+ 'English (South Africa)-Luke- (Male)': 'en-ZA-LukeNeural',
131
+ 'Spanish (Argentina)-Elena- (Female)': 'es-AR-ElenaNeural',
132
+ 'Spanish (Argentina)-Tomas- (Male)': 'es-AR-TomasNeural',
133
+ 'Spanish (Bolivia)-Marcelo- (Male)': 'es-BO-MarceloNeural',
134
+ 'Spanish (Bolivia)-Sofia- (Female)': 'es-BO-SofiaNeural',
135
+ 'Spanish (Colombia)-Gonzalo- (Male)': 'es-CO-GonzaloNeural',
136
+ 'Spanish (Colombia)-Salome- (Female)': 'es-CO-SalomeNeural',
137
+ 'Spanish (Costa Rica)-Juan- (Male)': 'es-CR-JuanNeural',
138
+ 'Spanish (Costa Rica)-Maria- (Female)': 'es-CR-MariaNeural',
139
+ 'Spanish (Cuba)-Belkys- (Female)': 'es-CU-BelkysNeural',
140
+ 'Spanish (Dominican Republic)-Emilio- (Male)': 'es-DO-EmilioNeural',
141
+ 'Spanish (Dominican Republic)-Ramona- (Female)': 'es-DO-RamonaNeural',
142
+ 'Spanish (Ecuador)-Andrea- (Female)': 'es-EC-AndreaNeural',
143
+ 'Spanish (Ecuador)-Luis- (Male)': 'es-EC-LuisNeural',
144
+ 'Spanish (Spain)-Alvaro- (Male)': 'es-ES-AlvaroNeural',
145
+ 'Spanish (Spain)-Elvira- (Female)': 'es-ES-ElviraNeural',
146
+ 'Spanish (Equatorial Guinea)-Teresa- (Female)': 'es-GQ-TeresaNeural',
147
+ 'Spanish (Guatemala)-Andres- (Male)': 'es-GT-AndresNeural',
148
+ 'Spanish (Guatemala)-Marta- (Female)': 'es-GT-MartaNeural',
149
+ 'Spanish (Honduras)-Carlos- (Male)': 'es-HN-CarlosNeural',
150
+ 'Spanish (Honduras)-Karla- (Female)': 'es-HN-KarlaNeural',
151
+ 'Spanish (Nicaragua)-Federico- (Male)': 'es-NI-FedericoNeural',
152
+ 'Spanish (Nicaragua)-Yolanda- (Female)': 'es-NI-YolandaNeural',
153
+ 'Spanish (Panama)-Margarita- (Female)': 'es-PA-MargaritaNeural',
154
+ 'Spanish (Panama)-Roberto- (Male)': 'es-PA-RobertoNeural',
155
+ 'Spanish (Peru)-Alex- (Male)': 'es-PE-AlexNeural',
156
+ 'Spanish (Peru)-Camila- (Female)': 'es-PE-CamilaNeural',
157
+ 'Spanish (Puerto Rico)-Karina- (Female)': 'es-PR-KarinaNeural',
158
+ 'Spanish (Puerto Rico)-Victor- (Male)': 'es-PR-VictorNeural',
159
+ 'Spanish (Paraguay)-Mario- (Male)': 'es-PY-MarioNeural',
160
+ 'Spanish (Paraguay)-Tania- (Female)': 'es-PY-TaniaNeural',
161
+ 'Spanish (El Salvador)-Lorena- (Female)': 'es-SV-LorenaNeural',
162
+ 'Spanish (El Salvador)-Rodrigo- (Male)': 'es-SV-RodrigoNeural',
163
+ 'Spanish (United States)-Alonso- (Male)': 'es-US-AlonsoNeural',
164
+ 'Spanish (United States)-Paloma- (Female)': 'es-US-PalomaNeural',
165
+ 'Spanish (Uruguay)-Mateo- (Male)': 'es-UY-MateoNeural',
166
+ 'Spanish (Uruguay)-Valentina- (Female)': 'es-UY-ValentinaNeural',
167
+ 'Spanish (Venezuela)-Paola- (Female)': 'es-VE-PaolaNeural',
168
+ 'Spanish (Venezuela)-Sebastian- (Male)': 'es-VE-SebastianNeural',
169
+ 'Estonian (Estonia)-Anu- (Female)': 'et-EE-AnuNeural',
170
+ 'Estonian (Estonia)-Kert- (Male)': 'et-EE-KertNeural',
171
+ 'Persian (Iran)-Dilara- (Female)': 'fa-IR-DilaraNeural',
172
+ 'Persian (Iran)-Farid- (Male)': 'fa-IR-FaridNeural',
173
+ 'Finnish (Finland)-Harri- (Male)': 'fi-FI-HarriNeural',
174
+ 'Finnish (Finland)-Noora- (Female)': 'fi-FI-NooraNeural',
175
+ 'French (Belgium)-Charline- (Female)': 'fr-BE-CharlineNeural',
176
+ 'French (Belgium)-Gerard- (Male)': 'fr-BE-GerardNeural',
177
+ 'French (Canada)-Sylvie- (Female)': 'fr-CA-SylvieNeural',
178
+ 'French (Canada)-Antoine- (Male)': 'fr-CA-AntoineNeural',
179
+ 'French (Canada)-Jean- (Male)': 'fr-CA-JeanNeural',
180
+ 'French (Switzerland)-Ariane- (Female)': 'fr-CH-ArianeNeural',
181
+ 'French (Switzerland)-Fabrice- (Male)': 'fr-CH-FabriceNeural',
182
+ 'Irish (Ireland)-Colm- (Male)': 'ga-IE-ColmNeural',
183
+ 'Irish (Ireland)-Orla- (Female)': 'ga-IE-OrlaNeural',
184
+ 'Galician (Spain)-Roi- (Male)': 'gl-ES-RoiNeural',
185
+ 'Galician (Spain)-Sabela- (Female)': 'gl-ES-SabelaNeural',
186
+ 'Gujarati (India)-Dhwani- (Female)': 'gu-IN-DhwaniNeural',
187
+ 'Gujarati (India)-Niranjan- (Male)': 'gu-IN-NiranjanNeural',
188
+ 'Hindi (India)-Madhur- (Male)': 'hi-IN-MadhurNeural',
189
+ 'Hindi (India)-Swara- (Female)': 'hi-IN-SwaraNeural',
190
+ 'Croatian (Croatia)-Gabrijela- (Female)': 'hr-HR-GabrijelaNeural',
191
+ 'Croatian (Croatia)-Srecko- (Male)': 'hr-HR-SreckoNeural',
192
+ 'Hungarian (Hungary)-Noemi- (Female)': 'hu-HU-NoemiNeural',
193
+ 'Hungarian (Hungary)-Tamas- (Male)': 'hu-HU-TamasNeural',
194
+ 'Icelandic (Iceland)-Gudrun- (Female)': 'is-IS-GudrunNeural',
195
+ 'Icelandic (Iceland)-Gunnar- (Male)': 'is-IS-GunnarNeural',
196
+ 'Javanese (Indonesia)-Dimas- (Male)': 'jv-ID-DimasNeural',
197
+ 'Javanese (Indonesia)-Siti- (Female)': 'jv-ID-SitiNeural',
198
+ 'Georgian (Georgia)-Eka- (Female)': 'ka-GE-EkaNeural',
199
+ 'Georgian (Georgia)-Giorgi- (Male)': 'ka-GE-GiorgiNeural',
200
+ 'Kazakh (Kazakhstan)-Aigul- (Female)': 'kk-KZ-AigulNeural',
201
+ 'Kazakh (Kazakhstan)-Daulet- (Male)': 'kk-KZ-DauletNeural',
202
+ 'Khmer (Cambodia)-Piseth- (Male)': 'km-KH-PisethNeural',
203
+ 'Khmer (Cambodia)-Sreymom- (Female)': 'km-KH-SreymomNeural',
204
+ 'Kannada (India)-Gagan- (Male)': 'kn-IN-GaganNeural',
205
+ 'Kannada (India)-Sapna- (Female)': 'kn-IN-SapnaNeural',
206
+ 'Lao (Laos)-Chanthavong- (Male)': 'lo-LA-ChanthavongNeural',
207
+ 'Lao (Laos)-Keomany- (Female)': 'lo-LA-KeomanyNeural',
208
+ 'Lithuanian (Lithuania)-Leonas- (Male)': 'lt-LT-LeonasNeural',
209
+ 'Lithuanian (Lithuania)-Ona- (Female)': 'lt-LT-OnaNeural',
210
+ 'Latvian (Latvia)-Everita- (Female)': 'lv-LV-EveritaNeural',
211
+ 'Latvian (Latvia)-Nils- (Male)': 'lv-LV-NilsNeural',
212
+ 'Macedonian (North Macedonia)-Aleksandar- (Male)': 'mk-MK-AleksandarNeural',
213
+ 'Macedonian (North Macedonia)-Marija- (Female)': 'mk-MK-MarijaNeural',
214
+ 'Malayalam (India)-Midhun- (Male)': 'ml-IN-MidhunNeural',
215
+ 'Malayalam (India)-Sobhana- (Female)': 'ml-IN-SobhanaNeural',
216
+ 'Mongolian (Mongolia)-Bataa- (Male)': 'mn-MN-BataaNeural',
217
+ 'Mongolian (Mongolia)-Yesui- (Female)': 'mn-MN-YesuiNeural',
218
+ 'Marathi (India)-Aarohi- (Female)': 'mr-IN-AarohiNeural',
219
+ 'Marathi (India)-Manohar- (Male)': 'mr-IN-ManoharNeural',
220
+ 'Maltese (Malta)-Grace- (Female)': 'mt-MT-GraceNeural',
221
+ 'Maltese (Malta)-Joseph- (Male)': 'mt-MT-JosephNeural',
222
+ 'Burmese (Myanmar)-Nilar- (Female)': 'my-MM-NilarNeural',
223
+ 'Burmese (Myanmar)-Thiha- (Male)': 'my-MM-ThihaNeural',
224
+ 'Nepali (Nepal)-Hemkala- (Female)': 'ne-NP-HemkalaNeural',
225
+ 'Nepali (Nepal)-Sagar- (Male)': 'ne-NP-SagarNeural',
226
+ 'Dutch (Belgium)-Arnaud- (Male)': 'nl-BE-ArnaudNeural',
227
+ 'Dutch (Belgium)-Dena- (Female)': 'nl-BE-DenaNeural',
228
+ 'Polish (Poland)-Marek- (Male)': 'pl-PL-MarekNeural',
229
+ 'Polish (Poland)-Zofia- (Female)': 'pl-PL-ZofiaNeural',
230
+ 'Pashto (Afghanistan)-Gul Nawaz- (Male)': 'ps-AF-Gul',}