liumaolin
commited on
Commit
·
46eb532
1
Parent(s):
efefd3d
更新MoYoYo语音模型配置,修改文本分割方法并新增角色模型
Browse files- 将多个角色的`text_split_method`从"cut3"修改为"cut0",以优化文本处理。
- 调整部分角色的优先级,以确保正确排序。
src/voice_dialogue/services/audio/generators/configs/moyoyo.py
CHANGED
|
@@ -12,6 +12,7 @@ BASE_PRETRAINED_FILES = {
|
|
| 12 |
|
| 13 |
# MoYoYo TTS配置列表
|
| 14 |
MOYOYO_TTS_CONFIGS = [
|
|
|
|
| 15 |
{
|
| 16 |
'repository': 'MoYoYoTech/tone-models',
|
| 17 |
'character_name': 'Doubao',
|
|
@@ -33,7 +34,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 33 |
'top_k': 5,
|
| 34 |
'top_p': 1,
|
| 35 |
'temperature': 1,
|
| 36 |
-
'text_split_method': "
|
| 37 |
'batch_size': 100,
|
| 38 |
'speed_factor': 1.1,
|
| 39 |
'split_bucket': True,
|
|
@@ -63,7 +64,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 63 |
'top_k': 5,
|
| 64 |
'top_p': 1,
|
| 65 |
'temperature': 1,
|
| 66 |
-
'text_split_method': "
|
| 67 |
'batch_size': 100,
|
| 68 |
'speed_factor': 1.1,
|
| 69 |
'split_bucket': True,
|
|
@@ -94,7 +95,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 94 |
'top_k': 5,
|
| 95 |
'top_p': 1,
|
| 96 |
'temperature': 1,
|
| 97 |
-
'text_split_method': "
|
| 98 |
'batch_size': 100,
|
| 99 |
'speed_factor': 1.1,
|
| 100 |
'split_bucket': True,
|
|
@@ -124,7 +125,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 124 |
'top_k': 5,
|
| 125 |
'top_p': 1,
|
| 126 |
'temperature': 1,
|
| 127 |
-
'text_split_method': "
|
| 128 |
'batch_size': 100,
|
| 129 |
'speed_factor': 1.1,
|
| 130 |
'split_bucket': True,
|
|
@@ -155,7 +156,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 155 |
'top_k': 5,
|
| 156 |
'top_p': 1,
|
| 157 |
'temperature': 1,
|
| 158 |
-
'text_split_method': "
|
| 159 |
'batch_size': 100,
|
| 160 |
'speed_factor': 1.1,
|
| 161 |
'split_bucket': True,
|
|
@@ -186,7 +187,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 186 |
'top_k': 5,
|
| 187 |
'top_p': 1,
|
| 188 |
'temperature': 1,
|
| 189 |
-
'text_split_method': "
|
| 190 |
'batch_size': 100,
|
| 191 |
'speed_factor': 1.1,
|
| 192 |
'split_bucket': True,
|
|
@@ -216,7 +217,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 216 |
'top_k': 5,
|
| 217 |
'top_p': 1,
|
| 218 |
'temperature': 1,
|
| 219 |
-
'text_split_method': "
|
| 220 |
'batch_size': 100,
|
| 221 |
'speed_factor': 1.1,
|
| 222 |
'split_bucket': True,
|
|
@@ -225,10 +226,42 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 225 |
'seed': 233333,
|
| 226 |
},
|
| 227 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 228 |
{
|
| 229 |
'repository': 'MoYoYoTech/tone-models',
|
| 230 |
'character_name': 'Maple',
|
| 231 |
-
'priority':
|
| 232 |
'cover_image': '',
|
| 233 |
'description': '轻松坦率',
|
| 234 |
'file_size': '241M',
|
|
@@ -246,7 +279,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 246 |
'top_k': 5,
|
| 247 |
'top_p': 1,
|
| 248 |
'temperature': 1,
|
| 249 |
-
'text_split_method': "
|
| 250 |
'batch_size': 100,
|
| 251 |
'speed_factor': 1.0,
|
| 252 |
'split_bucket': True,
|
|
@@ -258,7 +291,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 258 |
{
|
| 259 |
'repository': 'MoYoYoTech/tone-models',
|
| 260 |
'character_name': 'Cove',
|
| 261 |
-
'priority':
|
| 262 |
'cover_image': '',
|
| 263 |
'description': '沉稳直率',
|
| 264 |
'file_size': '241M',
|
|
@@ -276,7 +309,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 276 |
'top_k': 5,
|
| 277 |
'top_p': 1,
|
| 278 |
'temperature': 1,
|
| 279 |
-
'text_split_method': "
|
| 280 |
'batch_size': 100,
|
| 281 |
'speed_factor': 1.0,
|
| 282 |
'split_bucket': True,
|
|
@@ -288,7 +321,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 288 |
{
|
| 289 |
'repository': 'MoYoYoTech/tone-models',
|
| 290 |
'character_name': 'Juniper',
|
| 291 |
-
'priority':
|
| 292 |
'cover_image': '',
|
| 293 |
'description': '开放豁达',
|
| 294 |
'file_size': '241M',
|
|
@@ -306,7 +339,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 306 |
'top_k': 5,
|
| 307 |
'top_p': 1,
|
| 308 |
'temperature': 1,
|
| 309 |
-
'text_split_method': "
|
| 310 |
'batch_size': 100,
|
| 311 |
'speed_factor': 1.0,
|
| 312 |
'split_bucket': True,
|
|
@@ -318,7 +351,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 318 |
{
|
| 319 |
'repository': 'MoYoYoTech/tone-models',
|
| 320 |
'character_name': 'Ellen',
|
| 321 |
-
'priority':
|
| 322 |
'cover_image': '',
|
| 323 |
'description': '美国脱口秀主持人Ellen DeGeneres的AI语音模型,具有其标志性的幽默风格和独特的表达方式,适合制作娱乐和创意内容。',
|
| 324 |
'file_size': '241M',
|
|
@@ -336,37 +369,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 336 |
'top_k': 5,
|
| 337 |
'top_p': 1,
|
| 338 |
'temperature': 1,
|
| 339 |
-
'text_split_method': "
|
| 340 |
-
'batch_size': 100,
|
| 341 |
-
'speed_factor': 1.0,
|
| 342 |
-
'split_bucket': True,
|
| 343 |
-
'return_fragment': False,
|
| 344 |
-
'fragment_interval': 0.07,
|
| 345 |
-
'seed': 233333,
|
| 346 |
-
},
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
'repository': 'MoYoYoTech/tone-models',
|
| 350 |
-
'character_name': 'Bai Yansong',
|
| 351 |
-
'priority': 8,
|
| 352 |
-
'cover_image': '',
|
| 353 |
-
'description': '',
|
| 354 |
-
'file_size': '241M',
|
| 355 |
-
'is_chinese_voice': True,
|
| 356 |
-
'model_files': {
|
| 357 |
-
**BASE_PRETRAINED_FILES,
|
| 358 |
-
'gpt-weights': 'GPT_weights/bys_best_gpt.ckpt',
|
| 359 |
-
'sovits-weights': 'SoVITS_weights/bys_best_sovits.pth',
|
| 360 |
-
'reference_audio': 'ref_audios/bys_ref.wav',
|
| 361 |
-
},
|
| 362 |
-
'inference_parameters': {
|
| 363 |
-
'text_lang': "zh",
|
| 364 |
-
'prompt_text': "这一切都使年轻人觉得连自己都很难照顾,而且是贫于奔命,我还怎么去抚育下一代。",
|
| 365 |
-
'prompt_lang': "zh",
|
| 366 |
-
'top_k': 5,
|
| 367 |
-
'top_p': 1,
|
| 368 |
-
'temperature': 1,
|
| 369 |
-
'text_split_method': "cut4",
|
| 370 |
'batch_size': 100,
|
| 371 |
'speed_factor': 1.0,
|
| 372 |
'split_bucket': True,
|
|
@@ -375,6 +378,7 @@ MOYOYO_TTS_CONFIGS = [
|
|
| 375 |
'seed': 233333,
|
| 376 |
},
|
| 377 |
},
|
|
|
|
| 378 |
]
|
| 379 |
|
| 380 |
|
|
|
|
| 12 |
|
| 13 |
# MoYoYo TTS配置列表
|
| 14 |
MOYOYO_TTS_CONFIGS = [
|
| 15 |
+
# ================================ 中文 ================================
|
| 16 |
{
|
| 17 |
'repository': 'MoYoYoTech/tone-models',
|
| 18 |
'character_name': 'Doubao',
|
|
|
|
| 34 |
'top_k': 5,
|
| 35 |
'top_p': 1,
|
| 36 |
'temperature': 1,
|
| 37 |
+
'text_split_method': "cut0",
|
| 38 |
'batch_size': 100,
|
| 39 |
'speed_factor': 1.1,
|
| 40 |
'split_bucket': True,
|
|
|
|
| 64 |
'top_k': 5,
|
| 65 |
'top_p': 1,
|
| 66 |
'temperature': 1,
|
| 67 |
+
'text_split_method': "cut0",
|
| 68 |
'batch_size': 100,
|
| 69 |
'speed_factor': 1.1,
|
| 70 |
'split_bucket': True,
|
|
|
|
| 95 |
'top_k': 5,
|
| 96 |
'top_p': 1,
|
| 97 |
'temperature': 1,
|
| 98 |
+
'text_split_method': "cut0",
|
| 99 |
'batch_size': 100,
|
| 100 |
'speed_factor': 1.1,
|
| 101 |
'split_bucket': True,
|
|
|
|
| 125 |
'top_k': 5,
|
| 126 |
'top_p': 1,
|
| 127 |
'temperature': 1,
|
| 128 |
+
'text_split_method': "cut0",
|
| 129 |
'batch_size': 100,
|
| 130 |
'speed_factor': 1.1,
|
| 131 |
'split_bucket': True,
|
|
|
|
| 156 |
'top_k': 5,
|
| 157 |
'top_p': 1,
|
| 158 |
'temperature': 1,
|
| 159 |
+
'text_split_method': "cut0",
|
| 160 |
'batch_size': 100,
|
| 161 |
'speed_factor': 1.1,
|
| 162 |
'split_bucket': True,
|
|
|
|
| 187 |
'top_k': 5,
|
| 188 |
'top_p': 1,
|
| 189 |
'temperature': 1,
|
| 190 |
+
'text_split_method': "cut0",
|
| 191 |
'batch_size': 100,
|
| 192 |
'speed_factor': 1.1,
|
| 193 |
'split_bucket': True,
|
|
|
|
| 217 |
'top_k': 5,
|
| 218 |
'top_p': 1,
|
| 219 |
'temperature': 1,
|
| 220 |
+
'text_split_method': "cut0",
|
| 221 |
'batch_size': 100,
|
| 222 |
'speed_factor': 1.1,
|
| 223 |
'split_bucket': True,
|
|
|
|
| 226 |
'seed': 233333,
|
| 227 |
},
|
| 228 |
},
|
| 229 |
+
{
|
| 230 |
+
'repository': 'MoYoYoTech/tone-models',
|
| 231 |
+
'character_name': 'Bai Yansong',
|
| 232 |
+
'priority': 8,
|
| 233 |
+
'cover_image': '',
|
| 234 |
+
'description': '',
|
| 235 |
+
'file_size': '241M',
|
| 236 |
+
'is_chinese_voice': True,
|
| 237 |
+
'model_files': {
|
| 238 |
+
**BASE_PRETRAINED_FILES,
|
| 239 |
+
'gpt-weights': 'GPT_weights/bys_best_gpt.ckpt',
|
| 240 |
+
'sovits-weights': 'SoVITS_weights/bys_best_sovits.pth',
|
| 241 |
+
'reference_audio': 'ref_audios/bys_ref.wav',
|
| 242 |
+
},
|
| 243 |
+
'inference_parameters': {
|
| 244 |
+
'text_lang': "zh",
|
| 245 |
+
'prompt_text': "这一切都使年轻人觉得连自己都很难照顾,而且是贫于奔命,我还怎么去抚育下一代。",
|
| 246 |
+
'prompt_lang': "zh",
|
| 247 |
+
'top_k': 5,
|
| 248 |
+
'top_p': 1,
|
| 249 |
+
'temperature': 1,
|
| 250 |
+
'text_split_method': "cut0",
|
| 251 |
+
'batch_size': 100,
|
| 252 |
+
'speed_factor': 1.0,
|
| 253 |
+
'split_bucket': True,
|
| 254 |
+
'return_fragment': False,
|
| 255 |
+
'fragment_interval': 0.07,
|
| 256 |
+
'seed': 233333,
|
| 257 |
+
},
|
| 258 |
+
},
|
| 259 |
+
|
| 260 |
+
# ================================ 英文 ================================
|
| 261 |
{
|
| 262 |
'repository': 'MoYoYoTech/tone-models',
|
| 263 |
'character_name': 'Maple',
|
| 264 |
+
'priority': 6,
|
| 265 |
'cover_image': '',
|
| 266 |
'description': '轻松坦率',
|
| 267 |
'file_size': '241M',
|
|
|
|
| 279 |
'top_k': 5,
|
| 280 |
'top_p': 1,
|
| 281 |
'temperature': 1,
|
| 282 |
+
'text_split_method': "cut0",
|
| 283 |
'batch_size': 100,
|
| 284 |
'speed_factor': 1.0,
|
| 285 |
'split_bucket': True,
|
|
|
|
| 291 |
{
|
| 292 |
'repository': 'MoYoYoTech/tone-models',
|
| 293 |
'character_name': 'Cove',
|
| 294 |
+
'priority': 6,
|
| 295 |
'cover_image': '',
|
| 296 |
'description': '沉稳直率',
|
| 297 |
'file_size': '241M',
|
|
|
|
| 309 |
'top_k': 5,
|
| 310 |
'top_p': 1,
|
| 311 |
'temperature': 1,
|
| 312 |
+
'text_split_method': "cut0",
|
| 313 |
'batch_size': 100,
|
| 314 |
'speed_factor': 1.0,
|
| 315 |
'split_bucket': True,
|
|
|
|
| 321 |
{
|
| 322 |
'repository': 'MoYoYoTech/tone-models',
|
| 323 |
'character_name': 'Juniper',
|
| 324 |
+
'priority': 6,
|
| 325 |
'cover_image': '',
|
| 326 |
'description': '开放豁达',
|
| 327 |
'file_size': '241M',
|
|
|
|
| 339 |
'top_k': 5,
|
| 340 |
'top_p': 1,
|
| 341 |
'temperature': 1,
|
| 342 |
+
'text_split_method': "cut0",
|
| 343 |
'batch_size': 100,
|
| 344 |
'speed_factor': 1.0,
|
| 345 |
'split_bucket': True,
|
|
|
|
| 351 |
{
|
| 352 |
'repository': 'MoYoYoTech/tone-models',
|
| 353 |
'character_name': 'Ellen',
|
| 354 |
+
'priority': 5,
|
| 355 |
'cover_image': '',
|
| 356 |
'description': '美国脱口秀主持人Ellen DeGeneres的AI语音模型,具有其标志性的幽默风格和独特的表达方式,适合制作娱乐和创意内容。',
|
| 357 |
'file_size': '241M',
|
|
|
|
| 369 |
'top_k': 5,
|
| 370 |
'top_p': 1,
|
| 371 |
'temperature': 1,
|
| 372 |
+
'text_split_method': "cut0",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 373 |
'batch_size': 100,
|
| 374 |
'speed_factor': 1.0,
|
| 375 |
'split_bucket': True,
|
|
|
|
| 378 |
'seed': 233333,
|
| 379 |
},
|
| 380 |
},
|
| 381 |
+
|
| 382 |
]
|
| 383 |
|
| 384 |
|