blackteay/GPT-SoVITS
1
0
Fork 0
mirror of https://github.com/RVC-Boss/GPT-SoVITS.git synced 2025-04-05 19:41:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

funasr support cantonese

Browse Source 
funasr support cantonese
This commit is contained in:
RVC-Boss 2024-08-03 00:44:25 +08:00 committed by GitHub
parent 8cbd916222
commit 8a101474b5
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 29 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tools/asr/config.py
View File

@ -18,7 +18,7 @@ def check_fw_local_models():
asr_dict = { asr_dict = {
"达摩 ASR (中文)": { "达摩 ASR (中文)": {
'lang': ['zh'], 'lang': ['zh','yue'],
'size': ['large'], 'size': ['large'],
'path': 'funasr_asr.py', 'path': 'funasr_asr.py',
'precision': ['float32'] 'precision': ['float32']

45
tools/asr/funasr_asr.py
View File

@ -8,22 +8,6 @@ from tqdm import tqdm
# version_checker.check_for_update = lambda: None # version_checker.check_for_update = lambda: None
from funasr import AutoModel from funasr import AutoModel
path_asr = 'tools/asr/models/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch'
path_vad = 'tools/asr/models/speech_fsmn_vad_zh-cn-16k-common-pytorch'
path_punc = 'tools/asr/models/punc_ct-transformer_zh-cn-common-vocab272727-pytorch'
path_asr = path_asr if os.path.exists(path_asr) else "iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
path_vad = path_vad if os.path.exists(path_vad) else "iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
path_punc = path_punc if os.path.exists(path_punc) else "iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
model = AutoModel(
model = path_asr,
model_revision = "v2.0.4",
vad_model = path_vad,
vad_model_revision = "v2.0.4",
punc_model = path_punc,
punc_model_revision = "v2.0.4",
)
def only_asr(input_file): def only_asr(input_file):
try: try:
@ -42,6 +26,7 @@ def execute_asr(input_folder, output_folder, model_size, language):
for file_name in tqdm(input_file_names): for file_name in tqdm(input_file_names):
try: try:
print(file_name)
file_path = os.path.join(input_folder, file_name) file_path = os.path.join(input_folder, file_name)
text = model.generate(input=file_path)[0]["text"] text = model.generate(input=file_path)[0]["text"]
output.append(f"{file_path}|{output_file_name}|{language.upper()}|{text}") output.append(f"{file_path}|{output_file_name}|{language.upper()}|{text}")
@ -65,12 +50,38 @@ if __name__ == '__main__':
help="Output folder to store transcriptions.") help="Output folder to store transcriptions.")
parser.add_argument("-s", "--model_size", type=str, default='large', parser.add_argument("-s", "--model_size", type=str, default='large',
help="Model Size of FunASR is Large") help="Model Size of FunASR is Large")
parser.add_argument("-l", "--language", type=str, default='zh', choices=['zh'], parser.add_argument("-l", "--language", type=str, default='zh', choices=['zh','yue'],
help="Language of the audio files.") help="Language of the audio files.")
parser.add_argument("-p", "--precision", type=str, default='float16', choices=['float16','float32'], parser.add_argument("-p", "--precision", type=str, default='float16', choices=['float16','float32'],
help="fp16 or fp32")#还没接入 help="fp16 or fp32")#还没接入
cmd = parser.parse_args() cmd = parser.parse_args()
path_vad = 'tools/asr/models/speech_fsmn_vad_zh-cn-16k-common-pytorch'
path_punc = 'tools/asr/models/punc_ct-transformer_zh-cn-common-vocab272727-pytorch'
path_vad = path_vad if os.path.exists(path_vad) else "iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
path_punc = path_punc if os.path.exists(path_punc) else "iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
vad_model_revision=punc_model_revision="v2.0.4"
if(cmd.language=="zh"):
path_asr = 'tools/asr/models/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch'
path_asr = path_asr if os.path.exists(path_asr) else "iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
model_revision="v2.0.4"
else:
path_asr = 'tools/asr/models/speech_UniASR_asr_2pass-cantonese-CHS-16k-common-vocab1468-tensorflow1-online'
path_asr = path_asr if os.path.exists(path_asr) else "iic/speech_UniASR_asr_2pass-cantonese-CHS-16k-common-vocab1468-tensorflow1-online"
model_revision="master"
path_vad=path_punc=vad_model_revision=punc_model_revision=None###友情提示：粤语带VAD识别可能会有少量shape不对报错的，但是不带VAD可以.不带vad只能分阶段单独加标点。不过标点模型对粤语效果真的不行…
model = AutoModel(
model=path_asr,
model_revision=model_revision,
vad_model=path_vad,
vad_model_revision=vad_model_revision,
punc_model=path_punc,
punc_model_revision=punc_model_revision,
)
execute_asr( execute_asr(
input_folder = cmd.input_folder, input_folder = cmd.input_folder,
output_folder = cmd.output_folder, output_folder = cmd.output_folder,