blackteay/GPT-SoVITS
1
0
Fork 0
mirror of https://github.com/RVC-Boss/GPT-SoVITS.git synced 2025-10-15 21:26:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add FunASR Download

Browse Source
This commit is contained in:
XXXXRT666 2025-10-05 14:37:24 +01:00
parent fc7c652105
commit 495ef66177
1 changed files with 24 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
tools/asr/funasr_asr.py
View File

@ -4,9 +4,8 @@ import argparse
import os
import traceback
# from funasr.utils import version_checker
# version_checker.check_for_update = lambda: None
from funasr import AutoModel
from modelscope import snapshot_download
from tqdm import tqdm
funasr_models = {} # 存储模型避免重复加载
@ -16,40 +15,43 @@ def only_asr(input_file, language):
try:
model = create_model(language)
text = model.generate(input=input_file)[0]["text"]
except:
except Exception:
text = ""
print(traceback.format_exc())
return text
def create_model(language="zh"):
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 language == "zh":
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 = "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"
snapshot_download(
"iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
local_dir="tools/asr/models/speech_fsmn_vad_zh-cn-16k-common-pytorch",
)
snapshot_download(
"iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
local_dir="tools/asr/models/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
)
snapshot_download(
"iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
local_dir="tools/asr/models/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
)
model_revision = "v2.0.4"
elif language == "yue":
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"
snapshot_download(
"iic/speech_UniASR_asr_2pass-cantonese-CHS-16k-common-vocab1468-tensorflow1-online",
local_dir="tools/asr/models/speech_UniASR_asr_2pass-cantonese-CHS-16k-common-vocab1468-tensorflow1-online",
)
model_revision = "master"
path_vad = path_punc = None
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…
vad_model_revision = punc_model_revision = ""
model_revision = "master"
else:
raise ValueError("FunASR 不支持该语言" + ": " + language)
raise ValueError(f"{language} is not supported")
vad_model_revision = punc_model_revision = "v2.0.4"
if language in funasr_models:
return funasr_models[language]
@ -83,7 +85,7 @@ def execute_asr(input_folder, output_folder, model_size, language):
file_path = os.path.join(input_folder, file_name)
text = model.generate(input=file_path)[0]["text"]
output.append(f"{file_path}|{output_file_name}|{language.upper()}|{text}")
except:
except Exception:
print(traceback.format_exc())
output_folder = output_folder or "output/asr_opt"