zhuchaowe/maxkb
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: fix bailian stt

Browse Source
This commit is contained in:
wxg0103 2024-12-23 11:25:00 +08:00 committed by wxg
parent 1113e1ffec
commit 53f4e11145
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/model/stt.py
View File

@ -4,6 +4,7 @@ from typing import Dict
import dashscope import dashscope
from dashscope.audio.asr import (Recognition) from dashscope.audio.asr import (Recognition)
from pydub import AudioSegment
from setting.models_provider.base_model_provider import MaxKBBaseModel from setting.models_provider.base_model_provider import MaxKBBaseModel
from setting.models_provider.impl.base_stt import BaseSpeechToText from setting.models_provider.impl.base_stt import BaseSpeechToText
@ -40,7 +41,6 @@ class AliyunBaiLianSpeechToText(MaxKBBaseModel, BaseSpeechToText):
dashscope.api_key = self.api_key dashscope.api_key = self.api_key
recognition = Recognition(model=self.model, recognition = Recognition(model=self.model,
format='mp3', format='mp3',
sample_rate=16000,
callback=None) callback=None)
with tempfile.NamedTemporaryFile(delete=False) as temp_file: with tempfile.NamedTemporaryFile(delete=False) as temp_file:
# 将上传的文件保存到临时文件中 # 将上传的文件保存到临时文件中
@ -49,6 +49,13 @@ class AliyunBaiLianSpeechToText(MaxKBBaseModel, BaseSpeechToText):
temp_file_path = temp_file.name temp_file_path = temp_file.name
try: try:
audio = AudioSegment.from_file(temp_file_path)
if audio.channels != 1:
audio = audio.set_channels(1)
audio = audio.set_frame_rate(16000)
# 将转换后的音频文件保存到临时文件中
audio.export(temp_file_path, format='mp3')
# 识别临时文件 # 识别临时文件
result = recognition.call(temp_file_path) result = recognition.call(temp_file_path)
text = '' text = ''