Azure Sound to Text Configuration

caution

For azure-stt, you need to use the correct language codes, which are different from those used in openai-whisper. For example, for Chinese in azure-stt, you would use the code zh-CN. In the roles.json file, make sure to configure the correct language code for each role. You can find a list of all language codes supported by azure-stt here: Azure STT language BCP-47 codes

Configurable parameters:

Parameter Name	Parameter Description	Default Value
language	The language of the input audio. Click here to see all language codes: Azure STT language BCP-47 codes	zh-CN
key	Azure Speech Services Key. For more information, refer to: Azure Speech Services Overview	None
service_region	Azure Speech Services Region. For more information, refer to: Azure Speech Services Overview	None

Configuration example:

roles.json

{
 "1": {  
     "start_text": "你好，我是小兔兔，请问有什么我可以帮助你的吗？",
     "prompt": "你扮演一个孩子的小伙伴，名字叫小兔兔，性格和善，说话活泼可爱，对孩子充满爱心，经常赞赏和鼓励孩子，用5岁孩子容易理解语言提供有趣和创新的回答，每次回复根据聊天主题询问她的看法以激发她的思考和好奇心，现在她来到了你身边问了第一个问题:[你是谁]",
     "stt_type": "azure-stt",
     "stt_config": {
       "language": "zh-CN",
       "key": "ef0f2781b5a24bbbbbbbbbbbbbbbbbbbb",
       "service_region": "eastus"
     }
 }
}