如何在项目 appsettings.json 中配置多个 QnA 模型？

Question

我有一个使用 Dispatch 工具在 Luis & QnA 模型上运行的机器人。现在这里的问题是我已经制作了另一个 QnA 模型，我想在我的机器人项目中配置它，即我想更新我的 c# 项目的 appsettings.json 中的知识库 ID。 我怎样才能做到这一点？因为没有我的第二个 QnA 没有被选中。

{
  "Logging": {
    "LogLevel": {
      "Default": "Warning"
    }
  },

  "MicrosoftAppId": "xxxxxxxxxxxxx",
  "MicrosoftAppPassword": "xxxxxxxxxxxxxxxx",

  "QnAKnowledgebaseId": "xxxxxxxxxxxx",
  "QnAAuthKey": "xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
  "QnAEndpointHostName": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",

   "QnAKnowledgebaseId": "xxxxxxxxxxxx",
  "QnAAuthKey": "xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
  "QnAEndpointHostName": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",



  "LuisAppId": "xxxxxxxxxxxxxxxxxxxx",
  "LuisAPIKey": "xxxxxxxxxxxxxxxxxxxxx",
  "LuisAPIHostName": "xxxxxx",



 `enter code here`"AllowedHosts": "*"
}

Answer 1

您为什么不能在应用设置中将其设置为 MODELNAME_QnAKnowledgebaseId 等，然后使用简单的 switch case 或在您的机器人中调用它？

},

  "MicrosoftAppId": "xxxxxxxxxxxxx",
  "MicrosoftAppPassword": "xxxxxxxxxxxxxxxx",

  "MODEL1_QnAKnowledgebaseId": "xxxxxxxxxxxx",
  "MODEL1_QnAAuthKey": "xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
  "MODEL1_QnAEndpointHostName": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",

  "MODEL2_QnAKnowledgebaseId": "xxxxxxxxxxxx",
  "MODEL2_QnAAuthKey": "xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
  "MODEL2_QnAEndpointHostName": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",

  "LuisAppId": "xxxxxxxxxxxxxxxxxxxx",
  "LuisAPIKey": "xxxxxxxxxxxxxxxxxxxxx",
  "LuisAPIHostName": "xxxxxx",

 "AllowedHosts": "*"
}

然后在您的 BotServices 中：

public BotServices(IConfiguration configuration)
        {
            // Read the setting for cognitive services (LUIS, QnA) from the appsettings.json

            MODEL1_QnA = new QnAMaker(new QnAMakerEndpoint
            {
                KnowledgeBaseId = configuration["MODEL1_QnAKnowledgebaseId"],
                EndpointKey = configuration["MODEL1_QnAEndpointKey"],
                Host = configuration["MODEL1_QnAEndpointHostName"]
            });
        }