Skip to content

Latest commit

 

History

History
79 lines (60 loc) · 2.21 KB

README.md

File metadata and controls

79 lines (60 loc) · 2.21 KB
Raw

qa-bot

A smart qa bot that can answer questions based on existing documents.

Getting Started

Prerequisites

  1. Java 17+
  2. Milvus 2.2.0+

Installation

  1. Clone this repo
  2. Execute ./build.sh
  3. Upload and unzip the generated qa-bot-x.x.x.jar to the server

Configuration

Edit the config/application.yaml

  1. Config the markdown.files.location to the directory of the markdown files
  2. Config the markdown.files.scheduleEnabled to true if you want to auto update the markdown files
  3. Config the milvus.host and milvus.port to the Milvus server
  4. Config the milvus.useZillzCloud, milvus.zillizCloudUri and milvus.zillizCloudToken if you are using Zilliz Cloud
  5. Config other parameters as needed

Edit the qa-bot.conf

  1. Config the OPENAI_API_KEY to the OpenAI API key
  2. Config the HTTP_PROXY to the proxy server if needed
export OPENAI_API_KEY=sk-************************************************
export HTTP_PROXY=http{s}://${username}:${password}@${url}:${port}

Usage

  1. Start the service: ./scripts/startup.sh
  2. Stop the service: ./scripts/shutdown.sh
  3. Check the logs: tail -f /opt/logs/qa-bot.log
  4. Manually trigger the markdown files processing: curl http://${your-server-url}:9090/markdown/load
  5. Test the QA bot via browser: http://${your-server-url}:9090

Integrate the QA bot with your website

Refer apollo pr for an example.

<html>
    <head>
      ...
    </head>
    <body>
      ...
    </body>
    <!-- add qa bot -->
    <script>
      (function() {
        var link = document.createElement('link');
        link.rel = 'stylesheet';
        link.href = 'http://${your-server-url}:9090/qa-bot.css';
        document.head.appendChild(link);

        var script = document.createElement('script');
        script.src = 'http://${your-server-url}:9090/qa-bot.js';
        script.onload = function () {
          QABot.initialize({
            "serverUrl": "http://${your-server-url}:9090/qa",
            "documentSiteUrlPrefix": "https://${your-documentation-site-prefix-url}"
          });
        };
        document.body.appendChild(script);
      })();
    </script>
</html>