DeepSeekChatbotVB 1.0.2

DeepSeekChatbotVB

A console-based chatbot built with VB.NET, using DeepSeek API with text-to-speech (TTS) support, packaged as a .NET Global Tool.

v1.0.2 Latest Update: Added full note conversion (A#-G# to "A sharp"-"G sharp") optimized for Chinese TTS voices.

Version Notes: 1.0.0 → 1.0.2

✨ v1.0.2 Improvements:

• Extended voice processing to convert all musical notes (A#-G#) to spoken form ("A sharp"-"G sharp")
• Optimized for Chinese TTS voices which may mispronounce sharp symbols
• Replaced loop-based conversion with efficient Regex pattern matching

v1.0.1 Improvements:

• Fixed inefficient retry logic in voice saving: Now exits immediately on success instead of completing all retries
• Removed redundant save operation that always ran after retry loop
• Improved async/await pattern for better performance
• Added graceful error handling for voice save failures (shows warning instead of crashing)
• Added basic "C#" to "C sharp" conversion for voice output

Note: Starting from version 1.0.0, this .NET tool always work with the same configuration - simply create an empty folder with a deepseekvb-config.json file and run deepseekvb.

Features

• 🤖 Chat with DeepSeek AI directly from the console
• 🔊 Text-to-speech conversion using Edge TTS
• 📝 Conversation logging in YAML formats
• ⚙️ Configurable via JSON configuration file
• 🎯 Automatic retry for voice saving failures

Installation

Prerequisites

• .NET 10.0 or later

Install as .NET Global Tool

dotnet tool install --global DeepSeekChatbotVB

Update

dotnet tool update --global DeepSeekChatbotVB

Uninstall

dotnet tool uninstall --global DeepSeekChatbotVB

Usage

Quick Start

Create an empty folder, place your deepseekvb-config.json file inside it, and run:

deepseekvb

The tool will automatically create the necessary output directories (deepseekvb-voices) and files (deepseekvb-chatlog.yml) as needed.

Note: Without deepseekvb-config.json file, you will be prompted to enter your DeepSeek API key, and the chatbot will use the default configuration.

Configuration

Create a deepseekvb-config.json file in your working directory:

{
    "api_key": "sk-YOUR_DEEPSEEK_API_KEY",
    "system_prompt": [
        "You are a helpful assistant, mastering various scientific fields.",
        "Please answer the user's questions precisely in English."
    ],
    "voices": {
        "assistant": "zh-CN-YunxiNeural",
        "user": "zh-CN-YunyangNeural"
    },
    "model_params": {
        "temperature": 0.7,
        "max_tokens": 500,
        "frequency_penalty": 0.3,
        "presence_penalty": 0.3
    }
}

Configuration Options

Minimal Configuration Example

{
    "api_key": "sk-YOUR_DEEPSEEK_API_KEY",
    "system_prompt": ["You are a helpful assistant. Please reply in English."]
}

Output Files

• deepseekvb-chatlog.yml file:
  • Structured conversation log for programmatic access
• deepseekvb-voices directory:
  • user_input_{index}.mp3 (voice recording of user input)
  • bot_output_{index}.mp3 (voice recording of chatbot response)

Acknowledgments

• DeepSeek API for AI chat completions
• EdgeTTS.DotNet for text-to-speech
• YamlDotNet for YAML serialization

License

This project is licensed under the BSD-3-Clause License. See the LICENSE file for details.