Repository navigation
v0.4.1 - 配置热重载升级
v0.4.1 - 配置热重载升级
中文
v0.4.1 - 配置热重载升级
主要改进
配置热重载
- 支持转录服务商无重启切换（本地/Groq/SiliconFlow/Qwen）
- 批量配置更新机制，防止重复模型加载
- 热重载后自动加载本地模型
- 配置变更即时生效，无需重启应用
性能优化
- 消除重复模型实例化（节省 226MB 内存）
- 配置批量更新减少 IO 操作
- 事件监听器生命周期优化
稳定性增强
- 修复变量命名冲突导致的热键失效问题
- 改进事件监听器生命周期管理
- 优化控制器启动和停止流程
技术细节
架构改进
- BaseController 变量重命名：
_state→_state_manager（避免与 LifecycleComponent 冲突） - 事件监听器现在在
_do_start()注册，在_do_stop()清理 - 新增
ConfigService.set_settings_batch()批量配置 API - 统一参数命名：
whisper_engine_factory→speech_service_factory
修复的问题
- 修复热重载后热键失效（变量命名冲突）
- 修复热重载后事件监听器丢失
- 修复配置切换时重复实例化模型
- 修复本地提供商切换后模型未加载
使用方法
切换转录服务商
- 打开设置窗口
- 选择"转录设置"标签
- 切换"转录提供商"（本地/Groq/SiliconFlow/Qwen）
- 点击"应用"按钮
- 立即生效，无需重启
支持的服务商
- 本地：sherpa-onnx（Paraformer/Zipformer）
- Groq：whisper-large-v3-turbo
- SiliconFlow：FunAudioLLM/SenseVoiceSmall
- Qwen：qwen3-asr-flash
升级步骤
从 v0.4.0 升级
- 下载 SonicInput-v0.4.1-win64.exe
- 关闭旧版本，替换可执行文件
- 启动应用（配置自动生效）
注意事项
- 完全兼容 v0.4.0 配置
- 无需手动调整任何设置
- 直接替换即可使用
系统要求
最低配置
- Windows 10 64-bit
- Intel Core i3 或同等 CPU
- 4GB RAM
- 500MB 磁盘空间
推荐配置
- Windows 11 64-bit
- Intel Core i5 或 AMD Ryzen 5
- 8GB RAM
- 1GB 磁盘空间
技术支持
配置和日志
- 配置：
%AppData%\SonicInput\config.json - 日志：
%AppData%\SonicInput\logs\app.log - 模型：
%AppData%\SonicInput\models/
问题反馈
- GitHub Issues: 提交 Issue
English
v0.4.1 - Configuration Hot Reload Upgrade
Key Improvements
Configuration Hot Reload
- Support switching transcription providers without restart (Local/Groq/SiliconFlow/Qwen)
- Batch configuration update mechanism to prevent duplicate model loading
- Auto-load local model after hot reload
- Instant configuration changes without app restart
Performance Optimization
- Eliminate duplicate model instantiation (save 226MB memory)
- Batch configuration updates reduce IO operations
- Event listener lifecycle optimization
Stability Enhancements
- Fix hotkey failure caused by variable naming conflict
- Improve event listener lifecycle management
- Optimize controller startup and shutdown flow
Technical Details
Architecture Improvements
- BaseController variable rename:
_state→_state_manager(avoid conflict with LifecycleComponent) - Event listeners now register in
_do_start(), cleanup in_do_stop() - New
ConfigService.set_settings_batch()batch configuration API - Unified parameter naming:
whisper_engine_factory→speech_service_factory
Fixed Issues
- Fix hotkey failure after hot reload (variable naming conflict)
- Fix event listener loss after hot reload
- Fix duplicate model instantiation during config switch
- Fix model not loaded after switching to local provider
Usage
Switch Transcription Provider
- Open Settings window
- Select "Transcription Settings" tab
- Switch "Transcription Provider" (Local/Groq/SiliconFlow/Qwen)
- Click "Apply" button
- Takes effect immediately, no restart needed
Supported Providers
- Local: sherpa-onnx (Paraformer/Zipformer)
- Groq: whisper-large-v3-turbo
- SiliconFlow: FunAudioLLM/SenseVoiceSmall
- Qwen: qwen3-asr-flash
Upgrade Steps
From v0.4.0
- Download SonicInput-v0.4.1-win64.exe
- Close old version, replace executable
- Launch app (config auto-applied)
Notes
- Fully compatible with v0.4.0 config
- No manual settings adjustment needed
- Direct replacement ready to use
System Requirements
Minimum
- Windows 10 64-bit
- Intel Core i3 or equivalent CPU
- 4GB RAM
- 500MB disk space
Recommended
- Windows 11 64-bit
- Intel Core i5 or AMD Ryzen 5
- 8GB RAM
- 1GB disk space
Technical Support
Configuration and Logs
- Config:
%AppData%\SonicInput\config.json - Logs:
%AppData%\SonicInput\logs\app.log - Models:
%AppData%\SonicInput\models/
Issue Reporting
- GitHub Issues: Submit Issue
核心改进总结: v0.4.1 实现配置热重载功能，支持转录服务商无重启切换，节省内存，提升用户体验。
Core Improvements: v0.4.1 implements configuration hot reload, supports switching transcription providers without restart, saves memory, improves user experience.