Starstruk Logo MarkStarstruk

v2.0

MinishLab/model2vecv2.0Mar 31, 2026by lipku

AI Summary

Major refactoring introducing a plugin-based architecture for digital human models and TTS, reorganizing code into specific directories, and adding Aliyun QwenTTS support.

Key Highlights

  • Refactored code structure with models moved to `avatars` directory
  • Transitioned to plugin-based architecture for models and TTS
  • Added `streamout` class for transport with RTMP output
  • Consolidated common logic into `BaseAvatar` and `BaseAsr` base classes
  • Integrated Aliyun QwenTTS

Breaking Changes

  • Architecture changed to plugin-based models and TTS
  • Code reorganization affecting integration paths

New Features

  • RTMP output support
  • Aliyun QwenTTS integration

Full Release Notes

1，重构代码，数字人模型和音频特征代码移到avatars目录下
2，数字人模型、tts、传输方式改成plugin方式接入
3，传输方式整理成单独的类，放到streamout目录下，添加rtmp输出
4，数字人推理和回贴公共部分代码整合到BaseAvatar中，各模型只需要实现自己特有部分代码
5，音频特征的切分代码整合到BaseAsr中
6，tts添加阿里云qwentts