Starstruk Logo MarkStarstruk

v2.0

a16z-infra/JungleGymv2.0Mar 31, 2026by lipku

AI Summary

This release represents a major architectural refactor introducing a plugin-based system for digital human models and TTS. It reorganizes code into modular directories, standardizes base classes for inference and audio processing, and adds new output capabilities.

Key Highlights

  • Plugin-based architecture for digital human models, TTS, and transmission methods
  • Code reorganization into `avatars` and `streamout` directories
  • Added RTMP output support for transmission
  • Integrated Alibaba Cloud QwenTTS

New Features

  • RTMP output support
  • Alibaba Cloud QwenTTS integration
  • Plugin architecture for models and TTS
  • BaseAvatar and BaseAsr integration for common logic

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