Commits · 6de0996cd10179d827d919d23f50b75320d9d1fa · xuwx1 / LightX2V

08 Aug, 2025 1 commit
- Feature:Implement single-loading multi-GPU broadcast model loading logic · 6de0996c
  wangshankun authored Aug 08, 2025
  
  6de0996c
07 Aug, 2025 1 commit
- Bug Fix: Fix incomplete parallel loading of audio model · c7eb4631
  wangshankun authored Aug 07, 2025
  
  c7eb4631
06 Aug, 2025 3 commits
- update wan2.2 moe parallel · 9067043e
  helloyongyang authored Aug 06, 2025
  
  9067043e
- Support: audio r2v dist infer · dd958c79
  wangshankun authored Aug 06, 2025
  
  dd958c79
- Support bf16/fp16 inference and mixed-precision inference with fp32 for some layers · 3e215bad
  gushiqiao authored Aug 06, 2025
  
  3e215bad
05 Aug, 2025 2 commits
- style: remove unused import (#184) · c0b36010
  PengGao authored Aug 05, 2025
  
  c0b36010
- style: add ruff isort (#183) · f6d4fc85
  PengGao authored Aug 05, 2025
  
  f6d4fc85
30 Jul, 2025 1 commit
- wan2.2 audio driven · 6943aa52
  wangshankun authored Jul 30, 2025
  
  6943aa52
28 Jul, 2025 2 commits
- 重构并行模块 · 1a881d63
  helloyongyang authored Jul 28, 2025
  
  1a881d63
- Wan系列模型可以使用radial attention;hunyuan系使用旧方法 · 18e2b23a
  wangshankun authored Jul 28, 2025
  
  18e2b23a
03 Jul, 2025 3 commits
- ♻️ Refactor: Move audio inference files to 'infer/audio' subdirectory · b8084e83
  wangshankun authored Jul 03, 2025
  
  b8084e83
- Support:radial attention · 6060ff4f
  wangshankun authored Jul 03, 2025
  
  6060ff4f
- audio驱动wan视频生成 · e58dd9fe
  wangshankun authored Jun 25, 2025
  
  e58dd9fe