GitLab

Co-authored-by: root <root@pt-de4c35727a1b4d1b9f27f422f06026ec-worker-0.pt-de4c35727a1b4d1b9f27f422f06026ec.ns-devsft-3460edd0.svc.cluster.local>

1. 修复之前的代码合并冲突，并测试通过。

---------
Co-authored-by: Yang Yong (雍洋) <yongyang1030@163.com>

Refactor initialization to set run_device before loading adapter
checkpoint.

Co-authored-by: yihuiwen <yihuiwen@sensetime.com>

Thanks to HunyuanVideo Team and ModelTC Team.

---------
Co-authored-by: gushiqiao <975033167@qq.com>
Co-authored-by: gushiqiao <77222802+gushiqiao@users.noreply.github.com>
Co-authored-by: chendingyu <chendingyu1@sensetime.com>
Co-authored-by: XHPlus <xhplus@163.com>
Co-authored-by: wangshankun <wangshankun2011@hotmail.com>
Co-authored-by: STwangyingrui <86730325+STwangyingrui@users.noreply.github.com>
Co-authored-by: root <root@pt-80f094c20fc44a8cad096e5f3dbc962e-worker-0.pt-80f094c20fc44a8cad096e5f3dbc962e.ns-devsft-3460edd0.svc.cluster.local>

Feature:
    1. added mlu590 bfloat16, single-gpu and multi-gpus inference.
    2. added mlu590 int8 inference.

主要用于步数蒸馏实验
- 支持在 highnoise / lownoise 文件夹下指定特定的checkpoint用于加载
- 删除冗余的 Wan22MoeDistill 类
- 修改同时发起多个请求时，请求中没有 seed 参数导致的 bug

support self-forcing wan2.1 dmd

Co-authored-by: PengGao <peng.gaoc@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Yang Yong (雍洋) <yongyang1030@163.com>