Multi-gpus Context Parallel training support?

[yinian-lw]

Nowadays, the number of parameters in video generation models is increasing, and the video length is increasing. When training video models, it is difficult to fit a complete video sequence(200k~ tokens) on a single GPU. Some sequence parallel training technologies can solve this problem, such as the fastvideo training framework, but the imperfection of this framework makes it difficult to use. Can the diffusers framework support sequence parallel training?

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.