sample solution #10640

[Himasnhu-AT]

#10640

This pull request includes a new method to the FluxPipeline class in the src/diffusers/pipelines/flux/pipeline_flux_fill.py file. The addition of the parameters method allows for load distribution using torch.nn.parallel.dataparallel

• src/diffusers/pipelines/flux/pipeline_flux_fill.py: Added a parameters method to return a list of parameters from transformer, vae, text_encoder, and text_encoder_2.

Important

Wasn't able to test using testcases because im almost out of mem on my PC

USECASE:

pipe = FluxFillPipeline( ... )
pipe = torch.nn.DataParallel(pipe)

[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.

[a-r-r-o-w]

Closing as not planned since we want pipelines to eventually just be a container object without tying in information of models too closely. Explanation for how to do data parallel is here.