No controlnet pag inpaint model in pipeline

[john09282922]

Describe the bug

Hi, I would like to utilize pag version on sdxl controlnet inpaint model. But there is no pipeline. Is it just copy and past utilizing sdxl pag model?

Reproduction

na

Logs

na

System Info

na

Who can help?

No response

[hlky]

Hi @john09282922. Yes the changes for a pag controlnet sd xl inpaint pipeline can be copied from other pipelines, would you like to create a PR to add it? We are happy to assist with the PR.

[john09282922]

Sure, when changing some code line, I made this issue.

Traceback (most recent call last):
File "/home/user/Jungmin/zest/run_control_pag.py", line 183, in
images = ip_model.generate(
File "/home/user/Jungmin/zest/IP_Adapter/ip_adapter.py", line 451, in generate
images = self.pipe(
File "/home/user/anaconda3/envs/zest/lib/python3.9/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
File "/home/user/anaconda3/envs/zest/lib/python3.9/site-packages/diffusers/pipelines/pag/pipeline_pag_controlnet_sd_xl_inpaint.py", line 1757, in call
self._set_pag_attn_processor(
File "/home/user/anaconda3/envs/zest/lib/python3.9/site-packages/diffusers/pipelines/pag/pag_utils.py", line 86, in _set_pag_attn_processor
module.processor = pag_attn_proc
File "/home/user/anaconda3/envs/zest/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1984, in setattr
raise TypeError(
TypeError: cannot assign 'diffusers.models.attention_processor.PAGCFGIdentitySelfAttnProcessor2_0' as child module 'processor' (torch.nn.Module or None expected)

@hlky

[john09282922]

I fixed the code, it works. Can you tell me how to upload the file in PR

@hlky