Tsivilev Vladislav

getting the same problem, is there a way to run flux with xformers in diffusers?

model.attn_processors: `self.attn.processor=

the problem is that in this line https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformers/transformer_flux.py#L151 with xformers enabled `attention_outputs` is tensor but in lines bellow output expected to be a tuple with len 2 or 3 -...

with installation from source `pip install git+https://github.com/huggingface/diffusers` and latest version of xformers, I'm getting the same problem - XFormerAttnProcessor and wrong output for attention_ouput attentions - `transformer.attn_processors={'transformer_blocks.0.attn.processor': , 'transformer_blocks.1.attn.processor': ,...