max-fofanov

State_dict mismatch with the huggingface model

1

I stumbled upon a problem: the models have different amount of keys in their state_dicts. Can you please provide a conversion script?

Increasing RAM usage with enable_model_cpu_offload

7

### Describe the bug When using enable_model_cpu_offload on StableDiffusionXLPipeline, each consecutive __call__ takes more and more RAM. Also, after deleting pipe not all memory is freed ### Reproduction ```python import...