AttributeError: module 'diffusers' has no attribute 'UNetUnconditionalModel'
#2
by
j35t3r
- opened
I am getting this error: AttributeError: module 'diffusers' has no attribute 'UNetUnconditionalModel'
when starting the pipeline...
My package list:
diffusers 0.15.0
denoising-diffusion-pytorch 1.1.1
ema-pytorch 0.1.4
pytorch-fid 0.3.0
pytorch-lightning 1.8.0
torch 2.0.1
torch-tb-profiler 0.4.1
torchmetrics 0.11.4
torchvision 0.15.2
Extended error:
/home/H/lib/python3.9/site-packages/diffusers/pip β
β elines/pipeline_utils.py:1039 in from_pretrained β
β β
β 1036 β β β β loaded_sub_model = passed_class_obj[name] β
β 1037 β β β else: β
β 1038 β β β β # load sub model β
β β± 1039 β β β β loaded_sub_model = load_sub_model( β
β 1040 β β β β β library_name=library_name, β
β 1041 β β β β β class_name=class_name, β
β 1042 β β β β β importable_classes=importable_classes, β
β β
β /home/H/lib/python3.9/site-packages/diffusers/pip β
β elines/pipeline_utils.py:367 in load_sub_model β
β β
β 364 ): β
β 365 β """Helper method to load the module `name` from `library_name` and `class_name`""" β
β 366 β # retrieve class candidates β
β β± 367 β class_obj, class_candidates = get_class_obj_and_candidates( β
β 368 β β library_name, class_name, importable_classes, pipelines, is_pipeline_module β
β 369 β ) β
β 370 β
β β
β /home/H/lib/python3.9/site-packages/diffusers/pip β
β elines/pipeline_utils.py:320 in get_class_obj_and_candidates β
β β
β 317 β β # else we just import it from the library. β
β 318 β β library = importlib.import_module(library_name) β
β 319 β β β
β β± 320 β β class_obj = getattr(library, class_name) β
β 321 β β class_candidates = {c: getattr(library, c, None) for c in importable_classes.key β
β 322 β β
β 323 β return class_obj, class_candidates