ValueError: Trying to set a tensor of shape torch.Size([16, 2048, 1, 1]) in "weight" #2
Description
I'm getting this error when trying to run inference (3B/3B) on Ubuntu. It's a fresh install of Stable Cascade, running in a venv.
STAGE C READY
['transforms', 'clip_preprocess', 'gdf', 'sampling_configs', 'effnet_preprocess']
Traceback (most recent call last):
File "/home/mhaines/StableCascade/inference/trained_c_3b_stage_b_3b.py", line 43, in
models_b = core_b.setup_models(extras_b, skip_clip=True)
File "/home/mhaines/StableCascade/train/train_b.py", line 182, in setup_models
set_module_tensor_to_device(generator, param_name, "cpu", value=param)
File "/home/mhaines/StableCascade/venv/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 358, in set_module_tensor_to_device
raise ValueError(
ValueError: Trying to set a tensor of shape torch.Size([16, 2048, 1, 1]) in "weight" (which has shape torch.Size([16, 320, 1, 1])), this look incorrect.