Torch Issue: AttributeError: module 'torch' has no attribute 'uint1'

[Pivo987]

Hello, I have installed everything step by step and properly on Windows.
I am actually having this issue, could someone help me please?

(MagicQuill) C:\Users\aaaaaaa\OneDrive\Desktop\MagicQuill>python gradio_run.py
Total VRAM 8192 MB, total RAM 32689 MB
pytorch version: 2.1.2+cu118
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 3070 : native

Using pytorch cross attention
Traceback (most recent call last):
File "C:\Users\aaaaaaa\OneDrive\Desktop\MagicQuill\gradio_run.py", line 17, in
from MagicQuill.scribble_color_edit import ScribbleColorEditModel
File "C:\Users\aaaaaaa\OneDrive\Desktop\MagicQuill\MagicQuill\scribble_color_edit.py", line 9, in
from .brushnet_nodes import BrushNetLoader, BrushNet, BlendInpaint, get_files_with_extension
File "C:\Users\aaaaaaa\OneDrive\Desktop\MagicQuill\MagicQuill\brushnet_nodes.py", line 21, in
from .brushnet.brushnet import BrushNetModel
File "C:\Users\aaaaaaa\OneDrive\Desktop\MagicQuill\MagicQuill\brushnet\brushnet.py", line 18, in
from diffusers.models.modeling_utils import ModelMixin
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\diffusers\models\modeling_utils.py", line 35, in
from ..quantizers import DiffusersAutoQuantizer, DiffusersQuantizer
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\diffusers\quantizers_init_.py", line 15, in
from .auto import DiffusersAutoQuantizer
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\diffusers\quantizers\auto.py", line 31, in
from .torchao import TorchAoHfQuantizer
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\diffusers\quantizers\torchao_init_.py", line 15, in
from .torchao_quantizer import TorchAoHfQuantizer
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\diffusers\quantizers\torchao\torchao_quantizer.py", line 45, in
torch.uint1,
File "C:\Users\aaaaaaa\miniconda3\envs\MagicQuill\lib\site-packages\torch_init_.py", line 1833, in getattr
raise AttributeError(f"module '{name}' has no attribute '{name}'")
AttributeError: module 'torch' has no attribute 'uint1'

[armansh2210]

Same issue here as well

[GiannisCon]

@Pivo987 @armansh2210
i managed to resolve the issue by executing these commands:
first this:
pip uninstall torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2
then this:
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

[elssantosr]

@Pivo987 @armansh2210 i managed to resolve the issue by executing these commands: first this: pip uninstall torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 then this: pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

this evolves for the following error:
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
llava 1.2.2.post1 requires torch==2.1.2, but you have torch 2.5.1+cu118 which is incompatible.
llava 1.2.2.post1 requires torchvision==0.16.2, but you have torchvision 0.20.1+cu118 which is incompatible.

but even so, it goes thru the uint1 error.

thanks @GiannisCon

[GiannisCon]

@elssantosr
i get the same error as you but it runs, i can use the app when i execute the command:
python gradio_run.py

[vhrs6]

same issue

[Pivo987]

@Pivo987 @armansh2210 i managed to resolve the issue by executing these commands: first this: pip uninstall torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 then this: pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

this evolves for the following error: ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. llava 1.2.2.post1 requires torch==2.1.2, but you have torch 2.5.1+cu118 which is incompatible. llava 1.2.2.post1 requires torchvision==0.16.2, but you have torchvision 0.20.1+cu118 which is incompatible.

but even so, it goes thru the uint1 error.

thanks @GiannisCon

Same here. Error evolved, but now it is working at least. Thanks a lot @GiannisCon

[gjnave]

i belive the issue is that diffusers just updated yesteday.. been unable to get anything working but am attempting to try different versions of dependent packages.
I update one and it breaks another

[thepilip]

i get the same error as you but it runs, i can use the app when i execute the command:
python gradio_run.py

when i run gradio_run.py i get
OSError: Incorrect path_or_model_id: 'C:\Users\xxxxx\MagicQuill\models\llava-v1.5-7b-finetune-clean'. Please provide either the path to a local folder or the repo_id of a model on the Hub.

[zliucz]

@thepilip Hi, check this issue#54. Please download the llava checkpoint and place it in the correct path.