fix bug for torch.uint1-7 not support in torch<2.6

[baymax591]

What does this PR do?

fix bug for torch.uint1-7 not support in torch<2.6

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline?
• Did you read our philosophy doc (important for complex PRs)?
• Was this discussed/approved via a GitHub issue or the forum? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

before this PR

When I execute the following command

python -c "from diffusers import AutoPipelineForText2Image"

get the following error:

Traceback (most recent call last):
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/utils/import_utils.py", line 920, in _get_module
    return importlib.import_module("." + module_name, self.__name__)
  File "/root/miniconda3/envs/baymax/lib/python3.10/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1050, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1027, in _find_and_load
  File "<frozen importlib._bootstrap>", line 1006, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 688, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 883, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/pipelines/auto_pipeline.py", line 21, in <module>
    from ..models.controlnets import ControlNetUnionModel
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/models/controlnets/__init__.py", line 5, in <module>
    from .controlnet import ControlNetModel, ControlNetOutput
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/models/controlnets/controlnet.py", line 22, in <module>
    from ...loaders.single_file_model import FromOriginalModelMixin
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/loaders/single_file_model.py", line 23, in <module>
    from ..quantizers import DiffusersAutoQuantizer
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/quantizers/__init__.py", line 15, in <module>
    from .auto import DiffusersAutoQuantizer
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/quantizers/auto.py", line 31, in <module>
    from .torchao import TorchAoHfQuantizer
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/quantizers/torchao/__init__.py", line 15, in <module>
    from .torchao_quantizer import TorchAoHfQuantizer
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/quantizers/torchao/torchao_quantizer.py", line 45, in <module>
    torch.uint1,
  File "/root/miniconda3/envs/baymax/lib/python3.10/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'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "<frozen importlib._bootstrap>", line 1075, in _handle_fromlist
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/utils/import_utils.py", line 911, in __getattr__
    value = getattr(module, name)
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/utils/import_utils.py", line 910, in __getattr__
    module = self._get_module(self._class_to_module[name])
  File "/root/miniconda3/envs/baymax/lib/python3.10/site-packages/diffusers/utils/import_utils.py", line 922, in _get_module
    raise RuntimeError(
RuntimeError: Failed to import diffusers.pipelines.auto_pipeline because of the following error (look up to see its traceback):
module 'torch' has no attribute 'uint1'

The reason is that lower versions of torch do not support torch.uint1.
I added judgment on the torch version in the PR to avoid errors，and deleted redundant code.

[yiyixuxu]

oh thanks so much!
can we use is_torch_version?

diffusers/src/diffusers/utils/import_utils.py

Line 727 in c1e7fd5

def is_torch_version(operation: str, version: str):

[a-r-r-o-w]

Thank you! Apologies that I missed this 😅

I think this should be something like:

if version <= 2.5:
    # do not use any of the uintx
else:
    # use the uintx types as well

I believe the uintx data types were added in 2.5 and not 2.6

[a-r-r-o-w]

Also maybe good to have another branch that adds float8_e4m3fn and float8_e5m2 only if torch version is >= 2.2. int8 came in 2.0 I believe but since minimum diffusers requirement is torch>=1.4, another branch for it might be needed. LMK if I can help with any changes

[yiyixuxu]

are you able to run from diffusers import StableDiffusion3Pipeline?

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[yiyixuxu]

hi @baymax591
Thanks so much for the PR! we finished it up and get it merged in this PR #10368
(we normally ask before we take over a PR, because it is a bug, and we wanted to fix quickly! sorry!) your commit is there in the PR and you're an author :)

closing this one since it is merged