ImportError when running diffusers tutorial code in Ubuntu 20.04

### Describe the bug

When I try to run:

`from diffusers import DDPMPipeline
ddpm = DDPMPipeline.from_pretrained("google/ddpm-cat-256", use_safetensors=True).to("cuda")
image = ddpm(num_inference_steps=25).images[0]
image`

> (habitat) /home/zhangyiqing/IROS2025/habitat-lab $ /home/zhangyiqing/miniconda3/envs/habitat/bin/python /home/zhangyiqing/IROS2025/habitat-lab/dirtyHand/PPO_topdownmap.py
Traceback (most recent call last):
File "/home/zhangyiqing/IROS2025/habitat-lab/dirtyHand/PPO_topdownmap.py", line 7, in
ddpm = DDPMPipeline.from_pretrained("google/ddpm-cat-256", use_safetensors=True).to("cuda")
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 114, in _inner_fn
return fn(*args, **kwargs)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_utils.py", line 725, in from_pretrained
cached_folder = cls.download(
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 114, in _inner_fn
return fn(*args, **kwargs)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_utils.py", line 1390, in download
ignore_patterns = _get_ignore_patterns(
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_loading_utils.py", line 906, in _get_ignore_patterns
raise EnvironmentError(
OSError: Could not find the necessary `safetensors` weights in {'diffusion_pytorch_model.safetensors', 'diffusion_pytorch_model.bin'} (variant=None)

Change "use_safetensors" to None to download the model:

> Loading pipeline components...: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 30.22it/s]
Traceback (most recent call last):
File "/home/zhangyiqing/IROS2025/habitat-lab/dirtyHand/PPO_topdownmap.py", line 7, in
ddpm = DDPMPipeline.from_pretrained("google/ddpm-cat-256", use_safetensors=None).to("cuda")
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/huggingface_hub/utils/_validators.py", line 114, in _inner_fn
return fn(*args, **kwargs)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_utils.py", line 948, in from_pretrained
model = pipeline_class(**init_kwargs)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/ddpm/pipeline_ddpm.py", line 43, in __init__
self.register_modules(unet=unet, scheduler=scheduler)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_utils.py", line 165, in register_modules
library, class_name = _fetch_class_library_tuple(module)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_loading_utils.py", line 734, in _fetch_class_library_tuple
not_compiled_module = _unwrap_model(module)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/pipelines/pipeline_loading_utils.py", line 234, in _unwrap_model
if is_compiled_module(model):
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/diffusers/utils/torch_utils.py", line 88, in is_compiled_module
if is_torch_version("<", "2.0.0") or not hasattr(torch, "_dynamo"):
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/__init__.py", line 2214, in __getattr__
return importlib.import_module(f".{name}", __name__)
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/importlib/__init__.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "", line 1030, in _gcd_import
File "", line 1007, in _find_and_load
File "", line 986, in _find_and_load_unlocked
File "", line 680, in _load_unlocked
File "", line 850, in exec_module
File "", line 228, in _call_with_frames_removed
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/__init__.py", line 2, in
from . import convert_frame, eval_frame, resume_execution
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/convert_frame.py", line 48, in
from . import config, exc, trace_rules
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/trace_rules.py", line 52, in
from .variables import (
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/variables/__init__.py", line 38, in
from .higher_order_ops import (
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/variables/higher_order_ops.py", line 14, in
import torch.onnx.operators
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/onnx/__init__.py", line 49, in
from ._internal.exporter import ( # usort:skip. needs to be last to avoid circular import
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/onnx/_internal/exporter/__init__.py", line 13, in
from ._analysis import analyze
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/onnx/_internal/exporter/_analysis.py", line 14, in
import torch._export.serde.schema
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_export/__init__.py", line 33, in
from torch._export.non_strict_utils import make_constraints
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_export/non_strict_utils.py", line 16, in
from torch._dynamo.variables.builder import TrackedFake
File "/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/variables/builder.py", line 73, in
from ..trace_rules import (
ImportError: cannot import name 'is_callable_allowed' from partially initialized module 'torch._dynamo.trace_rules' (most likely due to a circular import) (/home/zhangyiqing/miniconda3/envs/habitat/lib/python3.9/site-packages/torch/_dynamo/trace_rules.py)

### Reproduction

from diffusers import DDPMPipeline
import matplotlib.pyplot as plt

from huggingface_hub.hf_api import HfFolder;
HfFolder.save_token("token")

ddpm = DDPMPipeline.from_pretrained("google/ddpm-cat-256", use_safetensors=True).to("cuda")
image = ddpm(num_inference_steps=25).images[0]

plt.imshow(image)

### Logs

_No response_

### System Info

Copy-and-paste the text below in your GitHub issue and FILL OUT the two last points.

- 🤗 Diffusers version: 0.31.0
- Platform: Linux-5.15.0-67-generic-x86_64-with-glibc2.31
- Running on Google Colab?: No
- Python version: 3.9.20
- PyTorch version (GPU?): 2.4.0.post301 (True)
- Flax version (CPU?/GPU?/TPU?): not installed (NA)
- Jax version: not installed
- JaxLib version: not installed
- Huggingface_hub version: 0.26.5
- Transformers version: not installed
- Accelerate version: 1.2.1
- PEFT version: not installed
- Bitsandbytes version: not installed
- Safetensors version: 0.4.5
- xFormers version: not installed
- Accelerator: NVIDIA GeForce RTX 4090, 24564 MiB
- Using GPU in script?:
- Using distributed or parallel set-up in script?:

### Who can help?

_No response_

Reproduce the DDPMPipeline example with the listed Python, PyTorch, diffusers, and huggingface_hub versions, then inspect diffusers/pipelines/pipeline_loading_utils.py and the torch._dynamo import chain shown in the traceback. Done means identifying the compatibility failure and documenting or validating a supported configuration for the Ubuntu 20.04 example.

Vom Indexierungsmodell aus dem Issue-Text verfasst.