error importing `is_compiled_module`

[5shekel]

https://github.com/tencent-ailab/IP-Adapter/blob/22da45667898fd237ab54d3681db53a9ae98bf1e/ip_adapter/utils.py#L9

after

pip install diffusers[torch]
#or
pip install diffusers

im getting

>>> from diffusers.utils import is_compiled_module
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: cannot import name 'is_compiled_module' from 'diffusers.utils' (C:\tools\miniconda3\envs\test\lib\site-packages\diffusers\utils\__init__.py)

simple print(dir(diffusers.utils)) cant find it.

>>> import diffusers.utils 
>>> print(dir(diffusers.utils))
['BACKENDS_MAPPING', 'BaseOutput', 'CONFIG_NAME', 'DEPRECATED_REVISION_ARGS', 'DIFFUSERS_CACHE', 'DIFFUSERS_DYNAMIC_MODULE_NAME', 'DummyObject', 'ENV_VARS_TRUE_AND_AUTO_VALUES', 'ENV_VARS_TRUE_VALUES', 'FLAX_WEIGHTS_NAME', 'HF_HUB_OFFLINE', 'HF_MODULES_CACHE', 'HUGGINGFACE_CO_RESOLVE_ENDPOINT', 'ONNX_EXTERNAL_WEIGHTS_NAME', 'ONNX_WEIGHTS_NAME', 'OptionalDependencyNotAvailable', 'PIL_INTERPOLATION', 'PushToHubMixin', 'SAFETENSORS_WEIGHTS_NAME', 'USE_JAX', 'USE_TF', 'USE_TORCH', 'WEIGHTS_NAME', '_LazyModule', '__builtins__', '__cached__', '__doc__', '__file__', '__loader__', '__name__', '__package__', '__path__', '__spec__', '__version__', '_add_variant', '_get_model_file', 'check_min_version', 'constants', 'deprecate', 'deprecation_utils', 'doc_utils', 'dummy_flax_and_transformers_objects', 'dummy_flax_objects', 'dummy_note_seq_objects', 'dummy_onnx_objects', 'dummy_torch_and_librosa_objects', 'dummy_torch_and_scipy_objects', 'dummy_torch_and_torchsde_objects', 'dummy_torch_and_transformers_and_k_diffusion_objects', 'dummy_torch_and_transformers_and_onnx_objects', 'dummy_torch_and_transformers_objects', 'dummy_transformers_and_torch_and_note_seq_objects', 'dynamic_modules_utils', 'export_to_gif', 'export_to_obj', 'export_to_ply', 'export_to_video', 'export_utils', 'extract_commit_hash', 'get_class_from_dynamic_module', 'get_logger', 'get_objects_from_module', 'http_user_agent', 'hub_utils', 'import_utils', 'is_accelerate_available', 'is_accelerate_version', 'is_bs4_available', 'is_flax_available', 'is_ftfy_available', 'is_inflect_available', 'is_invisible_watermark_available', 'is_k_diffusion_available', 'is_k_diffusion_version', 'is_librosa_available', 'is_note_seq_available', 'is_omegaconf_available', 'is_onnx_available', 'is_scipy_available', 'is_tensorboard_available', 'is_torch_available', 'is_torch_version', 'is_torchsde_available', 'is_transformers_available', 'is_transformers_version', 'is_unidecode_available', 'is_wandb_available', 'is_xformers_available', 'load_image', 'loading_utils', 'logger', 'logging', 'make_image_grid', 'numpy_to_pil', 'os', 'outputs', 'pil_utils', 'pt_to_pil', 'replace_example_docstring', 'requires_backends', 'version']

i see others using

from diffusers.utils.torch_utils import is_compiled_module

[xiaohu2015]

@5shekel hi, it seems that the new version of diffusers has do some changes.

you can use the new code, we refactor the code to handle this.

[TheLearner23]

@5shekel hi, it seems that the new version of diffusers has do some changes.

you can use the new code, we refactor the code to handle this.

Which version are you referring to? I have installed from source ("0.22.0.dev0" version) and dir(diffusers.utils) only lists the following. Could you please advise?!

['BACKENDS_MAPPING', 'BaseOutput', 'CONFIG_NAME', 'DEPRECATED_REVISION_ARGS', 'DIFFUSERS_CACHE', 'DIFFUSERS_DYNAMIC_MODULE_NAME', 'DummyObject', 'ENV_VARS_TRUE_AND_AUTO_VALUES', 'ENV_VARS_TRUE_VALUES', 'FLAX_WEIGHTS_NAME', 'HF_HUB_OFFLINE', 'HF_MODULES_CACHE', 'HUGGINGFACE_CO_RESOLVE_ENDPOINT', 'ONNX_EXTERNAL_WEIGHTS_NAME', 'ONNX_WEIGHTS_NAME', 'OptionalDependencyNotAvailable', 'PIL_INTERPOLATION', 'PushToHubMixin', 'SAFETENSORS_WEIGHTS_NAME', 'USE_JAX', 'USE_TF', 'USE_TORCH', 'WEIGHTS_NAME', '_LazyModule', 'builtins', 'cached', 'doc', 'file', 'loader', 'name', 'package', 'path', 'spec', 'version', '_add_variant', '_get_model_file', 'check_min_version', 'constants', 'convert_state_dict_to_diffusers', 'convert_state_dict_to_peft', 'deprecate', 'deprecation_utils', 'doc_utils', 'dummy_flax_and_transformers_objects', 'dummy_flax_objects', 'dummy_note_seq_objects', 'dummy_onnx_objects', 'dummy_torch_and_librosa_objects', 'dummy_torch_and_torchsde_objects', 'dummy_torch_and_transformers_and_k_diffusion_objects', 'dummy_torch_and_transformers_and_onnx_objects', 'dummy_transformers_and_torch_and_note_seq_objects', 'dynamic_modules_utils', 'export_to_gif', 'export_to_obj', 'export_to_ply', 'export_to_video', 'export_utils', 'extract_commit_hash', 'get_class_from_dynamic_module', 'get_logger', 'get_objects_from_module', 'http_user_agent', 'hub_utils', 'import_utils', 'is_accelerate_available', 'is_accelerate_version', 'is_bs4_available', 'is_flax_available', 'is_ftfy_available', 'is_inflect_available', 'is_invisible_watermark_available', 'is_k_diffusion_available', 'is_k_diffusion_version', 'is_librosa_available', 'is_note_seq_available', 'is_omegaconf_available', 'is_onnx_available', 'is_peft_available', 'is_scipy_available', 'is_tensorboard_available', 'is_torch_available', 'is_torch_version', 'is_torchsde_available', 'is_transformers_available', 'is_transformers_version', 'is_unidecode_available', 'is_wandb_available', 'is_xformers_available', 'load_image', 'loading_utils', 'logger', 'logging', 'make_image_grid', 'numpy_to_pil', 'os', 'outputs', 'peft_utils', 'pil_utils', 'pt_to_pil', 'recurse_remove_peft_layers', 'replace_example_docstring', 'requires_backends', 'state_dict_utils', 'version']

[Sivll]

hello! Do you have any news? still doesn't want to import is_compiled_module from anywhere...((( python 3.11.3

ImportError: cannot import name 'is_compiled_module' from 'diffusers.utils' (/Users/.../sd/MetalDiffusion/venv/lib/python3.11/site-packages/diffusers/utils/init.py)

may be I can download that py file, but from where?)

[xiaohu2015]

@5shekel @TheLearner23 hi, you can clone the new repo of https://github.com/tencent-ailab/IP-Adapter

[5shekel]

here is how i installed dependencies.
using python virtual env
tested on windows

python -m venv .venv
/.venv/scripts/activate #WINDOWS
#/.venv/bin/activate #LINUX
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118 
pip install diffusers transformers
pip install ipykernel #for jupyter notebook
pip install accelerate #for 

looks good!

[Sivll]

Am I able to run it in macOS terminal? I guess yes.

$ python3 -m venv venv $ source venv/bin/activate

Without dollar sign

[5shekel]

You don't Have cuda if you don't have Nvidia card. Maybe this can work in CPU. Idk

[Sivll]

There's AMD gpu on board.. But how to overcome it and run using CPU?

[5shekel]

update with your mac model so people can assist. i cant, sorry there is attempt[1] to run on mac new models, but idk about your situation. probebly better to open new issue. this is different [1] https://github.com/tencent-ailab/IP-Adapter/issues/53

[5shekel]

@Sivll try adding this [untested] check to the model loading cell

device = "cuda" if torch.cuda.is_available() else "cpu"

instead of device = "cude"

[Aime-ry]

To fix this, you can open the file located at /stable-diffusion-webui/extensions/sd-webui-controlnet/annotator/clipvision/init.py using a text editor or IDE (such as VSCode or PyCharm).

1、Navigate to line 81 and locate the line: clip_vision_h_uc = torch.load(clip_vision_h_uc)['uc']. 2、Modify this line to: clip_vision_h_uc = torch.load(clip_vision_h_uc, map_location=torch.device('cpu'))['uc']. 3、Save your changes and exit the editor. 4、Run your program again. This should prevent any CUDA-related errors.

is to add the command-line flag "--no-half" to your launch settings.