19wolf
commented
1 year ago I'm also having this issue. When I try to fix broken install, I get these errors:
Re-run the configure script to fix a broken install or to complete a major upgrade
[2023-08-11 17:35:32,614]::[InvokeAI]::INFO --> Initializing InvokeAI runtime directory
[2023-08-11 17:35:32,623]::[InvokeAI]::INFO --> Scanning /Users/19wolf/invokeai/models for new models
Traceback (most recent call last):
File "/Users/19wolf/invokeai/.venv/bin/invokeai-configure", line 8, in <module>
sys.exit(invokeai_configure())
^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/install/invokeai_configure.py", line 778, in main
models_to_download = default_user_selections(opt)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/install/invokeai_configure.py", line 540, in default_user_selections
installer = ModelInstall(config)
^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/install/model_install_backend.py", line 106, in __init__
self.mgr = model_manager or ModelManager(config.model_conf_path)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/model_management/model_manager.py", line 350, in __init__
self._read_models(config)
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/model_management/model_manager.py", line 373, in _read_models
self.scan_models_directory()
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/model_management/model_manager.py", line 938, in scan_models_directory
model_config: ModelConfigBase = model_class.probe_config(str(model_path))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/model_management/models/sdxl.py", line 56, in probe_config
checkpoint = read_checkpoint_meta(path)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/invokeai/backend/model_management/models/base.py", line 429, in read_checkpoint_meta
checkpoint = torch.load(path, map_location=torch.device("meta"))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/torch/serialization.py", line 815, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/19wolf/invokeai/.venv/lib/python3.11/site-packages/torch/serialization.py", line 1033, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
_pickle.UnpicklingError: invalid load key, '\x00'.
Millu
5chlupp
Is there an existing issue for this?
OS
macOS
GPU
mps
VRAM
64
What version did you experience this issue on?
3.0.2rc1
What happened?
Hello, since 3.0.1rc1 there is the problem that InvokeAI does not start anymore. No matter if you install it fresh or update an existing installation. Unfortunately, this is still the case with 3.0.2rc1. No matter if it is a Windows or Mac installation.
the error message is:
Of course, I always try number seven to fix the problem.
If you download the missing folder (Hugging Face), others folders are missing from the folder.
here is a link to the bug report that started with 3.0.1rc1 and documents it more. https://github.com/invoke-ai/InvokeAI/issues/4051
Are there other ways I can help or what I can do to fix it?
Screenshots
No response
Additional context
No response
Contact Details
No response