InvokeAI icon indicating copy to clipboard operation
InvokeAI copied to clipboard
verified publisher icon invoke-ai

[bug]: Cannot install any models in v6.9.0 in Windows

Open gogurtenjoyer opened this issue 2 months ago • 2 comments

Is there an existing issue for this problem?

  • [x] I have searched the existing issues

Install method

Invoke's Launcher

Operating system

Windows

GPU vendor

Nvidia (CUDA)

GPU model

RTX 4070 Ti Super

GPU VRAM

16GB

Version number

v6.9.0

Browser

Arc (Chromium)

System Information

{ "version": "6.9.0", "dependencies": { "absl-py" : "2.3.1", "accelerate" : "1.10.1", "annotated-types" : "0.7.0", "anyio" : "4.11.0", "asttokens" : "3.0.0", "attrs" : "25.4.0", "bidict" : "0.23.1", "bitsandbytes" : "0.48.1", "blake3" : "1.0.8", "certifi" : "2022.12.7", "cffi" : "2.0.0", "charset-normalizer" : "2.1.1", "click" : "8.3.0", "colorama" : "0.4.6", "coloredlogs" : "15.0.1", "compel" : "2.1.1", "contourpy" : "1.3.3", "CUDA" : "12.8", "cycler" : "0.12.1", "decorator" : "5.2.1", "Deprecated" : "1.2.18", "diffusers" : "0.33.0", "diskcache" : "5.6.3", "dnspython" : "2.8.0", "dynamicprompts" : "0.31.0", "einops" : "0.8.1", "executing" : "2.2.0", "fastapi" : "0.118.3", "fastapi-events" : "0.12.2", "filelock" : "3.19.1", "flatbuffers" : "25.9.23", "fonttools" : "4.60.1", "fsspec" : "2025.9.0", "gguf" : "0.17.1", "h11" : "0.16.0", "httptools" : "0.7.1", "huggingface-hub" : "0.35.3", "humanfriendly" : "10.0", "idna" : "3.4", "importlib_metadata" : "7.1.0", "inquirerpy" : "0.3.4", "invisible-watermark" : "0.2.0", "InvokeAI" : "6.9.0", "ipython" : "9.1.0", "ipython_pygments_lexers": "1.1.1", "jax" : "0.7.1", "jaxlib" : "0.7.1", "jedi" : "0.19.2", "Jinja2" : "3.1.6", "kiwisolver" : "1.4.9", "linkify-it-py" : "2.0.3", "llama_cpp_python" : "0.3.9", "markdown-it-py" : "3.0.0", "MarkupSafe" : "2.1.5", "matplotlib" : "3.10.7", "matplotlib-inline" : "0.1.7", "mdit-py-plugins" : "0.4.2", "mdurl" : "0.1.2", "mediapipe" : "0.10.14", "ml_dtypes" : "0.5.3", "mpmath" : "1.3.0", "networkx" : "3.5", "numpy" : "1.26.3", "onnx" : "1.16.1", "onnxruntime" : "1.19.2", "opencv-contrib-python" : "4.11.0.86", "opencv-python" : "4.11.0.86", "opt_einsum" : "3.4.0", "packaging" : "24.1", "parso" : "0.8.4", "pfzy" : "0.3.4", "picklescan" : "0.0.31", "pillow" : "11.3.0", "platformdirs" : "4.3.8", "prompt_toolkit" : "3.0.52", "protobuf" : "4.25.8", "psutil" : "7.1.1", "pure_eval" : "0.2.3", "pycparser" : "2.23", "pydantic" : "2.12.3", "pydantic-settings" : "2.11.0", "pydantic_core" : "2.41.4", "Pygments" : "2.19.1", "pyparsing" : "3.2.5", "PyPatchMatch" : "1.0.2", "pyreadline3" : "3.5.4", "python-dateutil" : "2.9.0.post0", "python-dotenv" : "1.1.1", "python-engineio" : "4.12.3", "python-multipart" : "0.0.20", "python-socketio" : "5.14.2", "PyWavelets" : "1.9.0", "PyYAML" : "6.0.3", "regex" : "2025.9.18", "requests" : "2.28.1", "rich" : "14.0.0", "safetensors" : "0.6.2", "scipy" : "1.16.2", "semver" : "3.0.4", "sentencepiece" : "0.2.0", "setuptools" : "70.2.0", "simple-websocket" : "1.1.0", "six" : "1.17.0", "sniffio" : "1.3.1", "sounddevice" : "0.5.3", "spandrel" : "0.4.1", "stack-data" : "0.6.3", "starlette" : "0.48.0", "sympy" : "1.14.0", "textual" : "3.3.0", "tokenizers" : "0.22.1", "torch" : "2.7.1+cu128", "torchsde" : "0.2.6", "torchvision" : "0.22.1+cu128", "tqdm" : "4.66.5", "traitlets" : "5.14.3", "trampoline" : "0.1.2", "transformers" : "4.57.1", "typing-inspection" : "0.4.2", "typing_extensions" : "4.15.0", "uc-micro-py" : "1.0.3", "urllib3" : "1.26.13", "uvicorn" : "0.38.0", "watchfiles" : "1.1.1", "wcwidth" : "0.2.14", "websockets" : "15.0.1", "wrapt" : "1.17.3", "wsproto" : "1.2.0", "zipp" : "3.19.2" }, "config": { "schema_version": "4.0.2", "legacy_models_yaml_path": null, "host": "0.0.0.0", "port": 9090, "allow_origins": [], "allow_credentials": true, "allow_methods": [""], "allow_headers": [""], "ssl_certfile": null, "ssl_keyfile": null, "log_tokenization": true, "patchmatch": true, "models_dir": "models", "convert_cache_dir": "models\.convert_cache", "download_cache_dir": "models\.download_cache", "legacy_conf_dir": "configs", "db_dir": "databases", "outputs_dir": "outputs", "custom_nodes_dir": "nodes", "style_presets_dir": "style_presets", "workflow_thumbnails_dir": "workflow_thumbnails", "log_handlers": ["console"], "log_format": "color", "log_level": "info", "log_sql": false, "log_level_network": "warning", "use_memory_db": false, "dev_reload": false, "profile_graphs": false, "profile_prefix": null, "profiles_dir": "profiles", "max_cache_ram_gb": null, "max_cache_vram_gb": null, "log_memory_usage": true, "device_working_mem_gb": 4, "enable_partial_loading": true, "keep_ram_copy_of_weights": true, "ram": null, "vram": null, "lazy_offload": true, "pytorch_cuda_alloc_conf": "backend:cudaMallocAsync", "device": "auto", "precision": "auto", "sequential_guidance": false, "attention_type": "auto", "attention_slice_size": "auto", "force_tiled_decode": false, "pil_compress_level": 6, "max_queue_size": 10000, "clear_queue_on_startup": false, "allow_nodes": null, "deny_nodes": null, "node_cache_size": 512, "hashing_algorithm": "random", "remote_api_tokens": [ {"url_regex": "civitai", "token": "REDACTED"}, {"url_regex": "huggingface", "token": "REDACTED"} ], "scan_models_on_startup": false, "unsafe_disable_picklescan": false, "allow_unknown_models": true }, "set_config_fields": [ "pytorch_cuda_alloc_conf", "pil_compress_level", "legacy_models_yaml_path", "hashing_algorithm", "enable_partial_loading", "remote_api_tokens", "log_tokenization", "host", "log_memory_usage", "device_working_mem_gb" ] }

What happened

Currently, it seems that models cannot be installed, with single file and diffusers style models failing in different ways:

  1. Single file (Flux GGUF model, in this case) - after hashing, the model install fails while attempting to unlink the temp file:
[2025-10-25 16:27:05,197]::[InvokeAI]::INFO --> Invoke running on http://0.0.0.0:9090 (Press CTRL+C to quit)
C:\Users\gogurt_disliker\Documents\invokeai\.venv\Lib\site-packages\huggingface_hub\utils\_deprecation.py:131: FutureWarning: 'get_toke
n_permission' (from 'huggingface_hub.hf_api') is deprecated and will be removed from version '1.0'. Permissions are more com
plex than when `get_token_permission` was first introduced. OAuth and fine-grain tokens allows for more detailed permissions
. If you need to know the permissions associated with a token, please use `whoami` and check the `'auth'` key.
  warnings.warn(warning_message, FutureWarning)
[2025-10-25 16:27:21,642]::[ModelInstallService]::INFO --> Queueing model install: https://civitai.com/api/download/models/1
719520?type=Model&format=GGUF&size=pruned&fp=fp8 (1 file)
[2025-10-25 16:27:21,642]::[InvokeAI]::INFO --> Started installation of https://civitai.com/api/download/models/1719520?type
=Model&format=GGUF&size=pruned&fp=fp8
[2025-10-25 16:27:22,942]::[DownloadQueueService]::INFO --> File download started: https://civitai.com/api/download/models/1
719520?type=Model&format=GGUF&size=pruned&fp=fp8
[2025-10-25 16:30:39,152]::[DownloadQueueService]::INFO --> Download complete: https://civitai.com/api/download/models/17195
20?type=Model&format=GGUF&size=pruned&fp=fp8
[2025-10-25 16:30:39,152]::[ModelInstallService]::INFO --> Model download complete: https://civitai.com/api/download/models/
1719520?type=Model&format=GGUF&size=pruned&fp=fp8
[2025-10-25 16:30:39,154]::[ModelInstallService]::INFO --> Model install started: https://civitai.com/api/download/models/17
19520?type=Model&format=GGUF&size=pruned&fp=fp8
Hashing pandoraFLUX_pandoraProAlphaQ4KM.gguf: 100%|████████████████████████████████████████████████| 1/1 [00:00<?, ?file/s]
[2025-10-25 16:30:43,748]::[ModelInstallService]::ERROR --> Model install error: https://civitai.com/api/download/models/171
9520?type=Model&format=GGUF&size=pruned&fp=fp8
PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'C:\\Users\\gogurt_disliker\\Documents\\invokeai\\models\\tmpinstall_cegf5k_1\\pandoraFLUX_pandoraProAlphaQ4KM.gguf'
Exception in thread Thread-3 (_install_next_item):
Traceback (most recent call last):
  File "C:\Users\gogurt_disliker\AppData\Roaming\uv\python\cpython-3.12.9-windows-x86_64-none\Lib\threading.py", line 1075, in _bootstr
ap_inner
    self.run()
  File "C:\Users\gogurt_disliker\AppData\Roaming\uv\python\cpython-3.12.9-windows-x86_64-none\Lib\threading.py", line 1012, in run
    self._target(*self._args, **self._kwargs)
  File "C:\Users\gogurt_disliker\Documents\invokeai\.venv\Lib\site-packages\invokeai\app\services\model_install\model_install_default.p
y", line 524, in _install_next_item
    rmtree(job._install_tmpdir)
  File "C:\Users\gogurt_disliker\AppData\Roaming\uv\python\cpython-3.12.9-windows-x86_64-none\Lib\shutil.py", line 781, in rmtree
    return _rmtree_unsafe(path, onexc)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\gogurt_disliker\AppData\Roaming\uv\python\cpython-3.12.9-windows-x86_64-none\Lib\shutil.py", line 635, in _rmtree_unsa
fe
    onexc(os.unlink, fullname, err)
  File "C:\Users\gogurt_disliker\AppData\Roaming\uv\python\cpython-3.12.9-windows-x86_64-none\Lib\shutil.py", line 633, in _rmtree_unsa
fe
    os.unlink(fullname)
PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'C:\\Users\\gogurt_disliker\\Documents\\invokeai\\models\\tmpinstall_cegf5k_1\\pandoraFLUX_pandoraProAlphaQ4KM.gguf'
C:\Users\gogurt_disliker\Documents\invokeai\.venv\Lib\site-packages\huggingface_hub\utils\_http.py:200: ResourceWarning: unclosed <ssl.
SSLSocket fd=3736, family=2, type=1, proto=0, laddr=('192.168.235.126', 62002), raddr=('18.244.202.118', 443)>
  _get_session_from_cache.cache_clear()
ResourceWarning: Enable tracemalloc to get the object allocation traceback

The 'file in use' error can also appear in the OS rather than Invoke when manually attempting to delete this temporary file.

  1. With multi-file diffusers style models, instead of an error, the install just hangs and the UI specifies it's still 'In Progress' - in this case, there's no errors in console, with the final line being [2025-10-25 16:32:22,430]::[ModelInstallService]::INFO --> Model download complete: Lykon/dreamshaper-8

This affects URL or HF-based downloads as well as models in Starter Models.

What you expected to happen

Models should be installable as in prior versions of Invoke AI.

How to reproduce the problem

In Invoke's model manager, install any model you'd like via any of the available download methods: URL (civitai, etc), HuggingFace repo, or any model from Starter Models that uses any of these methods.

Additional context

I'm guessing that this is the result of recent changes with regards to the model folder reorganization, or identification?

Discord username

gogurtenjoyer

gogurtenjoyer avatar Oct 25 '25 20:10 gogurtenjoyer

Just realized this is a duplicate of #8627 - sorry!

gogurtenjoyer avatar Oct 25 '25 20:10 gogurtenjoyer

Okay, this seems to be an issue with large files, so in particular, Flux models would be affected (but not Flux loras, typically). It's happening in Windows only because of how file locks are treated versus Mac and Linux. I'm working on a fix with an additional failsafe and will post more information in the PR.

gogurtenjoyer avatar Oct 30 '25 19:10 gogurtenjoyer

I confirm. I cannot install Flux Krea, PermissionError: [WinError 32] The process cannot access the file because it is being used by another process.

jvbrt4 avatar Nov 08 '25 10:11 jvbrt4

Install in place also exhibits the file locking issue

ihodgy avatar Nov 13 '25 23:11 ihodgy