MrForExample
Pytorch3D module not found (it does not install because it depends on old libs)
I can't install this node and when I launch ComfyUI it says
\custom_nodes\ComfyUI-3D-Pack\Gen_3D_Modules\TriplaneGaussian\utils\ops.py", line 8, in
And when I try to install pytorch3d Python tries to also install a requirement for pytorch3d called "gendr"
Also the problem = gendr is very old and also does not build (Using portable python 3.12.9)
When I try to install gendr it gets this
python_embeded\Lib\site-packages\torch\utils\cpp_extension.py", line 2522, in _run_ninja_build raise RuntimeError(message) from e RuntimeError: Error compiling objects for extension [end of output]
note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for gendr Running setup.py clean for gendr Failed to build gendr ERROR: Failed to build installable wheels for some pyproject.toml based projects (gendr)
i check their github page and there is no update and it works with torch 1.9.0 and CUDA 11.1 while almost all moderns ComfyUI installations and newer nodes use newer torch and CUDA
Don't know if its related to a new python, torch and CUDA
Any tips ?
Tks
what is your setup? linux installation guide for new cuda and torch is currenlty here: https://github.com/MrForExample/ComfyUI-3D-Pack/issues/464#issuecomment-2855473286 will be updated soon
One machine = Windows 11 with a Nvidia 3080TI Other machine = Windows server 2025 with Nvidia 4090 (I use this one mostly for video workloads since the 4090 is faster for that Both machines running ComfyUI portable with same versions, same Python version and almost all nodes, models and Loras so if I run a big workflow I can send it to my server build Is this only working for Linux ? Because in both machines 3D-Pack fails to import
I am only running Windows now here in my machines since I run a lot of stuff via RDP and I always had issues using remote clients to Linux machines
Tks
You can install pytorch3d from wheel
Download it from https://github.com/DenisKochetov/ComfyUI-3D-Pack-wheels And then pip install ./pytorch3d-0.7.8-cp312-cp312-win_amd64.whl
You can install pytorch3d from wheel
Download it from https://github.com/DenisKochetov/ComfyUI-3D-Pack-wheels And then pip install ./pytorch3d-0.7.8-cp312-cp312-win_amd64.whl
Hi tks ! But I still have issues installing utils3d from the requirements.txt It gets error message regarding pkgutil and the pack does not lot
Could you tell me what ComfyUI, Python and Torch versions you are using ? Or print here your pip install list
I saw some comments about recent Python versions having a lot of compatibility issues with older wheels, packs and libs
Tks
@CrisOrlandoBR We've prepared wheels for Windows with CUDA 12.8 and PyTorch 2.7.0. Please clone the latest version from the main branch and run python install.py. It should work out of the box — if you run into any errors, feel free to share them here.
@CrisOrlandoBR We've prepared wheels for Windows with CUDA 12.8 and PyTorch 2.7.0. Please clone the latest version from the
mainbranch and runpython install.py. It should work out of the box — if you run into any errors, feel free to share them here.
My big issue here is with utils3d It does not install at all (even trying to install it from command or from requirements.txt list
And the automatic install.py is giving an weird error here = I have torch 2.7.0 cu128 but it tries to use 12.4 wheels
Couldn't download folder _Build_Wheels/_Wheels_win_py312_torch2.7.0_cu124 from repo MrForExample/Comfy3D_Pre_Builds 404 {"message": "Not Found", "documentation_url": "https://docs.github.com/rest/repos/contents#get-repository-content", "status": "404"} ←[34m[Comfy3D] ←[0mTry building all required packages...←[0m
Any idea about what's causing this ?
Tks
D:\AMD_AI\ComfyUI_windows_portable\utils3d>..\python_embeded\python.exe -m pip install -e . Obtaining file:///D:/AMD_AI/ComfyUI_windows_portable/utils3d Installing build dependencies ... done Checking if build backend supports build_editable ... error error: subprocess-exited-with-error
× Checking if build backend supports build_editable did not run successfully.
│ exit code: 1
╰─> [31 lines of output]
Traceback (most recent call last):
File "D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\pip_vendor\pyproject_hooks_in_process_in_process.py", line 389, in
note: This error originates from a subprocess, and is likely not a problem with pip. error: subprocess-exited-with-error
× Checking if build backend supports build_editable did not run successfully. │ exit code: 1 ╰─> See above for output.
note: This error originates from a subprocess, and is likely not a problem with pip.
ps: I "fixed" by putting cuda 128 wheels inside 124 folder in the pre build wheels folder from the install It installed but even utils3d being installed it still shows a weird error in the console xecute management script for 'D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack'
ComfyUI-Manager: EXECUTE => ['D:\AMD_AI\ComfyUI_windows_portable\python_embeded\python.exe', '-s', '-m', 'pip', 'install', 'git+https://github.com/EasternJournalist/utils3d.git']
Execute management script for 'D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack'
Collecting git+https://github.com/EasternJournalist/utils3d.git
Cloning https://github.com/EasternJournalist/utils3d.git to c:\users\criso\appdata\local\temp\pip-req-build-_azvtnun
[!] Running command git clone --filter=blob:none --quiet https://github.com/EasternJournalist/utils3d.git 'C:\Users\criso\AppData\Local\Temp\pip-req-build-_azvtnun'
Resolved https://github.com/EasternJournalist/utils3d.git to commit c5daf6f6c244d251f252102d09e9b7bcef791a38
Installing build dependencies: started
Installing build dependencies: finished with status 'done'
Getting requirements to build wheel: started
Getting requirements to build wheel: finished with status 'error'
[!] error: subprocess-exited-with-error
[!]
[!] Getting requirements to build wheel did not run successfully.
[!] exit code: 1
[!]
[!] [31 lines of output]
[!] Traceback (most recent call last):
[!] File "D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\pip_vendor\pyproject_hooks_in_process_in_process.py", line 389, in
It looks like you're using an outdated version of the 3D-Pack — that 12.4 error used to happen in the latest version back then.
Please pull the latest version from main and try running:
python install.py
Then send the logs here — I believe we can get it installed properly.
And about utils3d
Still has compatibility issues D:\AMD_AI\ComfyUI_windows_portable>python_embeded\python.exe -m pip install utils3d --force-reinstall Collecting utils3d Using cached utils3d-0.1.3-py3-none-any.whl.metadata (2.3 kB) Collecting addict (from utils3d) Using cached addict-2.4.0-py3-none-any.whl.metadata (1.0 kB) Collecting matplotlib>=3 (from utils3d) Using cached matplotlib-3.10.3-cp312-cp312-win_amd64.whl.metadata (11 kB) Collecting numpy>1.18 (from utils3d) Using cached numpy-2.3.0-cp312-cp312-win_amd64.whl.metadata (60 kB) INFO: pip is looking at multiple versions of utils3d to determine which version is compatible with other requirements. This could take a while. Collecting utils3d Using cached utils3d-0.1.2-py3-none-any.whl.metadata (2.4 kB) Collecting jupyter<2.0.0,>=1.0.0 (from utils3d) Using cached jupyter-1.1.1-py2.py3-none-any.whl.metadata (2.0 kB) Collecting utils3d Using cached utils3d-0.1.1-py3-none-any.whl.metadata (2.4 kB) Using cached utils3d-0.1.0-py3-none-any.whl.metadata (2.4 kB) ERROR: Cannot install utils3d==0.1.0, utils3d==0.1.1, utils3d==0.1.2 and utils3d==0.1.3 because these package versions have conflicting dependencies.
The conflict is caused by: utils3d 0.1.3 depends on open3d<0.14.0 and >=0.13.0 utils3d 0.1.2 depends on open3d<0.14.0 and >=0.13.0 utils3d 0.1.1 depends on open3d<0.18.0 and >=0.17.0 utils3d 0.1.0 depends on open3d<0.19.0 and >=0.18.0
To fix this you could try to:
- loosen the range of package versions you've specified
- remove package versions to allow pip to attempt to solve the dependency conflict
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts
And if try to reinstall open3d, I get this
RROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. broken-source 0.8.0.dev0 requires attrs~=24.2.0, but you have attrs 25.3.0 which is incompatible. broken-source 0.8.0.dev0 requires numpy==1.26.4, but you have numpy 2.3.0 which is incompatible. broken-source 0.8.0.dev0 requires pydantic~=2.9.2, but you have pydantic 2.11.5 which is incompatible. broken-source 0.8.0.dev0 requires setuptools~=75.2.0, but you have setuptools 80.9.0 which is incompatible. broken-source 0.8.0.dev0 requires tqdm~=4.66.5, but you have tqdm 4.67.1 which is incompatible. browser-use 0.2.6 requires aiofiles>=24.1.0, but you have aiofiles 23.2.1 which is incompatible. browser-use 0.2.6 requires langchain-core==0.3.64, but you have langchain-core 0.3.65 which is incompatible. chromadb 1.0.12 requires fastapi==0.115.9, but you have fastapi 0.115.12 which is incompatible. colpali-engine 0.3.10 requires torch<2.7.0,>=2.5.0, but you have torch 2.7.0+cu128 which is incompatible. comfy-cli 1.4.1 requires click<=8.1.8, but you have click 8.2.1 which is incompatible. cupy-cuda12x 13.4.1 requires numpy<2.3,>=1.22, but you have numpy 2.3.0 which is incompatible. dctorch 0.1.2 requires numpy<2.0.0,>=1.22.3, but you have numpy 2.3.0 which is incompatible. descript-audiotools 0.7.2 requires protobuf<3.20,>=3.9.2, but you have protobuf 5.29.5 which is incompatible. dghs-imgutils 0.17.0 requires numpy<2, but you have numpy 2.3.0 which is incompatible. dghs-imgutils 0.17.0 requires shapely!=2.0.7, but you have shapely 2.0.7 which is incompatible. gensim 4.3.3 requires numpy<2.0,>=1.18.5, but you have numpy 2.3.0 which is incompatible. gensim 4.3.3 requires scipy<1.14.0,>=1.7.0, but you have scipy 1.15.3 which is incompatible. gpytoolbox 0.3.5 requires numpy<2.3,>=1.22.4, but you have numpy 2.3.0 which is incompatible. gpytoolbox 0.3.5 requires scipy<=1.15,>=1.7, but you have scipy 1.15.3 which is incompatible. gradio 3.48.0 requires markupsafe~=2.0, but you have markupsafe 3.0.2 which is incompatible. gradio 3.48.0 requires numpy~=1.0, but you have numpy 2.3.0 which is incompatible. gradio 3.48.0 requires websockets<12.0,>=10.0, but you have websockets 15.0.1 which is incompatible. gradio-client 0.6.1 requires websockets<12.0,>=10.0, but you have websockets 15.0.1 which is incompatible. graspologic 3.4.1 requires numpy<2.0.0,>=1.26.4, but you have numpy 2.3.0 which is incompatible. graspologic 3.4.1 requires scipy==1.12.0, but you have scipy 1.15.3 which is incompatible. grpcio-tools 1.73.0 requires protobuf<7.0.0,>=6.30.0, but you have protobuf 5.29.5 which is incompatible. gruut 2.2.3 requires networkx<3.0.0,>=2.5.0, but you have networkx 3.5 which is incompatible. gruut 2.2.3 requires numpy<2.0.0,>=1.19.0, but you have numpy 2.3.0 which is incompatible. hbutils 0.11.0 requires chardet<5,>=3.0.4, but you have chardet 5.1.0 which is incompatible. inference-cli 0.50.4 requires aiohttp<=3.10.11,>=3.9.0, but you have aiohttp 3.12.12 which is incompatible. inference-cli 0.50.4 requires click<8.2.0, but you have click 8.2.1 which is incompatible. inference-cli 0.50.4 requires numpy<2.3.0,>=2.0.0, but you have numpy 2.3.0 which is incompatible. inference-cli 0.50.4 requires pillow<12.0,>=11.0, but you have pillow 10.4.0 which is incompatible. inference-cli 0.50.4 requires supervision<=0.30.0,>=0.25.1, but you have supervision 0.6.0 which is incompatible. inference-gpu 0.50.4 requires aiohttp<=3.10.11,>=3.9.0, but you have aiohttp 3.12.12 which is incompatible. inference-gpu 0.50.4 requires click<8.2.0, but you have click 8.2.1 which is incompatible. inference-gpu 0.50.4 requires fastapi<0.111,>=0.100, but you have fastapi 0.115.12 which is incompatible. inference-gpu 0.50.4 requires numpy<2.3.0,>=2.0.0, but you have numpy 2.3.0 which is incompatible. inference-gpu 0.50.4 requires onnxruntime-gpu<1.22.0,>=1.15.1, but you have onnxruntime-gpu 1.22.0 which is incompatible. inference-gpu 0.50.4 requires packaging~=24.0, but you have packaging 25.0 which is incompatible. inference-gpu 0.50.4 requires pillow<12.0,>=11.0, but you have pillow 10.4.0 which is incompatible. inference-gpu 0.50.4 requires pydantic-settings<2.8, but you have pydantic-settings 2.9.1 which is incompatible. inference-gpu 0.50.4 requires supervision<=0.30.0,>=0.25.1, but you have supervision 0.6.0 which is incompatible. inference-gpu 0.50.4 requires typing_extensions<=4.12.2,>=4.8.0, but you have typing-extensions 4.14.0 which is incompatible. langchain-anthropic 0.3.15 requires anthropic<1,>=0.52.0, but you have anthropic 0.49.0 which is incompatible. langchain-aws 0.2.25 requires boto3>=1.37.24, but you have boto3 1.35.60 which is incompatible. langchain-core 0.3.65 requires packaging<25,>=23.2, but you have packaging 25.0 which is incompatible. langchain-google-genai 2.1.5 requires google-ai-generativelanguage<0.7.0,>=0.6.18, but you have google-ai-generativelanguage 0.6.15 which is incompatible. langchain-ollama 0.3.3 requires ollama<1.0.0,>=0.4.8, but you have ollama 0.3.3 which is incompatible. lightning 2.5.1.post0 requires packaging<25.0,>=20.0, but you have packaging 25.0 which is incompatible. matrix-client 0.4.0 requires urllib3~=1.21, but you have urllib3 2.4.0 which is incompatible. mediapipe 0.10.21 requires numpy<2, but you have numpy 2.3.0 which is incompatible. mediapipe 0.10.21 requires protobuf<5,>=4.25.3, but you have protobuf 5.29.5 which is incompatible. moderngl-window 2.4.6 requires numpy<2,>=1.16, but you have numpy 2.3.0 which is incompatible. numba 0.61.2 requires numpy<2.3,>=1.24, but you have numpy 2.3.0 which is incompatible. ollama 0.3.3 requires httpx<0.28.0,>=0.27.0, but you have httpx 0.28.1 which is incompatible. open-clip-torch 2.22.0 requires protobuf<4, but you have protobuf 5.29.5 which is incompatible. qdrant-client 1.14.2 requires portalocker<3.0.0,>=2.7.0, but you have portalocker 3.1.1 which is incompatible. selenium 4.33.0 requires typing_extensions~=4.13.2, but you have typing-extensions 4.14.0 which is incompatible. sensevoice-onnx 1.1.0 requires setuptools<=65.0, but you have setuptools 80.9.0 which is incompatible. streamlit 1.45.1 requires packaging<25,>=20, but you have packaging 25.0 which is incompatible. tensorflow 2.19.0 requires numpy<2.2.0,>=1.26.0, but you have numpy 2.3.0 which is incompatible. tensorflow-cpu 2.19.0 requires numpy<2.2.0,>=1.26.0, but you have numpy 2.3.0 which is incompatible. torchscale 0.3.0 requires fairscale==0.4.0, but you have fairscale 0.4.13 which is incompatible. torchtyping 0.1.5 requires typeguard<3,>=2.11.1, but you have typeguard 4.4.3 which is incompatible. transparent-background 1.3.4 requires timm>=1.0.3, but you have timm 0.6.13 which is incompatible.
And even after trying to reinstall open3d (to make utils3d work) 3D-Pack stills fails to import Tks
@CrisOrlandoBR yeah, you should definitely pull the latest version. Then run:
pip install -r requirements.txt
followed by:
python install.py
Make sure your NumPy version matches the expected one — it's important.
I did all this, updated all and still no dice
What your pip list so I can compare to mine to check if there is something missing or wrong version ?
Tks
@CrisOrlandoBR
(comfy_test) PS C:\Users\user> pip list
Package Version
--------------------------- -----------------------
accelerate 1.7.0
aiohappyeyeballs 2.6.1
aiohttp 3.12.9
aiosignal 1.3.2
annotated-types 0.7.0
antlr4-python3-runtime 4.9.3
attrs 25.3.0
av 14.4.0
ccimport 0.4.4
certifi 2025.4.26
cffi 1.17.1
chardet 5.2.0
charset-normalizer 3.4.2
click 8.2.1
cmake 4.0.2
colorama 0.4.6
coloredlogs 15.0.1
comfyui_frontend_package 1.19.9
comfyui_workflow_templates 0.1.14
contourpy 1.3.2
cryptography 45.0.3
cumm 0.7.11
custom_rasterizer 0.1
cycler 0.12.1
dataclasses-json 0.6.7
Deprecated 1.2.18
diff_gaussian_rasterization 0.0.0
diffusers 0.33.1
diso 0.1.4
easydict 1.13
einops 0.8.1
executing 2.2.0
fast_simplification 0.1.11
filelock 3.16.1
fire 0.7.0
flatbuffers 25.2.10
fonttools 4.58.1
frozenlist 1.6.2
fsspec 2024.10.0
ftfy 6.3.1
gitdb 4.0.12
GitPython 3.1.44
glcontext 3.0.0
gpytoolbox 0.3.5
hatchling 1.27.0
huggingface-hub 0.32.4
humanfriendly 10.0
idna 3.10
igraph 0.11.8
imageio 2.37.0
imageio-ffmpeg 0.6.0
importlib_metadata 8.7.0
iopath 0.1.10
jaxtyping 0.3.2
Jinja2 3.1.4
jsonschema 4.24.0
jsonschema-specifications 2025.4.1
kiui 0.2.16
kiwisolver 1.4.8
kornia 0.8.1
kornia_rs 0.1.9
lark 1.2.2
lazy_loader 0.4
libigl 2.6.1
lightning-utilities 0.14.3
llvmlite 0.44.0
markdown-it-py 3.0.0
MarkupSafe 2.1.5
marshmallow 3.26.1
matplotlib 3.10.3
matrix-client 0.4.0
mdurl 0.1.2
mmgp 3.4.7
moderngl 5.12.0
mpmath 1.3.0
multidict 6.4.4
mypy_extensions 1.1.0
nerfacc 0.5.3
networkx 3.4.2
ninja 1.11.1.4
numba 0.61.2
numpy 1.26.4
nvdiffrast 0.3.3
objprint 0.3.0
omegaconf 2.3.0
onnxruntime 1.22.0
open_clip_torch 2.32.0
opencv-python 4.11.0.86
opencv-python-headless 4.11.0.86
optimum-quanto 0.2.7
packaging 25.0
pathspec 0.12.1
pccm 0.4.16
peft 0.15.2
pillow 11.0.0
pip 25.1
platformdirs 4.3.8
pluggy 1.6.0
plyfile 1.1.2
pointnet2_ops 3.0.0
pooch 1.8.2
portalocker 3.1.1
propcache 0.3.1
protobuf 6.31.1
psutil 7.0.0
pybind11 2.13.6
pycparser 2.22
pydantic 2.11.5
pydantic_core 2.33.2
PyGithub 2.6.1
pygltflib 1.16.4
Pygments 2.19.1
pyhocon 0.3.61
PyJWT 2.10.1
PyMatting 1.1.14
PyMCubes 0.1.6
pymeshfix 0.17.1
pymeshlab 2023.12.post3
PyNaCl 1.5.0
pyparsing 3.2.3
pyreadline3 3.5.4
python-dateutil 2.9.0.post0
pytorch-lightning 2.5.1.post0
pytorch-msssim 1.0.0
pytorch3d 0.7.8
pyvista 0.45.2
pywin32 310
PyYAML 6.0.2
referencing 0.36.2
regex 2024.11.6
rembg 2.0.66
requests 2.32.3
rich 14.0.0
rpds-py 0.25.1
safetensors 0.5.3
scikit-image 0.25.2
scipy 1.15.0
scooby 0.10.1
scs 3.2.7.post2
sentencepiece 0.2.0
setuptools 78.1.1
shellingham 1.5.4
simple_knn 0.0.0
six 1.17.0
slangtorch 1.3.8
smmap 5.0.2
soundfile 0.13.1
spandrel 0.4.1
spconv 2.3.8
sympy 1.13.3
termcolor 3.1.0
texttable 1.7.0
tifffile 2025.6.1
timm 1.0.15
tokenizers 0.21.1
toml 0.10.2
torch 2.7.0+cu128
torch_scatter 2.1.2
torchaudio 2.8.0.dev20250604+cu128
torchmetrics 1.7.2
torchsde 0.2.6
torchtyping 0.1.5
torchvision 0.22.0+cu128
tqdm 4.67.1
trampoline 0.1.2
transformers 4.52.4
trimesh 4.6.11
triton-windows 3.3.1.post19
trove-classifiers 2025.5.9.12
typeguard 2.13.3
typer 0.16.0
typing_extensions 4.12.2
typing-inspect 0.9.0
typing-inspection 0.4.1
urllib3 1.26.20
utils3d 0.0.2
uv 0.7.11
varname 0.14.0
vox2seq 0.0.0
vtk 9.4.2
wadler_lindig 0.1.6
wcwidth 0.2.13
wheel 0.45.1
wrapt 1.17.2
xatlas 0.0.10
xformers 0.0.30
yarl 1.20.0
zipp 3.22.0
Tks, gonna compare your pip list to mine to see what is going wrong here
And I rechecked everything The issue looks to be with spconv Any tips to fix ? I also cloned spconv repo to build from source and stills names it self as spconv-cu126 and fails to import
spconv not working, will install from mapping Detected CUDA version: 12.8 Installing spconv-cu126... spconv installed but verification failed: DLL load failed while importing core_cc: The specified module could not be found. Detected CUDA version: 12.8 Installing spconv-cu126... spconv installed but verification failed: DLL load failed while importing core_cc: The specified module could not be found. [Comfy3D] [ERROR] Comfy3D install failed: Dependency installation has failed. Please install manually: https://github.com/MrForExample/ComfyUI-3D-Pack/tree/main/_Pre_Builds/README.md.
[ComfyUI-Manager] Startup script completed.
Hi
Even with only spconv-cu126 installed I still get this
Tks
[SPARSE] Backend: spconv, Attention: xformers
Warn!: Traceback (most recent call last):
File "D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\nodes.py", line 2124, in load_custom_node
module_spec.loader.exec_module(module)
File "
Warn!: Cannot import D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack module for custom nodes: DLL load failed while importing _C: The specified procedure could not be found.
Loading: ComfyUI-Manager (V3.32.8)
[ComfyUI-Manager] network_mode: public
ComfyUI Revision: 183 [7ea79ebb] *DETACHED | Released on '2025-06-15'
Import times for custom nodes: 0.0 seconds: D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\websocket_image_save.py 0.7 seconds: D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui-manager 6.7 seconds (IMPORT FAILED): D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack
Do you have vs build tools installed? Guide: https://github.com/bycloudai/InstallVSBuildToolsWindows
also try pip install msvc-runtime to check that they work.
if you still get error then please delete 3d-pack folder and try clean installation of the last version.
I did everything here and still fails with spconv (spconv-cu126 installed fine, but there is still a DLL issue) BTW: I had to change de requirements.txt bacause this
ERROR: Could not find a version that satisfies the requirement kiuikit (unavailable) (from versions: none) ERROR: No matching distribution found for kiuikit (unavailable)
It only resolved as kiui without kiui egg and I installed it from the repo
And spconv still does not install proper (even I being told its proper installed)
Tks
[Prompt Server] web root: D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\comfyui_frontend_package\static
D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\kiui\nn_init_.py:31: FutureWarning: torch.cuda.amp.custom_fwd(args...) is deprecated. Please use torch.amp.custom_fwd(args..., device_type='cuda') instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\kiui\nn_init_.py:37: FutureWarning: torch.cuda.amp.custom_bwd(args...) is deprecated. Please use torch.amp.custom_bwd(args..., device_type='cuda') instead.
@torch.cuda.amp.custom_bwd
Warn!: xFormers is available (Attention) Warn!: D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\slangtorch\slangtorch.py:3: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81. import pkg_resources
Warn!: D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\setuptools_distutils_msvccompiler.py:12: UserWarning: _get_vc_env is private; find an alternative (pypa/distutils#340) warnings.warn(
Warn!: D:\AMD_AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\utils\cpp_extension.py:2356: UserWarning: TORCH_CUDA_ARCH_LIST is not set, all archs for visible cards are included for compilation. If this is not desired, please set os.environ['TORCH_CUDA_ARCH_LIST']. warnings.warn(
[SPARSE] Backend: spconv, Attention: xformers
Warn!: Traceback (most recent call last):
File "D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\nodes.py", line 2124, in load_custom_node
module_spec.loader.exec_module(module)
File "
Warn!: Cannot import D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack module for custom nodes: DLL load failed while importing _C: The specified procedure could not be found.
Loading: ComfyUI-Manager (V3.32.8)
[ComfyUI-Manager] network_mode: public
ComfyUI Revision: 183 [7ea79ebb] *DETACHED | Released on '2025-06-15'
Import times for custom nodes: 0.0 seconds: D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\websocket_image_save.py 0.7 seconds: D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui-manager 22.0 seconds (IMPORT FAILED): D:\AMD_AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-3D-Pack
The log tells me to go to the README from Pre Builds, but the README does not tell a thing about spconv error
Tks
@CrisOrlandoBR the new error seems to be not with spconv but with diso try manual installation
pip uninstall diso
git clone https://github.com/ashawkey/diso.git
cd diso
pip install .
Hi, I can't find this diso . git URL Only found https://github.com/SarahWeiii/diso (its the main repo) Cloning and installing it fixed the node :)
I think the build scripts are searching for .git files that are not alive anymore, since both diso and kiui were having problems like this I had to the main repo pages to get updated repos to clone and install
Gonna install some extra nodes here to test some example files to make sure they will run fine
Tks !
@CrisOrlandoBR happy to help! Thanks a lot for your patience, hopefully your experience helps us make the install smoother for everyone.
Tks, working now but I got a error message from spconv running a example file Failed to build Trellis pipeline: No module named 'spconv.pytorch'
Checked spconv page and I can't find this issue
Tks
@CrisOrlandoBR I has issues installting spconv from their repo so that's why I recommended to uninstall it and install with
pip install spconv-cuda126
also I have uploaded wheel for windows https://github.com/MrForExample/Comfy3D_Pre_Builds/blob/main/_Build_Wheels/_Wheels_win_py312_torch2.7.0_cu128/spconv-2.3.8-cp312-cp312-win_amd64.whl
but it seems like there wheels are not suitable for your system
Actually I installed from wheel Gonna use --force-reinstall to make sure it did not miss stuff
Tks
Ok, the error is gone, another error message also from Trellis, but I have no idea what I would have to reinstall here
And all my CUDA are updated as the same from my GPU drivers
Tks [Comfy3D] StableGen Trellis Image To 3D
CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
My GPU is a 3080TI and I was able to run other 3d models from Comfy
Have no idea why its happening I installed Hunyuan 3D to make a simple text from image to 3d and the result finished with no errors
The issue is that when this error appears it does not say its about a specific module
Tks
@CrisOrlandoBR I recommend running in debug mode and checking the logs to see where it breaks.
set CUDA_LAUNCH_BLOCKING=1
set TORCH_SHOW_CPP_STACKTRACES=1
python main.py > log.txt 2>&1
I suspect it might be nvdiffrast or torch-scatter — try reinstalling them (either from wheels or from source).
Had issues with xformers then pytorch3d then torch-scatter
Reinstalled all packages that I could
and still got to:
Warn!: xFormers is available (Attention) Warn!: Traceback (most recent call last): File "D:\AI\ComfyUI_3D\ComfyUI\nodes.py", line 2124, in load_custom_node module_spec.loader.exec_module(module) File "<frozen importlib._bootstrap_external>", line 999, in exec_module File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed File "D:\AI\ComfyUI_3D\ComfyUI\custom_nodes\ComfyUI-3D-Pack\__init__.py", line 54, in <module> module = importlib.import_module(f".{nodes_filename}", package=__name__) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "importlib\__init__.py", line 90, in import_module File "<frozen importlib._bootstrap>", line 1387, in _gcd_import File "<frozen importlib._bootstrap>", line 1360, in _find_and_load File "<frozen importlib._bootstrap>", line 1331, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 935, in _load_unlocked File "<frozen importlib._bootstrap_external>", line 999, in exec_module File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed File "D:\AI\ComfyUI_3D\ComfyUI\custom_nodes\ComfyUI-3D-Pack\nodes.py", line 85, in <module> from Unique3D.scripts.project_mesh import multiview_color_projection, multiview_color_projection_texture, get_cameras_list, get_orbit_cameras_list File "D:\AI\ComfyUI_3D\ComfyUI\custom_nodes\ComfyUI-3D-Pack\Gen_3D_Modules\Unique3D\scripts\project_mesh.py", line 99, in <module> pix2faces_renderer = Pix2FacesRenderer() ^^^^^^^^^^^^^^^^^^^ File "D:\AI\ComfyUI_3D\ComfyUI\custom_nodes\ComfyUI-3D-Pack\Gen_3D_Modules\Unique3D\scripts\project_mesh.py", line 71, in __init__ self._glctx = dr.RasterizeCudaContext(device=device) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\nvdiffrast\torch\ops.py", line 184, in __init__ self.cpp_wrapper = _get_plugin().RasterizeCRStateWrapper(cuda_device_idx) ^^^^^^^^^^^^^ File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\nvdiffrast\torch\ops.py", line 125, in _get_plugin torch.utils.cpp_extension.load(name=plugin_name, sources=source_paths, extra_cflags=common_opts+cc_opts, extra_cuda_cflags=common_opts+['-lineinfo'], extra_ldflags=ldflags, with_cuda=True, verbose=False) File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\torch\utils\cpp_extension.py", line 1670, in load return _jit_compile( ^^^^^^^^^^^^^ File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\torch\utils\cpp_extension.py", line 2127, in _jit_compile _write_ninja_file_and_build_library( File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\torch\utils\cpp_extension.py", line 2237, in _write_ninja_file_and_build_library verify_ninja_availability() File "D:\AI\ComfyUI_3D\python_embeded\Lib\site-packages\torch\utils\cpp_extension.py", line 2298, in verify_ninja_availability raise RuntimeError("Ninja is required to load C++ extensions (pip install ninja to get it)") RuntimeError: Ninja is required to load C++ extensions (pip install ninja to get it)
Everytime it's such a pain to get this running
*Nevermind , manual install + pip install spconv-cu126 helped
@siraxe glad manual installation worked! comfyui-manager in a new version should do the same