Encountering errors? Join my bycloud's discord and ask in #general-tech-support discord.com/invite/NhJZGtH IF you get a "ModuleNotFoundError: No module named 'diff_gaussian_rasterization'" Error, download CUDA 11.8. developer.nvidia.com/cuda-11-8-0-download-archive It should be installed according to your system, and to get whether if you have installed correctly, you can double check with this method: github.com/bycloudai/SwapCudaVersionWindows
I had to run: conda remove cuda and then: conda_hook && conda deactivate && conda deactivate && conda deactivate && conda activate base && conda install -y cuda -c nvidia/label/cuda-11.8 to get 11.8 installed properly.
if i try to use train py i get this error: Traceback (most recent call last): File "train py", line 16, in from gaussian_renderer import render, network_gui File "C:\Users\mukatsuku\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
I got this ERROR from gaussian_renderer import render, network_gui File "C:\Users\moroc\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization' Who knows how to fix this?
Wow seems like a great leap forward in helping the not so smarty (like me) but because I am not so smarty, I appear to be getting the following when I try to point to my photos 'The system cannot find the path specified. ERROR:root:Feature extraction failed with code 1. Exiting. ' Checked the template and it is definitely right
Hi, had the same error. What helped for me was: Checking the path of the COLMAP that is in Environmental variables. Check if the version is correct (should be 3.8 or 3.8.1). My path is as such c:/Users/myProfilename/COLMAP-3.8-windows-cuda (not copied exactly check yours). Then in the command delete the last part before the slash ---> python convert.py -s input_data/fern --colmap_executable (COLMAP-3.8-windows-cuda\ -----> delete this)COLMAP.bat ---> python convert.py -s input_data/fern --colmap_executable COLMAP.bat Hope this helps : ))
Hello how are you? a query: in 3D Gaussian Splatting Windows, I was able to do everything, but when I get to point 3 of the step 3 tutorial: Train/optimize the images, when I insert: python train.py -s input_data/fern, it returns this error via console: Traceback (most recent call last): File "train.py", line 13, in import torch ModuleNotFoundError: No module named 'torch' What can be? thank you
Ohh this is amazing. Thanks for this. The other tutorial is exhausting. I guess my image resolution were too high? I got a " Encountered quite large input images (>1.6K pixels width), rescaling to 1.6K. If this is not desired, please explicitly specify '--resolution/-r' as 1""... however the process did continue. Thanks.
In step 3 I get the following error: (base) C:\Users\Usuario\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/fern Traceback (most recent call last): File "C:\Users\Usuario\pinokio\api\gaussian-splatting-Windows.git\train.py", line 13, in import torch ModuleNotFoundError: No module named 'torch'
For everyone struggling with this error: COLMAP-3.8-windows-cuda\COLMAP.bat The system cannot find the path specified. You probably don't have COLMAP setup correctly for some reason AND / OR the command is wrong. This is setup in Environmental Variables and then Path and then Edit (google this if you don't know how) For me the path I had was C:\Users\XXXXXXXXX\COLMAP-3.8-windows-cuda And so the command is not python convert.py -s input_data/somefunnyname--colmap_executable COLMAP-3.8-windows-cuda\COLMAP.bat but rather python convert.py -s input_data/somefunnyname --colmap_executable COLMAP.bat Hope this helps! :)
after using the command "python train.py -s input_data/Test" i get the following error: (gaussian-splatting) C:\Users\StriK3FoRC3\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/Test Traceback (most recent call last): File "train.py", line 16, in from gaussian_renderer import render, network_gui File "C:\Users\StriK3FoRC3\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
yeah, that would be nice to know. I wanted to test with my 4070 ti, NeRF works like a charm... but i also get an error during installation. Something went wrong after installing all the bits.
(gaussian-splatting) C:\Users\baseer.farooqui\pinokio\api\gaussian-splatting-Windows.git>python convert.py -s input_data/door --colmap_executable COLMAP-3.8-windows-cuda\COLMAP.bat The system cannot find the path specified. ERROR:root:Feature extraction failed with code 1. Exiting. [1:54 PM] any idea how to fix
So I watched all that, but I don't really know what it does? I appreciate it's only an installation tutorial, so I'm going to have a quick catch up on your videos now, in hope to find a "usage" video. If you have any specific recommendation for a usage video, please let me know. Thanks
I'm getting this message: RuntimeError: [enforce fail at C:\cb\pytorch_1000000000000\work\c10\core\impl\alloc_cpu.cpp:81] data. DefaultCPUAllocator: not enough memory: you tried to allocate 20294400 bytes. Is this Ram or space disk problem?
Hey thanks for the tutorial. I am encountering this error while installing the package. CondaVerificationError: The package for pytorch located at Users\pinokio\bin\miniconda\pkgs\pytorch-1.12.1-py3.7_cuda11.6_cudnn8_0 appears to be corrupted. The path 'Lib/site-packages/caffe2/experiments/python/tt_pad_op_test.py' specified in the package manifest cannot be found.
@@valeryandreev4535 i used a plug in for blender whoch used my ply as particles and it was cool but the quality was like 25% of what the gaussian looked like in polycam
in the 3th step it give me this err (gaussian-splatting) C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git>start viewers/bin/SIBR_GaussianViewer_app.exe . -m output\fern (gaussian-splatting) C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/fern Traceback (most recent call last): File "train.py", line 16, in from gaussian_renderer import render, network_gui File "C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
if i try to use train py i get this error: Traceback (most recent call last): File "train py", line 16, in from gaussian_renderer import render, network_gui File "C:\Users\mukatsuku\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
Encountering errors? Join my bycloud's discord and ask in #general-tech-support discord.com/invite/NhJZGtH
IF you get a "ModuleNotFoundError: No module named 'diff_gaussian_rasterization'" Error, download CUDA 11.8. developer.nvidia.com/cuda-11-8-0-download-archive
It should be installed according to your system, and to get whether if you have installed correctly, you can double check with this method: github.com/bycloudai/SwapCudaVersionWindows
these solutions are not working, pls offer another
I had to run:
conda remove cuda
and then:
conda_hook && conda deactivate && conda deactivate && conda deactivate && conda activate base && conda install -y cuda -c nvidia/label/cuda-11.8
to get 11.8 installed properly.
A big thanks man. I really wanted to try it out, I was waiting for an easy installation like this 👍
if i try to use train py i get this error:
Traceback (most recent call last):
File "train py", line 16, in
from gaussian_renderer import render, network_gui
File "C:\Users\mukatsuku\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in
from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer
ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
Hey did you find the solution for this error i am facing the same error ?
same one.
im facing this issue as well
same
same here, any solution yet ?
@@natgeo1416
Is there a way of exporting a usable file? Like the .Ply so we can import it into another program?
did you ever get an answer to this?
I got this ERROR
from gaussian_renderer import render, network_gui
File "C:\Users\moroc\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in
from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer
ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
Who knows how to fix this?
Wow seems like a great leap forward in helping the not so smarty (like me) but because I am not so smarty, I appear to be getting the following when I try to point to my photos
'The system cannot find the path specified.
ERROR:root:Feature extraction failed with code 1. Exiting. '
Checked the template and it is definitely right
Hi, had the same error.
What helped for me was:
Checking the path of the COLMAP that is in Environmental variables. Check if the version is correct (should be 3.8 or 3.8.1).
My path is as such c:/Users/myProfilename/COLMAP-3.8-windows-cuda (not copied exactly check yours).
Then in the command delete the last part before the slash ---> python convert.py -s input_data/fern --colmap_executable (COLMAP-3.8-windows-cuda\ -----> delete this)COLMAP.bat ---> python convert.py -s input_data/fern --colmap_executable COLMAP.bat
Hope this helps : ))
Thank you for that was really struggling with other tuts
Hello how are you? a query: in 3D Gaussian Splatting Windows, I was able to do everything, but when I get to point 3 of the step 3 tutorial: Train/optimize the images, when I insert: python train.py -s input_data/fern, it returns this error via console: Traceback (most recent call last):
File "train.py", line 13, in
import torch
ModuleNotFoundError: No module named 'torch'
What can be? thank you
same err...
same here
same
gaussian-rasterization setup error. Is there any way to improve it?
Ohh this is amazing. Thanks for this. The other tutorial is exhausting. I guess my image resolution were too high? I got a " Encountered quite large input images (>1.6K pixels width), rescaling to 1.6K.
If this is not desired, please explicitly specify '--resolution/-r' as 1""... however the process did continue. Thanks.
thanks for the great video. how much vram does my graphic card need for that? or is it run via CPU?
It needs 24gb vram
In step 3 I get the following error: (base) C:\Users\Usuario\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/fern
Traceback (most recent call last):
File "C:\Users\Usuario\pinokio\api\gaussian-splatting-Windows.git\train.py", line 13, in
import torch
ModuleNotFoundError: No module named 'torch'
You're amazing, thanks for this.
Thanks for toturial , If I only use rtx3060 8g Vram so does it worth Install AI on local windows? Too slow ? right?
For everyone struggling with this error:
COLMAP-3.8-windows-cuda\COLMAP.bat
The system cannot find the path specified.
You probably don't have COLMAP setup correctly for some reason AND / OR the command is wrong. This is setup in Environmental Variables and then Path and then Edit (google this if you don't know how)
For me the path I had was C:\Users\XXXXXXXXX\COLMAP-3.8-windows-cuda
And so the command is not python convert.py -s input_data/somefunnyname--colmap_executable COLMAP-3.8-windows-cuda\COLMAP.bat
but rather python convert.py -s input_data/somefunnyname --colmap_executable COLMAP.bat
Hope this helps! :)
after using the command "python train.py -s input_data/Test" i get the following error:
(gaussian-splatting) C:\Users\StriK3FoRC3\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/Test
Traceback (most recent call last):
File "train.py", line 16, in
from gaussian_renderer import render, network_gui
File "C:\Users\StriK3FoRC3\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in
from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer
ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
getting this too, do not know how to fix.
@@zerosaturn416 same error
this work!!!
pip install submodules\diff-gaussian-rasterization
pip install submodules\simple-knn
how to install@@dmitrybo
@@ismafikri7414 just copy in cmd windows and run.
so 12gb vram ? can't run or more time need to process
yeah, that would be nice to know. I wanted to test with my 4070 ti, NeRF works like a charm... but i also get an error during installation. Something went wrong after installing all the bits.
@@pandadayi did you figure out? / does it take longer but work?
@@nicolasdesola3332 I couldn't find out because during installation an error occurred. Couldn't test it.. 😓
Is there any version for 12 GB VRAM?
Did you end up finding a 12gb version
No, 24 its the Minimum... The New one Live render in 120vram xD
The New versión appear 3 weeks ago, so this is old
pinochio is hanging when i try to download the git files its just a spinning disk
(gaussian-splatting) C:\Users\baseer.farooqui\pinokio\api\gaussian-splatting-Windows.git>python convert.py -s input_data/door --colmap_executable COLMAP-3.8-windows-cuda\COLMAP.bat
The system cannot find the path specified.
ERROR:root:Feature extraction failed with code 1. Exiting.
[1:54 PM]
any idea how to fix
its due to the incorrect path of COLMAP.bat, locate the file and use the correct path.
ModuleNotFoundError: No module named 'torch'
same...
So I watched all that, but I don't really know what it does?
I appreciate it's only an installation tutorial, so I'm going to have a quick catch up on your videos now, in hope to find a "usage" video.
If you have any specific recommendation for a usage video, please let me know.
Thanks
ua-cam.com/video/C708Mh7EHZM/v-deo.html
are you braindead or a bot?
I'm getting this message: RuntimeError: [enforce fail at C:\cb\pytorch_1000000000000\work\c10\core\impl\alloc_cpu.cpp:81] data. DefaultCPUAllocator: not enough memory: you tried to allocate 20294400 bytes.
Is this Ram or space disk problem?
Works on my 4090, which gpu are u on?
Got it to work, thanks man!
Hello great video , it is possible to run the script cloud machine like google collab ? My pc is bad
I think the patreon link is broken
nick work!
why is there no tutorials on how to gaussian splat from ply files ahhhhhh
it gives error to me as well
thank you for your help, hao i can ran this :ModuleNotFoundError: No module named 'simple_knn'
How can i export as a mesh ?
24gb vram ???
Why ponokio? Why not just a .bat file. It’s so suspect, like what the hell is ponokio?
This is also my question, and my pinokio was completely unable to install
Hey thanks for the tutorial. I am encountering this error while installing the package. CondaVerificationError: The package for pytorch located at Users\pinokio\bin\miniconda\pkgs\pytorch-1.12.1-py3.7_cuda11.6_cudnn8_0
appears to be corrupted. The path 'Lib/site-packages/caffe2/experiments/python/tt_pad_op_test.py'
specified in the package manifest cannot be found.
continuously saying something went wrong...
The authentic programmer experience
Does one work with AMD?
I just don't understand AMD users, AI isn't for you...yet. hold on
yes it does. AMD user here
@@valeryandreev4535what is your graphics card Sir
Cant even use gaussian splats anywhere. They are cool but useless
you can if you know how to use particles in some engines... actually you can make quite cool stuff
@@valeryandreev4535 i used a plug in for blender whoch used my ply as particles and it was cool but the quality was like 25% of what the gaussian looked like in polycam
well this tutorial was crap
in the 3th step it give me this err
(gaussian-splatting) C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git>start viewers/bin/SIBR_GaussianViewer_app.exe . -m output\fern
(gaussian-splatting) C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git>python train.py -s input_data/fern
Traceback (most recent call last):
File "train.py", line 16, in
from gaussian_renderer import render, network_gui
File "C:\Users\giuse\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in
from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer
ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
this work!!!
pip install submodules\diff-gaussian-rasterization
pip install submodules\simple-knn
You saved me!@@dmitrybo
if i try to use train py i get this error: Traceback (most recent call last): File "train py", line 16, in from gaussian_renderer import render, network_gui File "C:\Users\mukatsuku\pinokio\api\gaussian-splatting-Windows.git\gaussian_renderer\__init__.py", line 14, in from diff_gaussian_rasterization import GaussianRasterizationSettings, GaussianRasterizer ModuleNotFoundError: No module named 'diff_gaussian_rasterization'
same
same
same
same. I have a 4090. any ideas?
wrong CUDA install? hmmm