Comments (6)
It seems to be caused by Viser.
Have you tried run command line version (i.e. without WebUI)
from gaussianeditor.
It seems to be caused by Viser. Have you tried run command line version (i.e. without WebUI)
Thanks for your reply. I have tried to run the command line and meet another issue, both on my windows 11 (RTX 3090, cuda 11.7) and ubuntu 20.04 system (Rtx 4090, cuda 11.8). I ensure I have installed pytorch with CUDA support and build the tiny-cuda-nn (include the pytorch extension).
python launch.py --config configs/edit-n2n.yaml --train --gpu 1 trainer.max_steps=2000 \ system.prompt_processor.prompt="Turn the scene into an arsenal of future weapons in the style of cyberpunk" \ system.max_densify_percent=0.03 system.anchor_weight_init_g0=0.0 system.anchor_weight_init=0.02 system.anchor_weight_multiplier=1.3 \ system.seg_prompt="man" system.loss.lambda_anchor_color=5 system.loss.lambda_anchor_geo=50 system.loss.lambda_anchor_scale=50 \ system.loss.lambda_anchor_opacity=50 system.densify_from_iter=100 system.densify_until_iter=5000 \ system.densification_interval=300 data.source=autodl-tmp/3DGS_upload/data/armour_museum_house_center-600F_896-20F-CF1_cyberpuck-1K6 \ system.gs_source=autodl-tmp/3DGS_upload/output/armour_museum_house_center-600F_896-20F-CF1-1K6_cyberpunck/point_cloud/iteration_50000/point_cloud.ply \ system.loggers.wandb.enable=true system.loggers.wandb.name="edit_scene_to_cyberpunk" Traceback (most recent call last): File "launch.py", line 250, in <module> main(args, extras) File "launch.py", line 74, in main import threestudio File "/root/GaussianEditor/threestudio/__init__.py", line 36, in <module> from . import data, models, systems File "/root/GaussianEditor/threestudio/data/__init__.py", line 1, in <module> from . import gs_load File "/root/GaussianEditor/threestudio/data/gs_load.py", line 11, in <module> from threestudio.utils.base import Updateable File "/root/GaussianEditor/threestudio/utils/__init__.py", line 1, in <module> from . import base File "/root/GaussianEditor/threestudio/utils/base.py", line 7, in <module> from threestudio.utils.misc import get_device, load_module_weights File "/root/GaussianEditor/threestudio/utils/misc.py", line 5, in <module> import tinycudann as tcnn File "/root/miniconda3/lib/python3.8/site-packages/tinycudann/__init__.py", line 9, in <module> from tinycudann.modules import free_temporary_memory, NetworkWithInputEncoding, Network, Encoding File "/root/miniconda3/lib/python3.8/site-packages/tinycudann/modules.py", line 19, in <module> raise EnvironmentError("Unknown compute capability. Ensure PyTorch with CUDA support is installed.") OSError: Unknown compute capability. Ensure PyTorch with CUDA support is installed.
from gaussianeditor.
I have never met this error.
Can you successfully import torch? Please check your torch version by torch.version and CUDA version with nvcc -V
from gaussianeditor.
I have never met this error. Can you successfully import torch? Please check your torch version by torch.version and CUDA version with nvcc -V
Sure. I have verified the installation of pytorch and tinycudann by:
`import torch
print(torch.version)
print(torch.cuda.is_available())
2.0.0+cu118
True`
and
`!python ./samples/mlp_learning_an_image_pytorch.py data/images/test.png data/config_hash.json
================================================================
This script replicates the behavior of the native CUDA example
mlp_learning_an_image.cu using tiny-cuda-nn's PyTorch extension.
Using PyTorch version 2.0.0+cu118 with CUDA 11.8
NetworkWithInputEncoding(n_input_dims=2, n_output_dims=3, seed=1337, dtype=torch.float16, hyperparams={'encoding': {'base_resolution': 16, 'hash': 'CoherentPrime', 'interpolation': 'Linear', 'log2_hashmap_size': 15, 'n_features_per_level': 2, 'n_levels': 16, 'otype': 'Grid', 'per_level_scale': 1.5, 'type': 'Hash'}, 'network': {'activation': 'ReLU', 'n_hidden_layers': 2, 'n_neurons': 64, 'otype': 'FullyFusedMLP', 'output_activation': 'None'}, 'otype': 'NetworkWithInputEncoding'})
/root/miniconda3/lib/python3.8/site-packages/torch/functional.py:504: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3483.)
return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
Writing 'reference.jpg'... done.
Beginning optimization with 10000000 training steps.
./samples/mlp_learning_an_image_pytorch.py:70: TracerWarning: torch.tensor results are registered as constants in the trace. You can safely ignore this warning if you use this function to create tensors out of constant variables that would be the same every time you call this function. In any other case, this might cause the trace to be incorrect.
xs = xs * torch.tensor([shape[1], shape[0]], device=xs.device).float()
Step#0: loss=68.5625 time=1252960[µs]
Writing '0.jpg'... done.
Step#10: loss=0.5341796875 time=20094[µs]
Writing '10.jpg'... done.
Step#100: loss=0.01119232177734375 time=181919[µs]
Writing '100.jpg'... done.
Step#1000: loss=0.0004858970642089844 time=2387365[µs]
Writing '1000.jpg'... done.`
from gaussianeditor.
It seems to be caused by Viser. Have you tried run command line version (i.e. without WebUI)
Thanks for your reply. I have tried to run the command line and meet another issue, both on my windows 11 (RTX 3090, cuda 11.7) and ubuntu 20.04 system (Rtx 4090, cuda 11.8). I ensure I have installed pytorch with CUDA support and build the tiny-cuda-nn (include the pytorch extension).
python launch.py --config configs/edit-n2n.yaml --train --gpu 1 trainer.max_steps=2000 \ system.prompt_processor.prompt="Turn the scene into an arsenal of future weapons in the style of cyberpunk" \ system.max_densify_percent=0.03 system.anchor_weight_init_g0=0.0 system.anchor_weight_init=0.02 system.anchor_weight_multiplier=1.3 \ system.seg_prompt="man" system.loss.lambda_anchor_color=5 system.loss.lambda_anchor_geo=50 system.loss.lambda_anchor_scale=50 \ system.loss.lambda_anchor_opacity=50 system.densify_from_iter=100 system.densify_until_iter=5000 \ system.densification_interval=300 data.source=autodl-tmp/3DGS_upload/data/armour_museum_house_center-600F_896-20F-CF1_cyberpuck-1K6 \ system.gs_source=autodl-tmp/3DGS_upload/output/armour_museum_house_center-600F_896-20F-CF1-1K6_cyberpunck/point_cloud/iteration_50000/point_cloud.ply \ system.loggers.wandb.enable=true system.loggers.wandb.name="edit_scene_to_cyberpunk" Traceback (most recent call last): File "launch.py", line 250, in <module> main(args, extras) File "launch.py", line 74, in main import threestudio File "/root/GaussianEditor/threestudio/__init__.py", line 36, in <module> from . import data, models, systems File "/root/GaussianEditor/threestudio/data/__init__.py", line 1, in <module> from . import gs_load File "/root/GaussianEditor/threestudio/data/gs_load.py", line 11, in <module> from threestudio.utils.base import Updateable File "/root/GaussianEditor/threestudio/utils/__init__.py", line 1, in <module> from . import base File "/root/GaussianEditor/threestudio/utils/base.py", line 7, in <module> from threestudio.utils.misc import get_device, load_module_weights File "/root/GaussianEditor/threestudio/utils/misc.py", line 5, in <module> import tinycudann as tcnn File "/root/miniconda3/lib/python3.8/site-packages/tinycudann/__init__.py", line 9, in <module> from tinycudann.modules import free_temporary_memory, NetworkWithInputEncoding, Network, Encoding File "/root/miniconda3/lib/python3.8/site-packages/tinycudann/modules.py", line 19, in <module> raise EnvironmentError("Unknown compute capability. Ensure PyTorch with CUDA support is installed.") OSError: Unknown compute capability. Ensure PyTorch with CUDA support is installed.
I have met the same error called OSError: Unknown compute capability. Ensure PyTorch with CUDA support is installed.
, but I tried to import torch in a test file it did work. Later, my friend told me the error was caused by the optional parameters --gpu
. After I set it to 0, and the problem was solved.
from gaussianeditor.
I have tried to install it on windows and seems everything has been correctly set up. However, when I try to run the webui.py, I meet bellow errors:
`(GaussianEditor) PS D:\01_DL\GaussianEditor> python webui.py --gs_source D:\01_DL\gaussian-splatting\output\armour_museum_house_center-600F_896-20F-CF1-1K6_cyberpunck\point_cloud\iteration_50000\point_cloud.ply --colmap_dir D:\01_DL\gaussian-splatting\data\finland\armour_museum_house_center-600F_896-20F-CF1_cyberpuck-1K6 Reading camera 502/502 C:\ProgramData\Miniconda3\envs\GaussianEditor\lib\site-packages\torch\functional.py:504: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ..\aten\src\ATen\native\TensorShape.cpp:3484.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] final text_encoder_type: bert-base-uncased Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertModel: ['cls.predictions.transform.dense.weight', 'cls.seq_relationship.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.seq_relationship.weight', 'cls.predictions.transform.LayerNorm.bias']
This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
Model loaded from C:\Users\dell.cache\huggingface\hub\models--ShilongLiu--GroundingDINO\snapshots\a94c9b567a2a374598f05c584e96798a170c56fb\groundingdino_swinb_cogcoor.pth
=> _IncompatibleKeys(missing_keys=[], unexpected_keys=['label_enc.weight'])
(viser) No client build found. Building now...
C:\ProgramData\Miniconda3\envs\GaussianEditor\lib\site-packages\nodeenv.py:48: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
from pkg_resources import parse_versionEnvironment already exists: D:\01_DL\GaussianEditor\extern\viser\src\viser\client.nodeenv
Traceback (most recent call last):
File "D:\01_DL\GaussianEditor\webui.py", line 1540, in
webui = WebUI(args)
File "D:\01_DL\GaussianEditor\webui.py", line 156, in init
self.server = viser.ViserServer(port=self.port)
File "D:\01_DL\GaussianEditor\extern\viser\src\viser_viser.py", line 326, in init
_client_autobuild.ensure_client_is_built()
File "D:\01_DL\GaussianEditor\extern\viser\src\viser_client_autobuild.py", line 57, in ensure_client_is_built
env_dir = _install_sandboxed_node()
File "D:\01_DL\GaussianEditor\extern\viser\src\viser_client_autobuild.py", line 81, in _install_sandboxed_node
subprocess.run(
File "C:\ProgramData\Miniconda3\envs\GaussianEditor\lib\subprocess.py", line 503, in run
with Popen(*popenargs, **kwargs) as process:
File "C:\ProgramData\Miniconda3\envs\GaussianEditor\lib\subprocess.py", line 971, in init
self._execute_child(args, executable, preexec_fn, close_fds,
File "C:\ProgramData\Miniconda3\envs\GaussianEditor\lib\subprocess.py", line 1456, in _execute_child
hp, ht, pid, tid = _winapi.CreateProcess(executable, args,
FileNotFoundError: [WinError 2] 系统找不到指定的文件。`System: Windows 10, Cuda 11.7, torch 2.0.1, Rtx 3090, Python 3.10.12
I also met this error after the second step of "add", after clicked ending 2d inpainting
System: Windows 11, Cuda 11.8, torch 2.0.1, Rtx 4080, Python 3.10
from gaussianeditor.
Related Issues (20)
- An elegant way to install origin viser instead of modified viser HOT 4
- AssertionError: please install our forked viser HOT 3
- How to get .ply files on other datasets HOT 1
- Delta Denoising Score in GaussianEditor HOT 3
- Problem with the webUI HOT 2
- A problem occurs during the webUI operation HOT 1
- How can I output a rendered video of the edited scene? HOT 1
- No such file or directory: '.cache/dpt/omnidata_dpt_depth_v2.ckpt' HOT 5
- Adding operation related question HOT 3
- Which checkpoint is used in Unet module of UNetMV2DConditionModel HOT 8
- Which part of the code the formula (6) in the paper corresponds to HOT 3
- FileNotFoundError: [Errno 2] No such file or directory: HOT 2
- During the add editing operation, can I import the existing 3D model? HOT 2
- FileNotFoundError: [Errno 2] No such file or directory: '~/projects/GaussianEditor/tmp_add/multiview_pred_images/rgb_000_front.png' HOT 3
- Adding mode will use different GPU on GPU clusters, resulting scene mix up when running multiple scenes parallelly
- Can i run it with a 16G GPU(RTX 2080 ti)? HOT 1
- What can apply_weights do? HOT 2
- export edited ply HOT 2
- npm error
- Question about paper GaussianEditor HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from gaussianeditor.