Error when running zipnerf method in nerfstudio

[sblat021]

I'm trying to run zipnerf and I'm getting the error below when running: ns-train zipnerf nerfstudio-data --data {DATA_DIR}. DATA_DIR is the directory containing the processed data generated with ns-process-data.

I also tried to run : ns-train zipnerf --data {DATA_DIR} ns-train zipnerf colmap --data {DATA_DIR/colmap}

But I'm getting errors on all of those. It seems that ns-train zipnerf nerfstudio-data --data {DATA_DIR} should be the correct command when using nerfstudio data, but there's no documentation on this yet.

ERROR:root:Path not found: /configs/360.gin Traceback (most recent call last): File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\runpy.py", line 194, in _run_module_as_main return _run_code(code, main_globals, None, File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\runpy.py", line 87, in _run_code exec(code, run_globals) File "C:\Users\micro\anaconda3\envs\nerfstudio101\Scripts\ns-train.exe\__main__.py", line 7, in <module> File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\scripts\train.py", line 262, in entrypoint main( File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\scripts\train.py", line 247, in main launch( File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\scripts\train.py", line 189, in launch main_func(local_rank=0, world_size=world_size, config=config) File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\scripts\train.py", line 99, in train_loop trainer.setup() File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\engine\trainer.py", line 149, in setup self.pipeline = self.config.pipeline.setup( File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\configs\base_config.py", line 54, in setup return self._target(self, **kwargs) File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\zipnerf_ns\zipnerf_pipeline.py", line 63, in __init__ self._model = config.model.setup( File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\configs\base_config.py", line 54, in setup return self._target(self, **kwargs) File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\nerfstudio\models\base_model.py", line 85, in __init__ self.populate_modules() # populate the modules File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\zipnerf_ns\zipnerf_model.py", line 56, in populate_modules gin.parse_config_files_and_bindings(gin_files, None) File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\gin\config.py", line 2497, in parse_config_files_and_bindings includes_and_imports = parse_config_file(config_file, skip_unknown) File "C:\Users\micro\anaconda3\envs\nerfstudio101\lib\site-packages\gin\config.py", line 2457, in parse_config_file raise IOError(err_str.format(config_file, prefixes)) OSError: Unable to open file: /configs/360.gin. Searched config paths: [''].

Steps to reproduce the behavior:

1. Install nerfstudio and all dependencies
2. Install zipnerf method using : pip install git+https://github.com/SuLvXiangXin/zipnerf-pytorch
3. Install torch-scatter using: pip install torch-scatter -f https://data.pyg.org/whl/torch-2.1.2+${CUDA}.html
4. navigate to project directory
5. run ns-train zipnerf nerfstudio-data --data {DATA_DIR} or ns-train zipnerf --data {DATA_DIR}

Expected behavior: Should start training zipnerf using processed data.

Note: It looks like it is looking for some 360.gin file, which doesn't exist in nerfstudio's processed data. Did anyone have any luck getting zipnerf to run with nerfstudio?

[michaelrubloff]

I actually know how to fix this! I had this exact issue. I ended up cloning the original pytorch-Zipnerf's repository and that 360config file was in there. I just dropped it into the main nerfstudio folder and it started working.

I also was running into an error where it was looking for the sparse set within COLMAP and I fixed that error by simply just copying the sparse folder from within the COLMAP folder from the dataset I was trying to train and that fixed that issue too

[sblat021]

When you say main nerfstudio folder, do you mean: \anaconda3\envs\nerfstudio\Lib\site-packages\nerfstudio ? I installed nerfstudio using pip.

[michaelrubloff]

No, it was just the nerfstudio folder where your training data is hosted. So for me it's just user/name/nerfstudio

[ichsan2895]

Seems bug.. I follow this way from zipnerf-pytorch's website, and it works out of box.

pip install nerfstudio==1.0.1

cd zipnerf-pytorch
git clone https://github.com/SuLvXiangXin/zipnerf-pytorch.git
pip install ./extensions/cuda
pip install torch-scatter -f https://data.pyg.org/whl/torch-2.0.0+${CUDA}.html

pip install -e . 
ns-install-cli

Note: I use Nerfstudio v1.0.1 which have not yet merged between zipnerf-pytorch and nerfstudio

[sblat021]

unfortunately that doesn't work for me either.. it's failing when i pip install ./extensions/cuda

`Processing c:\apps\zipnerf-pytorch\extensions\cuda Preparing metadata (setup.py) ... error error: subprocess-exited-with-error

× python setup.py egg_info did not run successfully. │ exit code: 1 ╰─> [12 lines of output] Traceback (most recent call last): File "", line 2, in File "", line 34, in File "C:\Apps\zipnerf-pytorch\extensions\cuda\setup.py", line 35, in CUDAExtension( File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 1074, in CUDAExtension library_dirs += library_paths(cuda=True) File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 1208, in library_paths paths.append(_join_cuda_home(lib_dir)) File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 2407, in _join_cuda_home raise OSError('CUDA_HOME environment variable is not set. ' OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root. [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip. error: metadata-generation-failed

× Encountered error while generating package metadata. ╰─> See above for output.

note: This is an issue with the package mentioned above, not pip. hint: See above for details.`

[yggdrasil-917]

Anyone gotten this to work recently? I've been having similar issues setting this up on both Windows and Linux. Even zipnerf-pytorch project is failing for me on the same step as mentioned above. Any ideas or insights would be appreciated.

[Renbry]

Now with Nerfstudio v1.0.2 I have a different issue than mentioned - Run : ns-train zipnerf It asks if I want to install the package, Yes.

Errors out on several Windows machines with :

ERROR: Could not install packages due to an OSError: [WinError 5] Access is denied: 'C:\Users\renbr\anaconda3\envs\nerfstudio\Lib\site-packages\cv2\cv2.pyd' Consider using the --user option or check the permissions.

Error installing method.

Anyone know what to try next ? it's a fresh install (1hr old!) and happens across all my windows machines.

[Jing1Ling]

unfortunately that doesn't work for me either.. it's failing when i pip install ./extensions/cuda

`Processing c:\apps\zipnerf-pytorch\extensions\cuda Preparing metadata (setup.py) ... error error: subprocess-exited-with-error

× python setup.py egg_info did not run successfully. │ exit code: 1 ╰─> [12 lines of output] Traceback (most recent call last): File "", line 2, in File "", line 34, in File "C:\Apps\zipnerf-pytorch\extensions\cuda\setup.py", line 35, in CUDAExtension( File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 1074, in CUDAExtension library_dirs += library_paths(cuda=True) File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 1208, in library_paths paths.append(_join_cuda_home(lib_dir)) File "C:\Users\micro\anaconda3\envs\zipnerf\lib\site-packages\torch\utils\cpp_extension.py", line 2407, in _join_cuda_home raise OSError('CUDA_HOME environment variable is not set. ' OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root. [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip. error: metadata-generation-failed

× Encountered error while generating package metadata. ╰─> See above for output.

note: This is an issue with the package mentioned above, not pip. hint: See above for details.`

@sblat021 'CUDA_HOME environment variable is not set'. It seems that the CUDA environment variables are not set correctly. Does the following command return true when you execute it? import torch torch.cuda.is_available()

[tosemfdk]

there is another method just make colmap output( sparse data) using nerfstudio's processed data (transform.json) check out here https://github.com/NVlabs/instant-ngp/issues/1080

[sumanttyagi]

is this issue resolved please share the solution

[ichsan2895]

is this issue resolved please share the solution

Have you try this way? It works with Linux. I never test on Windows

pip install nerfstudio==1.0.3
pip install git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch

git clone https://github.com/SuLvXiangXin/zipnerf-pytorch.git
cd zipnerf-pytorch
pip install ./extensions/cuda
pip install torch-scatter -f https://data.pyg.org/whl/torch-2.0.0+${CUDA}.html

pip install -e . 
ns-install-cli

Then you will able to run this command

ns-train zipnerf nerfstudio-data --data path/to/your/data