You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am not sure is there the right place to post this issue. I am using apptainer container in a HPC cluster, this is the only way for me to set up nerfstudio on that cluster due to limit permission.
After the container building finished, the python inside the container cannot import nerfstudio and even cannot import numpy, then I figured out some env variables are missing. And after I added them, all function works and I can train and render successfully. But for each cmd I run, it got stuck and took minutes to get the 1st line of output, even for the simple help command.
To Reproduce
build the container bash apptainer build --sandbox nerstudio_sandbox docker://dromni/nerfstudio:1.1.3
launch the container, --env is needed because I found this variables are missing for some reason and without them python cannot loacte pkgs. apptainer exec -e --nv --writable\ -B $SCRATCH/nerfstudio/:/workspace/nerfstudio\ --env PYTHONPATH=/home/user/.local/lib/python3.10/site-packages:/home/user/nerfstudio \ $SCRATCH/nerfstudio_sandbox env HOME=/home/user bas
run any command, and wait minutes to get output. bash ns-train --help Expected behavior
At least the help command wont take so long to execute.
Additional context
My apptainer version is 1.3.2, after the 1st line of output shows on the screen, everything seems work in normal speed. So I guess there are some issues behind the command-line setup.
The text was updated successfully, but these errors were encountered:
Describe the bug
Hi there,
I am not sure is there the right place to post this issue. I am using apptainer container in a HPC cluster, this is the only way for me to set up nerfstudio on that cluster due to limit permission.
After the container building finished, the python inside the container cannot import nerfstudio and even cannot import numpy, then I figured out some env variables are missing. And after I added them, all function works and I can train and render successfully. But for each cmd I run, it got stuck and took minutes to get the 1st line of output, even for the simple help command.
To Reproduce
bash apptainer build --sandbox nerstudio_sandbox docker://dromni/nerfstudio:1.1.3
--env
is needed because I found this variables are missing for some reason and without them python cannot loacte pkgs.apptainer exec -e --nv --writable\ -B $SCRATCH/nerfstudio/:/workspace/nerfstudio\ --env PYTHONPATH=/home/user/.local/lib/python3.10/site-packages:/home/user/nerfstudio \ $SCRATCH/nerfstudio_sandbox env HOME=/home/user bas
bash ns-train --help
Expected behavior
At least the help command wont take so long to execute.
Additional context
My apptainer version is 1.3.2, after the 1st line of output shows on the screen, everything seems work in normal speed. So I guess there are some issues behind the command-line setup.
The text was updated successfully, but these errors were encountered: