multinerf
multinerf copied to clipboard
google-research
dataset problem
hi,i meet some problem when run this :
python -m train --gin_configs=configs/360.gin --gin_bindings="Config.data_dir = 'my_dataset_dir'" --gin_bindings="Config.checkpoint_dir = 'my_dataset_dir/checkpoints'" --logtostderr
how can i do next ,please give me some advice ,thanks!!!
Did you try to use your custome data？ Maybe you should use the dataset of class llff instead of blender？
Did you successfully run COLMAP? There should be a subdirectory my_dataset_dir/sparse/0/ if so, which should prevent the code from reaching this error.
I'm getting the same No such file or directory: 'my_dataset_dir/transforms.json' issue. I ran COLMAP via scripts/local_colmap_and_resize.sh, but sparse is empty because the command that populates it (I think) fails on my machine; the Mapper.ba_global_function_tolerance option is unrecognized. This probably has to do with the fact that I'm running an older version of COLMAP (3.5) because of issues I had getting COLMAP running on my OS (macOS Monterey).
After installing suite-sparse I was able to upgrade to the latest version of COLMAP (3.7) and to run the aforementioned command successfully. sparse/0 now exists in my system.
- https://github.com/colmap/colmap/issues/935#issuecomment-827627508
I'm currently training multinerf. Hopefully, the steps I presented will help you get further @RuiYangQuan. Thanks @bmild for the explanation.
This problem exists in google colab also, any solution with respect to that?, I tried installing colamp 3.8 but it wasn't useful
