won't work on nvidia jetson AGX

[AndreV84]

`~/llff$ bash zip2mpis.sh /home/nvidia/llff/data/testscene.zip test --height 360
Archive: /home/nvidia/llff/data/testscene.zip
replace test/images/IMG_4030.JPG? [y]es, [n]o, [A]ll, [N]one, [r]ename: N
Don't need to run COLMAP
Post-colmap
Cameras 5
Images # 20
Points (9993, 3) Visibility (9993, 20)
Depth stats 13.73132579288042 119.12624530357981 30.405215378877163
Done with imgs2poses
2022-05-24 17:07:04.359776: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0
WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them.
WARNING:tensorflow:
The TensorFlow contrib module will not be included in TensorFlow 2.0.
For more information, please see:

• https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md
• https://github.com/tensorflow/addons
• https://github.com/tensorflow/io (for I/O related ops)
  If you depend on functionality not listed there, please file an issue.

usage: imgs2mpis.py [-h] [--checkpoint CHECKPOINT] [--factor FACTOR] [--width WIDTH] [--height HEIGHT] [--numplanes NUMPLANES] [--psvs] [--no_mpis] scenedir mpidir
imgs2mpis.py: error: argument --height: expected one argument
Path components [False, False, False, False, True]
Saved to test/outputs/test_path.txt
make: 'cuda_renderer' is up to date.
Loading test/mpis_--height
loading -892274752 mpis
65535 x 0
0 0 65535 -892274752
Big request (host) 0 MB
terminate called after throwing an instance of 'std::bad_array_new_length'
what(): std::bad_array_new_length
zip2mpis.sh: line 9: 3502 Aborted (core dumped) cuda_renderer/cuda_renderer $2/mpis_$3 $2/outputs/test_path.txt $2/outputs/test_vid.mp4 -1 .8 18
`