some package in docker image building file seems outdated

[JKYang01]

My cuda version is 12.2 I cannot find the base image nvidia/cuda:12.2-cunn8-devel-ubuntu22.04 so I installed with nvidia/cuda:12.2.2-devel-ubuntu22.04 instead
in the next steps I encountered the following error:

164.1 Installed /usr/local/lib/python3.10/dist-packages/qudida-0.0.4-py3.10.egg
164.1 error: fsspec 2023.1.0 is installed but fsspec>=2023.5.0 is required by {'huggingface-hub'}
------
Dockerfile:19
--------------------
  17 |     WORKDIR /workspace/nougat
  18 |     
  19 | >>> RUN python3 setup.py install
  20 |     
  21 |     EXPOSE 8503
--------------------
ERROR: failed to solve: process "/bin/sh -c python3 setup.py install" did not complete successfully: exit code: 1

198.4 Installed /usr/local/lib/python3.10/dist-packages/tifffile-2024.2.12-py3.10.egg
198.4 error: networkx 2.6.3 is installed but networkx>=2.8 is required by {'scikit-image'}
------
Dockerfile:19
--------------------
  17 |     WORKDIR /workspace/nougat
  18 |     
  19 | >>> RUN python3 setup.py install
  20 |     
  21 |     EXPOSE 8503
--------------------
ERROR: failed to solve: process "/bin/sh -c python3 setup.py install" did not complete successfully: exit code: 1

also I got a lot of warning message saying
/usr/lib/python3/dist-packages/pkg_resources/__init__.py:116: PkgResourcesDeprecationWarning: python-headless-3.4.10.37 is an invalid version and will not be supported in a future release
is this normal?

[JKYang01]

I tried to update those two packages to match the version and the image did build,
however when I try to run the image the new error come up:

docker run -it -d -p 5050:8503 --gpus all nougat_run
6c00e4782882c264b7ab63022b29dda2d80ae65a4f011f7e935f0169470fd3aa
docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as 'legacy'
nvidia-container-cli: initialization error: load library failed: libnvidia-ml.so.1: cannot open shared object file: no such file or directory: unknown

[fqassemi]

a fix is posted here:
#208