Support PyTorch 2.6.0 / Add zarr as optional dependency

[github-actions]

Run tests on latest PyTorch release
Issue/s resolved: #1774
TODO:

• update .github/workflows/ci.yaml to include n-1 Pytorch version
• update Nvidia and AMD Docker images on gitlab CI
  Auto-generated by [create-pull-request][1]
  [1]: https://github.com/peter-evans/create-pull-request

[mrfh92]

Steps done

• (added optional zarr support to setup.py as it makes sense to to this if we need to change things on codebase anyway)
• created branch pytorch261 on codebase (based on previous branch pytorch251)
  • in this new branch, I have adapted the installation instructions for PyTorch accordingly (both for AMD and Nvidia);
  • changing the base images was not necessary as CUDA and ROCm versions in these images were fine
  • (in addition, I have added the optional zarr dependency to the pip install command for installing Heat)
• changed the version of pytorchxyz in the heat/support branch to 261 to let the CI for this PR run with the yaml file from pytorch261-branch already
• after everything mentioned above was done in the right way (i.e. whithout demanding option cu126 from the wheels channel cu124 etc. ) the run succeeded ---> we are ready for PyTorch 2.6.0 on main
• approved in order to let CI matrix run
• since everything was fine, I set pytorch261 in the yaml files in the branches heat/base (CI runs on main) and heat/pr (CI runs for PRs) on codebase
• ... and merged this PR

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.26%. Comparing base (d66e404) to head (eda2c72).
Report is 5 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1775   +/-   ##
=======================================
  Coverage   92.26%   92.26%           
=======================================
  Files          84       84           
  Lines       12447    12447           
=======================================
  Hits        11484    11484           
  Misses        963      963           

Flag	Coverage Δ
unit	92.26% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.