GPU workflow should run tests on the GPU

[jatkinson1000]

As raised by #427 whilst there is now a GPU workflow to run examples (integration tests), many of them are still being run on CPU by default.

To run on GPU would require passing --device='cuda' from the commandline call to the python files and pt2ts script.

Really we should look to update the script to run everything possible on the GPU.

cc'ing @niccolozanotti as the original author of the workflow in case he has thoughts/comments.

[niccolozanotti]

Ah yes, this was not considered in #410.

Really we should look to update the script to run everything possible on the GPU.

Definitely.
I could take this on next week.

[jatkinson1000]

If you're up for that that would be great, thanks @niccolozanotti!

[zhenkunl]

That's a good idea. For integration tests, CMakeLists.txt in all examples need to be adjusted as well.