Extend the device test suite: gradient, op-arithmetic and template tests

[timmysilv]

I ran the plugin test suite against this branch so I could see which tests fail right away on which devices (done, see here, braket and quantuminspire are failing for their own reasons), and skip them for those devices. Everything is now passing as expected.

Context:
The device test suite doesn't cover a lot of PennyLane's features, so I'm adding more coverage.

Description of the Change:

• modified CI to install interfaces if needed (basically always except for default.qubit.autograd)
• updated test_measurements.py to test some basic arithmetic ops by comparing the result to default.qubit
• Added 5 basic differentiation tests for each interface: test_basic_grad, test_backprop_state, test_parameter_shift, test_probs, test_multi_meas test_hessian. They do what they sound like they do 😄
• Added a test for every single template in PennyLane 🐳 I just grabbed the example from each template's docstring, got the result with default.qubit, and I'm asserting that every other device also gets that result. If the docstring didn't have one (or it just returned 1 when there are better examples), I tried to get one from the tests for that template. I'd also try to swap qml.state() with qml.probs() whenever possible so the test would work with finite-shot devices.

Benefits:
More confidence in our plugin devices!

Possible Drawbacks:

• Slower device tests
• New failures to manage?
• Some tests needed explicit skips from certain devices. I opened a story to track those explicit skips being done in the device test suite upgrade epic

[sc-57488]

[github-actions]

Hello. You may have forgotten to update the changelog!
Please edit doc/releases/changelog-dev.md with:

• A one-to-two sentence description of the change. You may include a small working example for new features.
• A link back to this PR.
• Your name (or GitHub username) in the contributors section.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.66%. Comparing base (b2b0c36) to head (07b432f).
Report is 3 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5273      +/-   ##
==========================================
- Coverage   99.67%   99.66%   -0.01%     
==========================================
  Files         404      404              
  Lines       37833    37554     -279     
==========================================
- Hits        37709    37429     -280     
- Misses        124      125       +1     

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

[astralcai]

Looks good! Just a few small comments there.

[astralcai]

Looks Good! 🎸

[timmysilv]

this change will require branch protection updates because of device-tests (default.qubit.autograd, None, jax,tf,torch, 3.9), cc @albi3ro