Updating torch to 2.3.0 (#6258)

**Context:** We want to make PL compatible with torch 2.3.0 (including
GPU tests) after updating Numpy from 1.x to 2.x.

**Description of the Change:** As above.

**Benefits:** We are sure that PL is compatible with torch 2.3.0.

**Possible Drawbacks:** None that I can think of right now.

**Related GitHub Issues:** None.

**Related Shortcut Stories:** [sc-61391]

.github/workflows/tests-gpu.yml

@@ -15,7 +15,7 @@ concurrency:
  cancel-in-progress: true
 
 env:
- TORCH_VERSION: 2.2.0
+ TORCH_VERSION: 2.3.0
 
 jobs:
  gpu-tests:

doc/releases/changelog-dev.md

@@ -7,7 +7,8 @@
 <h3>Improvements 🛠</h3>
 
 * PennyLane is now compatible with NumPy 2.0.
- [(#6061)](https://github.com/PennyLaneAI/pennylane/pull/6061)
+ [(#6061)](https://github.com/PennyLaneAI/pennylane/pull/6061)
+ [(#6258)](https://github.com/PennyLaneAI/pennylane/pull/6258)
 
 * `qml.qchem.excitations` now optionally returns fermionic operators.
  [(#6171)](https://github.com/PennyLaneAI/pennylane/pull/6171)