refac: build on LuxCore

[MartinuzziFrancesco]

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the SciML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

This PR refactors ReservoirComputing.jl to build on LuxCore. Closes #206, #202, and #191

[MartinuzziFrancesco]

Some naming things to decide:

• signature fro train! either
  • train!(rc, train_data, target_data, ps, st, sr::StandardRidge=StandardRidge(0.0)) (current)
  • train!(rc, train_data, target_data, sr::StandardRidge=StandardRidge(0.0), ps, st)
• name of the collect states function (there a problem with the states nomenclature too, considering Lux as st as states)
  • collectstates (current)
  • collect_reservoirstates
  • collect_reservoir_states
• collection layer
  • Collect() (current)
  • CollectLayer()
• readout layer
  • Readout()
  • LinearReadout() (current)
• train! or fit!?

[MartinuzziFrancesco]

Not really sure about the Downgrade failure here, if I run import Pkg; Pkg.test(;coverage=true, julia_args=["--check-bounds=yes", "--compiled-modules=yes", "--depwarn=yes"], force_latest_compatible_version=false, allow_reresolve=false) in Julia 1.10.10 locally everything is fine

[MartinuzziFrancesco]

This should be more or less ready to go, there are some things I need to smooth out but its current state it's far, far better compared to the current version 0.11.4. Pending Downgrade solution I would merge this and improve eventual minor issues directly on main. After merging I will need to address some minor init breaking changes before tagging 0.12, which is essentially a v1.0 candidate

ping @ChrisRackauckas