Solver

Solvers assemble fields, constraints, and optional enforced-constraint pipelines into a scalar loss suitable for optimization.

For a conceptual overview (loss evaluation, enforced pipelines, training loop behavior), see Guides → Solvers and training.

Note

Key notes:

• Use FunctionalSolver to sum constraint losses into \(L=\sum_i \ell_i\).
• Use enforced constraint pipelines to enforce conditions by construction (no penalty term).