BFGS (Broyden–Fletcher–Goldfarb–Shanno) minimizer. More...
#include <bfgs.hpp>
Inheritance diagram for cpu_mlp::BFGS< V, M, Solver >:
Collaboration diagram for cpu_mlp::BFGS< V, M, Solver >:
Public Member Functions | |
| requires (UseDefaultSolver) | |
| requires (!UseDefaultSolver) | |
| void | setInitialHessian (const M &b) |
| Sets the initial approximate Hessian matrix. More... | |
| V | solve (V x, VecFun< V, double > &f, GradFun< V > &Gradient) override |
| Solves the optimization problem using BFGS method. More... | |
 Public Member Functions inherited from cpu_mlp::FullBatchMinimizer< V, M > | |
| virtual | ~FullBatchMinimizer ()=default |
| virtual void | setHessian (const HessFun< V, M > &) |
| Sets the Hessian function (for Second Order methods). More... | |
| template<auto LossFn, typename DataType > | |
| V | solve_with_enzyme (V x, DataType *data) |
| Helper to solve directly using an Enzyme-compatible raw function. More... | |
| void | setMaxIterations (int max_iters) |
| Sets the maximum number of iterations. More... | |
| void | setMaxLineIters (int max_line) |
| Sets the maximum number of iterations for the line search. More... | |
| void | setArmijoMaxIter (int max_armijo) |
| Sets the maximum iterations for Armijo condition check (alias for setMaxLineIters). More... | |
| void | setTolerance (double tol) |
| Sets the tolerance for convergence. More... | |
| unsigned int | iterations () const |
| Returns the number of iterations performed. More... | |
| double | tolerance () const |
| Returns the tolerance used. More... | |
| void | setRecorder (::IterationRecorder< CpuBackend > *recorder) |
| Attach a recorder for loss/grad history. More... | |
Protected Types | |
| using | SolverT = typename std::conditional< UseDefaultSolver, Solver, Solver & >::type |
Static Protected Attributes | |
| static constexpr bool | UseDefaultSolver = std::is_same_v<Solver, DefaultSolverT<M>> |
Additional Inherited Members | |
| Protected Member Functions inherited from cpu_mlp::FullBatchMinimizer< V, M > | |
| double | line_search (V x, V p, VecFun< V, double > &f, GradFun< V > &Gradient) |
| Backtracking Line Search satisfying Wolfe Conditions. More... | |
| Protected Attributes inherited from cpu_mlp::FullBatchMinimizer< V, M > | |
| unsigned int | _max_iters = 1000 |
| unsigned int | _iters = 0 |
| double | _tol = 1e-10 |
| ::IterationRecorder< CpuBackend > * | recorder_ = nullptr |
| Optional recorder for diagnostics. More... | |
| double | c1 = 1e-4 |
| double | c2 = 0.9 |
| double | rho = 0.5 |
| double | max_line_iters = 50 |
Detailed Description
template<typename V, typename M, typename Solver = DefaultSolverT<M>>
class cpu_mlp::BFGS< V, M, Solver >
BFGS (Broyden–Fletcher–Goldfarb–Shanno) minimizer.
Member Typedef Documentation
◆ SolverT
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
protected |
Member Function Documentation
◆ requires() [1/2]
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
inline |
◆ requires() [2/2]
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
inline |
◆ setInitialHessian()
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
inlinevirtual |
Sets the initial approximate Hessian matrix.
- Parameters
-
b Initial Hessian approximation (usually Identity).
Reimplemented from cpu_mlp::FullBatchMinimizer< V, M >.
◆ solve()
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
inlineoverridevirtual |
Solves the optimization problem using BFGS method.
- Parameters
-
x Initial guess for the parameters. f Objective function to minimize. Gradient Function to compute the gradient of f.
- Returns
- Optimized parameter vector.
Implements cpu_mlp::FullBatchMinimizer< V, M >.
Here is the call graph for this function:
Member Data Documentation
◆ UseDefaultSolver
template<typename V , typename M , typename Solver = DefaultSolverT<M>>
|
staticconstexprprotected |
The documentation for this class was generated from the following file:
- src/minimizer/bfgs.hpp
