Template Function sequant::opt::detail::reconstructed_batched_peak

• Defined in File cost_model.hpp

Function Documentation

template<typename TIdxs, typename IdxToSz>
double sequant::opt::detail::reconstructed_batched_peak(TensorNetwork const &network, TIdxs const &tidxs, IdxToSz &&idxsz, std::function<bool(Index const&)> const &is_batchable, std::function<std::size_t(Index const&)> const &batch_target_size, std::function<bool(Tensor const&)> const &is_volatile_leaf)

Independent memory-simulation recomputation of the chosen batched reconstruction’s model-A peak. Builds PeakBatchedModel, runs solve_single_term for the back-pointer table, and recomputes the subtree peak by direct memory simulation (NOT the DP’s max/+ formula). Must EQUAL peak_cost_batched; a mismatch signals a DP/reconstruction bug.