pygmtools.pytorch_backend.cao_fast_solver

pygmtools.pytorch_backend.cao_fast_solver(K, X, num_graph, num_node, max_iter, lambda_init, lambda_step, lambda_max, iter_boost)[source]

Pytorch implementation of CAO solver in fast config (mode=”pc”)

Parameters

  • K – affinity matrix, (m, m, n*n, n*n)

  • X – initial matching, (m, m, n, n)

  • num_graph – number of graphs, int

  • num_node – number of nodes, int

Returns

X, (m, m, n, n)