rogerwwww
commented
1 year ago 您好，可以试试neural solver里面的pca gm和cie，不过很可能需要你给训练标签做监督训练。
Hi there, you may try pygm.pca_gm and pygm.cie which are more memory efficient. However, it is likely that you need the labels to do supervised training.
各位大佬，我想问下pygmtool目前有什么可以匹配大图的方法呀，我想匹配的图有上千个节点，直接使用build_aff_mat会直接爆显存甚至内存
Guys, I would like to ask what method pygmtool currently has to match large graphs. The graph I want to match has thousands of nodes. Using build_aff_mat directly will lead to out of memory for GPU mem or even the RAM.