Closed sabetAI closed 4 years ago
Were you able to get to a resolution? I am getting the same error, trying to load a 700M edges Graph. The graph file is only 13GB and I am using a huge RAM machine.
node2vec caches intermediate results to produce the output in a reasonable time, which is very memory intensive, especially if there are many high degree nodes. A solution is to run the program on smaller subgraphs of the original graph.
I'm running node2vec on the wikipedia graph, but run into memory errors when node2vec tries to load the entire graph into memory. Why does the graph consume SO MUCH memory, and how can I avoid loading the entire thing into RAM all at once?