Li Zeng
Li Zeng
when I want to login, there is some special characters in the password. It fails.
Does byteps support GNN training? Or graph computing?
请问现在graph-learn是否支持spark分布式训练?多机多GPU分布式训练? 是否支持TransformerConv,KGCN等较新的算法? egograph是否可以对接pyg,或者像subgraph那样通过GSL转换数据集从而支持pyg? 训练/预测时各个点是独立训练的么,除了需要共享参数? 每个点维护自身的k跳邻居信息,这种做法是否会导致较大的内存开销?(是否有量化的统计数据) PS: graphscope中GLE就对应graph-learn,这两边有区别么?会实时同步?
If you have any question,you can ask questions here. And you also can ask questions there: [http://www.gstore.cn/pcsite/index.html#/medical](http://www.gstore.cn/pcsite/index.html#/medical) (Currently, only for Chinese)
Which framework does Grape use for distributed computing? Does it support encryption, if not, hwo can we implement encryption or use other rpc framework like grpc?
in the ending of split function in ne.cpp, the edge cut is transformed to vertex cut. Is the transformed partition still good? Is the transformation idea from any paper?
And the performance may be something different?