Search papers, labs, and topics across Lattice.
University of Bayreuth
1
0
3
2
Partitioning graphs using node embeddings, rather than the graph structure itself, unlocks 100x speedups in distributed GNN training.