Search papers, labs, and topics across Lattice.
1
0
2
Achieve faster, Byzantine-robust distributed learning by combining double momentum with variance reduction, eliminating the need for large batch sizes.
