Implements a message passing interface (MPI) wrapper that makes it easy to do massively parallel computations inside the Torch deep-learning framework.
Primary LanguageC++OtherNOASSERTION
No issues in this repository yet.