Improved transpose operator's performance
Use specialized GPU kernels on tensors when the permutation can be reduced to {0,2,1}, {2,1,0} or {1,0}.
Change: 151147354
Loading
Please sign in to comment
Use specialized GPU kernels on tensors when the permutation can be reduced to {0,2,1}, {2,1,0} or {1,0}.
Change: 151147354