pytorch/pytorch
http://ift.tt/eA8V8J
Linear algebra operations such as svd() and symeig() only support a single matrix input (e.g., a matrix of size M x N ). I request to extend these functions to support running these operations on batches in parallel on a single GPU machine. That is, the input will take the form of B x M x N and the function ...
from Algebra
Linear algebra operations such as svd() and symeig() only support a single matrix input (e.g., a matrix of size M x N ). I request to extend these functions to support running these operations on batches in parallel on a single GPU machine. That is, the input will take the form of B x M x N and the function ...
from Algebra
