torch.symeig
def forward(self,x):
x = self.fc1(x)
x = get_x(x)
def get_x(x):
#using x make matrix
for j in range(total_batch):
evals[j],evecs[j] = torch.symeig(sim_total[j],eigenvectors=True)
Is there a way the task should be processed simultaneously on the GPU?
i want to simultaneously only torch.symeig part !!