Closed BlueSkyLT closed 1 year ago
I suppose there is a tensor that is not on GPU, can you guess which one?
The issue still existed.
def _non_adaptive_forward(self, x, y):
N, C, T, V = x.size()
for i in range(self.num_subset):
A1 = self.A[i]
A2 = x.view(N, C * T, V)
> z = self.conv_d[i](torch.matmul(A2, A1).view(N, C, T, V))
# A1 is on CPU, A2 is on GPU (same as `x`)
A temp fix could be
def _non_adaptive_forward(self, x, y):
_device = x.device
N, C, T, V = x.size()
for i in range(self.num_subset):
A1 = self.A[i].to(_device)
A2 = x.view(N, C * T, V)
z = self.conv_d[i](torch.matmul(A2, A1).view(N, C, T, V))
When running the test, all tests passed except for
Similar to #46 , if I force it on CPU then it will pass