x = torch.tensor([0.2,0.3,0.4,0.5],device=device).float()
y = torch.tensor(np.exp([0.2,0.3,0.4,0.5]),device=device).float()
for i in range(1024):
Interp1d()(x, y, torch.tensor([0.2,0.3,1.],device=device))
After iteration 1023 I get the following error message:
/opt/conda/lib/python3.7/site-packages/torchinterp1d/interp1d.py in forward(ctx, x, y, xnew, out)
113 # must be inserted to preserve order. And we want the index of the
114 # preceeding value.
--> 115 ind -= 1
116 # we clamp the index, because the number of intervals is x.shape-1,
117 # and the left neighbour should hence be at most number of intervals
RuntimeError: CUDA error: an illegal memory access was encountered
Example:
After iteration 1023 I get the following error message:
Thank you for the package and your time!