When trying to implement a custom autograd.Function, I’m calling torch.from_numpy(x) inside the backward() definition of the custom function. However, this doesn’t work. Replacing torch.from_numpy(x) with torch.zeros(…) works. Copying over all values one-by-one from numpy array ‘x’ to the fresh torch.zeros(…) tensor works. What am I doing wrong? Also, the example for custom modules from numpy and scipy doesn’t work.
Could you provide details about how it doesn’t work (an error message), and could you provide a minimal code example that causes it?
There is nor error message, it just says “Segmentation fault (core dumped)”.
My own code tried to call torch.from_numpy(…) on a numpy array (in the definition of backward() of my custom function), instead of the torch.FloatTensor(…) in the documentation example for custom parameterized modules from numpy/scipy.
I also tried running the example in http://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html#parametrized-example and that gave the same segfault.
What version of pytorch are you using? (You can check with
I ran the example on pytorch 0.2 and master and couldn’t reproduce the segfault.
PyTorch version: 0.2.0_2
Numpy version: 1.13.1
Tried that example again, still giving segfault.
Could you try upgrading your pytorch and seeing if that helps? I’m running 0.2.0_4.
Upgraded to 0.2.0_4, same segfault.
Could you provide a minimal example to reproduce the segfault please so that we can look into it in more details locally.
adapted from the example in http://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html:
import torch print(torch.__version__) class DummyFunction(torch.autograd.Function): def forward(self, x): self.save_for_backward(x) y = x.numpy() + 1 return torch.from_numpy(y) def backward(self, y_grad): x = self.saved_tensors x_grad = y_grad.clone() x_grad = x_grad.numpy() + 1 return torch.from_numpy(x_grad) # return torch.FloatTensor(x_grad) input = torch.autograd.Variable(torch.rand(5), requires_grad=True) output = DummyFunction()(input) print(output) output.backward(torch.rand(5)) print(input.grad)
This code outputs:
0.2.0_4 Variable containing: 1.4786 1.8203 1.4139 1.2448 1.8730 [torch.FloatTensor of size 5] Segmentation fault (core dumped)
However, if I replace torch.from_numpy(…) with
ret = torch.zeros(x_grad.shape) retnp = ret.numpy() retnp += x_grad
and return ret, it works
Your original code works with my install from master.
How did you installed pytorch?
installed it from conda for python 2.7, no cuda (Linux)
asked a colleague to reproduce (Mac, no CUDA, pip-installed, 0.2.0_3), outputs
Segmentation Fault: 11
(edit: colleague ran example at http://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html#parametrized-example)
also setup a fresh amazon instance, only installed python, pip, ipython and ran:
pip install http://download.pytorch.org/whl/cu75/torch-0.2.0.post3-cp27-cp27mu-manylinux1_x86_64.whl
(Python 2.7, no CUDA, pip-installed 0.2.0_3 (didn’t install torchvision))
The dummy example I pasted above also gives segfault on the fresh machine.
Okay, I was able to repro the problem. Building from source (master) or the v0.3.0 branch makes the segfault go away on the machine I reproduced on. @purrfegt could you try installing from source on a machine and check if the segfault is still there?
running into problems with cmake, no time to resolve now.
Do you think the segfault will be resolved in prebuilt v0.3.0? Maybe this needs a test if there isn’t any yet.
I tested on v0.3.0 and the segfault isn’t there, so it’ll probably be fine
We’ll be testing pre-built v0.3.0 binaries on the tutorials so the segfault will be caught if it still exists.
ok thanks. Any estimate when v0.3 is arriving? I think I can live with my workaround for now, not using it in anything crucial.