RuntimeError in default main experiment

[trdavidson]

Hi Rianne,

I'm trying to run the default experiment on cpu with a small latent space dimension (z=5):

python main_experiment.py -d mnist --flow no_flow -nc --z_size 5

Which unfortunately gives the following error:

Traceback (most recent call last):
  File "main_experiment.py", line 278, in <module>
    run(args, kwargs)
  File "main_experiment.py", line 189, in run
    tr_loss = train(epoch, train_loader, model, optimizer, args)
  File ".../sylvester-flows/optimization/training.py", line 39, in train
    loss.backward()
  File "//anaconda/envs/dl/lib/python3.6/site-packages/torch/tensor.py", line 102, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph)
  File "//anaconda/envs/dl/lib/python3.6/site-packages/torch/autograd/__init__.py", line 90, in backward
    allow_unreachable=True)  # allow_unreachable flag
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation

I am using PyTorch version 1.0.0 and did not modify the code.

[riannevdberg]

Hi Tim,

I've just checked and for pytorch 0.3.0 this error does not occur either on cpu or gpu.

As described in the README file under requirements, the code won't work with pytorch 1.0.0 without making at least some changes to the loss function (particularly nn.BCELoss). Last time I checked it will silently ignore the fact that default flags have been changed for later versions, and throw errors in other parts of the code that seem unrelated. I currently don't have time to change the code to be compatible with pytorch 1.0.0, but I will hopefully find some time soon.