• OS:
  Ubuntu 16.04

• PyTorch version:
  master, at commit aaa0288

• How you installed PyTorch (conda, pip, source):
  Using git and built using the dockerfile for cuda 9 and cudnn v7

• Python version: 3.6.4

• CUDA/cuDNN version:
  cuda 9.0 and cudnn 7

• GPU models and configuration:
  Geforce 1080

• GCC version (if compiling from source):
  gcc 5.4.0

• Script to reproduce the bug:
  the tutorial script for cifar10 classification: http://pytorch.org/tutorials/_downloads/cifar10_tutorial.py

At the end of the script execution, I have the following error message:
Error in atexit._run_exitfuncs: Traceback (most recent call last): File "/opt/conda/envs/pytorch-py3.6/lib/python3.6/multiprocessing/popen_fork.py", line 35, in poll pid, sts = os.waitpid(self.pid, flag) File "/opt/conda/envs/pytorch-py3.6/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 172, in handler _error_if_any_worker_fails() RuntimeError: DataLoader worker (pid 710) is killed by signal: Terminated. Exception ignored in: <bound method DataLoaderIter.__del__ of <torch.utils.data.dataloader.DataLoaderIter object at 0x7f296ab343c8>> Traceback (most recent call last): File "/opt/conda/envs/pytorch-py3.6/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 322, in __del__ File "/opt/conda/envs/pytorch-py3.6/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 310, in _shutdown_workers File "/opt/conda/envs/pytorch-py3.6/lib/python3.6/multiprocessing/queues.py", line 337, in get ImportError: sys.meta_path is None, Python is likely shutting down