Exception in DataLoader function

1

Hello, I am training my deep learning model and keep receiving this warning/error. I am not sure what it means but it seems to point to the DataLoader function. The code continues to run even with these errors.

Can someone tell me what this means and how I can fix it ?

Exception ignored in: <bound method _DataLoaderIter.__del__ of <torch.utils.data.dataloader._DataLoaderIter object at 0x7f47480d6f98>>
Traceback (most recent call last):
  File "/home/kong/anaconda3/envs/social/lib/python3.5/site-packages/torch/utils/data/dataloader.py", line 349, in __del__
    self._shutdown_workers()
  File "/home/kong/anaconda3/envs/social/lib/python3.5/site-packages/torch/utils/data/dataloader.py", line 328, in _shutdown_workers
    self.worker_result_queue.get()
  File "/home/kong/anaconda3/envs/social/lib/python3.5/multiprocessing/queues.py", line 337, in get
    return ForkingPickler.loads(res)
  File "/home/kong/anaconda3/envs/social/lib/python3.5/site-packages/torch/multiprocessing/reductions.py", line 70, in rebuild_storage_fd
    fd = df.detach()
  File "/home/kong/anaconda3/envs/social/lib/python3.5/multiprocessing/resource_sharer.py", line 58, in detach
    return reduction.recv_handle(conn)
  File "/home/kong/anaconda3/envs/social/lib/python3.5/multiprocessing/reduction.py", line 181, in recv_handle
    return recvfds(s, 1)[0]
  File "/home/kong/anaconda3/envs/social/lib/python3.5/multiprocessing/reduction.py", line 152, in recvfds
    msg, ancdata, flags, addr = sock.recvmsg(1, socket.CMSG_LEN(bytes_size))
ConnectionResetError: [Errno 104] Connection reset by peer
2

no one knows what the error means ?

3

If could be related to this issue, which was recently fixed in this PR.
As it’s quite new, you could try to build PyTorch from source and check, if your issue still occurs.
You can find the build instructions here.
Let me know, if you get stuck.

1 Like