gistfile1.txt

*****************************************
Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
*****************************************
(37131) -- Epoch 0 --
(37131) -- DistributedDataParallel --
(37130) -- Epoch 0 --
(37130) -- DistributedDataParallel --
(37131) device: cuda:1, forward size: torch.Size([4, 2])
(37130) device: cuda:0, forward size: torch.Size([4, 2])
tensor([[-0.0597, -0.4675],
        [ 0.1235, -0.1510],
        [ 0.3206, -1.0219],
        [ 1.7456, -1.2779]], device='cuda:1')
tensor([[ 0.0539,  0.6684],
        [-0.3144, -0.4963],
        [-0.3424, -1.4020],
        [ 1.4635, -0.7477]], device='cuda:0')
Parameter containing:
tensor([[-0.6661, -0.1387],
        [-0.3396, -0.1886]], device='cuda:0', requires_grad=True)
Parameter containing:
tensor([[-0.6661, -0.1387],
        [-0.3396, -0.1886]], device='cuda:1', requires_grad=True)
tensor([[-0.1286, -0.1444],
        [ 0.2783,  0.2004],
        [ 0.4226,  0.3807],
        [-0.8711, -0.3560]], device='cuda:0', grad_fn=<MmBackward>)
tensor([[ 0.1046,  0.1084],
        [-0.0613, -0.0135],
        [-0.0718,  0.0838],
        [-0.9855, -0.3518]], device='cuda:1', grad_fn=<MmBackward>)
(37130) device: cuda:0, loss: 0.6541517972946167
(37131) device: cuda:1, loss: 0.6011970639228821
(37131) (37130) tensor([[ 0.1275, -0.2672],
        [-0.1275,  0.2672]], device='cuda:0')
tensor([[ 0.1275, -0.2672],
        [-0.1275,  0.2672]], device='cuda:1')