W1102 22:11:21.695000 1 torch/distributed/run.py:851] 
W1102 22:11:21.695000 1 torch/distributed/run.py:851] *****************************************
W1102 22:11:21.695000 1 torch/distributed/run.py:851] 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. 
W1102 22:11:21.695000 1 torch/distributed/run.py:851] *****************************************
Using Device: cpu, Backend: gloo
Using Device: cpu, Backend: gloo
Distributed Training for WORLD_SIZE: 4, RANK: 0, LOCAL_RANK: 0
Distributed Training for WORLD_SIZE: 4, RANK: 1, LOCAL_RANK: 1
Using dataset from shared PVC at /mnt/shared
Using dataset from shared PVC at /mnt/shared
Train Epoch: 1 [0/60000 (0%)]	Loss: 2.301982
Train Epoch: 1 [1000/60000 (7%)]	Loss: 2.057940
Train Epoch: 1 [2000/60000 (13%)]	Loss: 2.035207
Train Epoch: 1 [3000/60000 (20%)]	Loss: 0.984869
Train Epoch: 1 [4000/60000 (27%)]	Loss: 0.889687
Train Epoch: 1 [5000/60000 (33%)]	Loss: 0.676185
Train Epoch: 1 [6000/60000 (40%)]	Loss: 0.716787
Train Epoch: 1 [7000/60000 (47%)]	Loss: 0.699388
Train Epoch: 1 [8000/60000 (53%)]	Loss: 0.508775
Train Epoch: 1 [9000/60000 (60%)]	Loss: 0.561437
Train Epoch: 1 [10000/60000 (67%)]	Loss: 0.410902
Train Epoch: 1 [11000/60000 (73%)]	Loss: 0.516288
Train Epoch: 1 [12000/60000 (80%)]	Loss: 0.491655
Train Epoch: 1 [13000/60000 (87%)]	Loss: 0.491751
Train Epoch: 1 [14000/60000 (93%)]	Loss: 0.364018
Train Epoch: 2 [0/60000 (0%)]	Loss: 0.514253
Train Epoch: 2 [1000/60000 (7%)]	Loss: 0.416691
Train Epoch: 2 [2000/60000 (13%)]	Loss: 0.389681
Train Epoch: 2 [3000/60000 (20%)]	Loss: 0.359761
Train Epoch: 2 [4000/60000 (27%)]	Loss: 0.396558
Train Epoch: 2 [5000/60000 (33%)]	Loss: 0.361857
Train Epoch: 2 [6000/60000 (40%)]	Loss: 0.516661
Train Epoch: 2 [7000/60000 (47%)]	Loss: 0.503106
Train Epoch: 2 [8000/60000 (53%)]	Loss: 0.315789
Train Epoch: 2 [9000/60000 (60%)]	Loss: 0.366495
Train Epoch: 2 [10000/60000 (67%)]	Loss: 0.326843
Train Epoch: 2 [11000/60000 (73%)]	Loss: 0.332871
Train Epoch: 2 [12000/60000 (80%)]	Loss: 0.314597
Train Epoch: 2 [13000/60000 (87%)]	Loss: 0.350471
Train Epoch: 2 [14000/60000 (93%)]	Loss: 0.281177
Training is finished