
=== Start adding workers ===
=> Add worker SGDMWorker(index=0, momentum=0.9)
=> Add worker SGDMWorker(index=1, momentum=0.9)
=> Add worker SGDMWorker(index=2, momentum=0.9)
=> Add worker SGDMWorker(index=3, momentum=0.9)
=> Add worker SGDMWorker(index=4, momentum=0.9)
=> Add worker SGDMWorker(index=5, momentum=0.9)
=> Add worker SGDMWorker(index=6, momentum=0.9)
=> Add worker SGDMWorker(index=7, momentum=0.9)
=> Add worker SGDMWorker(index=8, momentum=0.9)
=> Add worker SGDMWorker(index=9, momentum=0.9)
=> Add worker SGDMWorker(index=10, momentum=0.9)
=> Add worker ByzantineWorker(index=11)

=== Start adding graph ===
TwoCliquesWithByzantine(m=5,b=1)

Train epoch 1
[E 1B0  |    384/60000 (  1%) ] Loss: 2.3142 top1=  7.6705

=== Peeking data label distribution E1B0 ===
Worker 0 has targets: tensor([0, 0, 0, 0, 0], device='cuda:0')
Worker 1 has targets: tensor([1, 1, 1, 0, 1], device='cuda:0')
Worker 2 has targets: tensor([1, 1, 2, 1, 1], device='cuda:0')
Worker 3 has targets: tensor([2, 2, 3, 2, 2], device='cuda:0')
Worker 4 has targets: tensor([3, 3, 3, 3, 3], device='cuda:0')
Worker 5 has targets: tensor([4, 4, 4, 4, 4], device='cuda:0')
Worker 6 has targets: tensor([5, 5, 5, 5, 5], device='cuda:0')
Worker 7 has targets: tensor([6, 6, 6, 5, 6], device='cuda:0')
Worker 8 has targets: tensor([6, 7, 7, 6, 6], device='cuda:0')
Worker 9 has targets: tensor([7, 7, 8, 7, 7], device='cuda:0')
Worker 10 has targets: tensor([8, 8, 9, 8, 8], device='cuda:0')
Worker 11 has targets: tensor([9, 9, 9, 9, 9], device='cuda:0')



=== Log global consensus distance @ E1B0 ===
consensus_distance=0.000



=== Log clique consensus distance @ E1B0 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.000



=== Log mixing matrix @ E1B0 ===
[[0.167 0.167 0.167 0.167 0.167 0.    0.    0.    0.    0.    0.167 0.   ]
 [0.167 0.233 0.2   0.2   0.2   0.    0.    0.    0.    0.    0.    0.   ]
 [0.167 0.2   0.233 0.2   0.2   0.    0.    0.    0.    0.    0.    0.   ]
 [0.167 0.2   0.2   0.233 0.2   0.    0.    0.    0.    0.    0.    0.   ]
 [0.167 0.2   0.2   0.2   0.233 0.    0.    0.    0.    0.    0.    0.   ]
 [0.    0.    0.    0.    0.    0.233 0.2   0.2   0.2   0.167 0.    0.   ]
 [0.    0.    0.    0.    0.    0.2   0.233 0.2   0.2   0.167 0.    0.   ]
 [0.    0.    0.    0.    0.    0.2   0.2   0.233 0.2   0.167 0.    0.   ]
 [0.    0.    0.    0.    0.    0.2   0.2   0.2   0.233 0.167 0.    0.   ]
 [0.    0.    0.    0.    0.    0.167 0.167 0.167 0.167 0.167 0.167 0.   ]
 [0.167 0.    0.    0.    0.    0.    0.    0.    0.    0.167 0.417 0.25 ]
 [0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.25  0.75 ]]


[E 1B10 |   4224/60000 (  7%) ] Loss: 2.0681 top1= 42.0455

=== Log global consensus distance @ E1B10 ===
consensus_distance=0.008



=== Log clique consensus distance @ E1B10 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.031


[E 1B20 |   8064/60000 ( 13%) ] Loss: 1.6681 top1= 51.9886

=== Log global consensus distance @ E1B20 ===
consensus_distance=0.034



=== Log clique consensus distance @ E1B20 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.147



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.0841 top1= 33.6238


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.6715 top1= 38.7420


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.4264 top1= 33.7340

Train epoch 2
[E 2B0  |    384/60000 (  1%) ] Loss: 1.2072 top1= 69.0341

=== Log global consensus distance @ E2B0 ===
consensus_distance=0.066



=== Log clique consensus distance @ E2B0 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.282


[E 2B10 |   4224/60000 (  7%) ] Loss: 0.9086 top1= 76.7045

=== Log global consensus distance @ E2B10 ===
consensus_distance=0.085



=== Log clique consensus distance @ E2B10 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.366


[E 2B20 |   8064/60000 ( 13%) ] Loss: 0.6728 top1= 84.6591

=== Log global consensus distance @ E2B20 ===
consensus_distance=0.092



=== Log clique consensus distance @ E2B20 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.397



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.4727 top1= 66.1358


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.7258 top1= 39.3630


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.7133 top1= 35.6871

Train epoch 3
[E 3B0  |    384/60000 (  1%) ] Loss: 0.4442 top1= 90.6250

=== Log global consensus distance @ E3B0 ===
consensus_distance=0.092



=== Log clique consensus distance @ E3B0 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.398


[E 3B10 |   4224/60000 (  7%) ] Loss: 0.4001 top1= 90.0568

=== Log global consensus distance @ E3B10 ===
consensus_distance=0.084



=== Log clique consensus distance @ E3B10 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.365


[E 3B20 |   8064/60000 ( 13%) ] Loss: 0.3676 top1= 89.2045

=== Log global consensus distance @ E3B20 ===
consensus_distance=0.074



=== Log clique consensus distance @ E3B20 ===
clique1_consensus_distance=0.001
clique2_consensus_distance=0.320



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.0387 top1= 73.7981


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.1692 top1= 40.7552


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.4190 top1= 40.2644

Train epoch 4
[E 4B0  |    384/60000 (  1%) ] Loss: 0.2493 top1= 93.4659

=== Log global consensus distance @ E4B0 ===
consensus_distance=0.064



=== Log clique consensus distance @ E4B0 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.277


[E 4B10 |   4224/60000 (  7%) ] Loss: 0.2821 top1= 93.1818

=== Log global consensus distance @ E4B10 ===
consensus_distance=0.055



=== Log clique consensus distance @ E4B10 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.238


[E 4B20 |   8064/60000 ( 13%) ] Loss: 0.2753 top1= 92.6136

=== Log global consensus distance @ E4B20 ===
consensus_distance=0.047



=== Log clique consensus distance @ E4B20 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.203



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=0.8414 top1= 75.8914


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.8908 top1= 46.0437


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.1796 top1= 43.1591

Train epoch 5
[E 5B0  |    384/60000 (  1%) ] Loss: 0.1923 top1= 94.6023

=== Log global consensus distance @ E5B0 ===
consensus_distance=0.041



=== Log clique consensus distance @ E5B0 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.177


[E 5B10 |   4224/60000 (  7%) ] Loss: 0.2340 top1= 94.0341

=== Log global consensus distance @ E5B10 ===
consensus_distance=0.036



=== Log clique consensus distance @ E5B10 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.156


[E 5B20 |   8064/60000 ( 13%) ] Loss: 0.2295 top1= 92.8977

=== Log global consensus distance @ E5B20 ===
consensus_distance=0.032



=== Log clique consensus distance @ E5B20 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.138



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=0.7307 top1= 77.2035


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.7164 top1= 49.8998

