
=== Start adding workers ===
=> Add worker SGDMWorker(index=0, momentum=0.0)
=> Add worker SGDMWorker(index=1, momentum=0.0)
=> Add worker SGDMWorker(index=2, momentum=0.0)
=> Add worker SGDMWorker(index=3, momentum=0.0)
=> Add worker SGDMWorker(index=4, momentum=0.0)
=> Add worker SGDMWorker(index=5, momentum=0.0)
=> Add worker SGDMWorker(index=6, momentum=0.0)
=> Add worker SGDMWorker(index=7, momentum=0.0)
=> Add worker SGDMWorker(index=8, momentum=0.0)
=> Add worker SGDMWorker(index=9, momentum=0.0)
=> Add worker SGDMWorker(index=10, momentum=0.0)
=> Add worker ByzantineWorker(index=11)
=> Add worker ByzantineWorker(index=12)

=== Start adding graph ===
TwoCliquesWithByzantine(m=5,b=2)

Train epoch 1
[E 1B0  |    416/60000 (  1%) ] Loss: 2.3051 top1=  8.5227

=== Peeking data label distribution E1B0 ===
Worker 0 has targets: tensor([9, 9, 2, 5, 8], device='cuda:0')
Worker 1 has targets: tensor([3, 4, 1, 1, 9], device='cuda:0')
Worker 2 has targets: tensor([5, 0, 6, 6, 2], device='cuda:0')
Worker 3 has targets: tensor([4, 6, 5, 0, 1], device='cuda:0')
Worker 4 has targets: tensor([7, 8, 5, 1, 0], device='cuda:0')
Worker 5 has targets: tensor([4, 9, 4, 4, 7], device='cuda:0')
Worker 6 has targets: tensor([9, 9, 7, 4, 9], device='cuda:0')
Worker 7 has targets: tensor([6, 3, 4, 0, 2], device='cuda:0')
Worker 8 has targets: tensor([8, 1, 1, 0, 6], device='cuda:0')
Worker 9 has targets: tensor([8, 3, 8, 7, 8], device='cuda:0')
Worker 10 has targets: tensor([7, 2, 7, 0, 7], device='cuda:0')
Worker 11 has targets: tensor([4, 5, 3, 6, 7], device='cuda:0')
Worker 12 has targets: tensor([8, 0, 3, 7, 2], 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


[E 1B10 |   4576/60000 (  8%) ] Loss: 2.2960 top1= 11.6477

=== Log global consensus distance @ E1B10 ===
consensus_distance=0.000



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


[E 1B20 |   8736/60000 ( 15%) ] Loss: 2.2861 top1= 11.6477

=== Log global consensus distance @ E1B20 ===
consensus_distance=0.000



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



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.2851 top1=  9.9559


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.2806 top1= 12.3898


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.2804 top1= 12.3197

Train epoch 2
[E 2B0  |    416/60000 (  1%) ] Loss: 2.2721 top1= 13.3523

=== Log global consensus distance @ E2B0 ===
consensus_distance=0.003



=== Log clique consensus distance @ E2B0 ===
clique1_consensus_distance=0.000
clique2_consensus_distance=0.000


[E 2B10 |   4576/60000 (  8%) ] Loss: 2.2711 top1= 14.2045

=== Log global consensus distance @ E2B10 ===
consensus_distance=0.030



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


[E 2B20 |   8736/60000 ( 15%) ] Loss: 2.3071 top1= 15.6250

=== Log global consensus distance @ E2B20 ===
consensus_distance=0.274



=== Log clique consensus distance @ E2B20 ===
clique1_consensus_distance=0.012
clique2_consensus_distance=0.012



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.4862 top1= 19.2608


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.3522 top1= 19.2608


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.3526 top1= 19.2608

Train epoch 3
[E 3B0  |    416/60000 (  1%) ] Loss: 2.3813 top1= 18.4659

=== Log global consensus distance @ E3B0 ===
consensus_distance=2.505



=== Log clique consensus distance @ E3B0 ===
clique1_consensus_distance=0.110
clique2_consensus_distance=0.110


[E 3B10 |   4576/60000 (  8%) ] Loss: 2.7201 top1= 18.7500

=== Log global consensus distance @ E3B10 ===
consensus_distance=12.836



=== Log clique consensus distance @ E3B10 ===
clique1_consensus_distance=0.638
clique2_consensus_distance=0.638


[E 3B20 |   8736/60000 ( 15%) ] Loss: 3.3341 top1= 23.0114

=== Log global consensus distance @ E3B20 ===
consensus_distance=20.439



=== Log clique consensus distance @ E3B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.640



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.1294 top1= 33.2432


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=2.0169 top1= 38.2712


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=2.0237 top1= 38.7119

Train epoch 4
[E 4B0  |    416/60000 (  1%) ] Loss: 3.0601 top1= 17.6136

=== Log global consensus distance @ E4B0 ===
consensus_distance=29.880



=== Log clique consensus distance @ E4B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.640


[E 4B10 |   4576/60000 (  8%) ] Loss: 2.9777 top1= 22.4432

=== Log global consensus distance @ E4B10 ===
consensus_distance=41.158



=== Log clique consensus distance @ E4B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.641


[E 4B20 |   8736/60000 ( 15%) ] Loss: 3.7521 top1= 24.1477

=== Log global consensus distance @ E4B20 ===
consensus_distance=54.267



=== Log clique consensus distance @ E4B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.641



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.8725 top1= 41.4663


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.5849 top1= 47.2456


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.5928 top1= 47.1454

Train epoch 5
[E 5B0  |    416/60000 (  1%) ] Loss: 3.5018 top1= 24.4318

=== Log global consensus distance @ E5B0 ===
consensus_distance=69.213



=== Log clique consensus distance @ E5B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.642


[E 5B10 |   4576/60000 (  8%) ] Loss: 3.3149 top1= 30.6818

=== Log global consensus distance @ E5B10 ===
consensus_distance=85.995



=== Log clique consensus distance @ E5B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.643


[E 5B20 |   8736/60000 ( 15%) ] Loss: 4.1286 top1= 32.1023

=== Log global consensus distance @ E5B20 ===
consensus_distance=104.608



=== Log clique consensus distance @ E5B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.643



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.7173 top1= 44.5513


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4176 top1= 51.0917


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.4557 top1= 48.2071

Train epoch 6
[E 6B0  |    416/60000 (  1%) ] Loss: 3.5116 top1= 31.2500

=== Log global consensus distance @ E6B0 ===
consensus_distance=125.061



=== Log clique consensus distance @ E6B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.644


[E 6B10 |   4576/60000 (  8%) ] Loss: 3.2844 top1= 34.6591

=== Log global consensus distance @ E6B10 ===
consensus_distance=147.349



=== Log clique consensus distance @ E6B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.646


[E 6B20 |   8736/60000 ( 15%) ] Loss: 11.9198 top1= 31.5341

=== Log global consensus distance @ E6B20 ===
consensus_distance=171.480



=== Log clique consensus distance @ E6B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.647



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.5309 top1= 50.2304


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.3733 top1= 53.8862


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.3587 top1= 54.1767

Train epoch 7
[E 7B0  |    416/60000 (  1%) ] Loss: 4.5763 top1= 33.8068

=== Log global consensus distance @ E7B0 ===
consensus_distance=197.364



=== Log clique consensus distance @ E7B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.648


[E 7B10 |   4576/60000 (  8%) ] Loss: 2.9662 top1= 34.6591

=== Log global consensus distance @ E7B10 ===
consensus_distance=225.159



=== Log clique consensus distance @ E7B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.649


[E 7B20 |   8736/60000 ( 15%) ] Loss: 3.6870 top1= 34.9432

=== Log global consensus distance @ E7B20 ===
consensus_distance=254.784



=== Log clique consensus distance @ E7B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.650



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.8926 top1= 55.2484


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.3317 top1= 56.5204


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.3358 top1= 55.9395

Train epoch 8
[E 8B0  |    416/60000 (  1%) ] Loss: 3.2890 top1= 36.6477

=== Log global consensus distance @ E8B0 ===
consensus_distance=286.250



=== Log clique consensus distance @ E8B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.651


[E 8B10 |   4576/60000 (  8%) ] Loss: 2.8973 top1= 36.6477

=== Log global consensus distance @ E8B10 ===
consensus_distance=319.554



=== Log clique consensus distance @ E8B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.652


[E 8B20 |   8736/60000 ( 15%) ] Loss: 3.4223 top1= 36.9318

=== Log global consensus distance @ E8B20 ===
consensus_distance=354.693



=== Log clique consensus distance @ E8B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.653



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.6143 top1= 57.0212


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.3579 top1= 52.6142


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.3705 top1= 52.7244

Train epoch 9
[E 9B0  |    416/60000 (  1%) ] Loss: 3.2518 top1= 38.9205

=== Log global consensus distance @ E9B0 ===
consensus_distance=391.664



=== Log clique consensus distance @ E9B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.654


[E 9B10 |   4576/60000 (  8%) ] Loss: 2.8277 top1= 38.0682

=== Log global consensus distance @ E9B10 ===
consensus_distance=430.477



=== Log clique consensus distance @ E9B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.656


[E 9B20 |   8736/60000 ( 15%) ] Loss: 3.5157 top1= 36.9318

=== Log global consensus distance @ E9B20 ===
consensus_distance=471.119



=== Log clique consensus distance @ E9B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.657



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.7455 top1= 52.8746


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4161 top1= 47.5461


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.4225 top1= 48.3974

Train epoch 10
[E10B0  |    416/60000 (  1%) ] Loss: 3.5481 top1= 40.3409

=== Log global consensus distance @ E10B0 ===
consensus_distance=513.601



=== Log clique consensus distance @ E10B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.657


[E10B10 |   4576/60000 (  8%) ] Loss: 3.0048 top1= 35.2273

=== Log global consensus distance @ E10B10 ===
consensus_distance=557.918



=== Log clique consensus distance @ E10B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.659


[E10B20 |   8736/60000 ( 15%) ] Loss: 3.5276 top1= 36.0795

=== Log global consensus distance @ E10B20 ===
consensus_distance=604.072



=== Log clique consensus distance @ E10B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.659



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=1.9020 top1= 51.0417


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4587 top1= 45.8734


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.4509 top1= 47.5962

Train epoch 11
[E11B0  |    416/60000 (  1%) ] Loss: 3.4062 top1= 39.2045

=== Log global consensus distance @ E11B0 ===
consensus_distance=652.060



=== Log clique consensus distance @ E11B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.660


[E11B10 |   4576/60000 (  8%) ] Loss: 3.0690 top1= 35.2273

=== Log global consensus distance @ E11B10 ===
consensus_distance=701.883



=== Log clique consensus distance @ E11B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.661


[E11B20 |   8736/60000 ( 15%) ] Loss: 3.4432 top1= 35.5114

=== Log global consensus distance @ E11B20 ===
consensus_distance=753.546



=== Log clique consensus distance @ E11B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.662



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.0713 top1= 46.2440


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4970 top1= 44.4912


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.4816 top1= 46.8850

Train epoch 12
[E12B0  |    416/60000 (  1%) ] Loss: 3.8638 top1= 39.7727

=== Log global consensus distance @ E12B0 ===
consensus_distance=807.041



=== Log clique consensus distance @ E12B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.663


[E12B10 |   4576/60000 (  8%) ] Loss: 3.2384 top1= 33.8068

=== Log global consensus distance @ E12B10 ===
consensus_distance=862.368



=== Log clique consensus distance @ E12B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.663


[E12B20 |   8736/60000 ( 15%) ] Loss: 3.8174 top1= 34.3750

=== Log global consensus distance @ E12B20 ===
consensus_distance=919.542



=== Log clique consensus distance @ E12B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.664



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.1337 top1= 46.9551


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4941 top1= 44.4812


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.4818 top1= 46.1538

Train epoch 13
[E13B0  |    416/60000 (  1%) ] Loss: 3.7710 top1= 37.7841

=== Log global consensus distance @ E13B0 ===
consensus_distance=978.546



=== Log clique consensus distance @ E13B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.666


[E13B10 |   4576/60000 (  8%) ] Loss: 3.3216 top1= 35.7955

=== Log global consensus distance @ E13B10 ===
consensus_distance=1039.385



=== Log clique consensus distance @ E13B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.667


[E13B20 |   8736/60000 ( 15%) ] Loss: 3.7854 top1= 33.2386

=== Log global consensus distance @ E13B20 ===
consensus_distance=1102.065



=== Log clique consensus distance @ E13B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.668



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.3784 top1= 45.3025


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.4963 top1= 45.0621


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.5172 top1= 45.0821

Train epoch 14
[E14B0  |    416/60000 (  1%) ] Loss: 4.3632 top1= 38.3523

=== Log global consensus distance @ E14B0 ===
consensus_distance=1166.577



=== Log clique consensus distance @ E14B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.670


[E14B10 |   4576/60000 (  8%) ] Loss: 3.5623 top1= 36.0795

=== Log global consensus distance @ E14B10 ===
consensus_distance=1232.924



=== Log clique consensus distance @ E14B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.671


[E14B20 |   8736/60000 ( 15%) ] Loss: 3.9302 top1= 34.3750

=== Log global consensus distance @ E14B20 ===
consensus_distance=1301.105



=== Log clique consensus distance @ E14B20 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.672



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=2.2412 top1= 49.5292


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=1.5546 top1= 44.1306


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=1.5619 top1= 43.4696

Train epoch 15
[E15B0  |    416/60000 (  1%) ] Loss: 4.7446 top1= 37.5000

=== Log global consensus distance @ E15B0 ===
consensus_distance=1371.127



=== Log clique consensus distance @ E15B0 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.673


[E15B10 |   4576/60000 (  8%) ] Loss: 4.5397 top1= 34.6591

=== Log global consensus distance @ E15B10 ===
consensus_distance=1442.991



=== Log clique consensus distance @ E15B10 ===
clique1_consensus_distance=0.640
clique2_consensus_distance=0.674


[E15B20 |   8736/60000 ( 15%) ] Loss: 81839107.1607 top1= 27.2727

=== Log global consensus distance @ E15B20 ===
consensus_distance=508849152.000



=== Log clique consensus distance @ E15B20 ===
clique1_consensus_distance=0.448
clique2_consensus_distance=0.489



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 16
[E16B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E16B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E16B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E16B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E16B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E16B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E16B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E16B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E16B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 17
[E17B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E17B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E17B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E17B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E17B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E17B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E17B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E17B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E17B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 18
[E18B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E18B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E18B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E18B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E18B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E18B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E18B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E18B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E18B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 19
[E19B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E19B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E19B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E19B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E19B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E19B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E19B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E19B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E19B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 20
[E20B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E20B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E20B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E20B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E20B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E20B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E20B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E20B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E20B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 21
[E21B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E21B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E21B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E21B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E21B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E21B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E21B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E21B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E21B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 22
[E22B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E22B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E22B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E22B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E22B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E22B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E22B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E22B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E22B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 23
[E23B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E23B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E23B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E23B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E23B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E23B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E23B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E23B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E23B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 24
[E24B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E24B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E24B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E24B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E24B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E24B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E24B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E24B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E24B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 25
[E25B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E25B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E25B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E25B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E25B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E25B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E25B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E25B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E25B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 26
[E26B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E26B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E26B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E26B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E26B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E26B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E26B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E26B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E26B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 27
[E27B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E27B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E27B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E27B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E27B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E27B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E27B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E27B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E27B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 28
[E28B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E28B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E28B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E28B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E28B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E28B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E28B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E28B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E28B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 29
[E29B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E29B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E29B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E29B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E29B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E29B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E29B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E29B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E29B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 30
[E30B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E30B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E30B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E30B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E30B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E30B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E30B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E30B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E30B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 31
[E31B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E31B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E31B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E31B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E31B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E31B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E31B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E31B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E31B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 32
[E32B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E32B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E32B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E32B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E32B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E32B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E32B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E32B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E32B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 33
[E33B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E33B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E33B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E33B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E33B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E33B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E33B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E33B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E33B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 34
[E34B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E34B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E34B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E34B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E34B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E34B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E34B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E34B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E34B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 35
[E35B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E35B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E35B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E35B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E35B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E35B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E35B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E35B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E35B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 36
[E36B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E36B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E36B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E36B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E36B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E36B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E36B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E36B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E36B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 37
[E37B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E37B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E37B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E37B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E37B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E37B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E37B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E37B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E37B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 38
[E38B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E38B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E38B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E38B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E38B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E38B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E38B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E38B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E38B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 39
[E39B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E39B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E39B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E39B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E39B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E39B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E39B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E39B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E39B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

Train epoch 40
[E40B0  |    416/60000 (  1%) ] Loss: nan top1=  7.3864

=== Log global consensus distance @ E40B0 ===
consensus_distance=nan



=== Log clique consensus distance @ E40B0 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E40B10 |   4576/60000 (  8%) ] Loss: nan top1=  9.9432

=== Log global consensus distance @ E40B10 ===
consensus_distance=nan



=== Log clique consensus distance @ E40B10 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan


[E40B20 |   8736/60000 ( 15%) ] Loss: nan top1=  9.6591

=== Log global consensus distance @ E40B20 ===
consensus_distance=nan



=== Log clique consensus distance @ E40B20 ===
clique1_consensus_distance=nan
clique2_consensus_distance=nan



=> Averaged model (Global Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique1 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057


=> Averaged model (Clique2 Average Validation Accuracy) | Eval Loss=nan top1=  9.8057

