hello
hello
Labels

📌S Retain class distribution for seed 3:
Class 0: 4500
Class 1: 4500
Class 2: 4500
Class 3: 4500
Class 4: 4500
Class 5: 4500
Class 6: 4500
Class 7: 4500
Class 8: 4500
Class 9: 4500

📌S Forget class distribution for seed 3:
Class 0: 500
Class 1: 500
Class 2: 500
Class 3: 500
Class 4: 500
Class 5: 500
Class 6: 500
Class 7: 500
Class 8: 500
Class 9: 500
78090990

📊 Updated class distribution:
Retain set:
  Class 0: 4625
  Class 1: 4625
  Class 2: 4625
  Class 3: 4625
  Class 4: 4625
  Class 5: 4625
  Class 6: 4625
  Class 7: 4625
  Class 8: 4625
  Class 9: 4625
Forget set:
  Class 0: 375
  Class 1: 375
  Class 2: 375
  Class 3: 375
  Class 4: 375
  Class 5: 375
  Class 6: 375
  Class 7: 375
  Class 8: 375
  Class 9: 375
hello
hello
⚠️ Warning: Retain train loader may not be shuffled.
Training Epoch: 1 [256/46250]	Loss: 2.3222	LR: 0.000000
Training Epoch: 1 [512/46250]	Loss: 2.3548	LR: 0.000552
Training Epoch: 1 [768/46250]	Loss: 2.3035	LR: 0.001105
Training Epoch: 1 [1024/46250]	Loss: 2.3243	LR: 0.001657
Training Epoch: 1 [1280/46250]	Loss: 2.2841	LR: 0.002210
Training Epoch: 1 [1536/46250]	Loss: 2.2512	LR: 0.002762
Training Epoch: 1 [1792/46250]	Loss: 2.2518	LR: 0.003315
Training Epoch: 1 [2048/46250]	Loss: 2.2245	LR: 0.003867
Training Epoch: 1 [2304/46250]	Loss: 2.1813	LR: 0.004420
Training Epoch: 1 [2560/46250]	Loss: 2.1777	LR: 0.004972
Training Epoch: 1 [2816/46250]	Loss: 2.1800	LR: 0.005525
Training Epoch: 1 [3072/46250]	Loss: 2.0327	LR: 0.006077
Training Epoch: 1 [3328/46250]	Loss: 2.0592	LR: 0.006630
Training Epoch: 1 [3584/46250]	Loss: 2.0126	LR: 0.007182
Training Epoch: 1 [3840/46250]	Loss: 1.9534	LR: 0.007735
Training Epoch: 1 [4096/46250]	Loss: 1.9302	LR: 0.008287
Training Epoch: 1 [4352/46250]	Loss: 1.9196	LR: 0.008840
Training Epoch: 1 [4608/46250]	Loss: 1.8084	LR: 0.009392
Training Epoch: 1 [4864/46250]	Loss: 1.8314	LR: 0.009945
Training Epoch: 1 [5120/46250]	Loss: 1.7595	LR: 0.010497
Training Epoch: 1 [5376/46250]	Loss: 1.8600	LR: 0.011050
Training Epoch: 1 [5632/46250]	Loss: 1.8732	LR: 0.011602
Training Epoch: 1 [5888/46250]	Loss: 1.7738	LR: 0.012155
Training Epoch: 1 [6144/46250]	Loss: 1.7684	LR: 0.012707
Training Epoch: 1 [6400/46250]	Loss: 1.7622	LR: 0.013260
Training Epoch: 1 [6656/46250]	Loss: 1.8515	LR: 0.013812
Training Epoch: 1 [6912/46250]	Loss: 1.6687	LR: 0.014365
Training Epoch: 1 [7168/46250]	Loss: 1.7382	LR: 0.014917
Training Epoch: 1 [7424/46250]	Loss: 1.6141	LR: 0.015470
Training Epoch: 1 [7680/46250]	Loss: 1.7155	LR: 0.016022
Training Epoch: 1 [7936/46250]	Loss: 1.6387	LR: 0.016575
Training Epoch: 1 [8192/46250]	Loss: 1.7568	LR: 0.017127
Training Epoch: 1 [8448/46250]	Loss: 1.6740	LR: 0.017680
Training Epoch: 1 [8704/46250]	Loss: 1.7728	LR: 0.018232
Training Epoch: 1 [8960/46250]	Loss: 1.7022	LR: 0.018785
Training Epoch: 1 [9216/46250]	Loss: 1.6002	LR: 0.019337
Training Epoch: 1 [9472/46250]	Loss: 1.6765	LR: 0.019890
Training Epoch: 1 [9728/46250]	Loss: 1.7109	LR: 0.020442
Training Epoch: 1 [9984/46250]	Loss: 1.6271	LR: 0.020994
Training Epoch: 1 [10240/46250]	Loss: 1.6610	LR: 0.021547
Training Epoch: 1 [10496/46250]	Loss: 1.8135	LR: 0.022099
Training Epoch: 1 [10752/46250]	Loss: 1.6435	LR: 0.022652
Training Epoch: 1 [11008/46250]	Loss: 1.5706	LR: 0.023204
Training Epoch: 1 [11264/46250]	Loss: 1.6812	LR: 0.023757
Training Epoch: 1 [11520/46250]	Loss: 1.7318	LR: 0.024309
Training Epoch: 1 [11776/46250]	Loss: 1.7305	LR: 0.024862
Training Epoch: 1 [12032/46250]	Loss: 1.5268	LR: 0.025414
Training Epoch: 1 [12288/46250]	Loss: 1.7765	LR: 0.025967
Training Epoch: 1 [12544/46250]	Loss: 1.6020	LR: 0.026519
Training Epoch: 1 [12800/46250]	Loss: 1.8469	LR: 0.027072
Training Epoch: 1 [13056/46250]	Loss: 1.7238	LR: 0.027624
Training Epoch: 1 [13312/46250]	Loss: 1.5678	LR: 0.028177
Training Epoch: 1 [13568/46250]	Loss: 1.9026	LR: 0.028729
Training Epoch: 1 [13824/46250]	Loss: 1.7749	LR: 0.029282
Training Epoch: 1 [14080/46250]	Loss: 1.7389	LR: 0.029834
Training Epoch: 1 [14336/46250]	Loss: 1.7089	LR: 0.030387
Training Epoch: 1 [14592/46250]	Loss: 1.6111	LR: 0.030939
Training Epoch: 1 [14848/46250]	Loss: 1.8192	LR: 0.031492
Training Epoch: 1 [15104/46250]	Loss: 1.6690	LR: 0.032044
Training Epoch: 1 [15360/46250]	Loss: 1.7221	LR: 0.032597
Training Epoch: 1 [15616/46250]	Loss: 1.6949	LR: 0.033149
Training Epoch: 1 [15872/46250]	Loss: 1.6934	LR: 0.033702
Training Epoch: 1 [16128/46250]	Loss: 1.4882	LR: 0.034254
Training Epoch: 1 [16384/46250]	Loss: 1.4360	LR: 0.034807
Training Epoch: 1 [16640/46250]	Loss: 1.5858	LR: 0.035359
Training Epoch: 1 [16896/46250]	Loss: 1.4963	LR: 0.035912
Training Epoch: 1 [17152/46250]	Loss: 1.4138	LR: 0.036464
Training Epoch: 1 [17408/46250]	Loss: 1.4123	LR: 0.037017
Training Epoch: 1 [17664/46250]	Loss: 1.4862	LR: 0.037569
Training Epoch: 1 [17920/46250]	Loss: 1.4563	LR: 0.038122
Training Epoch: 1 [18176/46250]	Loss: 1.5707	LR: 0.038674
Training Epoch: 1 [18432/46250]	Loss: 1.5446	LR: 0.039227
Training Epoch: 1 [18688/46250]	Loss: 1.5000	LR: 0.039779
Training Epoch: 1 [18944/46250]	Loss: 1.3925	LR: 0.040331
Training Epoch: 1 [19200/46250]	Loss: 1.6054	LR: 0.040884
Training Epoch: 1 [19456/46250]	Loss: 1.5257	LR: 0.041436
Training Epoch: 1 [19712/46250]	Loss: 1.5949	LR: 0.041989
Training Epoch: 1 [19968/46250]	Loss: 1.5439	LR: 0.042541
Training Epoch: 1 [20224/46250]	Loss: 1.5687	LR: 0.043094
Training Epoch: 1 [20480/46250]	Loss: 1.4780	LR: 0.043646
Training Epoch: 1 [20736/46250]	Loss: 1.4898	LR: 0.044199
Training Epoch: 1 [20992/46250]	Loss: 1.5078	LR: 0.044751
Training Epoch: 1 [21248/46250]	Loss: 1.5523	LR: 0.045304
Training Epoch: 1 [21504/46250]	Loss: 1.5025	LR: 0.045856
Training Epoch: 1 [21760/46250]	Loss: 1.5301	LR: 0.046409
Training Epoch: 1 [22016/46250]	Loss: 1.7239	LR: 0.046961
Training Epoch: 1 [22272/46250]	Loss: 1.7349	LR: 0.047514
Training Epoch: 1 [22528/46250]	Loss: 1.6662	LR: 0.048066
Training Epoch: 1 [22784/46250]	Loss: 1.7032	LR: 0.048619
Training Epoch: 1 [23040/46250]	Loss: 1.6411	LR: 0.049171
Training Epoch: 1 [23296/46250]	Loss: 1.4974	LR: 0.049724
Training Epoch: 1 [23552/46250]	Loss: 1.5000	LR: 0.050276
Training Epoch: 1 [23808/46250]	Loss: 1.4433	LR: 0.050829
Training Epoch: 1 [24064/46250]	Loss: 1.5272	LR: 0.051381
Training Epoch: 1 [24320/46250]	Loss: 1.3511	LR: 0.051934
Training Epoch: 1 [24576/46250]	Loss: 1.5955	LR: 0.052486
Training Epoch: 1 [24832/46250]	Loss: 1.5087	LR: 0.053039
Training Epoch: 1 [25088/46250]	Loss: 1.4007	LR: 0.053591
Training Epoch: 1 [25344/46250]	Loss: 1.5004	LR: 0.054144
Training Epoch: 1 [25600/46250]	Loss: 1.4535	LR: 0.054696
Training Epoch: 1 [25856/46250]	Loss: 1.4676	LR: 0.055249
Training Epoch: 1 [26112/46250]	Loss: 1.4314	LR: 0.055801
Training Epoch: 1 [26368/46250]	Loss: 1.4669	LR: 0.056354
Training Epoch: 1 [26624/46250]	Loss: 1.5070	LR: 0.056906
Training Epoch: 1 [26880/46250]	Loss: 1.5832	LR: 0.057459
Training Epoch: 1 [27136/46250]	Loss: 1.4458	LR: 0.058011
Training Epoch: 1 [27392/46250]	Loss: 1.5259	LR: 0.058564
Training Epoch: 1 [27648/46250]	Loss: 1.3454	LR: 0.059116
Training Epoch: 1 [27904/46250]	Loss: 1.3808	LR: 0.059669
Training Epoch: 1 [28160/46250]	Loss: 1.5743	LR: 0.060221
Training Epoch: 1 [28416/46250]	Loss: 1.2353	LR: 0.060773
Training Epoch: 1 [28672/46250]	Loss: 1.4592	LR: 0.061326
Training Epoch: 1 [28928/46250]	Loss: 1.3313	LR: 0.061878
Training Epoch: 1 [29184/46250]	Loss: 1.2315	LR: 0.062431
Training Epoch: 1 [29440/46250]	Loss: 1.4110	LR: 0.062983
Training Epoch: 1 [29696/46250]	Loss: 1.3734	LR: 0.063536
Training Epoch: 1 [29952/46250]	Loss: 1.4993	LR: 0.064088
Training Epoch: 1 [30208/46250]	Loss: 1.3646	LR: 0.064641
Training Epoch: 1 [30464/46250]	Loss: 1.4179	LR: 0.065193
Training Epoch: 1 [30720/46250]	Loss: 1.2892	LR: 0.065746
Training Epoch: 1 [30976/46250]	Loss: 1.3309	LR: 0.066298
Training Epoch: 1 [31232/46250]	Loss: 1.2986	LR: 0.066851
Training Epoch: 1 [31488/46250]	Loss: 1.4934	LR: 0.067403
Training Epoch: 1 [31744/46250]	Loss: 1.2525	LR: 0.067956
Training Epoch: 1 [32000/46250]	Loss: 1.3167	LR: 0.068508
Training Epoch: 1 [32256/46250]	Loss: 1.4435	LR: 0.069061
Training Epoch: 1 [32512/46250]	Loss: 1.5293	LR: 0.069613
Training Epoch: 1 [32768/46250]	Loss: 1.3129	LR: 0.070166
Training Epoch: 1 [33024/46250]	Loss: 1.2870	LR: 0.070718
Training Epoch: 1 [33280/46250]	Loss: 1.3778	LR: 0.071271
Training Epoch: 1 [33536/46250]	Loss: 1.3783	LR: 0.071823
Training Epoch: 1 [33792/46250]	Loss: 1.3313	LR: 0.072376
Training Epoch: 1 [34048/46250]	Loss: 1.4528	LR: 0.072928
Training Epoch: 1 [34304/46250]	Loss: 1.4126	LR: 0.073481
Training Epoch: 1 [34560/46250]	Loss: 1.4961	LR: 0.074033
Training Epoch: 1 [34816/46250]	Loss: 1.4428	LR: 0.074586
Training Epoch: 1 [35072/46250]	Loss: 1.2985	LR: 0.075138
Training Epoch: 1 [35328/46250]	Loss: 1.3563	LR: 0.075691
Training Epoch: 1 [35584/46250]	Loss: 1.2302	LR: 0.076243
Training Epoch: 1 [35840/46250]	Loss: 1.2699	LR: 0.076796
Training Epoch: 1 [36096/46250]	Loss: 1.3790	LR: 0.077348
Training Epoch: 1 [36352/46250]	Loss: 1.1738	LR: 0.077901
Training Epoch: 1 [36608/46250]	Loss: 1.2715	LR: 0.078453
Training Epoch: 1 [36864/46250]	Loss: 1.3783	LR: 0.079006
Training Epoch: 1 [37120/46250]	Loss: 1.3884	LR: 0.079558
Training Epoch: 1 [37376/46250]	Loss: 1.1867	LR: 0.080110
Training Epoch: 1 [37632/46250]	Loss: 1.1410	LR: 0.080663
Training Epoch: 1 [37888/46250]	Loss: 1.2387	LR: 0.081215
Training Epoch: 1 [38144/46250]	Loss: 1.3517	LR: 0.081768
Training Epoch: 1 [38400/46250]	Loss: 1.2664	LR: 0.082320
Training Epoch: 1 [38656/46250]	Loss: 1.1845	LR: 0.082873
Training Epoch: 1 [38912/46250]	Loss: 1.3347	LR: 0.083425
Training Epoch: 1 [39168/46250]	Loss: 1.3167	LR: 0.083978
Training Epoch: 1 [39424/46250]	Loss: 1.0059	LR: 0.084530
Training Epoch: 1 [39680/46250]	Loss: 1.1988	LR: 0.085083
Training Epoch: 1 [39936/46250]	Loss: 1.2285	LR: 0.085635
Training Epoch: 1 [40192/46250]	Loss: 1.2658	LR: 0.086188
Training Epoch: 1 [40448/46250]	Loss: 1.4026	LR: 0.086740
Training Epoch: 1 [40704/46250]	Loss: 1.5444	LR: 0.087293
Training Epoch: 1 [40960/46250]	Loss: 1.2127	LR: 0.087845
Training Epoch: 1 [41216/46250]	Loss: 1.2738	LR: 0.088398
Training Epoch: 1 [41472/46250]	Loss: 1.2235	LR: 0.088950
Training Epoch: 1 [41728/46250]	Loss: 1.3259	LR: 0.089503
Training Epoch: 1 [41984/46250]	Loss: 1.2033	LR: 0.090055
Training Epoch: 1 [42240/46250]	Loss: 1.2784	LR: 0.090608
Training Epoch: 1 [42496/46250]	Loss: 1.1880	LR: 0.091160
Training Epoch: 1 [42752/46250]	Loss: 1.3231	LR: 0.091713
Training Epoch: 1 [43008/46250]	Loss: 1.2814	LR: 0.092265
Training Epoch: 1 [43264/46250]	Loss: 1.3012	LR: 0.092818
Training Epoch: 1 [43520/46250]	Loss: 1.3748	LR: 0.093370
Training Epoch: 1 [43776/46250]	Loss: 1.2700	LR: 0.093923
Training Epoch: 1 [44032/46250]	Loss: 1.6458	LR: 0.094475
Training Epoch: 1 [44288/46250]	Loss: 1.2873	LR: 0.095028
Training Epoch: 1 [44544/46250]	Loss: 1.1560	LR: 0.095580
Training Epoch: 1 [44800/46250]	Loss: 1.3064	LR: 0.096133
Training Epoch: 1 [45056/46250]	Loss: 1.3333	LR: 0.096685
Training Epoch: 1 [45312/46250]	Loss: 1.2865	LR: 0.097238
Training Epoch: 1 [45568/46250]	Loss: 1.0744	LR: 0.097790
Training Epoch: 1 [45824/46250]	Loss: 1.2719	LR: 0.098343
Training Epoch: 1 [46080/46250]	Loss: 1.2416	LR: 0.098895
Training Epoch: 1 [46250/46250]	Loss: 1.1715	LR: 0.099448
Epoch 1 - Average Train Loss: 1.5478, Train Accuracy: 0.4397
Epoch 1 training time consumed: 18.14s
Evaluating Network.....
Test set: Epoch: 1, Average loss: 0.0095, Accuracy: 0.4277, Time consumed:0.90s
Saving weights file to checkpoint/retrain/ResNet18/Saturday_02_August_2025_11h_20m_52s/ResNet18-Cifar10-seed3-ret25-1-best.pth
Valid (Test) Dl:  10000
Train Dl:  50000
Retain Train Dl:  46250
Forget Train Dl:  3750
Retain Valid Dl:  46250
Forget Valid Dl:  3750
retain_prob Distribution: 10000 samples
test_prob Distribution: 10000 samples
forget_prob Distribution: 3750 samples
Set1 Distribution: 3750 samples
Set2 Distribution: 3750 samples
Set1 Distribution: 3750 samples
Set2 Distribution: 3750 samples
Set1 Distribution: 10000 samples
Set2 Distribution: 10000 samples
Set1 Distribution: 10000 samples
Set2 Distribution: 10000 samples
Test Accuracy: 42.5
Retain Accuracy: 43.61606979370117
Zero-Retain Forget (ZRF): 0.8996286392211914
Membership Inference Attack (MIA): 0.46826666666666666
Forget vs Retain Membership Inference Attack (MIA): 0.5246666666666666
Forget vs Test Membership Inference Attack (MIA): 0.5233333333333333
Test vs Retain Membership Inference Attack (MIA): 0.58425
Train vs Test Membership Inference Attack (MIA): 0.49825
Forget Set Accuracy (Df): 43.39702606201172
Method Execution Time: 915.38 seconds
