Arch: resnet50_pt
Bs trn: 128
Bs val: 128
Hidden dim: 256
Dataset: celebA
Resample class: 
Slice with: rep
Rep cluster method: gmm
Num anchor: 32
Num positive: 32
Num negative: 32
Num negative easy: 0
Weight anc by loss: False
Weight pos by loss: False
Weight neg by loss: False
Anc loss temp: 0.5
Pos loss temp: 0.5
Neg loss temp: 0.5
Data wide pos: False
Target sample ratio: 1
Balance targets: False
Additional negatives: False
Hard negative factor: 0
Full contrastive: False
Train encoder: False
No projection head: False
Projection dim: 128
Batch factor: None
Temperature: 0.05
Single pos: False
Supervised linear scale up: False
Supervised update delay: 0
Contrastive weight: 0.5
Classifier update interval: 8
Optim: sgd
Max epoch: 5
Lr: 0.0001
Momentum: 0.9
Weight decay: 0.1
Weight decay c: 0.1
Stopping window: 30
Load encoder: 
Freeze encoder: False
Finetune epochs: 0
Clip grad norm: False
Lr scheduler classifier: 
Lr scheduler: 
Grad clip grad norm: False
Erm: False
Erm only: False
Pretrained spurious path: 
Max epoch s: 1
Bs trn s: 32
Lr s: 0.001
Momentum s: 0.9
Weight decay s: 0.0005
Slice temp: 10
Log loss interval: 10
Checkpoint interval: 50
Grad checkpoint interval: 50
Log visual interval: 100
Log grad visual interval: 50
Verbose: True
Seed: 37
Replicate: 0
No cuda: False
Resume: False
New slice: False
Num workers: 12
Evaluate: False
Data cmap: hsv
Test cmap: 
P correlation: 0.9
P corr by class: None
Train classes: ['blond', 'nonblond']
Train class ratios: None
Test shift: random
Flipped: False
Q: 0.7
Pretrained bmodel: False
Cosine: False
Exp: stage_one_erm
Supervised contrast: True
Prioritize spurious pos: False
Contrastive type: cnc
Compute auroc: False
Model type: resnet50_pt_cnc
Criterion: cross_entropy
Pretrained: False
Max grad norm: 1.0
Adam epsilon: 1e-08
Warmup steps: 0
Max grad norm s: 1.0
Adam epsilon s: 1e-08
Warmup steps s: 0
Grad max grad norm: 1.0
Grad adam epsilon: 1e-08
Grad warmup steps: 0
Device: cuda
Img file type: .png
Display image: False
Image path: ./images/celebA/celebA/config/contrastive_umaps
Log interval: 1
Log path: ./logs/celebA/config
Results path: ./results/celebA/config
Model path: ./model/celebA/config
Loss factor: 1
Supersample labels: False
Subsample labels: False
Weigh slice samples by loss: True
Val split: 0.2
Spurious train split: 0.2
Subsample groups: False
Train method: sc
Max robust acc: -1
Max robust epoch: -1
Max robust group acc: (None, None)
Root dir: ./datasets/data/CelebA/
Target name: Blond_Hair
Confounder names: ['Male']
Image mean: 0.449
Image std: 0.226
Augment data: False
Task: celebA
Num classes: 2
Experiment configs: config
Experiment name: cnc-celebA-sw=re-na=32-np=32-nn=32-nne=0-tsr=1-t=0.05-bf=None-cw=0.5-sud=0-me=5-bst=128-o=sgd-lr=0.0001-mo=0.9-wd=0.1-wdc=0.1-spur-me=1-bst=32-lr=0.001-mo=0.9-wd=0.0005-sts=0.2-s=37-r=0
Mi resampled: None

Loading checkpoints for train split:
[-1 -1 -1 ... -1 -1 -1]
<class 'numpy.ndarray'>
[0 1 2 3] [71629 66874 22880  1387]
Loading checkpoints for val split:
[-1 -1 -1 ... -1  1 -1]
<class 'numpy.ndarray'>
[0 1 2 3] [8535 8276 2874  182]
Loading checkpoints for test split:
[-1 -1 -1 ... -1 -1  1]
<class 'numpy.ndarray'>
[0 1 2 3] [9767 7535 2480  180]
Train dataset:
    Blond_Hair = 0, Male = 0 : n = 71629
    Blond_Hair = 0, Male = 1 : n = 66874
    Blond_Hair = 1, Male = 0 : n = 22880
    Blond_Hair = 1, Male = 1 : n = 1387
Val dataset:
    Blond_Hair = 0, Male = 0 : n = 8535
    Blond_Hair = 0, Male = 1 : n = 8276
    Blond_Hair = 1, Male = 0 : n = 2874
    Blond_Hair = 1, Male = 1 : n = 182
Test dataset:
    Blond_Hair = 0, Male = 0 : n = 9767
    Blond_Hair = 0, Male = 1 : n = 7535
    Blond_Hair = 1, Male = 0 : n = 2480
    Blond_Hair = 1, Male = 1 : n = 180
Pretrained model loaded from 
Epoch:   1 | Train Loss: 0.000 | Train Acc: 85.073 | Val Loss: 0.003 | Val Acc: 84.618
Training:
Accuracies by groups:
0, 0  acc: 71613 / 71629 =  99.978
0, 1  acc: 66854 / 66874 =  99.970
1, 0  acc:     6 / 22880 =   0.026
1, 1  acc:     0 /  1387 =   0.000
--------------------------------------
Average acc: 138473 / 162770 =  85.073
Robust  acc:     0 /  1387 =   0.000
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8535 /  8535 = 100.000
0, 1  acc:  8276 /  8276 = 100.000
1, 0  acc:     0 /  2874 =   0.000
1, 1  acc:     0 /   182 =   0.000
------------------------------------
Average acc: 16811 / 19867 =  84.618
Robust  acc:     0 /  2874 =   0.000
------------------------------------
Save biased model at epoch 0
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch0_seed37.pt
New max average-worst acc gap: 84.61770775658127
bias model - Saving best checkpoint at epoch 0
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_worst_avg_gap_best_epoch0_seed37.pt
-------------------------------------------
Avg Test Loss: 0.003 | Avg Test Acc: 86.675
Robust Acc: 0.000 | Best Acc: 100.000
-------------------------------------
Training, Epoch 0:
Accuracies by groups:
0, 0  acc:  9767 /  9767 = 100.000
0, 1  acc:  7535 /  7535 = 100.000
1, 0  acc:     0 /  2480 =   0.000
1, 1  acc:     0 /   180 =   0.000
------------------------------------
Average acc: 17302 / 19962 =  86.675
Robust  acc:     0 /  2480 =   0.000
------------------------------------
Accuracies by groups:
0, 0  acc:  9767 /  9767 = 100.000
0, 1  acc:  7535 /  7535 = 100.000
1, 0  acc:     0 /  2480 =   0.000
1, 1  acc:     0 /   180 =   0.000
------------------------------------
Average acc: 17302 / 19962 =  86.675
Robust  acc:     0 /  2480 =   0.000
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9767 /  9767 = 100.000
0, 1  acc:  7535 /  7535 = 100.000
1, 0  acc:     0 /  2480 =   0.000
1, 1  acc:     0 /   180 =   0.000
------------------------------------
Average acc: 17302 / 19962 =  86.675
Robust  acc:     0 /  2480 =   0.000
------------------------------------
Epoch:   2 | Train Loss: 0.001 | Train Acc: 86.915 | Val Loss: 0.002 | Val Acc: 91.926
Training:
Accuracies by groups:
0, 0  acc: 71377 / 71629 =  99.648
0, 1  acc: 66872 / 66874 =  99.997
1, 0  acc:  3215 / 22880 =  14.052
1, 1  acc:     8 /  1387 =   0.577
--------------------------------------
Average acc: 141472 / 162770 =  86.915
Robust  acc:     8 /  1387 =   0.577
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8285 /  8535 =  97.071
0, 1  acc:  8276 /  8276 = 100.000
1, 0  acc:  1696 /  2874 =  59.012
1, 1  acc:     6 /   182 =   3.297
------------------------------------
Average acc: 18263 / 19867 =  91.926
Robust  acc:     6 /   182 =   3.297
------------------------------------
Save biased model at epoch 1
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch1_seed37.pt
New max average-worst acc gap: 88.62960666453895
bias model - Saving best checkpoint at epoch 1
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_worst_avg_gap_best_epoch1_seed37.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 92.451
Robust Acc: 6.111 | Best Acc: 99.973
------------------------------------
Training, Epoch 1:
Accuracies by groups:
0, 0  acc:  9582 /  9767 =  98.106
0, 1  acc:  7533 /  7535 =  99.973
1, 0  acc:  1329 /  2480 =  53.589
1, 1  acc:    11 /   180 =   6.111
------------------------------------
Average acc: 18455 / 19962 =  92.451
Robust  acc:    11 /   180 =   6.111
------------------------------------
Accuracies by groups:
0, 0  acc:  9582 /  9767 =  98.106
0, 1  acc:  7533 /  7535 =  99.973
1, 0  acc:  1329 /  2480 =  53.589
1, 1  acc:    11 /   180 =   6.111
------------------------------------
Average acc: 18455 / 19962 =  92.451
Robust  acc:    11 /   180 =   6.111
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9582 /  9767 =  98.106
0, 1  acc:  7533 /  7535 =  99.973
1, 0  acc:  1329 /  2480 =  53.589
1, 1  acc:    11 /   180 =   6.111
------------------------------------
Average acc: 18455 / 19962 =  92.451
Robust  acc:    11 /   180 =   6.111
------------------------------------
Epoch:   3 | Train Loss: 0.000 | Train Acc: 93.134 | Val Loss: 0.001 | Val Acc: 93.960
Training:
Accuracies by groups:
0, 0  acc: 69417 / 71629 =  96.912
0, 1  acc: 66753 / 66874 =  99.819
1, 0  acc: 15265 / 22880 =  66.718
1, 1  acc:   159 /  1387 =  11.464
--------------------------------------
Average acc: 151594 / 162770 =  93.134
Robust  acc:   159 /  1387 =  11.464
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8147 /  8535 =  95.454
0, 1  acc:  8257 /  8276 =  99.770
1, 0  acc:  2233 /  2874 =  77.697
1, 1  acc:    30 /   182 =  16.484
------------------------------------
Average acc: 18667 / 19867 =  93.960
Robust  acc:    30 /   182 =  16.484
------------------------------------
Save biased model at epoch 2
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch2_seed37.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 94.279
Robust Acc: 20.000 | Best Acc: 99.867
-------------------------------------
Training, Epoch 2:
Accuracies by groups:
0, 0  acc:  9444 /  9767 =  96.693
0, 1  acc:  7525 /  7535 =  99.867
1, 0  acc:  1815 /  2480 =  73.185
1, 1  acc:    36 /   180 =  20.000
------------------------------------
Average acc: 18820 / 19962 =  94.279
Robust  acc:    36 /   180 =  20.000
------------------------------------
Accuracies by groups:
0, 0  acc:  9444 /  9767 =  96.693
0, 1  acc:  7525 /  7535 =  99.867
1, 0  acc:  1815 /  2480 =  73.185
1, 1  acc:    36 /   180 =  20.000
------------------------------------
Average acc: 18820 / 19962 =  94.279
Robust  acc:    36 /   180 =  20.000
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9444 /  9767 =  96.693
0, 1  acc:  7525 /  7535 =  99.867
1, 0  acc:  1815 /  2480 =  73.185
1, 1  acc:    36 /   180 =  20.000
------------------------------------
Average acc: 18820 / 19962 =  94.279
Robust  acc:    36 /   180 =  20.000
------------------------------------
Epoch:   4 | Train Loss: 0.000 | Train Acc: 94.195 | Val Loss: 0.001 | Val Acc: 94.503
Training:
Accuracies by groups:
0, 0  acc: 68827 / 71629 =  96.088
0, 1  acc: 66621 / 66874 =  99.622
1, 0  acc: 17559 / 22880 =  76.744
1, 1  acc:   315 /  1387 =  22.711
--------------------------------------
Average acc: 153322 / 162770 =  94.195
Robust  acc:   315 /  1387 =  22.711
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8151 /  8535 =  95.501
0, 1  acc:  8253 /  8276 =  99.722
1, 0  acc:  2334 /  2874 =  81.211
1, 1  acc:    37 /   182 =  20.330
------------------------------------
Average acc: 18775 / 19867 =  94.503
Robust  acc:    37 /   182 =  20.330
------------------------------------
Save biased model at epoch 3
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch3_seed37.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 94.880
Robust Acc: 26.111 | Best Acc: 99.788
-------------------------------------
Training, Epoch 3:
Accuracies by groups:
0, 0  acc:  9454 /  9767 =  96.795
0, 1  acc:  7519 /  7535 =  99.788
1, 0  acc:  1920 /  2480 =  77.419
1, 1  acc:    47 /   180 =  26.111
------------------------------------
Average acc: 18940 / 19962 =  94.880
Robust  acc:    47 /   180 =  26.111
------------------------------------
Accuracies by groups:
0, 0  acc:  9454 /  9767 =  96.795
0, 1  acc:  7519 /  7535 =  99.788
1, 0  acc:  1920 /  2480 =  77.419
1, 1  acc:    47 /   180 =  26.111
------------------------------------
Average acc: 18940 / 19962 =  94.880
Robust  acc:    47 /   180 =  26.111
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9454 /  9767 =  96.795
0, 1  acc:  7519 /  7535 =  99.788
1, 0  acc:  1920 /  2480 =  77.419
1, 1  acc:    47 /   180 =  26.111
------------------------------------
Average acc: 18940 / 19962 =  94.880
Robust  acc:    47 /   180 =  26.111
------------------------------------
Epoch:   5 | Train Loss: 0.000 | Train Acc: 94.591 | Val Loss: 0.001 | Val Acc: 94.725
Training:
Accuracies by groups:
0, 0  acc: 68768 / 71629 =  96.006
0, 1  acc: 66585 / 66874 =  99.568
1, 0  acc: 18246 / 22880 =  79.747
1, 1  acc:   367 /  1387 =  26.460
--------------------------------------
Average acc: 153966 / 162770 =  94.591
Robust  acc:   367 /  1387 =  26.460
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8145 /  8535 =  95.431
0, 1  acc:  8251 /  8276 =  99.698
1, 0  acc:  2383 /  2874 =  82.916
1, 1  acc:    40 /   182 =  21.978
------------------------------------
Average acc: 18819 / 19867 =  94.725
Robust  acc:    40 /   182 =  21.978
------------------------------------
Save biased model at epoch 4
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch4_seed37.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 95.316
Robust Acc: 28.889 | Best Acc: 99.761
-------------------------------------
Training, Epoch 4:
Accuracies by groups:
0, 0  acc:  9452 /  9767 =  96.775
0, 1  acc:  7517 /  7535 =  99.761
1, 0  acc:  2006 /  2480 =  80.887
1, 1  acc:    52 /   180 =  28.889
------------------------------------
Average acc: 19027 / 19962 =  95.316
Robust  acc:    52 /   180 =  28.889
------------------------------------
Accuracies by groups:
0, 0  acc:  9452 /  9767 =  96.775
0, 1  acc:  7517 /  7535 =  99.761
1, 0  acc:  2006 /  2480 =  80.887
1, 1  acc:    52 /   180 =  28.889
------------------------------------
Average acc: 19027 / 19962 =  95.316
Robust  acc:    52 /   180 =  28.889
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9452 /  9767 =  96.775
0, 1  acc:  7517 /  7535 =  99.761
1, 0  acc:  2006 /  2480 =  80.887
1, 1  acc:    52 /   180 =  28.889
------------------------------------
Average acc: 19027 / 19962 =  95.316
Robust  acc:    52 /   180 =  28.889
------------------------------------
replace: True
Checkpoint saved at ./model/celebA/config/bias-end_seed37.pt
training biased model is done
