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: 31
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=31-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: 85.007 | Val Loss: 0.003 | Val Acc: 84.618
Training:
Accuracies by groups:
0, 0  acc: 71492 / 71629 =  99.809
0, 1  acc: 66805 / 66874 =  99.897
1, 0  acc:    66 / 22880 =   0.288
1, 1  acc:     3 /  1387 =   0.216
--------------------------------------
Average acc: 138366 / 162770 =  85.007
Robust  acc:     3 /  1387 =   0.216
--------------------------------------
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_seed31.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_seed31.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: 88.747 | Val Loss: 0.001 | Val Acc: 92.616
Training:
Accuracies by groups:
0, 0  acc: 70902 / 71629 =  98.985
0, 1  acc: 66852 / 66874 =  99.967
1, 0  acc:  6674 / 22880 =  29.170
1, 1  acc:    26 /  1387 =   1.875
--------------------------------------
Average acc: 144454 / 162770 =  88.747
Robust  acc:    26 /  1387 =   1.875
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8246 /  8535 =  96.614
0, 1  acc:  8272 /  8276 =  99.952
1, 0  acc:  1873 /  2874 =  65.170
1, 1  acc:     9 /   182 =   4.945
------------------------------------
Average acc: 18400 / 19867 =  92.616
Robust  acc:     9 /   182 =   4.945
------------------------------------
Save biased model at epoch 1
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch1_seed31.pt
New max average-worst acc gap: 87.67084076139294
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_seed31.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 92.992
Robust Acc: 8.889 | Best Acc: 99.987
------------------------------------
Training, Epoch 1:
Accuracies by groups:
0, 0  acc:  9545 /  9767 =  97.727
0, 1  acc:  7534 /  7535 =  99.987
1, 0  acc:  1468 /  2480 =  59.194
1, 1  acc:    16 /   180 =   8.889
------------------------------------
Average acc: 18563 / 19962 =  92.992
Robust  acc:    16 /   180 =   8.889
------------------------------------
Accuracies by groups:
0, 0  acc:  9545 /  9767 =  97.727
0, 1  acc:  7534 /  7535 =  99.987
1, 0  acc:  1468 /  2480 =  59.194
1, 1  acc:    16 /   180 =   8.889
------------------------------------
Average acc: 18563 / 19962 =  92.992
Robust  acc:    16 /   180 =   8.889
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9545 /  9767 =  97.727
0, 1  acc:  7534 /  7535 =  99.987
1, 0  acc:  1468 /  2480 =  59.194
1, 1  acc:    16 /   180 =   8.889
------------------------------------
Average acc: 18563 / 19962 =  92.992
Robust  acc:    16 /   180 =   8.889
------------------------------------
Epoch:   3 | Train Loss: 0.000 | Train Acc: 93.450 | Val Loss: 0.001 | Val Acc: 94.171
Training:
Accuracies by groups:
0, 0  acc: 69132 / 71629 =  96.514
0, 1  acc: 66719 / 66874 =  99.768
1, 0  acc: 16058 / 22880 =  70.184
1, 1  acc:   200 /  1387 =  14.420
--------------------------------------
Average acc: 152109 / 162770 =  93.450
Robust  acc:   200 /  1387 =  14.420
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8116 /  8535 =  95.091
0, 1  acc:  8256 /  8276 =  99.758
1, 0  acc:  2308 /  2874 =  80.306
1, 1  acc:    29 /   182 =  15.934
------------------------------------
Average acc: 18709 / 19867 =  94.171
Robust  acc:    29 /   182 =  15.934
------------------------------------
Save biased model at epoch 2
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch2_seed31.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 94.485
Robust Acc: 19.444 | Best Acc: 99.827
-------------------------------------
Training, Epoch 2:
Accuracies by groups:
0, 0  acc:  9422 /  9767 =  96.468
0, 1  acc:  7522 /  7535 =  99.827
1, 0  acc:  1882 /  2480 =  75.887
1, 1  acc:    35 /   180 =  19.444
------------------------------------
Average acc: 18861 / 19962 =  94.485
Robust  acc:    35 /   180 =  19.444
------------------------------------
Accuracies by groups:
0, 0  acc:  9422 /  9767 =  96.468
0, 1  acc:  7522 /  7535 =  99.827
1, 0  acc:  1882 /  2480 =  75.887
1, 1  acc:    35 /   180 =  19.444
------------------------------------
Average acc: 18861 / 19962 =  94.485
Robust  acc:    35 /   180 =  19.444
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9422 /  9767 =  96.468
0, 1  acc:  7522 /  7535 =  99.827
1, 0  acc:  1882 /  2480 =  75.887
1, 1  acc:    35 /   180 =  19.444
------------------------------------
Average acc: 18861 / 19962 =  94.485
Robust  acc:    35 /   180 =  19.444
------------------------------------
Epoch:   4 | Train Loss: 0.000 | Train Acc: 94.244 | Val Loss: 0.001 | Val Acc: 94.493
Training:
Accuracies by groups:
0, 0  acc: 68883 / 71629 =  96.166
0, 1  acc: 66629 / 66874 =  99.634
1, 0  acc: 17597 / 22880 =  76.910
1, 1  acc:   292 /  1387 =  21.053
--------------------------------------
Average acc: 153401 / 162770 =  94.244
Robust  acc:   292 /  1387 =  21.053
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8189 /  8535 =  95.946
0, 1  acc:  8259 /  8276 =  99.795
1, 0  acc:  2294 /  2874 =  79.819
1, 1  acc:    31 /   182 =  17.033
------------------------------------
Average acc: 18773 / 19867 =  94.493
Robust  acc:    31 /   182 =  17.033
------------------------------------
Save biased model at epoch 3
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch3_seed31.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 94.885
Robust Acc: 23.333 | Best Acc: 99.801
-------------------------------------
Training, Epoch 3:
Accuracies by groups:
0, 0  acc:  9492 /  9767 =  97.184
0, 1  acc:  7520 /  7535 =  99.801
1, 0  acc:  1887 /  2480 =  76.089
1, 1  acc:    42 /   180 =  23.333
------------------------------------
Average acc: 18941 / 19962 =  94.885
Robust  acc:    42 /   180 =  23.333
------------------------------------
Accuracies by groups:
0, 0  acc:  9492 /  9767 =  97.184
0, 1  acc:  7520 /  7535 =  99.801
1, 0  acc:  1887 /  2480 =  76.089
1, 1  acc:    42 /   180 =  23.333
------------------------------------
Average acc: 18941 / 19962 =  94.885
Robust  acc:    42 /   180 =  23.333
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9492 /  9767 =  97.184
0, 1  acc:  7520 /  7535 =  99.801
1, 0  acc:  1887 /  2480 =  76.089
1, 1  acc:    42 /   180 =  23.333
------------------------------------
Average acc: 18941 / 19962 =  94.885
Robust  acc:    42 /   180 =  23.333
------------------------------------
Epoch:   5 | Train Loss: 0.000 | Train Acc: 94.691 | Val Loss: 0.001 | Val Acc: 94.866
Training:
Accuracies by groups:
0, 0  acc: 68889 / 71629 =  96.175
0, 1  acc: 66590 / 66874 =  99.575
1, 0  acc: 18279 / 22880 =  79.891
1, 1  acc:   370 /  1387 =  26.676
--------------------------------------
Average acc: 154128 / 162770 =  94.691
Robust  acc:   370 /  1387 =  26.676
--------------------------------------
Validating:
Accuracies by groups:
0, 0  acc:  8139 /  8535 =  95.360
0, 1  acc:  8249 /  8276 =  99.674
1, 0  acc:  2416 /  2874 =  84.064
1, 1  acc:    43 /   182 =  23.626
------------------------------------
Average acc: 18847 / 19867 =  94.866
Robust  acc:    43 /   182 =  23.626
------------------------------------
Save biased model at epoch 4
replace: True
Checkpoint saved at ./model/celebA/config/stage_one_erm_model_b_epoch4_seed31.pt
-------------------------------------------
Avg Test Loss: 0.001 | Avg Test Acc: 95.341
Robust Acc: 28.333 | Best Acc: 99.708
-------------------------------------
Training, Epoch 4:
Accuracies by groups:
0, 0  acc:  9446 /  9767 =  96.713
0, 1  acc:  7513 /  7535 =  99.708
1, 0  acc:  2022 /  2480 =  81.532
1, 1  acc:    51 /   180 =  28.333
------------------------------------
Average acc: 19032 / 19962 =  95.341
Robust  acc:    51 /   180 =  28.333
------------------------------------
Accuracies by groups:
0, 0  acc:  9446 /  9767 =  96.713
0, 1  acc:  7513 /  7535 =  99.708
1, 0  acc:  2022 /  2480 =  81.532
1, 1  acc:    51 /   180 =  28.333
------------------------------------
Average acc: 19032 / 19962 =  95.341
Robust  acc:    51 /   180 =  28.333
------------------------------------
Testing:
Accuracies by groups:
0, 0  acc:  9446 /  9767 =  96.713
0, 1  acc:  7513 /  7535 =  99.708
1, 0  acc:  2022 /  2480 =  81.532
1, 1  acc:    51 /   180 =  28.333
------------------------------------
Average acc: 19032 / 19962 =  95.341
Robust  acc:    51 /   180 =  28.333
------------------------------------
replace: True
Checkpoint saved at ./model/celebA/config/bias-end_seed31.pt
training biased model is done
