text
stringlengths 5
1.13k
|
|---|
2020-02-03 04:28:36 Iteration 250 Training Loss: 1.831e-01 Loss in Target Net: 4.958e-02 |
2020-02-03 04:29:02 Iteration 300 Training Loss: 1.789e-01 Loss in Target Net: 4.869e-02 |
2020-02-03 04:29:33 Iteration 350 Training Loss: 1.795e-01 Loss in Target Net: 4.543e-02 |
2020-02-03 04:30:10 Iteration 400 Training Loss: 1.813e-01 Loss in Target Net: 4.880e-02 |
2020-02-03 04:30:45 Iteration 450 Training Loss: 1.765e-01 Loss in Target Net: 4.997e-02 |
2020-02-03 04:31:18 Iteration 500 Training Loss: 1.749e-01 Loss in Target Net: 4.660e-02 |
2020-02-03 04:32:01 Iteration 550 Training Loss: 1.799e-01 Loss in Target Net: 4.575e-02 |
2020-02-03 04:32:42 Iteration 600 Training Loss: 1.839e-01 Loss in Target Net: 4.792e-02 |
2020-02-03 04:33:21 Iteration 650 Training Loss: 1.727e-01 Loss in Target Net: 4.872e-02 |
2020-02-03 04:34:11 Iteration 700 Training Loss: 1.727e-01 Loss in Target Net: 4.390e-02 |
2020-02-03 04:34:47 Iteration 750 Training Loss: 1.696e-01 Loss in Target Net: 6.173e-02 |
2020-02-03 04:35:15 Iteration 800 Training Loss: 1.755e-01 Loss in Target Net: 5.195e-02 |
2020-02-03 04:35:50 Iteration 850 Training Loss: 1.721e-01 Loss in Target Net: 4.913e-02 |
2020-02-03 04:36:23 Iteration 900 Training Loss: 1.745e-01 Loss in Target Net: 4.809e-02 |
2020-02-03 04:36:59 Iteration 950 Training Loss: 1.702e-01 Loss in Target Net: 4.600e-02 |
2020-02-03 04:37:35 Iteration 1000 Training Loss: 1.736e-01 Loss in Target Net: 4.021e-02 |
2020-02-03 04:38:09 Iteration 1050 Training Loss: 1.767e-01 Loss in Target Net: 4.759e-02 |
2020-02-03 04:38:41 Iteration 1100 Training Loss: 1.689e-01 Loss in Target Net: 5.575e-02 |
2020-02-03 04:39:16 Iteration 1150 Training Loss: 1.684e-01 Loss in Target Net: 5.229e-02 |
2020-02-03 04:39:51 Iteration 1200 Training Loss: 1.695e-01 Loss in Target Net: 4.776e-02 |
2020-02-03 04:40:25 Iteration 1250 Training Loss: 1.704e-01 Loss in Target Net: 4.688e-02 |
2020-02-03 04:40:57 Iteration 1300 Training Loss: 1.697e-01 Loss in Target Net: 4.796e-02 |
2020-02-03 04:41:27 Iteration 1350 Training Loss: 1.704e-01 Loss in Target Net: 4.697e-02 |
2020-02-03 04:41:56 Iteration 1400 Training Loss: 1.692e-01 Loss in Target Net: 5.288e-02 |
2020-02-03 04:42:35 Iteration 1450 Training Loss: 1.725e-01 Loss in Target Net: 5.079e-02 |
2020-02-03 04:43:12 Iteration 1499 Training Loss: 1.718e-01 Loss in Target Net: 4.362e-02 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-03 04:43:26, Epoch 0, Iteration 7, loss 0.180 (0.472), acc 94.231 (91.400) |
2020-02-03 04:45:21, Epoch 30, Iteration 7, loss 0.000 (0.000), acc 100.000 (100.000) |
Target Label: 6, Poison label: 8, Prediction:6, Target's Score:[-3.8153315, -3.0058737, -1.5918738, -1.4286675, -0.03679273, -1.3335235, 11.116996, -2.1915617, 4.045632, -1.3245122], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-03 04:46:42 Epoch 59, Val iteration 0, acc 92.800 (92.800) |
2020-02-03 04:46:57 Epoch 59, Val iteration 19, acc 93.000 (92.950) |
* Prec: 92.9500015258789 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 18 |
TARGET INDEX: 17 |
DPN92 0 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='2', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=1, num_per_class=50, original_grad=True, poison_decay_ites=[], poison_decay_ratio=0.1, poison_epsilon=0.1, poison_ites=1500, poison_label=8, poison_lr=0.04, poison_momentum=0.9, poison_num=5, poison_opt='adam', resume_poison_ite=0, retrain_bsize=64, retrain_epochs=60, retrain_lr=0.0001, retrain_momentum=0.9, retrain_opt='adam', retrain_wd=0.0005, subs_chk_name=['ckpt-%s-4800-dp0.200-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.250-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.300-droplayer0.000.t7'], subs_dp=[0.2, 0.25, 0.3], subset_group=0, substitute_nets=['DPN92', 'SENet18', 'ResNet50', 'ResNeXt29_2x64d'], target_index=18, target_label=6, target_net=['DPN92'], test_chk_name='ckpt-%s-4800.t7', tol=1e-06, train_data_path='datasets/CIFAR10_TRAIN_Split.pth') |
Path: chk-black-end2end/mean/1500/18 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-02 11:35:59 Iteration 0 Training Loss: 1.046e+00 Loss in Target Net: 1.428e+00 |
2020-02-02 11:36:15 Iteration 50 Training Loss: 2.816e-01 Loss in Target Net: 2.163e-01 |
2020-02-02 11:36:32 Iteration 100 Training Loss: 2.427e-01 Loss in Target Net: 1.310e-01 |
2020-02-02 11:36:48 Iteration 150 Training Loss: 2.191e-01 Loss in Target Net: 8.669e-02 |
2020-02-02 11:37:06 Iteration 200 Training Loss: 2.174e-01 Loss in Target Net: 1.088e-01 |
2020-02-02 11:37:22 Iteration 250 Training Loss: 2.082e-01 Loss in Target Net: 8.732e-02 |
2020-02-02 11:37:39 Iteration 300 Training Loss: 2.073e-01 Loss in Target Net: 7.680e-02 |
2020-02-02 11:37:56 Iteration 350 Training Loss: 2.015e-01 Loss in Target Net: 5.847e-02 |
2020-02-02 11:38:12 Iteration 400 Training Loss: 2.025e-01 Loss in Target Net: 5.016e-02 |
2020-02-02 11:38:29 Iteration 450 Training Loss: 2.053e-01 Loss in Target Net: 5.692e-02 |
2020-02-02 11:38:45 Iteration 500 Training Loss: 1.946e-01 Loss in Target Net: 4.327e-02 |
2020-02-02 11:39:02 Iteration 550 Training Loss: 1.951e-01 Loss in Target Net: 5.445e-02 |
2020-02-02 11:39:19 Iteration 600 Training Loss: 1.986e-01 Loss in Target Net: 3.833e-02 |
2020-02-02 11:39:35 Iteration 650 Training Loss: 1.932e-01 Loss in Target Net: 4.431e-02 |
2020-02-02 11:39:51 Iteration 700 Training Loss: 1.958e-01 Loss in Target Net: 4.011e-02 |
2020-02-02 11:40:08 Iteration 750 Training Loss: 1.866e-01 Loss in Target Net: 3.513e-02 |
2020-02-02 11:40:25 Iteration 800 Training Loss: 1.907e-01 Loss in Target Net: 3.662e-02 |
2020-02-02 11:40:42 Iteration 850 Training Loss: 1.882e-01 Loss in Target Net: 4.450e-02 |
2020-02-02 11:40:58 Iteration 900 Training Loss: 1.892e-01 Loss in Target Net: 4.302e-02 |
2020-02-02 11:41:14 Iteration 950 Training Loss: 1.874e-01 Loss in Target Net: 3.900e-02 |
2020-02-02 11:41:32 Iteration 1000 Training Loss: 1.951e-01 Loss in Target Net: 4.012e-02 |
2020-02-02 11:41:48 Iteration 1050 Training Loss: 1.901e-01 Loss in Target Net: 3.677e-02 |
2020-02-02 11:42:04 Iteration 1100 Training Loss: 1.888e-01 Loss in Target Net: 4.291e-02 |
2020-02-02 11:42:21 Iteration 1150 Training Loss: 1.952e-01 Loss in Target Net: 2.984e-02 |
2020-02-02 11:42:37 Iteration 1200 Training Loss: 1.900e-01 Loss in Target Net: 3.208e-02 |
2020-02-02 11:42:53 Iteration 1250 Training Loss: 1.912e-01 Loss in Target Net: 3.000e-02 |
2020-02-02 11:43:10 Iteration 1300 Training Loss: 1.896e-01 Loss in Target Net: 3.608e-02 |
2020-02-02 11:43:26 Iteration 1350 Training Loss: 1.893e-01 Loss in Target Net: 3.941e-02 |
2020-02-02 11:43:43 Iteration 1400 Training Loss: 1.873e-01 Loss in Target Net: 3.885e-02 |
2020-02-02 11:44:01 Iteration 1450 Training Loss: 1.891e-01 Loss in Target Net: 4.242e-02 |
2020-02-02 11:44:19 Iteration 1499 Training Loss: 1.911e-01 Loss in Target Net: 4.432e-02 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-02 11:44:28, Epoch 0, Iteration 7, loss 0.384 (0.405), acc 90.385 (91.000) |
2020-02-02 11:45:26, Epoch 30, Iteration 7, loss 0.000 (0.000), acc 100.000 (100.000) |
Target Label: 6, Poison label: 8, Prediction:8, Target's Score:[0.7341862, 0.7691202, -1.7899215, -1.5533844, -3.215847, -4.7919846, 2.165239, -1.3069195, 9.918374, -0.6283417], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-02 11:46:26 Epoch 59, Val iteration 0, acc 93.600 (93.600) |
2020-02-02 11:46:34 Epoch 59, Val iteration 19, acc 92.800 (92.840) |
* Prec: 92.84000129699707 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 8 |
TARGET INDEX: 18 |
DPN92 1 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='3', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=1, num_per_class=50, original_grad=True, poison_decay_ites=[], poison_decay_ratio=0.1, poison_epsilon=0.1, poison_ites=1500, poison_label=8, poison_lr=0.04, poison_momentum=0.9, poison_num=5, poison_opt='adam', resume_poison_ite=0, retrain_bsize=64, retrain_epochs=60, retrain_lr=0.0001, retrain_momentum=0.9, retrain_opt='adam', retrain_wd=0.0005, subs_chk_name=['ckpt-%s-4800-dp0.200-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.250-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.300-droplayer0.000.t7'], subs_dp=[0.2, 0.25, 0.3], subset_group=0, substitute_nets=['DPN92', 'SENet18', 'ResNet50', 'ResNeXt29_2x64d'], target_index=19, target_label=6, target_net=['DPN92'], test_chk_name='ckpt-%s-4800.t7', tol=1e-06, train_data_path='datasets/CIFAR10_TRAIN_Split.pth') |
Path: chk-black-end2end/mean/1500/19 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-02 11:33:40 Iteration 0 Training Loss: 1.087e+00 Loss in Target Net: 1.606e+00 |
2020-02-02 11:33:55 Iteration 50 Training Loss: 2.862e-01 Loss in Target Net: 2.828e-01 |
2020-02-02 11:34:13 Iteration 100 Training Loss: 2.613e-01 Loss in Target Net: 1.831e-01 |
2020-02-02 11:34:30 Iteration 150 Training Loss: 2.492e-01 Loss in Target Net: 1.400e-01 |
2020-02-02 11:34:46 Iteration 200 Training Loss: 2.298e-01 Loss in Target Net: 1.633e-01 |
2020-02-02 11:35:02 Iteration 250 Training Loss: 2.301e-01 Loss in Target Net: 1.130e-01 |
2020-02-02 11:35:19 Iteration 300 Training Loss: 2.289e-01 Loss in Target Net: 1.547e-01 |
Subsets and Splits