-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathcommands.sh
89 lines (80 loc) · 9.7 KB
/
commands.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
##### Semantic Car with Stripe BG
##### Neurotoxin
python training.py --params utils/params_car_stripe_bg.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 100
##### Baseline
python training.py --params utils/params_car_stripe_bg.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 100
#####Chameleon
python training.py --params utils/params_car_stripe_bg.yaml --GPU_id "0" --size_of_secret_dataset 100
#####Anticipate
python training.py --params utils/params_car_stripe_bg.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1
##### Green Cars
##### Neurotoxin
python training.py --params utils/params_green_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 100 --persistence_diff 0.001
##### Baseline
python training.py --params utils/params_green_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 100 --persistence_diff 0.001
#####Chameleon
python training.py --params utils/params_green_car.yaml --GPU_id "0" --size_of_secret_dataset 100 --persistence_diff 0.001
#####Anticipate
python training.py --params utils/params_green_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1 --persistence_diff 0.001
##### Car with Racing Stripes
##### Neurotoxin
python training.py --params utils/params_racing_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 100
##### Baseline
python training.py --params utils/params_racing_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 100
#####Chameleon
python training.py --params utils/params_racing_car.yaml --GPU_id "0" --size_of_secret_dataset 100
#####Anticipate
python training.py --params utils/params_racing_car.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1
##### Pixel Pattern CIFAR10
##### Chameleon
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 1 --size_of_secret_dataset 25 --persistence_diff 0.03 --contrastive_loss_scale_weight 6.0
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.1 --size_of_secret_dataset 25 --persistence_diff 0.03 --contrastive_loss_scale_weight 6.0
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.5 --size_of_secret_dataset 25 --persistence_diff 0.03 --contrastive_loss_scale_weight 9.0
##### Neurotoxin
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.5 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
##### Baseline
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.5 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
##### Anticipate
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03 --anticipate 1
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03 --anticipate 1
python training.py --params utils/params_pixel_trigger_cifar10.yaml --GPU_id "0" --pattern_type 2 --pattern_diff 0.5 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03 --anticipate 1
##### Pixel Pattern EMNIST
##### Chameleon
python training.py --params utils/params_pixel_trigger_emnist.yaml --GPU_id "0" --pattern_type 1 --size_of_secret_dataset 25 --persistence_diff 0.03 --contrastive_loss_scale_weight 3.0
##### Baseline
python training.py --params utils/params_pixel_trigger_emnist.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
##### Neurotoxin
python training.py --params utils/params_pixel_trigger_emnist.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03
##### Anticipate
python training.py --params utils/params_pixel_trigger_emnist.yaml --GPU_id "0" --pattern_type 1 --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --persistence_diff 0.03 --anticipate 1
##### Edge-case Cifar10
##### Chameleon
python training.py --params utils/params_edge_case_cifar10.yaml --GPU_id "0" --size_of_secret_dataset 25 --edge_case 1
##### Baseline
python training.py --params utils/params_edge_case_cifar10.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --edge_case 1
##### Baseline
python training.py --params utils/params_edge_case_cifar10.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --edge_case 1
##### Anticipate
python training.py --params utils/params_edge_case_cifar10.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1 --edge_case 1
##### Semantic Car with Stripe BG cifar100
##### Neurotoxin
python training.py --params utils/params_car_stripe_bg_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 25
##### Baseline
python training.py --params utils/params_car_stripe_bg_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25
##### Chameleon
python training.py --params utils/params_car_stripe_bg_cifar100.yaml --GPU_id "0" --size_of_secret_dataset 100
##### Anticipate
python training.py --params utils/params_car_stripe_bg_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1
##### Black bicycle cifar100
##### Neurotoxin
python training.py --params utils/params_black_bicycle_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 0.99 --lr_gamma 0.01 --size_of_secret_dataset 100
##### Baseline
python training.py --params utils/params_black_bicycle_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 5 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 100
##### Chameleon
python training.py --params utils/params_black_bicycle_cifar100.yaml --GPU_id "0" --size_of_secret_dataset 100
##### Anticipate
python training.py --params utils/params_black_bicycle_cifar100.yaml --GPU_id "0" --is_frozen_params 0 --retrain_poison_contrastive 0 --retrain_poison 3 --poison_lr 0.1 --gradmask_ratio 1 --lr_gamma 0.01 --size_of_secret_dataset 25 --anticipate 1