-
Notifications
You must be signed in to change notification settings - Fork 0
/
train.sh
executable file
·109 lines (84 loc) · 12.2 KB
/
train.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
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
#!/bin/bash
#python train.py -net_arch RNN_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Biv2_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_indepentv2_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 4 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 8 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 16 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 20 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_mask -num_layers 24 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_hh_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_hp_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_ph_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pp_mask -num_layers 12 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 2 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 3 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 4 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 5 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 6 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 1 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 2 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 3 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 4 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 6 -opt options/train/train_demo.yml
#python train.py -net_arch PNN -opt options/train/train_demo.yml
#python train.py -net_arch DRPNN -opt options/train/train_demo.yml
#python train.py -net_arch PanNet -opt options/train/train_demo.yml
#python train.py -net_arch BDPN -opt options/train/train_demo.yml
#python train.py -net_arch TFNET -opt options/train/train_demo.yml
#python train.py -net_arch DILATED -opt options/train/train_demo.yml
#python train.py -net_arch DiCNN1 -opt options/train/train_demo.yml
#python train.py -net_arch MSDCNN -opt options/train/train_demo.yml
#python train.py -net_arch MSDCNN -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Share/PanSharpening_Methods/PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch MSDCNN -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
#python train.py -net_arch BDPN -loss myloss -opt options/train/train_demo.yml -in_channels 8 -datamode LRHRLRPAN -lrpath ../PanSharp_dataset/WV2/train/MTF/8bands/LRMS_npy
#python train.py -net_arch BDPN -loss myloss -opt options/train/train_demo.yml -in_channels 4 -datamode LRHRLRPAN -lrpath /media/clh/Share/PanSharpening_Methods/PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch BDPN -loss myloss -opt options/train/train_demo.yml -in_channels 4 -datamode LRHRLRPAN -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
#python train.py -net_arch BDPN -loss l1 -opt options/train/train_demo.yml -in_channels 8 -lrpath ../PanSharp_dataset/WV2/train/MTF/8bands/LRMS_npy
#python train.py -net_arch BDPN -loss l1 -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Share/PanSharpening_Methods/PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch BDPN -loss l1 -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
#python train.py -net_arch BIEDN -opt options/train/train_demo.yml -in_channels 8 -datamode LRHRLRPAN -lrpath ../PanSharp_dataset/WV2/train/MTF/8bands/LRMS_npy
#python train.py -net_arch BIEDN -opt options/train/train_demo.yml -in_channels 4 -datamode LRHRLRPAN -lrpath /media/clh/Share/PanSharpening_Methods/PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch BIEDN -opt options/train/train_demo.yml -in_channels 4 -datamode LRHRLRPAN -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
#python train.py -net_arch PanNet -opt options/train/train_demo.yml -in_channels 8 -lrpath ../PanSharp_dataset/WV2/train/MTF/8bands/LRMS_npy
#python train.py -net_arch PanNet -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Share/PanSharpening_Methods/PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch PanNet -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 8 -lrpath ../PanSharp_dataset/WV2/train/MTF/8bands/LRMS_npy
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath /media/clh/Files/PanSharp_dataset_20200922/IK/train/MTF/4bands/LRMS_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/1bands_rand/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/2bands_rand/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/3bands_rand/LRMS_npy -panpath ../PanSharp_dadtaset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
#python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_rand/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/3bands_sort/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB_20200922/train/MTF/2bands_sort/LRMS_npy -panpath ../PanSharp_dataset/QB_20200922/train/MTF/4bands_sort/LRPAN_npy
############ ablation study on MIX-dataset ############
# CUDA_VISIBLE_DEVICES=0, nohup python train.py -net_arch RNN_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml -in_channels 4 > RNN_MIX4.txt 2>&1 &
# CUDA_VISIBLE_DEVICES=0, nohup python train.py -net_arch RNN_Residual_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml -in_channels 4 > RNN_RL_MIX4.txt 2>&1 &
# CUDA_VISIBLE_DEVICES=0, nohup python train.py -net_arch RNN_Residual_Biv2_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml -in_channels 4 > BiRNN_LR_MIX4.txt 2>&1 &
# CUDA_VISIBLE_DEVICES=0, nohup python train.py -net_arch RNN_Residual_Bi_pan_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml -in_channels 4 > ArbRPN_MIX4.txt 2>&1 &
# CUDA_VISIBLE_DEVICES=0, nohup python train.py -net_arch RNN_Residual_Bi_pan_indepentv2_mask -num_resblock 1 -num_cycle 5 -opt options/train/train_demo.yml -in_channels 4 -lrpath ../PanSharp_dataset/QB/train/MTF/MIX_4_rand/LRMS_npy -panpath ../PanSharp_dataset/QB/train/MTF/4bands/LRPAN_npy > ArbRPN_SD_MIX4.txt 2>&1 &
########### train with different training strategies #############
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 > mix4sort.txt 2>&1 &
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name FIX4Rand -lrpath ../PanSharp_dataset/QB/train/MTF/4bands_rand/LRMS_npy -panpath ../PanSharp_dataset/QB/train/MTF/4bands/LRPAN_npy > FIX4Rand.txt 2>&1 &
# python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name mix4sort_batch16_lr0002 > mix4sort_batch16_lr0002.txt 2>&1 &
# nohup python train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name train_strategies/rangFIX1-4 > rangFIX1-4.txt 2>&1 &
# nohup python train.py -net_arch paraBiRNN -num_resblock 3 -num_cycle 1 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name paraBiRNN > paraBiRNN.txt 2>&1 &
# nohup python train_iterations.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name train_strategies/rangFIX1-4-iteration > rangFIX1-4-iteration.txt 2>&1 &
# nohup python train.py -net_arch paraBiRNN -num_resblock 3 -num_cycle 1 -loss myloss -opt options/train/train_demo.yml -in_channels 4 -save_name paraBiRNN > paraBiRNN.txt 2>&1 &
# train.py -net_arch RNN_Residual_Bi_pan_fb_mask -num_resblock 1 -num_cycle 5 -loss myloss -opt options/train/train_demo.yml -in_channels 4 > mix4sort.txt 2>&1 &
# sleep 210m
python train.py -net_arch TFNET -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/TFNet
python train.py -net_arch PNN -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/PNN
python train.py -net_arch MSDCNN -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/MSDCNN
python train.py -net_arch DRPNN -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/DRPNN
python train.py -net_arch DiCNN1 -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/DICNN1
python train.py -net_arch PanNet -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/PANNET
python train.py -net_arch DILATED -loss l1 -opt options/train/train_demo.yml -in_channels 4 -save_name SOTA_NBU_WV4/DILATED
python train.py -net_arch ArbRPN -loss myloss -num_resblock 3 -num_cycle 5 -opt options/train/train_demo.yml -save_name SOTA_NBU_WV4/ArbRPN_woInit