-
Notifications
You must be signed in to change notification settings - Fork 126
/
modelcard.sh
163 lines (124 loc) · 10.7 KB
/
modelcard.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
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
function dataset_to_memory() {
# README: copy data into memory
tar --use-compress-program=pigz -cvf ImageNet_ILSVRC2012.tar.pz ImageNet_ILSVRC2012/
sudo mount -t tmpfs -o size=150G tmpfs .media/memfs/
tar --use-compress-program=pigz -xvf ImageNet_ILSVRC2012.tar.pz -C /media/memfs/ # 5min
}
function classification() {
# VMambav0-T ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav0_tiny_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMambav0-S ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav0_small_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMambav0-B ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav0_base_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-T[s2l5] ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2_tiny_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-T ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2v_tiny_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-S ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2_small_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-S[s1l20] ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2v_small_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-B ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2_base_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
# VMamba-B[s1l20] ======================================================
torchrun --nproc_per_node 8 --master_port 12340 main.py --cfg configs/vssm/vmambav2v_base_224.yaml --data-path /media/memfs/ImageNet_ILSVRC2012 --output ../../out
}
function detection() {
mkdir -p detection/data
ln -s /media/Disk1/Dataset/MSCOCO2017 detection/data/coco
cd detection
# VMambav0-T ======================================================
export CKPT="publish/vssm0/classification/vssm0_tiny/vssm0_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/mask_rcnn_vssm_fpn_coco_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMambav0-S ======================================================
export CKPT="publish/vssm0/classification/vssm0_small/vssm0_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/mask_rcnn_vssm_fpn_coco_small.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMambav0-B ======================================================
export CKPT="publish/vssm0/classification/vssm0_base/vssm0_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/mask_rcnn_vssm_fpn_coco_base.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T[s2l5] ======================================================
export CKPT="publish/vssm1/classification/vssm1_tiny/vssm1_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_tiny1.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T ======================================================
export CKPT="publish/vssm2/classification/vssm2_tiny/vssm2_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-S ======================================================
export CKPT="publish/vssm1/classification/vssm1_small/vssm1_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_small.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-B ======================================================
export CKPT="publish/vssm1/classification/vssm1_base/vssm1_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_base.py 8 --cfg-options model.backbone.pretrained=$CKPT
}
function detection_ms3x() {
mkdir -p detection/data
ln -s /media/Disk1/Dataset/MSCOCO2017 detection/data/coco
cd detection
# VMambav0-T ======================================================
export CKPT="publish/vssm0/classification/vssm0_tiny/vssm0_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/mask_rcnn_vssm_fpn_coco_tiny_ms_3x.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMambav0-S ======================================================
export CKPT="publish/vssm0/classification/vssm0_small/vssm0_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/mask_rcnn_vssm_fpn_coco_small_ms_3x.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T[s2l5] ======================================================
export CKPT="publish/vssm1/classification/vssm1_tiny/vssm1_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_tiny1_ms_3x.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T ======================================================
export CKPT="publish/vssm2/classification/vssm2_tiny/vssm2_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_tiny_ms_3x.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-S ======================================================
export CKPT="publish/vssm1/classification/vssm1_small/vssm1_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/mask_rcnn_vssm_fpn_coco_small_ms_3x.py 8 --cfg-options model.backbone.pretrained=$CKPT
}
function segmentation() {
mkdir -p segmentation/data/ade
ln -s /media/Disk1/Dataset/ADEChallengeData2016 segmentation/data/ade
cd detection
# VMambav0-T ======================================================
export CKPT="publish/vssm0/classification/vssm0_tiny/vssm0_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMambav0-S ======================================================
export CKPT="publish/vssm0/classification/vssm0_small/vssm0_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_small.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMambav0-B ======================================================
export CKPT="publish/vssm0/classification/vssm0_base/vssm0_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_base.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T[s2l5] ======================================================
export CKPT="publish/vssm1/classification/vssm1_tiny/vssm1_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_tiny1.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-T ======================================================
export CKPT="publish/vssm2/classification/vssm2_tiny/vssm2_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-S ======================================================
export CKPT="publish/vssm1/classification/vssm1_small/vssm1_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_small.py 8 --cfg-options model.backbone.pretrained=$CKPT
# VMamba-B ======================================================
export CKPT="publish/vssm1/classification/vssm1_base/vssm1_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_train.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_base.py 8 --cfg-options model.backbone.pretrained=$CKPT
}
function segmentation_test_tta() {
mkdir -p segmentation/data/ade
ln -s /media/Disk1/Dataset/ADEChallengeData2016 segmentation/data/ade
cd segmentation
# VMambav0-T ======================================================
export CKPT="publish/vssm0/segmentation/vssm0_tiny/vssm0_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMambav0-S ======================================================
export CKPT="publish/vssm0/segmentation/vssm0_small/vssm0_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_small.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMambav0-B ======================================================
export CKPT="publish/vssm0/segmentation/vssm0_base/vssm0_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm/upernet_vssm_4xb4-160k_ade20k-512x512_base.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMamba-T[s2l5] ======================================================
export CKPT="publish/vssm1/segmentation/vssm1_tiny/vssm1_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_tiny1.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMamba-T ======================================================
export CKPT="publish/vssm2/segmentation/vssm2_tiny/vssm2_tiny_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_tiny.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMamba-S ======================================================
export CKPT="publish/vssm1/segmentation/vssm1_small/vssm1_small_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_small.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
# VMamba-B ======================================================
export CKPT="publish/vssm1/segmentation/vssm1_base/vssm1_base_ckpt_epoch.pth"
PORT=12345 bash ./tools/dist_test.sh configs/vssm1/upernet_vssm_4xb4-160k_ade20k-512x512_base.py 8 --cfg-options model.backbone.pretrained=$CKPT --tta
}