Hugging Face's logo

Datasets:
monetjoe
/
cv_backbones

Tasks:
Image Classification
Feature Extraction
Modalities:
Text
Formats:
json
Languages:
English
Size:
< 1K
Tags:
code
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
admin commited on
Commit
fe78cd0
·
1 Parent(s): da3654b

upd md

Browse files
Files changed (1) hide show
  1. README.md +107 -120
README.md CHANGED
@@ -18,7 +18,7 @@ This repository consolidates the collection of backbone networks for pre-trained
18
 
19
  ## Viewer
20
  <https://huggingface.co/spaces/monet-joe/cv-backbones>
21
- | ver | type | input_size(NxN) | url(https://download.pytorch.org/models/*) |
22
  | :-----------------------------------------: | :-------------------------------: | :-------------------------------: | :--------------------------------------------------------------------------------------------------------------: |
23
  | backbone name(for example wide_resnet101_2) | backbone type(for example resnet) | input image size(for example 224) | url of pretrained model .pth file(for example https://download.pytorch.org/models/wide_resnet101_2-d733dc28.pth) |
24
 
@@ -42,127 +42,114 @@ for weights in backbones["IMAGENET1K_V2"]:
42
 
43
  ## Param count
44
  ### IMAGENET1K_V1
45
- | Backbone | Params(M) |
46
- | :--: | :--: |
47
- | SqueezeNet1_0 | 1.2 |
48
- | SqueezeNet1_1 | 1.2 |
49
- | ShuffleNet_V2_X0_5 | 1.4 |
50
- | MNASNet0_5 | 2.2 |
51
- | ShuffleNet_V2_X1_0 | 2.3 |
52
- | MobileNet_V3_Small | 2.5 |
53
- | MNASNet0_75 | 3.2 |
54
- | MobileNet_V2 | 3.5 |
55
- | ShuffleNet_V2_X1_5 | 3.5 |
56
- | RegNet_Y_400MF | 4.3 |
57
- | MNASNet1_0 | 4.4 |
58
- | EfficientNet_B0 | 5.3 |
59
- | MobileNet_V3_Large | 5.5 |
60
- | RegNet_X_400MF | 5.5 |
61
- | MNASNet1_3 | 6.3 |
62
- | RegNet_Y_800MF | 6.4 |
63
- | GoogLeNet | 6.6 |
64
- | RegNet_X_800MF | 7.3 |
65
- | ShuffleNet_V2_X2_0 | 7.4 |
66
- | EfficientNet_B1 | 7.8 |
67
- | DenseNet121 | 8 |
68
- | EfficientNet_B2 | 9.1 |
69
- | RegNet_X_1_6GF | 9.2 |
70
- | RegNet_Y_1_6GF | 11.2 |
71
- | ResNet18 | 11.7 |
72
- | EfficientNet_B3 | 12.2 |
73
- | DenseNet169 | 14.1 |
74
- | RegNet_X_3_2GF | 15.3 |
75
- | EfficientNet_B4 | 19.3 |
76
- | RegNet_Y_3_2GF | 19.4 |
77
- | DenseNet201 | 20 |
78
- | EfficientNet_V2_S | 21.5 |
79
- | ResNet34 | 21.8 |
80
- | ResNeXt50_32X4D | 25 |
81
- | ResNet50 | 25.6 |
82
- | Inception_V3 | 27.2 |
83
- | Swin_T | 28.3 |
84
- | Swin_V2_T | 28.4 |
85
- | ConvNeXt_Tiny | 28.6 |
86
- | DenseNet161 | 28.7 |
87
- | EfficientNet_B5 | 30.4 |
88
- | MaxVit_T | 30.9 |
89
- | RegNet_Y_8GF | 39.4 |
90
- | RegNet_X_8GF | 39.6 |
91
- | EfficientNet_B6 | 43 |
92
- | ResNet101 | 44.5 |
93
- | Swin_S | 49.6 |
94
- | Swin_V2_S | 49.7 |
95
- | ConvNeXt_Small | 50.2 |
96
- | EfficientNet_V2_M | 54.1 |
97
- | RegNet_X_16GF | 54.3 |
98
- | ResNet152 | 60.2 |
99
- | AlexNet | 61.1 |
100
- | EfficientNet_B7 | 66.3 |
101
- | Wide_ResNet50_2 | 68.9 |
102
- | ResNeXt101_64X4D | 83.5 |
103
- | RegNet_Y_16GF | 83.6 |
104
- | RegNet_Y_16GF_Weights.IMAGENET1K_SWAG_E2E_V1 | 83.6 |
105
- | RegNet_Y_16GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 83.6 |
106
- | ViT_B_16 | 86.6 |
107
- | ViT_B_16_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 86.6 |
108
- | ViT_B_16_Weights.IMAGENET1K_SWAG_E2E_V1 | 86.9 |
109
- | Swin_B | 87.8 |
110
- | Swin_V2_B | 87.9 |
111
- | ViT_B_32 | 88.2 |
112
- | ConvNeXt_Base | 88.6 |
113
- | ResNeXt101_32X8D | 88.8 |
114
- | RegNet_X_32GF | 107.8 |
115
- | EfficientNet_V2_L | 118.5 |
116
- | Wide_ResNet101_2 | 126.9 |
117
- | VGG11_BN | 132.9 |
118
- | VGG11 | 132.9 |
119
- | VGG13 | 133 |
120
- | VGG13_BN | 133.1 |
121
- | VGG16_BN | 138.4 |
122
- | VGG16 | 138.4 |
123
- | VGG16_Weights.IMAGENET1K_FEATURES | 138.4 |
124
- | VGG19_BN | 143.7 |
125
- | VGG19 | 143.7 |
126
- | RegNet_Y_32GF | 145 |
127
- | RegNet_Y_32GF_Weights.IMAGENET1K_SWAG_E2E_V1 | 145 |
128
- | RegNet_Y_32GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 145 |
129
- | ConvNeXt_Large | 197.8 |
130
- | ViT_L_16 | 304.3 |
131
- | ViT_L_16_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 304.3 |
132
- | ViT_L_16_Weights.IMAGENET1K_SWAG_E2E_V1 | 305.2 |
133
- | ViT_L_32 | 306.5 |
134
- | ViT_H_14_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 632 |
135
- | ViT_H_14_Weights.IMAGENET1K_SWAG_E2E_V1 | 633.5 |
136
- | RegNet_Y_128GF_Weights.IMAGENET1K_SWAG_E2E_V1 | 644.8 |
137
- | RegNet_Y_128GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 | 644.8 |
138
 
139
  ### IMAGENET1K_V2
140
- | Backbone | Params(M) |
141
- | :--: | :--: |
142
- | MobileNet_V2 | 3.5 |
143
- | RegNet_Y_400MF | 4.3 |
144
- | MobileNet_V3_Large | 5.5 |
145
- | RegNet_X_400MF | 5.5 |
146
- | RegNet_Y_800MF | 6.4 |
147
- | RegNet_X_800MF | 7.3 |
148
- | EfficientNet_B1 | 7.8 |
149
- | RegNet_X_1_6GF | 9.2 |
150
- | RegNet_Y_1_6GF | 11.2 |
151
- | RegNet_X_3_2GF | 15.3 |
152
- | RegNet_Y_3_2GF | 19.4 |
153
- | ResNeXt50_32X4D | 25 |
154
- | ResNet50 | 25.6 |
155
- | RegNet_Y_8GF | 39.4 |
156
- | RegNet_X_8GF | 39.6 |
157
- | ResNet101 | 44.5 |
158
- | RegNet_X_16GF | 54.3 |
159
- | ResNet152 | 60.2 |
160
- | Wide_ResNet50_2 | 68.9 |
161
- | RegNet_Y_16GF | 83.6 |
162
- | ResNeXt101_32X8D | 88.8 |
163
- | RegNet_X_32GF | 107.8 |
164
- | Wide_ResNet101_2 | 126.9 |
165
- | RegNet_Y_32GF | 145 |
166
 
167
  ## Mirror
168
  <https://www.modelscope.cn/datasets/monetjoe/cv_backbones>
 
18
 
19
  ## Viewer
20
  <https://huggingface.co/spaces/monet-joe/cv-backbones>
21
+ | ver | type | input_size(NxN) | url(https://download.pytorch.org/models/*.pth) |
22
  | :-----------------------------------------: | :-------------------------------: | :-------------------------------: | :--------------------------------------------------------------------------------------------------------------: |
23
  | backbone name(for example wide_resnet101_2) | backbone type(for example resnet) | input image size(for example 224) | url of pretrained model .pth file(for example https://download.pytorch.org/models/wide_resnet101_2-d733dc28.pth) |
24
 
 
42
 
43
  ## Param count
44
  ### IMAGENET1K_V1
45
+ | Backbone | Params(M) |
46
+ | :----------------: | :-------: |
47
+ | SqueezeNet1_0 | 1.2 |
48
+ | SqueezeNet1_1 | 1.2 |
49
+ | ShuffleNet_V2_X0_5 | 1.4 |
50
+ | MNASNet0_5 | 2.2 |
51
+ | ShuffleNet_V2_X1_0 | 2.3 |
52
+ | MobileNet_V3_Small | 2.5 |
53
+ | MNASNet0_75 | 3.2 |
54
+ | MobileNet_V2 | 3.5 |
55
+ | ShuffleNet_V2_X1_5 | 3.5 |
56
+ | RegNet_Y_400MF | 4.3 |
57
+ | MNASNet1_0 | 4.4 |
58
+ | EfficientNet_B0 | 5.3 |
59
+ | MobileNet_V3_Large | 5.5 |
60
+ | RegNet_X_400MF | 5.5 |
61
+ | MNASNet1_3 | 6.3 |
62
+ | RegNet_Y_800MF | 6.4 |
63
+ | GoogLeNet | 6.6 |
64
+ | RegNet_X_800MF | 7.3 |
65
+ | ShuffleNet_V2_X2_0 | 7.4 |
66
+ | EfficientNet_B1 | 7.8 |
67
+ | DenseNet121 | 8 |
68
+ | EfficientNet_B2 | 9.1 |
69
+ | RegNet_X_1_6GF | 9.2 |
70
+ | RegNet_Y_1_6GF | 11.2 |
71
+ | ResNet18 | 11.7 |
72
+ | EfficientNet_B3 | 12.2 |
73
+ | DenseNet169 | 14.1 |
74
+ | RegNet_X_3_2GF | 15.3 |
75
+ | EfficientNet_B4 | 19.3 |
76
+ | RegNet_Y_3_2GF | 19.4 |
77
+ | DenseNet201 | 20 |
78
+ | EfficientNet_V2_S | 21.5 |
79
+ | ResNet34 | 21.8 |
80
+ | ResNeXt50_32X4D | 25 |
81
+ | ResNet50 | 25.6 |
82
+ | Inception_V3 | 27.2 |
83
+ | Swin_T | 28.3 |
84
+ | Swin_V2_T | 28.4 |
85
+ | ConvNeXt_Tiny | 28.6 |
86
+ | DenseNet161 | 28.7 |
87
+ | EfficientNet_B5 | 30.4 |
88
+ | MaxVit_T | 30.9 |
89
+ | RegNet_Y_8GF | 39.4 |
90
+ | RegNet_X_8GF | 39.6 |
91
+ | EfficientNet_B6 | 43 |
92
+ | ResNet101 | 44.5 |
93
+ | Swin_S | 49.6 |
94
+ | Swin_V2_S | 49.7 |
95
+ | ConvNeXt_Small | 50.2 |
96
+ | EfficientNet_V2_M | 54.1 |
97
+ | RegNet_X_16GF | 54.3 |
98
+ | ResNet152 | 60.2 |
99
+ | AlexNet | 61.1 |
100
+ | EfficientNet_B7 | 66.3 |
101
+ | Wide_ResNet50_2 | 68.9 |
102
+ | ResNeXt101_64X4D | 83.5 |
103
+ | RegNet_Y_16GF | 83.6 |
104
+ | ViT_B_16 | 86.6 |
105
+ | Swin_B | 87.8 |
106
+ | Swin_V2_B | 87.9 |
107
+ | ViT_B_32 | 88.2 |
108
+ | ConvNeXt_Base | 88.6 |
109
+ | ResNeXt101_32X8D | 88.8 |
110
+ | RegNet_X_32GF | 107.8 |
111
+ | EfficientNet_V2_L | 118.5 |
112
+ | Wide_ResNet101_2 | 126.9 |
113
+ | VGG11_BN | 132.9 |
114
+ | VGG11 | 132.9 |
115
+ | VGG13 | 133 |
116
+ | VGG13_BN | 133.1 |
117
+ | VGG16_BN | 138.4 |
118
+ | VGG16 | 138.4 |
119
+ | VGG19_BN | 143.7 |
120
+ | VGG19 | 143.7 |
121
+ | RegNet_Y_32GF | 145 |
122
+ | ConvNeXt_Large | 197.8 |
123
+ | ViT_L_16 | 304.3 |
124
+ | ViT_L_32 | 306.5 |
 
 
 
 
 
 
 
 
 
 
 
 
 
125
 
126
  ### IMAGENET1K_V2
127
+ | Backbone | Params(M) |
128
+ | :----------------: | :-------: |
129
+ | MobileNet_V2 | 3.5 |
130
+ | RegNet_Y_400MF | 4.3 |
131
+ | MobileNet_V3_Large | 5.5 |
132
+ | RegNet_X_400MF | 5.5 |
133
+ | RegNet_Y_800MF | 6.4 |
134
+ | RegNet_X_800MF | 7.3 |
135
+ | EfficientNet_B1 | 7.8 |
136
+ | RegNet_X_1_6GF | 9.2 |
137
+ | RegNet_Y_1_6GF | 11.2 |
138
+ | RegNet_X_3_2GF | 15.3 |
139
+ | RegNet_Y_3_2GF | 19.4 |
140
+ | ResNeXt50_32X4D | 25 |
141
+ | ResNet50 | 25.6 |
142
+ | RegNet_Y_8GF | 39.4 |
143
+ | RegNet_X_8GF | 39.6 |
144
+ | ResNet101 | 44.5 |
145
+ | RegNet_X_16GF | 54.3 |
146
+ | ResNet152 | 60.2 |
147
+ | Wide_ResNet50_2 | 68.9 |
148
+ | RegNet_Y_16GF | 83.6 |
149
+ | ResNeXt101_32X8D | 88.8 |
150
+ | RegNet_X_32GF | 107.8 |
151
+ | Wide_ResNet101_2 | 126.9 |
152
+ | RegNet_Y_32GF | 145 |
153
 
154
  ## Mirror
155
  <https://www.modelscope.cn/datasets/monetjoe/cv_backbones>