File size: 5,682 Bytes
17fbb28 c51087c c880bb4 c51087c 70e3f22 c51087c 17fbb28 18bce8b 70b7d54 a8c98ff 4d780cd 21c63d6 4b388ed b9a5ff6 70b7d54 3482f7b b9a5ff6 18bce8b 3482f7b e09029c 18bce8b 70b7d54 52bbeff a60b344 3482f7b 70b7d54 da3654b fe78cd0 da3654b fe78cd0 ce78729 5782f7d 254257c 596dd4d |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 |
---
license: mit
task_categories:
- image-classification
- feature-extraction
language:
- en
tags:
- code
pretty_name: Vi-Backbones
size_categories:
- n<1K
---
# Dataset Card for "monetjoe/cv_backbones"
This repository consolidates the collection of backbone networks for pre-trained computer vision models available on the PyTorch official website. It mainly includes various Convolutional Neural Networks (CNNs) and Vision Transformer models pre-trained on the ImageNet1K dataset. The entire collection is divided into two subsets, V1 and V2, encompassing multiple classic and advanced versions of visual models. These pre-trained backbone networks provide users with a robust foundation for transfer learning in tasks such as image recognition, object detection, and image segmentation. Simultaneously, it offers a convenient choice for researchers and practitioners to flexibly apply these pre-trained models in different scenarios.
## Data structure
| ver | type | input_size | url |
| :-----------: | :-----------: | :--------------: | :-------------------------------: |
| backbone name | backbone type | input image size | url of pretrained model .pth file |
## Maintenance
```bash
git clone [email protected]:datasets/monetjoe/cv_backbones
cd cv_backbones
```
## Usage
### ImageNet V1
```python
from datasets import load_dataset
backbones = load_dataset("monetjoe/cv_backbones", name="default", split="train")
for weights in backbones:
print(weights)
```
### ImageNet V2
```python
from datasets import load_dataset
backbones = load_dataset("monetjoe/cv_backbones", name="default", split="test")
for weights in backbones:
print(weights)
```
## Param counts of different backbones
### IMAGENET1K_V1
| Backbone | Params(M) |
| :----------------: | :-------: |
| SqueezeNet1_0 | 1.2 |
| SqueezeNet1_1 | 1.2 |
| ShuffleNet_V2_X0_5 | 1.4 |
| MNASNet0_5 | 2.2 |
| ShuffleNet_V2_X1_0 | 2.3 |
| MobileNet_V3_Small | 2.5 |
| MNASNet0_75 | 3.2 |
| MobileNet_V2 | 3.5 |
| ShuffleNet_V2_X1_5 | 3.5 |
| RegNet_Y_400MF | 4.3 |
| MNASNet1_0 | 4.4 |
| EfficientNet_B0 | 5.3 |
| MobileNet_V3_Large | 5.5 |
| RegNet_X_400MF | 5.5 |
| MNASNet1_3 | 6.3 |
| RegNet_Y_800MF | 6.4 |
| GoogLeNet | 6.6 |
| RegNet_X_800MF | 7.3 |
| ShuffleNet_V2_X2_0 | 7.4 |
| EfficientNet_B1 | 7.8 |
| DenseNet121 | 8 |
| EfficientNet_B2 | 9.1 |
| RegNet_X_1_6GF | 9.2 |
| RegNet_Y_1_6GF | 11.2 |
| ResNet18 | 11.7 |
| EfficientNet_B3 | 12.2 |
| DenseNet169 | 14.1 |
| RegNet_X_3_2GF | 15.3 |
| EfficientNet_B4 | 19.3 |
| RegNet_Y_3_2GF | 19.4 |
| DenseNet201 | 20 |
| EfficientNet_V2_S | 21.5 |
| ResNet34 | 21.8 |
| ResNeXt50_32X4D | 25 |
| ResNet50 | 25.6 |
| Inception_V3 | 27.2 |
| Swin_T | 28.3 |
| Swin_V2_T | 28.4 |
| ConvNeXt_Tiny | 28.6 |
| DenseNet161 | 28.7 |
| EfficientNet_B5 | 30.4 |
| MaxVit_T | 30.9 |
| RegNet_Y_8GF | 39.4 |
| RegNet_X_8GF | 39.6 |
| EfficientNet_B6 | 43 |
| ResNet101 | 44.5 |
| Swin_S | 49.6 |
| Swin_V2_S | 49.7 |
| ConvNeXt_Small | 50.2 |
| EfficientNet_V2_M | 54.1 |
| RegNet_X_16GF | 54.3 |
| ResNet152 | 60.2 |
| AlexNet | 61.1 |
| EfficientNet_B7 | 66.3 |
| Wide_ResNet50_2 | 68.9 |
| ResNeXt101_64X4D | 83.5 |
| RegNet_Y_16GF | 83.6 |
| ViT_B_16 | 86.6 |
| Swin_B | 87.8 |
| Swin_V2_B | 87.9 |
| ViT_B_32 | 88.2 |
| ConvNeXt_Base | 88.6 |
| ResNeXt101_32X8D | 88.8 |
| RegNet_X_32GF | 107.8 |
| EfficientNet_V2_L | 118.5 |
| Wide_ResNet101_2 | 126.9 |
| VGG11_BN | 132.9 |
| VGG11 | 132.9 |
| VGG13 | 133 |
| VGG13_BN | 133.1 |
| VGG16_BN | 138.4 |
| VGG16 | 138.4 |
| VGG19_BN | 143.7 |
| VGG19 | 143.7 |
| RegNet_Y_32GF | 145 |
| ConvNeXt_Large | 197.8 |
| ViT_L_16 | 304.3 |
| ViT_L_32 | 306.5 |
### IMAGENET1K_V2
| Backbone | Params(M) |
| :----------------: | :-------: |
| MobileNet_V2 | 3.5 |
| RegNet_Y_400MF | 4.3 |
| MobileNet_V3_Large | 5.5 |
| RegNet_X_400MF | 5.5 |
| RegNet_Y_800MF | 6.4 |
| RegNet_X_800MF | 7.3 |
| EfficientNet_B1 | 7.8 |
| RegNet_X_1_6GF | 9.2 |
| RegNet_Y_1_6GF | 11.2 |
| RegNet_X_3_2GF | 15.3 |
| RegNet_Y_3_2GF | 19.4 |
| ResNeXt50_32X4D | 25 |
| ResNet50 | 25.6 |
| RegNet_Y_8GF | 39.4 |
| RegNet_X_8GF | 39.6 |
| ResNet101 | 44.5 |
| RegNet_X_16GF | 54.3 |
| ResNet152 | 60.2 |
| Wide_ResNet50_2 | 68.9 |
| RegNet_Y_16GF | 83.6 |
| ResNeXt101_32X8D | 88.8 |
| RegNet_X_32GF | 107.8 |
| Wide_ResNet101_2 | 126.9 |
| RegNet_Y_32GF | 145 |
## Mirror
<https://www.modelscope.cn/datasets/monetjoe/cv_backbones>
## Reference
[1] <https://pytorch.org/vision/main/_modules><br>
[2] <https://pytorch.org/vision/main/models.html> |