This repository contains deep learning models. Keep updating....
-
Top-1, Top-5 are accuracy in imagenet val which I get from Internet.
-
C -> Caffe, M -> MXNet, C2 -> Caffe2 T -> Tensorflow
| Model | Top-1 | Top-5 | Caffe | MXNet | Caffe2 | TF |
|---|---|---|---|---|---|---|
| alexnet | C_57.1 | C_80.2 | 243.86M | |||
| caffenet | C_57.4 M_54.5 | C_80.4 M_78.3 | 243.86M | 233M | ||
| NIN | C_59.36 M_58.8 | M_81.3 | 29M | 29M | ||
| SqueezeNet1.0 | M_55.4 | C_>=80.3 M_78.8 | 4.8M | 4.8M | ||
| SqueezeNet1.1 2.4x less computation than 1.0 | C_sameAs1.0 | C_>=80.3 | 4.7M | 4.7M | ||
| VGG16 | M_71.0 | C_92.5 M_89.9 T_71.5 | 528M | 528M | ||
| VGG19 | M_71.0 | C_92.5 M_89.8 T_89.8 | 548M | 548M | * | |
| GoogleNet(Inception V1) | C_68.7 T_69.8 | C_88.9 T_89.6 | 53.53M | * | ||
| Inception-v2(Inception-bn) | M_72.5 T_73.9 | M_90.8 T_91.8 | 134.6M ImageNet21k | 43M ImageNet10k | * | |
| Inception-V3 | M_76.88 T_78.0 | M_93.344 T_93.9 | 95.6M | * | ||
| Inception V4 | M_79.83 T_80.2 | M_94.69 T_95.2 | [164M] | * | ||
| Inception-ResNet-v2 | M_80.13 T_80.4 | M_95.18 T_95.3 | [213.33MB] | * | ||
| ResNet-18 | C_69 M_69.52 | M_89 M_89.08 | 45M | 45M | ||
| ResNet-34 | M_72.8 | M_91.14 | 83M | |||
| ResNet-50 | C_75.3 M_75.61 T_75.2 | C_92.2 M_92.76 T_92.2 | 98M | 98M | * | |
| ResNet-101 | C_76.4 M_77.32 T_76.4 | C_92.9 M_93.42 T_92.9 | 171M | 170M | * | |
| ResNet-152 | C_77.0 M_77.75 T_76.8 | C_93.3 M_93.58 T_93.2 | 231M | 230M | * | |
| ResNet-200 | M_77.86 | M_93.84 | 247M | |||
| ResNet V2 50 | T_75.6 | T_92.8 | * | |||
| ResNet V2 101 | T_77.0 | T_93.7 | * | |||
| ResNet V2 152 | T_77.8 | T_94.1 | * | |||
| ResNeXt-50 | C_78.1 M_76.89 | C_94.1 M_93.32 | 95.76M | 96M | ||
| ResNeXt-101 | C_79.8 M_78.28 | C_95.1 M_94.08 | 169.10M | 169M | ||
| ResNeXt-152 | C_80.1 | C_95.2 | 229.53M | |||
| RexNeXt-101-64x4d | C_80.5 M_79.11 | C_95.2 M_94.30 | 319M | |||
| WRN-50-2-bottleneck | C_77.87 | C_93.87 | 263.1M | |||
| MobileNet | C_70.81 M_71.24 T_66.51 | C_89.85 M_90.15 T_87.09 | 16.2M | 16.2M | * | |
| DenseNet 121(k=32) | C_74.91 | C_92.19 | 30.8M | |||
| DenseNet 161(k=48) | C_77.64 | C_93.79 | 110M | |||
| DenseNet 169(k=32) | C_76.09 | C_93.14 | 54.6M | |||
| DenseNet 201(k=32) | C_77.31 | C_93.64 | 77.3M | sym | ||
| CRU-Net-56 @x14 32x4d | M_78.1 | 98M | ||||
| CRU-Net-56 @x14 136x1d | M_78.3 | 98M | ||||
| CRU-Net-116 @x28x14 32x4d | M_79.4 | 168M | ||||
| CRU-Net-116, wider @x28x14 64x4d | M_79.7 | 318M | ||||
| CRU-Net-56, tiny @x14 32x4d | M_77.1 | 48MB | ||||
| DPN-92 | M_79.27 | M_94.63 | 145M | |||
| DPN-98 | M_79.85 | M_94.85 | 236M | |||
| DPN-131 | M_80.07 | M_94.88 | 304M | |||
| DPN-107* | M_80.25 | M_95.06 | 333M | |||
| PolyNet | C_81.289 | C_95.75 | 463M |
- Anything helps this repo, including discussion, testing, promotion and of course your awesome code.