diff --git a/README.md b/README.md index 5c41a1e4a53e0540f2feb42cdd3d25863ce35d24..2ddbb254341a7a3b0f924bda682c058f0c58a10c 100644 --- a/README.md +++ b/README.md @@ -7,6 +7,8 @@ PaddleClas is a toolset for image classification tasks prepared for the industry and academia. It helps users train better computer vision models and apply them in real scenarios. **Recent update** + +- 2021.05.14 Add `SwinTransformer` series pretrained models, whose Top-1 Acc on ImageNet-1k dataset reaches 87.19%. - 2021.04.15 Add `MixNet` and `ReXNet` pretrained models, `MixNet_L`'s Top-1 Acc on ImageNet-1k reaches 78.6% and `ReXNet_3_0` reaches 82.09%. - 2021.03.02 Add support for model quantization. - 2021.02.01 Add `RepVGG` pretrained models, whose Top-1 Acc on ImageNet-1k dataset reaches 79.65%. @@ -63,10 +65,11 @@ PaddleClas is a toolset for image classification tasks prepared for the industry - [Inception series](#Inception_series) - [EfficientNet and ResNeXt101_wsl series](#EfficientNet_and_ResNeXt101_wsl_series) - [ResNeSt and RegNet series](#ResNeSt_and_RegNet_series) - - [Transformer series](#Transformer) + - [ViT and DeiT series](#ViT_and_DeiT) - [RepVGG series](#RepVGG) - [MixNet series](#MixNet) - [ReXNet series](#ReXNet) + - [SwinTransformer series](#SwinTransformer) - [Others](#Others) - HS-ResNet: arxiv link: [https://arxiv.org/pdf/2010.07621.pdf](https://arxiv.org/pdf/2010.07621.pdf). Code and models are coming soon! - Model training/evaluation @@ -353,10 +356,10 @@ Accuracy and inference time metrics of ResNeSt and RegNet series models are show | RegNetX_4GF | 0.785 | 0.9416 | 6.46478 | 11.19862 | 8 | 22.1 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetX_4GF_pretrained.pdparams) | - -### Transformer series + +### ViT and DeiT series -Accuracy and inference time metrics of ViT and DeiT series models are shown as follows. More detailed information can be refered to [Transformer series tutorial](./docs/en/models/Transformer_en.md). +Accuracy and inference time metrics of ViT and DeiT series models are shown as follows. More detailed information can be refered to [ViT and DeiT series tutorial](./docs/en/models/ViT_and_DeiT_en.md). | Model | Top-1 Acc | Top-5 Acc | time(ms)
bs=1 | time(ms)
bs=4 | Flops(G) | Params(M) | Download Address | @@ -430,6 +433,25 @@ Accuracy and inference time metrics of ReXNet series models are shown as follows | ReXNet_2_0 | 0.8122 | 0.9536 | | | 1.561 | 16.449 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ReXNet_2_0_pretrained.pdparams) | | ReXNet_3_0 | 0.8209 | 0.9612 | | | 3.445 | 34.833 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ReXNet_3_0_pretrained.pdparams) | + + +### SwinTransformer + +Accuracy and inference time metrics of SwinTransformer series models are shown as follows. More detailed information can be refered to [SwinTransformer series tutorial](./docs/en/models/SwinTransformer_en.md). + +| Model | Top-1 Acc | Top-5 Acc | time(ms)
bs=1 | time(ms)
bs=4 | Flops(G) | Params(M) | Download Address | +| ---------- | --------- | --------- | ---------------- | ---------------- | -------- | --------- | ------------------------------------------------------------ | +| SwinTransformer_tiny_patch4_window7_224 | 0.8069 | 0.9534 | | | 4.5 | 28 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_tiny_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_small_patch4_window7_224 | 0.8275 | 0.9613 | | | 8.7 | 50 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_small_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_base_patch4_window7_224 | 0.8300 | 0.9626 | | | 15.4 | 88 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_base_patch4_window12_384 | 0.8439 | 0.9693 | | | 47.1 | 88 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window12_384_pretrained.pdparams) | +| SwinTransformer_base_patch4_window7_224[1] | 0.8487 | 0.9746 | | | 15.4 | 88 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window7_224_22kto1k_pretrained.pdparams) | +| SwinTransformer_base_patch4_window12_384[1] | 0.8642 | 0.9807 | | | 47.1 | 88 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window12_384_22kto1k_pretrained.pdparams) | +| SwinTransformer_large_patch4_window7_224[1] | 0.8596 | 0.9783 | | | 34.5 | 197 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window7_224_22kto1k_pretrained.pdparams) | +| SwinTransformer_large_patch4_window12_384[1] | 0.8719 | 0.9823 | | | 103.9 | 197 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window12_384_22kto1k_pretrained.pdparams) | + +[1]: Based on imagenet22k dataset pre-training, and then in imagenet1k dataset transfer learning. + ### Others diff --git a/README_cn.md b/README_cn.md index 602c9aacb7b2e10bd96b483dfc16490d30d13c37..e3291c263873339d80959a845c25e444b482633f 100644 --- a/README_cn.md +++ b/README_cn.md @@ -8,7 +8,8 @@ **近期更新** -- 2021.04.15 添加`MixNet`和`ReXNet`系列模型，在ImageNet-1k上`MixNet_L` 模型Top1 Acc可达78.6%，`ReXNet_3_0`模型可达82.09% +- 2021.05.14 添加`SwinTransformer` 系列模型，在ImageNet-1k上，Top1 Acc可达87.19% +- 2021.04.15 添加`MixNet`和`ReXNet`系列模型，在ImageNet-1k上`MixNet_L`模型Top1 Acc可达78.6%，`ReXNet_3_0`模型可达82.09% - 2021.03.02 添加分类模型量化方法与使用教程。 - 2021.02.01 添加`RepVGG`系列模型，在ImageNet-1k上Top-1 Acc可达79.65%。 - 2021.01.27 添加`ViT`与`DeiT`模型，在ImageNet-1k上，`ViT`模型Top-1 Acc可达85.13%，`DeiT`模型可达85.1%。 @@ -65,10 +66,11 @@ - [Inception系列](#Inception系列) - [EfficientNet与ResNeXt101_wsl系列](#EfficientNet与ResNeXt101_wsl系列) - [ResNeSt与RegNet系列](#ResNeSt与RegNet系列) - - [Transformer系列](#Transformer系列) + - [ViT与DeiT系列](#ViT_and_DeiT系列) - [RepVGG系列](#RepVGG系列) - [MixNet系列](#MixNet系列) - [ReXNet系列](#ReXNet系列) + - [SwinTransformer系列](#SwinTransformer系列) - [其他模型](#其他模型) - HS-ResNet: arxiv文章链接: [https://arxiv.org/pdf/2010.07621.pdf](https://arxiv.org/pdf/2010.07621.pdf)。 代码和预训练模型即将开源，敬请期待。 - 模型训练/评估 @@ -358,10 +360,10 @@ ResNeSt与RegNet系列模型的精度、速度指标如下表所示，更多关 | RegNetX_4GF | 0.785 | 0.9416 | 6.46478 | 11.19862 | 8 | 22.1 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetX_4GF_pretrained.pdparams) | - -### Transformer系列 + +### ViT_and_DeiT系列 -ViT（Vision Transformer）与DeiT（Data-efficient Image Transformers）系列模型的精度、速度指标如下表所示. 更多关于该系列模型的介绍可以参考： [Transformer系列模型文档](./docs/zh_CN/models/Transformer.md)。 +ViT（Vision Transformer）与DeiT（Data-efficient Image Transformers）系列模型的精度、速度指标如下表所示. 更多关于该系列模型的介绍可以参考： [ViT_and_DeiT系列模型文档](./docs/zh_CN/models/ViT_and_DeiT.md)。 | 模型 | Top-1 Acc | Top-5 Acc | time(ms)
bs=1 | time(ms)
bs=4 | Flops(G) | Params(M) | 下载地址 | @@ -434,6 +436,25 @@ ViT（Vision Transformer）与DeiT（Data-efficient Image Transformers）系列 | ReXNet_2_0 | 0.8122 | 0.9536 | | | 1.561 | 16.449 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ReXNet_2_0_pretrained.pdparams) | | ReXNet_3_0 | 0.8209 | 0.9612 | | | 3.445 | 34.833 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ReXNet_3_0_pretrained.pdparams) | + + +### SwinTransformer系列 + +关于SwinTransformer系列模型的精度、速度指标如下表所示，更多介绍可以参考：[SwinTransformer系列模型文档](./docs/zh_CN/models/SwinTransformer.md)。 + +| 模型 | Top-1 Acc | Top-5 Acc | time(ms)
bs=1 | time(ms)
bs=4 | Flops(G) | Params(M) | 下载地址 | +| ---------- | --------- | --------- | ---------------- | ---------------- | -------- | --------- | ------------------------------------------------------------ | +| SwinTransformer_tiny_patch4_window7_224 | 0.8069 | 0.9534 | | | 4.5 | 28 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_tiny_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_small_patch4_window7_224 | 0.8275 | 0.9613 | | | 8.7 | 50 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_small_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_base_patch4_window7_224 | 0.8300 | 0.9626 | | | 15.4 | 88 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window7_224_pretrained.pdparams) | +| SwinTransformer_base_patch4_window12_384 | 0.8439 | 0.9693 | | | 47.1 | 88 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window12_384_pretrained.pdparams) | +| SwinTransformer_base_patch4_window7_224[1] | 0.8487 | 0.9746 | | | 15.4 | 88 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window7_224_22kto1k_pretrained.pdparams) | +| SwinTransformer_base_patch4_window12_384[1] | 0.8642 | 0.9807 | | | 47.1 | 88 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window12_384_22kto1k_pretrained.pdparams) | +| SwinTransformer_large_patch4_window7_224[1] | 0.8596 | 0.9783 | | | 34.5 | 197 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window7_224_22kto1k_pretrained.pdparams) | +| SwinTransformer_large_patch4_window12_384[1] | 0.8719 | 0.9823 | | | 103.9 | 197 | [下载链接](https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window12_384_22kto1k_pretrained.pdparams) | + +[1]：基于ImageNet22k数据集预训练，然后在ImageNet1k数据集迁移学习得到。 + ### 其他模型 diff --git a/configs/SwinTransformer/SwinTransformer_large_patch4_window12_384.yaml b/configs/SwinTransformer/SwinTransformer_large_patch4_window12_384.yaml new file mode 100644 index 0000000000000000000000000000000000000000..4d05754d142ee6268b6b5b632eb3c2bf7cfd1eeb --- /dev/null +++ b/configs/SwinTransformer/SwinTransformer_large_patch4_window12_384.yaml @@ -0,0 +1,72 @@ +mode: 'train' +ARCHITECTURE: + name: 'SwinTransformer_large_patch4_window12_384' + +pretrained_model: "" +model_save_dir: "./output/" +classes_num: 1000 +total_images: 1281167 +save_interval: 1 +validate: True +valid_interval: 1 +epochs: 120 +topk: 5 +image_shape: [3, 384, 384] + +use_mix: False +ls_epsilon: -1 + +LEARNING_RATE: + function: 'Piecewise' + params: + lr: 0.1 + decay_epochs: [30, 60, 90] + gamma: 0.1 + +OPTIMIZER: + function: 'Momentum' + params: + momentum: 0.9 + regularizer: + function: 'L2' + factor: 0.000100 + +TRAIN: + batch_size: 256 + num_workers: 4 + file_list: "./dataset/ILSVRC2012/train_list.txt" + data_dir: "./dataset/ILSVRC2012/" + shuffle_seed: 0 + transforms: + - DecodeImage: + to_rgb: True + channel_first: False + - RandCropImage: + size: 224 + - RandFlipImage: + flip_code: 1 + - NormalizeImage: + scale: 1./255. + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: '' + - ToCHWImage: + +VALID: + batch_size: 128 + num_workers: 4 + file_list: "./dataset/ILSVRC2012/val_list.txt" + data_dir: "./dataset/ILSVRC2012/" + shuffle_seed: 0 + transforms: + - DecodeImage: + to_rgb: True + channel_first: False + - ResizeImage: + size: [384, 384] + - NormalizeImage: + scale: 1.0/255.0 + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: '' + - ToCHWImage: diff --git a/configs/SwinTransformer/SwinTransformer_large_patch4_window7_224.yaml b/configs/SwinTransformer/SwinTransformer_large_patch4_window7_224.yaml new file mode 100644 index 0000000000000000000000000000000000000000..f2481da1121611f1c278115fbe28b18e41a7286c --- /dev/null +++ b/configs/SwinTransformer/SwinTransformer_large_patch4_window7_224.yaml @@ -0,0 +1,74 @@ +mode: 'train' +ARCHITECTURE: + name: 'SwinTransformer_large_patch4_window7_224' + +pretrained_model: "" +model_save_dir: "./output/" +classes_num: 1000 +total_images: 1281167 +save_interval: 1 +validate: True +valid_interval: 1 +epochs: 120 +topk: 5 +image_shape: [3, 224, 224] + +use_mix: False +ls_epsilon: -1 + +LEARNING_RATE: + function: 'Piecewise' + params: + lr: 0.1 + decay_epochs: [30, 60, 90] + gamma: 0.1 + +OPTIMIZER: + function: 'Momentum' + params: + momentum: 0.9 + regularizer: + function: 'L2' + factor: 0.000100 + +TRAIN: + batch_size: 256 + num_workers: 4 + file_list: "./dataset/ILSVRC2012/train_list.txt" + data_dir: "./dataset/ILSVRC2012/" + shuffle_seed: 0 + transforms: + - DecodeImage: + to_rgb: True + channel_first: False + - RandCropImage: + size: 224 + - RandFlipImage: + flip_code: 1 + - NormalizeImage: + scale: 1./255. + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: '' + - ToCHWImage: + +VALID: + batch_size: 64 + num_workers: 4 + file_list: "./dataset/ILSVRC2012/val_list.txt" + data_dir: "./dataset/ILSVRC2012/" + shuffle_seed: 0 + transforms: + - DecodeImage: + to_rgb: True + channel_first: False + - ResizeImage: + resize_short: 256 + - CropImage: + size: 224 + - NormalizeImage: + scale: 1.0/255.0 + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: '' + - ToCHWImage: diff --git a/docs/en/models/SwinTransformer_en.md b/docs/en/models/SwinTransformer_en.md new file mode 100644 index 0000000000000000000000000000000000000000..11d45d6c401c57f31d586b6c740d968b304c3574 --- /dev/null +++ b/docs/en/models/SwinTransformer_en.md @@ -0,0 +1,22 @@ +# SwinTransformer + +## Overview +Swin Transformer a new vision Transformer, that capably serves as a general-purpose backbone for computer vision. It is a hierarchical Transformer whose representation is computed with shifted windows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping local windows while also allowing for cross-window connection. [Paper](https://arxiv.org/abs/2103.14030)。 + + +## Accuracy, FLOPS and Parameters + +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| SwinTransformer_tiny_patch4_window7_224 | 0.8069 | 0.9534 | 0.812 | 0.955 | 4.5 | 28 | +| SwinTransformer_small_patch4_window7_224 | 0.8275 | 0.9613 | 0.832 | 0.962 | 8.7 | 50 | +| SwinTransformer_base_patch4_window7_224 | 0.8300 | 0.9626 | 0.835 | 0.965 | 15.4 | 88 | +| SwinTransformer_base_patch4_window12_384 | 0.8439 | 0.9693 | 0.845 | 0.970 | 47.1 | 88 | +| SwinTransformer_base_patch4_window7_224[1] | 0.8487 | 0.9746 | 0.852 | 0.975 | 15.4 | 88 | +| SwinTransformer_base_patch4_window12_384[1] | 0.8642 | 0.9807 | 0.864 | 0.980 | 47.1 | 88 | +| SwinTransformer_large_patch4_window7_224[1] | 0.8596 | 0.9783 | 0.863 | 0.979 | 34.5 | 197 | +| SwinTransformer_large_patch4_window12_384[1] | 0.8719 | 0.9823 | 0.873 | 0.982 | 103.9 | 197 | + +[1]: Based on imagenet22k dataset pre-training, and then in imagenet1k dataset transfer learning. + +**Note**: The difference of precision with reference from the difference of data preprocessing. diff --git a/docs/en/models/Transformer_en.md b/docs/en/models/ViT_and_DeiT_en.md similarity index 83% rename from docs/en/models/Transformer_en.md rename to docs/en/models/ViT_and_DeiT_en.md index 13d00fca7e9414a88009e0e324993bbcf1c9d908..ac275d9b4a5e9c653d0bd30c1a322505440f441c 100644 --- a/docs/en/models/Transformer_en.md +++ b/docs/en/models/ViT_and_DeiT_en.md @@ -9,27 +9,27 @@ DeiT(Data-efficient Image Transformers) series models were proposed by Facebook ## Accuracy, FLOPS and Parameters -| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | -|:--:|:--:|:--:|:--:|:--:|:--:| -| ViT_small_patch16_224 | 0.7769 | 0.9342 | 0.7785 | 0.9342 | | -| ViT_base_patch16_224 | 0.8195 | 0.9617 | 0.8178 | 0.9613 | | -| ViT_base_patch16_384 | 0.8414 | 0.9717 | 0.8420 | 0.9722 | | -| ViT_base_patch32_384 | 0.8176 | 0.9613 | 0.8166 | 0.9613 | | -| ViT_large_patch16_224 | 0.8323 | 0.9650 | 0.8306 | 0.9644 | | -| ViT_large_patch16_384 | 0.8513 | 0.9736 | 0.8517 | 0.9736 | | -| ViT_large_patch32_384 | 0.8153 | 0.9608 | 0.815 | - | | - - -| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | -|:--:|:--:|:--:|:--:|:--:|:--:| -| DeiT_tiny_patch16_224 | 0.718 | 0.910 | 0.722 | 0.911 | | -| DeiT_small_patch16_224 | 0.796 | 0.949 | 0.799 | 0.950 | | -| DeiT_base_patch16_224 | 0.817 | 0.957 | 0.818 | 0.956 | | -| DeiT_base_patch16_384 | 0.830 | 0.962 | 0.829 | 0.972 | | -| DeiT_tiny_distilled_patch16_224 | 0.741 | 0.918 | 0.745 | 0.919 | | -| DeiT_small_distilled_patch16_224 | 0.809 | 0.953 | 0.812 | 0.954 | | -| DeiT_base_distilled_patch16_224 | 0.831 | 0.964 | 0.834 | 0.965 | | -| DeiT_base_distilled_patch16_384 | 0.851 | 0.973 | 0.852 | 0.972 | | +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| ViT_small_patch16_224 | 0.7769 | 0.9342 | 0.7785 | 0.9342 | | | +| ViT_base_patch16_224 | 0.8195 | 0.9617 | 0.8178 | 0.9613 | | | +| ViT_base_patch16_384 | 0.8414 | 0.9717 | 0.8420 | 0.9722 | | | +| ViT_base_patch32_384 | 0.8176 | 0.9613 | 0.8166 | 0.9613 | | | +| ViT_large_patch16_224 | 0.8323 | 0.9650 | 0.8306 | 0.9644 | | | +| ViT_large_patch16_384 | 0.8513 | 0.9736 | 0.8517 | 0.9736 | | | +| ViT_large_patch32_384 | 0.8153 | 0.9608 | 0.815 | - | | | + + +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| DeiT_tiny_patch16_224 | 0.718 | 0.910 | 0.722 | 0.911 | | | +| DeiT_small_patch16_224 | 0.796 | 0.949 | 0.799 | 0.950 | | | +| DeiT_base_patch16_224 | 0.817 | 0.957 | 0.818 | 0.956 | | | +| DeiT_base_patch16_384 | 0.830 | 0.962 | 0.829 | 0.972 | | | +| DeiT_tiny_distilled_patch16_224 | 0.741 | 0.918 | 0.745 | 0.919 | | | +| DeiT_small_distilled_patch16_224 | 0.809 | 0.953 | 0.812 | 0.954 | | | +| DeiT_base_distilled_patch16_224 | 0.831 | 0.964 | 0.834 | 0.965 | | | +| DeiT_base_distilled_patch16_384 | 0.851 | 0.973 | 0.852 | 0.972 | | | Params, FLOPs, Inference speed and other information are coming soon. diff --git a/docs/zh_CN/models/SwinTransformer.md b/docs/zh_CN/models/SwinTransformer.md new file mode 100644 index 0000000000000000000000000000000000000000..1d9561ba8e5ad95b75470e4d9af305b519780368 --- /dev/null +++ b/docs/zh_CN/models/SwinTransformer.md @@ -0,0 +1,22 @@ +# SwinTransformer + +## 概述 +Swin Transformer 是一种新的视觉Transformer网络，可以用作计算机视觉领域的通用骨干网路。SwinTransformer由移动窗口（shifted windows）表示的层次Transformer结构组成。移动窗口将自注意计算限制在非重叠的局部窗口上，同时允许跨窗口连接，从而提高了网络性能。[论文地址](https://arxiv.org/abs/2103.14030)。 + + +## 精度、FLOPS和参数量 + +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| SwinTransformer_tiny_patch4_window7_224 | 0.8069 | 0.9534 | 0.812 | 0.955 | 4.5 | 28 | +| SwinTransformer_small_patch4_window7_224 | 0.8275 | 0.9613 | 0.832 | 0.962 | 8.7 | 50 | +| SwinTransformer_base_patch4_window7_224 | 0.8300 | 0.9626 | 0.835 | 0.965 | 15.4 | 88 | +| SwinTransformer_base_patch4_window12_384 | 0.8439 | 0.9693 | 0.845 | 0.970 | 47.1 | 88 | +| SwinTransformer_base_patch4_window7_224[1] | 0.8487 | 0.9746 | 0.852 | 0.975 | 15.4 | 88 | +| SwinTransformer_base_patch4_window12_384[1] | 0.8642 | 0.9807 | 0.864 | 0.980 | 47.1 | 88 | +| SwinTransformer_large_patch4_window7_224[1] | 0.8596 | 0.9783 | 0.863 | 0.979 | 34.5 | 197 | +| SwinTransformer_large_patch4_window12_384[1] | 0.8719 | 0.9823 | 0.873 | 0.982 | 103.9 | 197 | + +[1]：基于ImageNet22k数据集预训练，然后在ImageNet1k数据集迁移学习得到。 + +**注**：与Reference的精度差异源于数据预处理不同。 diff --git a/docs/zh_CN/models/Transformer.md b/docs/zh_CN/models/ViT_and_DeiT.md similarity index 83% rename from docs/zh_CN/models/Transformer.md rename to docs/zh_CN/models/ViT_and_DeiT.md index de037177acf3987cfb60c60e68a64ef15b82ec15..d14491d2a2b44d16d5fba4c316afb757543f21ac 100644 --- a/docs/zh_CN/models/Transformer.md +++ b/docs/zh_CN/models/ViT_and_DeiT.md @@ -11,26 +11,26 @@ DeiT（Data-efficient Image Transformers）系列模型是由FaceBook在2020年 ## 精度、FLOPS和参数量 -| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | -|:--:|:--:|:--:|:--:|:--:|:--:| -| ViT_small_patch16_224 | 0.7769 | 0.9342 | 0.7785 | 0.9342 | | -| ViT_base_patch16_224 | 0.8195 | 0.9617 | 0.8178 | 0.9613 | | -| ViT_base_patch16_384 | 0.8414 | 0.9717 | 0.8420 | 0.9722 | | -| ViT_base_patch32_384 | 0.8176 | 0.9613 | 0.8166 | 0.9613 | | -| ViT_large_patch16_224 | 0.8323 | 0.9650 | 0.8306 | 0.9644 | | -| ViT_large_patch16_384 | 0.8513 | 0.9736 | 0.8517 | 0.9736 | | -| ViT_large_patch32_384 | 0.8153 | 0.9608 | 0.815 | - | | - - -| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | -|:--:|:--:|:--:|:--:|:--:|:--:| -| DeiT_tiny_patch16_224 | 0.718 | 0.910 | 0.722 | 0.911 | | -| DeiT_small_patch16_224 | 0.796 | 0.949 | 0.799 | 0.950 | | -| DeiT_base_patch16_224 | 0.817 | 0.957 | 0.818 | 0.956 | | -| DeiT_base_patch16_384 | 0.830 | 0.962 | 0.829 | 0.972 | | -| DeiT_tiny_distilled_patch16_224 | 0.741 | 0.918 | 0.745 | 0.919 | | -| DeiT_small_distilled_patch16_224 | 0.809 | 0.953 | 0.812 | 0.954 | | -| DeiT_base_distilled_patch16_224 | 0.831 | 0.964 | 0.834 | 0.965 | | -| DeiT_base_distilled_patch16_384 | 0.851 | 0.973 | 0.852 | 0.972 | | +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| ViT_small_patch16_224 | 0.7769 | 0.9342 | 0.7785 | 0.9342 | | | +| ViT_base_patch16_224 | 0.8195 | 0.9617 | 0.8178 | 0.9613 | | | +| ViT_base_patch16_384 | 0.8414 | 0.9717 | 0.8420 | 0.9722 | | | +| ViT_base_patch32_384 | 0.8176 | 0.9613 | 0.8166 | 0.9613 | | | +| ViT_large_patch16_224 | 0.8323 | 0.9650 | 0.8306 | 0.9644 | | | +| ViT_large_patch16_384 | 0.8513 | 0.9736 | 0.8517 | 0.9736 | | | +| ViT_large_patch32_384 | 0.8153 | 0.9608 | 0.815 | - | | | + + +| Models | Top1 | Top5 | Reference
top1 | Reference
top5 | FLOPS
(G) | Params
(M) | +|:--:|:--:|:--:|:--:|:--:|:--:|:--:| +| DeiT_tiny_patch16_224 | 0.718 | 0.910 | 0.722 | 0.911 | | | +| DeiT_small_patch16_224 | 0.796 | 0.949 | 0.799 | 0.950 | | | +| DeiT_base_patch16_224 | 0.817 | 0.957 | 0.818 | 0.956 | | | +| DeiT_base_patch16_384 | 0.830 | 0.962 | 0.829 | 0.972 | | | +| DeiT_tiny_distilled_patch16_224 | 0.741 | 0.918 | 0.745 | 0.919 | | | +| DeiT_small_distilled_patch16_224 | 0.809 | 0.953 | 0.812 | 0.954 | | | +| DeiT_base_distilled_patch16_224 | 0.831 | 0.964 | 0.834 | 0.965 | | | +| DeiT_base_distilled_patch16_384 | 0.851 | 0.973 | 0.852 | 0.972 | | | 关于Params、FLOPs、Inference speed等信息，敬请期待。 diff --git a/paddleclas.py b/paddleclas.py index c4314ec6db2b0b56f84f00c48f0191dac635a15c..d8fcf9368473f11b5379a80a7d20bc4290677c0f 100644 --- a/paddleclas.py +++ b/paddleclas.py @@ -32,7 +32,10 @@ __dir__ = os.path.dirname(__file__) sys.path.append(os.path.join(__dir__, '')) import argparse import shutil +import textwrap +from difflib import SequenceMatcher +from prettytable import PrettyTable import cv2 import numpy as np import tarfile @@ -45,57 +48,148 @@ __all__ = ['PaddleClas'] BASE_DIR = os.path.expanduser("~/.paddleclas/") BASE_INFERENCE_MODEL_DIR = os.path.join(BASE_DIR, 'inference_model') BASE_IMAGES_DIR = os.path.join(BASE_DIR, 'images') - -model_names = { - 'Xception71', 'SE_ResNeXt101_32x4d', 'ShuffleNetV2_x0_5', 'ResNet34', - 'ShuffleNetV2_x2_0', 'ResNeXt101_32x4d', 'HRNet_W48_C_ssld', - 'ResNeSt50_fast_1s1x64d', 'MobileNetV2_x2_0', 'MobileNetV3_large_x1_0', - 'Fix_ResNeXt101_32x48d_wsl', 'MobileNetV2_ssld', 'ResNeXt101_vd_64x4d', - 'ResNet34_vd_ssld', 'MobileNetV3_small_x1_0', 'VGG11', - 'ResNeXt50_vd_32x4d', 'MobileNetV3_large_x1_25', - 'MobileNetV3_large_x1_0_ssld', 'MobileNetV2_x0_75', - 'MobileNetV3_small_x0_35', 'MobileNetV1_x0_75', 'MobileNetV1_ssld', - 'ResNeXt50_32x4d', 'GhostNet_x1_3_ssld', 'Res2Net101_vd_26w_4s', - 'ResNet152', 'Xception65', 'EfficientNetB0', 'ResNet152_vd', 'HRNet_W18_C', - 'Res2Net50_14w_8s', 'ShuffleNetV2_x0_25', 'HRNet_W64_C', - 'Res2Net50_vd_26w_4s_ssld', 'HRNet_W18_C_ssld', 'ResNet18_vd', - 'ResNeXt101_32x16d_wsl', 'SE_ResNeXt50_32x4d', 'SqueezeNet1_1', - 'SENet154_vd', 'SqueezeNet1_0', 'GhostNet_x1_0', 'ResNet50_vc', 'DPN98', - 'HRNet_W48_C', 'DenseNet264', 'SE_ResNet34_vd', 'HRNet_W44_C', - 'MobileNetV3_small_x1_25', 'MobileNetV1_x0_5', 'ResNet200_vd', 'VGG13', - 'EfficientNetB3', 'EfficientNetB2', 'ShuffleNetV2_x0_33', - 'MobileNetV3_small_x0_75', 'ResNeXt152_vd_32x4d', 'ResNeXt101_32x32d_wsl', - 'ResNet18', 'MobileNetV3_large_x0_35', 'Res2Net50_26w_4s', - 'MobileNetV2_x0_5', 'EfficientNetB0_small', 'ResNet101_vd_ssld', - 'EfficientNetB6', 'EfficientNetB1', 'EfficientNetB7', 'ResNeSt50', - 'ShuffleNetV2_x1_0', 'MobileNetV3_small_x1_0_ssld', 'InceptionV4', - 'GhostNet_x0_5', 'SE_HRNet_W64_C_ssld', 'ResNet50_ACNet_deploy', - 'Xception41', 'ResNet50', 'Res2Net200_vd_26w_4s_ssld', - 'Xception41_deeplab', 'SE_ResNet18_vd', 'SE_ResNeXt50_vd_32x4d', - 'HRNet_W30_C', 'HRNet_W40_C', 'VGG19', 'Res2Net200_vd_26w_4s', - 'ResNeXt101_32x8d_wsl', 'ResNet50_vd', 'ResNeXt152_64x4d', 'DarkNet53', - 'ResNet50_vd_ssld', 'ResNeXt101_64x4d', 'MobileNetV1_x0_25', - 'Xception65_deeplab', 'AlexNet', 'ResNet101', 'DenseNet121', - 'ResNet50_vd_v2', 'Res2Net50_vd_26w_4s', 'ResNeXt101_32x48d_wsl', - 'MobileNetV3_large_x0_5', 'MobileNetV2_x0_25', 'DPN92', 'ResNet101_vd', - 'MobileNetV2_x1_5', 'DPN131', 'ResNeXt50_vd_64x4d', 'ShuffleNetV2_x1_5', - 'ResNet34_vd', 'MobileNetV1', 'ResNeXt152_vd_64x4d', 'DPN107', 'VGG16', - 'ResNeXt50_64x4d', 'RegNetX_4GF', 'DenseNet161', 'GhostNet_x1_3', - 'HRNet_W32_C', 'Fix_ResNet50_vd_ssld_v2', 'Res2Net101_vd_26w_4s_ssld', - 'DenseNet201', 'DPN68', 'EfficientNetB4', 'ResNeXt152_32x4d', - 'InceptionV3', 'ShuffleNetV2_swish', 'GoogLeNet', 'ResNet50_vd_ssld_v2', - 'SE_ResNet50_vd', 'MobileNetV2', 'ResNeXt101_vd_32x4d', - 'MobileNetV3_large_x0_75', 'MobileNetV3_small_x0_5', 'DenseNet169', - 'EfficientNetB5', 'DeiT_base_distilled_patch16_224', - 'DeiT_base_distilled_patch16_384', 'DeiT_base_patch16_224', - 'DeiT_base_patch16_384', 'DeiT_small_distilled_patch16_224', - 'DeiT_small_patch16_224', 'DeiT_tiny_distilled_patch16_224', - 'DeiT_tiny_patch16_224', 'ViT_base_patch16_224', 'ViT_base_patch16_384', - 'ViT_base_patch32_384', 'ViT_large_patch16_224', 'ViT_large_patch16_384', - 'ViT_large_patch32_384', 'ViT_small_patch16_224' +model_series = { + "AlexNet": ["AlexNet"], + "DarkNet": ["DarkNet53"], + "DeiT": [ + 'DeiT_base_distilled_patch16_224', 'DeiT_base_distilled_patch16_384', + 'DeiT_base_patch16_224', 'DeiT_base_patch16_384', + 'DeiT_small_distilled_patch16_224', 'DeiT_small_patch16_224', + 'DeiT_tiny_distilled_patch16_224', 'DeiT_tiny_patch16_224' + ], + "DenseNet": [ + "DenseNet121", "DenseNet161", "DenseNet169", "DenseNet201", + "DenseNet264" + ], + "DPN": ["DPN68", "DPN92", "DPN98", "DPN107", "DPN131"], + "EfficientNet": [ + "EfficientNetB0", "EfficientNetB0_small", "EfficientNetB1", + "EfficientNetB2", "EfficientNetB3", "EfficientNetB4", "EfficientNetB5", + "EfficientNetB6", "EfficientNetB7" + ], + "GhostNet": + ["GhostNet_x0_5", "GhostNet_x1_0", "GhostNet_x1_3", "GhostNet_x1_3_ssld"], + "HRNet": [ + "HRNet_W18_C", "HRNet_W30_C", "HRNet_W32_C", "HRNet_W40_C", + "HRNet_W44_C", "HRNet_W48_C", "HRNet_W64_C", "HRNet_W18_C_ssld", + "HRNet_W48_C_ssld" + ], + "Inception": ["GoogLeNet", "InceptionV3", "InceptionV4"], + "MobileNetV1": [ + "MobileNetV1_x0_25", "MobileNetV1_x0_5", "MobileNetV1_x0_75", + "MobileNetV1", "MobileNetV1_ssld" + ], + "MobileNetV2": [ + "MobileNetV2_x0_25", "MobileNetV2_x0_5", "MobileNetV2_x0_75", + "MobileNetV2", "MobileNetV2_x1_5", "MobileNetV2_x2_0", + "MobileNetV2_ssld" + ], + "MobileNetV3": [ + "MobileNetV3_small_x0_35", "MobileNetV3_small_x0_5", + "MobileNetV3_small_x0_75", "MobileNetV3_small_x1_0", + "MobileNetV3_small_x1_25", "MobileNetV3_large_x0_35", + "MobileNetV3_large_x0_5", "MobileNetV3_large_x0_75", + "MobileNetV3_large_x1_0", "MobileNetV3_large_x1_25", + "MobileNetV3_small_x1_0_ssld", "MobileNetV3_large_x1_0_ssld" + ], + "RegNet": ["RegNetX_4GF"], + "Res2Net": [ + "Res2Net50_14w_8s", "Res2Net50_26w_4s", "Res2Net50_vd_26w_4s", + "Res2Net200_vd_26w_4s", "Res2Net101_vd_26w_4s", + "Res2Net50_vd_26w_4s_ssld", "Res2Net101_vd_26w_4s_ssld", + "Res2Net200_vd_26w_4s_ssld" + ], + "ResNeSt": ["ResNeSt50", "ResNeSt50_fast_1s1x64d"], + "ResNet": [ + "ResNet18", "ResNet18_vd", "ResNet34", "ResNet34_vd", "ResNet50", + "ResNet50_vc", "ResNet50_vd", "ResNet50_vd_v2", "ResNet101", + "ResNet101_vd", "ResNet152", "ResNet152_vd", "ResNet200_vd", + "ResNet34_vd_ssld", "ResNet50_vd_ssld", "ResNet50_vd_ssld_v2", + "ResNet101_vd_ssld", "Fix_ResNet50_vd_ssld_v2", "ResNet50_ACNet_deploy" + ], + "ResNeXt": [ + "ResNeXt50_32x4d", "ResNeXt50_vd_32x4d", "ResNeXt50_64x4d", + "ResNeXt50_vd_64x4d", "ResNeXt101_32x4d", "ResNeXt101_vd_32x4d", + "ResNeXt101_32x8d_wsl", "ResNeXt101_32x16d_wsl", + "ResNeXt101_32x32d_wsl", "ResNeXt101_32x48d_wsl", + "Fix_ResNeXt101_32x48d_wsl", "ResNeXt101_64x4d", "ResNeXt101_vd_64x4d", + "ResNeXt152_32x4d", "ResNeXt152_vd_32x4d", "ResNeXt152_64x4d", + "ResNeXt152_vd_64x4d" + ], + "SENet": [ + "SENet154_vd", "SE_HRNet_W64_C_ssld", "SE_ResNet18_vd", + "SE_ResNet34_vd", "SE_ResNet50_vd", "SE_ResNeXt50_32x4d", + "SE_ResNeXt50_vd_32x4d", "SE_ResNeXt101_32x4d" + ], + "ShuffleNetV2": [ + "ShuffleNetV2_swish", "ShuffleNetV2_x0_25", "ShuffleNetV2_x0_33", + "ShuffleNetV2_x0_5", "ShuffleNetV2_x1_0", "ShuffleNetV2_x1_5", + "ShuffleNetV2_x2_0" + ], + "SqueezeNet": ["SqueezeNet1_0", "SqueezeNet1_1"], + "SwinTransformer": [ + "SwinTransformer_large_patch4_window7_224_22kto1k", + "SwinTransformer_large_patch4_window12_384_22kto1k", + "SwinTransformer_base_patch4_window7_224_22kto1k", + "SwinTransformer_base_patch4_window12_384_22kto1k", + "SwinTransformer_base_patch4_window12_384", + "SwinTransformer_base_patch4_window7_224", + "SwinTransformer_small_patch4_window7_224", + "SwinTransformer_tiny_patch4_window7_224" + ], + "VGG": ["VGG11", "VGG13", "VGG16", "VGG19"], + "VisionTransformer": [ + "ViT_base_patch16_224", "ViT_base_patch16_384", "ViT_base_patch32_384", + "ViT_large_patch16_224", "ViT_large_patch16_384", + "ViT_large_patch32_384", "ViT_small_patch16_224" + ], + "Xception": [ + "Xception41", "Xception41_deeplab", "Xception65", "Xception65_deeplab", + "Xception71" + ] } +class ModelNameError(Exception): + """ ModelNameError + """ + + def __init__(self, message=''): + super().__init__(message) + + +def print_info(): + table = PrettyTable(['Series', 'Name']) + for series in model_series: + names = textwrap.fill(" ".join(model_series[series]), width=100) + table.add_row([series, names]) + print('Inference models that Paddle provides are listed as follows:') + print(table) + + +def get_model_names(): + model_names = [] + for series in model_series: + model_names += model_series[series] + return model_names + + +def similar_architectures(name='', names=[], thresh=0.1, topk=10): + """ + inferred similar architectures + """ + scores = [] + for idx, n in enumerate(names): + if n.startswith('__'): + continue + score = SequenceMatcher(None, n.lower(), name.lower()).quick_ratio() + if score > thresh: + scores.append((idx, score)) + scores.sort(key=lambda x: x[1], reverse=True) + similar_names = [names[s[0]] for s in scores[:min(topk, len(scores))]] + return similar_names + + def download_with_progressbar(url, save_path): response = requests.get(url, stream=True) total_size_in_bytes = int(response.headers.get('content-length', 0)) @@ -227,17 +321,26 @@ def parse_args(mMain=True, add_help=True): class PaddleClas(object): - print('Inference models that Paddle provides are listed as follows:\n\n{}'. - format(model_names), '\n') + print_info() def __init__(self, **kwargs): + model_names = get_model_names() process_params = parse_args(mMain=False, add_help=False) process_params.__dict__.update(**kwargs) if not os.path.exists(process_params.model_file): if process_params.model_name is None: - raise Exception( + raise ModelNameError( 'Please input model name that you want to use!') + + similar_names = similar_architectures(process_params.model_name, + model_names) + model_list = ', '.join(similar_names) + if process_params.model_name not in similar_names: + err = "{} is not exist! Maybe you want: [{}]" \ + "".format(process_params.model_name, model_list) + raise ModelNameError(err) + if process_params.model_name in model_names: url = 'https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/{}_infer.tar'.format( process_params.model_name) diff --git a/ppcls/modeling/architectures/__init__.py b/ppcls/modeling/architectures/__init__.py index 9865e47fbaa8d077b440f393da1115571a0bebd7..a0869259725cc28e97ce6a5af47bd0dd3620dc77 100644 --- a/ppcls/modeling/architectures/__init__.py +++ b/ppcls/modeling/architectures/__init__.py @@ -47,6 +47,6 @@ from .vision_transformer import ViT_small_patch16_224, ViT_base_patch16_224, ViT from .distilled_vision_transformer import DeiT_tiny_patch16_224, DeiT_small_patch16_224, DeiT_base_patch16_224, DeiT_tiny_distilled_patch16_224, DeiT_small_distilled_patch16_224, DeiT_base_distilled_patch16_224, DeiT_base_patch16_384, DeiT_base_distilled_patch16_384 from .distillation_models import ResNet50_vd_distill_MobileNetV3_large_x1_0, ResNeXt101_32x16d_wsl_distill_ResNet50_vd from .repvgg import RepVGG_A0, RepVGG_A1, RepVGG_A2, RepVGG_B0, RepVGG_B1, RepVGG_B2, RepVGG_B3, RepVGG_B1g2, RepVGG_B1g4, RepVGG_B2g2, RepVGG_B2g4, RepVGG_B3g2, RepVGG_B3g4 -from .swin_transformer import SwinTransformer_tiny_patch4_window7_224, SwinTransformer_small_patch4_window7_224, SwinTransformer_base_patch4_window7_224, SwinTransformer_base_patch4_window12_384 +from .swin_transformer import SwinTransformer_tiny_patch4_window7_224, SwinTransformer_small_patch4_window7_224, SwinTransformer_base_patch4_window7_224, SwinTransformer_base_patch4_window12_384, SwinTransformer_large_patch4_window7_224, SwinTransformer_large_patch4_window12_384 from .mixnet import MixNet_S, MixNet_M, MixNet_L from .rexnet import ReXNet_1_0, ReXNet_1_3, ReXNet_1_5, ReXNet_2_0, ReXNet_3_0 diff --git a/ppcls/modeling/architectures/swin_transformer.py b/ppcls/modeling/architectures/swin_transformer.py index 4b65ab5528c9eaa13b605902e37ca9d7fafa336b..97efbd1f341d69fd601bfb86697d4f1cb06ca022 100644 --- a/ppcls/modeling/architectures/swin_transformer.py +++ b/ppcls/modeling/architectures/swin_transformer.py @@ -759,3 +759,24 @@ def SwinTransformer_base_patch4_window12_384(**args): drop_path_rate=0.5, # NOTE: do not appear in offical code **args) return model + + +def SwinTransformer_large_patch4_window7_224(**args): + model = SwinTransformer( + embed_dim=192, + depths=[2, 2, 18, 2], + num_heads=[6, 12, 24, 48], + window_size=7, + **args) + return model + + +def SwinTransformer_large_patch4_window12_384(**args): + model = SwinTransformer( + img_size=384, + embed_dim=192, + depths=[2, 2, 18, 2], + num_heads=[6, 12, 24, 48], + window_size=12, + **args) + return model diff --git a/requirements.txt b/requirements.txt index 73d5e78bceb7b8712d2db6c01cbfd935c8857cb4..ec8806def5a58b45507076e820b0fad02025a11c 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,3 +1,4 @@ +prettytable ujson opencv-python==4.1.2.30 pillow