From 7c5d4c72edb3a4f4157e2d60c0f33e4eeb94963b Mon Sep 17 00:00:00 2001 From: cuicheng01 <45199522+cuicheng01@users.noreply.github.com> Date: Sat, 19 Sep 2020 17:44:03 +0800 Subject: [PATCH 1/2] Update README.md --- README.md | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/README.md b/README.md index 16e77cde..20f0ee1e 100644 --- a/README.md +++ b/README.md @@ -8,9 +8,9 @@ PaddleClas is a toolset for image classification tasks prepared for the industry **Recent update** -- 2020.09.17 Add `HRNet_W48_C_ssld` pretrained model, whose Top-1 Acc on ImageNet1k dataset reaches 83.62%. Add `ResNet34_vd_ssld` pretrained model, whose Top-1 Acc on ImageNet1k dataset reaches 79.72%. -- 2020.09.07 Add `HRNet_W18_C_ssld` pretrained model, whose Top-1 Acc on ImageNet1k dataset reaches 81.16%. -- 2020.07.14 Add `Res2Net200_vd_26w_4s_ssld` pretrained model, whose Top-1 Acc on ImageNet1k dataset reaches 85.13%. Add `Fix_ResNet50_vd_ssld_v2` pretrained model, whose Top-1 Acc on ImageNet1k dataset reaches 84.00%. +- 2020.09.17 Add `HRNet_W48_C_ssld` pretrained model, whose Top-1 Acc on ImageNet-1k dataset reaches 83.62%. Add `ResNet34_vd_ssld` pretrained model, whose Top-1 Acc on ImageNet-1k dataset reaches 79.72%. +- 2020.09.07 Add `HRNet_W18_C_ssld` pretrained model, whose Top-1 Acc on ImageNet-1k dataset reaches 81.16%. +- 2020.07.14 Add `Res2Net200_vd_26w_4s_ssld` pretrained model, whose Top-1 Acc on ImageNet-1k dataset reaches 85.13%. Add `Fix_ResNet50_vd_ssld_v2` pretrained model, whose Top-1 Acc on ImageNet-1k dataset reaches 84.00%. - 2020.06.17 Add English documents. - 2020.06.12 Add support for training and evaluation on Windows or CPU. - 2020.05.17 Add support for mixed precision training. @@ -19,9 +19,9 @@ PaddleClas is a toolset for image classification tasks prepared for the industry ## Features -- Rich model zoo. Based on the ImageNet1k classification dataset, PaddleClas provides 24 series of classification network structures and training configurations, 122 models' pretrained weights and their evaluation metrics. +- Rich model zoo. Based on the ImageNet-1k classification dataset, PaddleClas provides 24 series of classification network structures and training configurations, 122 models' pretrained weights and their evaluation metrics. -- SSLD Knowledge Distillation. Based on this SSLD distillation strategy, the accuracy of the distilled model is generally increased by more than 3%. +- SSLD Knowledge Distillation. Based on this SSLD distillation strategy, the top-1 acc of the distilled model is generally increased by more than 3%. - Data augmentation: PaddleClas provides detailed introduction of 8 data augmentation algorithms such as AutoAugment, Cutout, Cutmix, code reproduction and effect evaluation in a unified experimental environment. @@ -77,7 +77,7 @@ PaddleClas is a toolset for image classification tasks prepared for the industry <a name="Model_zoo_overview"></a> ### Model zoo overview -Based on the ImageNet1k classification dataset, the 24 classification network structures supported by PaddleClas and the corresponding 122 image classification pretrained models are shown below. Training trick, a brief introduction to each series of network structures, and performance evaluation will be shown in the corresponding chapters. The evaluation environment is as follows. +Based on the ImageNet-1k classification dataset, the 24 classification network structures supported by PaddleClas and the corresponding 122 image classification pretrained models are shown below. Training trick, a brief introduction to each series of network structures, and performance evaluation will be shown in the corresponding chapters. The evaluation environment is as follows. * CPU evaluation environment is based on Snapdragon 855 (SD855). * The GPU evaluation speed is measured by running 500 times under the FP32+TensorRT configuration (excluding the warmup time of the first 10 times). @@ -127,7 +127,7 @@ Accuracy and inference time metrics of ResNet and Vd series models are shown as Accuracy and inference time metrics of Mobile series models are shown as follows. More detailed information can be refered to [Mobile series tutorial](./docs/en/models/Mobile_en.md). -| Model | Top-1 Acc | Top-5 Acc | SD855 time(ms)<br>bs=1 | Flops(G) | Params(M) | 模型大å°(M) | Download Address | +| Model | Top-1 Acc | Top-5 Acc | SD855 time(ms)<br>bs=1 | Flops(G) | Params(M) | Model storage size(M) | Download Address | |----------------------------------|-----------|-----------|------------------------|----------|-----------|---------|-----------------------------------------------------------------------------------------------------------| | MobileNetV1_<br>x0_25 | 0.5143 | 0.7546 | 3.21985 | 0.07 | 0.46 | 1.9 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV1_x0_25_pretrained.tar) | | MobileNetV1_<br>x0_5 | 0.6352 | 0.8473 | 9.579599 | 0.28 | 1.31 | 5.2 | [Download link](https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV1_x0_5_pretrained.tar) | -- GitLab From 84b4ad6d2521dd15bbc27f8018cbf0c78686d578 Mon Sep 17 00:00:00 2001 From: cuicheng01 <45199522+cuicheng01@users.noreply.github.com> Date: Sat, 19 Sep 2020 17:52:59 +0800 Subject: [PATCH 2/2] Update README_cn.md --- README_cn.md | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/README_cn.md b/README_cn.md index 07cc1a4e..cdbc9efa 100644 --- a/README_cn.md +++ b/README_cn.md @@ -7,12 +7,12 @@ 飞桨图åƒåˆ†ç±»å¥—件PaddleClas是飞桨为工业界和å¦æœ¯ç•Œæ‰€å‡†å¤‡çš„一个图åƒåˆ†ç±»ä»»åŠ¡çš„工具集,助力使用者è®ç»ƒå‡ºæ›´å¥½çš„视觉模型和应用è½åœ°ã€‚ **近期更新** -- 2020.09.17 æ·»åŠ HRNet_W48_C_ssld模型,在ImageNet上Top-1 Accå¯è¾¾0.836ï¼›æ·»åŠ ResNet34_vd_ssld模型,在ImageNet上Top-1 Accå¯è¾¾0.797。 -- 2020.09.07 æ·»åŠ HRNet_W18_C_ssld模型,在ImageNet上Top-1 Accå¯è¾¾0.81162ï¼›æ·»åŠ MobileNetV3_small_x0_35_ssld模型,在ImageNet上Top-1 Accå¯è¾¾0.5555。 -- 2020.07.14 æ·»åŠ Res2Net200_vd_26w_4s_ssld模型,在ImageNet上Top-1 Accå¯è¾¾85.13%ï¼›æ·»åŠ Fix_ResNet50_vd_ssld_v2模型,在ImageNet上Top-1 Accå¯è¾¾84.0%。 +- 2020.09.17 æ·»åŠ `HRNet_W48_C_ssld `模型,在ImageNet-1k上Top-1 Accå¯è¾¾83.62%ï¼›æ·»åŠ `ResNet34_vd_ssld `模型,在ImageNet-1k上Top-1 Accå¯è¾¾79.72%。 +- 2020.09.07 æ·»åŠ `HRNet_W18_C_ssld `模型,在ImageNet-1k上Top-1 Accå¯è¾¾81.16%ï¼›æ·»åŠ `MobileNetV3_small_x0_35_ssld `模型,在ImageNet-1k上Top-1 Accå¯è¾¾55.55%。 +- 2020.07.14 æ·»åŠ `Res2Net200_vd_26w_4s_ssld `模型,在ImageNet-1k上Top-1 Accå¯è¾¾85.13%ï¼›æ·»åŠ `Fix_ResNet50_vd_ssld_v2 `模型,在ImageNet-1k上Top-1 Accå¯è¾¾84.0%。 - 2020.06.17 æ·»åŠ è‹±æ–‡æ–‡æ¡£ã€‚ - 2020.06.12 æ·»åŠ å¯¹windowså’ŒCPU环境的è®ç»ƒä¸Žè¯„估支æŒã€‚ -- 2020.05.17 æ·»åŠ æ··åˆç²¾åº¦è®ç»ƒï¼ŒåŸºäºŽResNet50æ¨¡åž‹ï¼Œç²¾åº¦å‡ ä¹Žæ— æŸçš„情况下,è®ç»ƒæ—¶é—´å¯ä»¥å‡å°‘约40%。 +- 2020.05.17 æ·»åŠ æ··åˆç²¾åº¦è®ç»ƒï¼ŒåŸºäºŽ `ResNet50 `æ¨¡åž‹ï¼Œç²¾åº¦å‡ ä¹Žæ— æŸçš„情况下,è®ç»ƒæ—¶é—´å¯ä»¥å‡å°‘约40%。 - [more](./docs/zh_CN/update_history.md) @@ -24,7 +24,7 @@ - æ•°æ®å¢žå¹¿ï¼šæ”¯æŒAutoAugmentã€Cutoutã€Cutmixç‰8ç§æ•°æ®å¢žå¹¿ç®—法详细介ç»ã€ä»£ç å¤çŽ°å’Œåœ¨ç»Ÿä¸€å®žéªŒçŽ¯å¢ƒä¸‹çš„效果评估。 -- 10万类图åƒåˆ†ç±»é¢„è®ç»ƒæ¨¡åž‹ï¼šç™¾åº¦è‡ªç ”并开æºäº†åŸºäºŽ10万类数æ®é›†è®ç»ƒçš„ResNet50_vd模型,在一些实际场景ä¸ï¼Œä½¿ç”¨è¯¥é¢„è®ç»ƒæ¨¡åž‹çš„识别准确率最多å¯ä»¥æå‡30%。 +- 10万类图åƒåˆ†ç±»é¢„è®ç»ƒæ¨¡åž‹ï¼šç™¾åº¦è‡ªç ”并开æºäº†åŸºäºŽ10万类数æ®é›†è®ç»ƒçš„ `ResNet50_vd `模型,在一些实际场景ä¸ï¼Œä½¿ç”¨è¯¥é¢„è®ç»ƒæ¨¡åž‹çš„识别准确率最多å¯ä»¥æå‡30%。 - 多ç§è®ç»ƒæ–¹æ¡ˆï¼ŒåŒ…括多机è®ç»ƒã€æ··åˆç²¾åº¦è®ç»ƒç‰ã€‚ -- GitLab