更换文档检测模型

2024-08-27 14:42:45 +08:00
parent aea6f19951
commit 1514e09c40
2072 changed files with 254336 additions and 4967 deletions
--- a/paddle_detection/configs/ppyoloe/README.md
+++ b/paddle_detection/configs/ppyoloe/README.md
@@ -0,0 +1,334 @@
+English | [简体中文](README_cn.md)
+
+# PP-YOLOE
+
+## Latest News
+- Release PP-YOLOE+ model: **(2022.08)**
+  - Pre training model using large-scale data set obj365
+  - In the backbone, add the alpha parameter to the block branch
+  - Optimize the end-to-end inference speed and improve the training convergence speed
+
+## Legacy model
+- Please refer to：[PP-YOLOE 2022.03](./README_legacy.md) for details
+
+## Table of Contents
+- [Introduction](#Introduction)
+- [Model Zoo](#Model-Zoo)
+- [Getting Start](#Getting-Start)
+- [Appendix](#Appendix)
+
+## Introduction
+PP-YOLOE is an excellent single-stage anchor-free model based on PP-YOLOv2, surpassing a variety of popular YOLO models. PP-YOLOE has a series of models, named s/m/l/x, which are configured through width multiplier and depth multiplier. PP-YOLOE avoids using special operators, such as Deformable Convolution or Matrix NMS, to be deployed friendly on various hardware. For more details, please refer to our [report](https://arxiv.org/abs/2203.16250).
+
+<div align="center">
+  <img src="../../docs/images/ppyoloe_plus_map_fps.png" width=500 />
+</div>
+
+PP-YOLOE+_l achieves 53.3 mAP on COCO test-dev2017 dataset with 78.1 FPS on Tesla V100. While using TensorRT FP16, PP-YOLOE+_l can be further accelerated to 149.2 FPS. PP-YOLOE+_s/m/x also have excellent accuracy and speed performance, which can be found in [Model Zoo](#Model-Zoo)
+
+PP-YOLOE is composed of following methods:
+- Scalable backbone and neck
+- [Task Alignment Learning](https://arxiv.org/abs/2108.07755)
+- Efficient Task-aligned head with [DFL](https://arxiv.org/abs/2006.04388) and [VFL](https://arxiv.org/abs/2008.13367)
+- [SiLU(Swish) activation function](https://arxiv.org/abs/1710.05941)
+
+## Model Zoo
+
+### Model Zoo on COCO
+
+|     Model      | Epoch | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>val<br>0.5:0.95 | Box AP<sup>test<br>0.5:0.95 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |                                       download                                       |                   config                    |
+|:--------------:|:-----:|:-------:|:----------:|:----------:| :-------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------:| :---------------------: |:------------------------------------------------------------------------------------:|:-------------------------------------------:|
+|  PP-YOLOE+_s   |  80   |     8      |     8      | cspresnet-s |     640     |            43.7            |            43.9             |   7.93    |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_s_80e_coco.yml) |
+|  PP-YOLOE+_m   |  80   |     8      |     8      | cspresnet-m |     640     |            49.8            |            50.0             |   23.43   |  49.91   |   123.4   |  208.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_m_80e_coco.yml) |
+|   PP-YOLOE+_m(distill)   |  80   |     8     |    8     | cspresnet-m |    640     |            **51.0**            |            51.2             |   23.43   |  49.91   |   123.4   |  208.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco_distill.pdparams) | [config](distill/ppyoloe_plus_crn_m_80e_coco_distill.yml) |
+|  PP-YOLOE+_l   |  80   |     8      |     8      | cspresnet-l |     640     |            52.9            |            53.3             |   52.20   |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_l_80e_coco.yml) |
+|   PP-YOLOE+_l(distill)   |  80   |     8     |    8     | cspresnet-l |    640     |            **54.0**            |            54.4             |   52.20   |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco_distill.pdparams) | [config](distill/ppyoloe_plus_crn_l_80e_coco_distill.yml) |
+|  PP-YOLOE+_x   |  80   |     8      |     8      | cspresnet-x |     640     |            54.7            |            54.9             |   98.42   |  206.59  |   45.0    |   95.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_x_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_x_80e_coco.yml) |
+
+**Note:**
+- M and L models use distillation, please refer to [distill](distill) for details.
+
+#### Tiny model
+
+|   Model  | Epoch | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>val<br>0.5:0.95 | Box AP<sup>test<br>0.5:0.95 | Params(M) | FLOPs(G) | T4 TensorRT FP16(FPS) | download  | config  |
+|:--------:|:-----:|:----------:|:----------:|:----------:|:-----------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------------:| :------: |:--------:|
+|   PP-YOLOE+_t-aux(640) |  300   |     8     |    8     | cspresnet-t  |    640  |            39.9       |            56.6             |   4.85   |  19.15 |  344.8   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(640)-relu |  300   |     8     |    8     | cspresnet-t  |    640  |            36.4       |            53.0        |   3.60   |  12.17 |   476.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_relu_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_relu_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(320) |  300   |     8     |    8     | cspresnet-t  |    320  |            33.3       |            48.5             |   4.85   |  4.80 |   729.9   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_320_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_320_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(320)-relu |  300   |     8     |    8     | cspresnet-t  |    320  |            30.1       |            44.7        |   3.60   |  3.04 |   984.8   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.yml) |
+
+
+### Comprehensive Metrics
+|          Model           | Epoch | AP<sup>0.5:0.95 | AP<sup>0.5 | AP<sup>0.75  | AP<sup>small | AP<sup>medium | AP<sup>large | AR<sup>small | AR<sup>medium | AR<sup>large |
+|:------------------------:|:-----:|:---------------:|:----------:|:------------:|:------------:| :-----------: |:------------:|:------------:|:-------------:|:------------:|
+|       PP-YOLOE+_s        |  80   |      43.7       |    60.6    |    47.9     |     26.5     |     47.5      |     59.0     |     46.7     |     71.4      |     81.7     |
+|       PP-YOLOE+_m        |  80   |      49.8       |    67.1    |    54.5     |     31.8     |     53.9      |     66.2     |     53.3     |     75.0      |     84.6     |
+|       PP-YOLOE+_m(distill)|  80   |      51.0       |    68.1    |    55.8     |     32.5     |     55.7      |     67.4     |     51.9     |     76.1      |     86.4     |
+|       PP-YOLOE+_l        |  80   |      52.9       |    70.1    |    57.9     |     35.2     |     57.5      |     69.1     |     56.0     |     77.9      |     86.9     |
+|       PP-YOLOE+_l(distill)|  80   |      54.0       |    71.2    |    59.2     |     36.1     |     58.8      |     70.4     |     55.0     |     78.7      |     87.7     |
+|       PP-YOLOE+_x        |  80   |      54.7       |    72.0    |    59.9     |     37.9     |     59.3      |     70.4     |     57.0     |     78.7      |     87.2     |
+
+**Note:**
+- M and L models use distillation, please refer to [distill](distill) for details.
+
+
+### End-to-end Speed
+|    Model    | AP<sup>0.5:0.95 | TRT-FP32(fps) | TRT-FP16(fps) |
+|:-----------:|:---------------:|:-------------:|:-------------:|
+| PP-YOLOE+_s |      43.7       |     44.44     |     47.85     |
+| PP-YOLOE+_m |      49.8       |     39.06     |     43.86     |
+| PP-YOLOE+_l |      52.9       |     34.01     |     42.02     |
+| PP-YOLOE+_x |      54.7       |     26.88     |     36.76     |
+
+**Notes:**
+
+- PP-YOLOE is trained on COCO train2017 dataset and evaluated on val2017 & test-dev2017 dataset.
+- The model weights in the table of Comprehensive Metrics are **the same as** that in the original Model Zoo, and evaluated on **val2017**.
+- PP-YOLOE used 8 GPUs for mixed precision training, if **GPU number** or **mini-batch size** is changed, **learning rate** should be adjusted according to the formula **lr<sub>new</sub> = lr<sub>default</sub> * (batch_size<sub>new</sub> * GPU_number<sub>new</sub>) / (batch_size<sub>default</sub> * GPU_number<sub>default</sub>)**.
+- PP-YOLOE inference speed is tesed on single Tesla V100 with batch size as 1, **CUDA 10.2**, **CUDNN 7.6.5**, **TensorRT 6.0.1.8** in TensorRT mode.
+- Refer to [Speed testing](#Speed-testing) to reproduce the speed testing results of PP-YOLOE.
+- If you set `--run_benchmark=True`，you should install these dependencies at first, `pip install pynvml psutil GPUtil`.
+- End-to-end speed test includes pre-processing + inference + post-processing and NMS time, using **Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz**, **single Tesla V100**, **CUDA 11.2**, **CUDNN 8.2.0**, **TensorRT 8.0.1.6**.
+
+### Model Zoo on Objects365
+|     Model      | Epoch | Machine number | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  download  |  config  |
+|:---------------:|:-----:|:-----------:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :--------:|:--------:|
+|   PP-YOLOE+_s   |  60   |  3 |  8     |    8     | cspresnet-s |    640     |   18.1  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_s_60e_objects365.yml) |
+|   PP-YOLOE+_m   |  60   |   4 |  8     |    8     | cspresnet-m |    640     |   25.0  |  23.43  |  49.91   |   123.4       |  208.3  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_m_60e_objects365.yml) |
+|   PP-YOLOE+_l   |  60   |   3 |  8     |    8     | cspresnet-l |    640     |   30.8  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_l_60e_objects365.yml) |
+|   PP-YOLOE+_x   |  60   |  4 |   8     |    8     | cspresnet-x |    640     |   32.7  |  98.42 |  206.59      |   45.0        |  95.2  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_x_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_x_60e_objects365.yml) |
+
+
+**Notes:**
+- The Details for multiple machine and multi-gpu training, see [DistributedTraining](../../docs/tutorials/DistributedTraining_en.md)
+- For Objects365 dataset download, please refer to [objects365 official website](http://www.objects365.org/overview.html). The specific category list can be downloaded from [objects365_detection_label_list.txt](https://bj.bcebos.com/v1/paddledet/data/objects365/objects365_detection_label_list.txt) organized by PaddleDetection team. It should be stored in `dataset/objects365/`, and each line represents one category. The categories need to be read when exporting the model or doing inference. If the json file is not exist, you can make the following changes to `configs/datasets/objects365_detection.yml`:
+```
+TestDataset:
+  !ImageFolder
+    # anno_path: annotations/zhiyuan_objv2_val.json
+    anno_path: objects365_detection_label_list.txt
+    dataset_dir: dataset/objects365/
+```
+
+
+### Model Zoo on VOC
+
+|     Model      | Epoch | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  download  |  config  |
+|:---------------:|:-----:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :-------: |:--------:|
+|   PP-YOLOE+_s   |  30   |     8     |    8     | cspresnet-s |    640     |   86.7  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_30e_voc.pdparams) | [config](./voc/ppyoloe_plus_crn_s_30e_voc.yml) |
+|   PP-YOLOE+_l   |  30   |     8     |    8     | cspresnet-l |    640     |   89.0  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_30e_voc.pdparams) | [config](./voc/ppyoloe_plus_crn_l_30e_voc.yml) |
+
+
+### Feature Models
+
+The PaddleDetection team provides configs and weights of various feature detection models based on PP-YOLOE, which users can download for use:
+
+|Scenarios | Related Datasets | Links|
+| :--------: | :---------: | :------: |
+|Pedestrian Detection | CrowdHuman | [pphuman](../pphuman) |
+|Vehicle Detection | BDD100K, UA-DETRAC | [ppvehicle](../ppvehicle) |
+|Small Object Detection | VisDrone、DOTA、xView |  [smalldet](../smalldet) |
+|Densely Packed Object Detection | SKU110k | [application](./application) |
+|Rotated Object Detection | DOTA | [PP-YOLOE-R](../rotate/ppyoloe_r/) |
+
+
+## Getting Start
+
+### Datasets and Metrics
+
+PaddleDetection team provides **COCO and VOC dataset** , decompress and place it under `PaddleDetection/dataset/`:
+
+```
+wget https://bj.bcebos.com/v1/paddledet/data/coco.tar
+# tar -xvf coco.tar
+
+wget https://bj.bcebos.com/v1/paddledet/data/voc.zip
+# unzip voc.zip
+```
+
+**Note:**
+  - For the format of COCO style dataset, please refer to [format-data](https://cocodataset.org/#format-data) and [format-results](https://cocodataset.org/#format-results).
+  - For the evaluation metric of COCO, please refer to [detection-eval](https://cocodataset.org/#detection-eval), and install  [cocoapi](https://github.com/cocodataset/cocoapi) at first.
+  - For the evaluation metric of VOC, please refer to [VOC2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/index.html).
+
+### Custom dataset
+
+1.For the annotation of custom dataset, please refer to [DetAnnoTools](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/docs/tutorials/data/DetAnnoTools_en.md);
+
+2.For training preparation of custom dataset，please refer to [PrepareDataSet](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/docs/tutorials/data/PrepareDetDataSet_en.md).
+
+
+### Training
+
+Training PP-YOLOE+ on 8 GPUs with following command
+
+```bash
+python -m paddle.distributed.launch --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml --eval --amp
+```
+
+**Notes:**
+- If you need to evaluate while training, please add `--eval`.
+- PP-YOLOE+ supports mixed precision training, please add `--amp`.
+- PaddleDetection supports multi-machine distributed training, you can refer to [DistributedTraining tutorial](../../docs/tutorials/DistributedTraining_en.md).
+
+
+### Evaluation
+
+Evaluating PP-YOLOE+ on COCO val2017 dataset in single GPU with following commands:
+
+```bash
+CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams
+```
+
+For evaluation on COCO test-dev2017 dataset, please download COCO test-dev2017 dataset from [COCO dataset download](https://cocodataset.org/#download) and decompress to COCO dataset directory and configure `EvalDataset` like `configs/ppyolo/ppyolo_test.yml`.
+
+### Inference
+
+Inference images in single GPU with following commands, use `--infer_img` to inference a single image and `--infer_dir` to inference all images in the directory.
+
+```bash
+# inference single image
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams --infer_img=demo/000000014439_640x640.jpg
+
+# inference all images in the directory
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams --infer_dir=demo
+```
+
+### Exporting models
+
+For deployment on GPU or speed testing, model should be first exported to inference model using `tools/export_model.py`.
+
+**Exporting PP-YOLOE+ for Paddle Inference without TensorRT**, use following command
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams
+```
+
+**Exporting PP-YOLOE+ for Paddle Inference with TensorRT** for better performance, use following command with extra `-o trt=True` setting.
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+```
+
+If you want to export PP-YOLOE model to **ONNX format**, use following command refer to [PaddleDetection Model Export as ONNX Format Tutorial](../../deploy/EXPORT_ONNX_MODEL_en.md).
+
+```bash
+# export inference model
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml --output_dir=output_inference -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+
+# install paddle2onnx
+pip install paddle2onnx
+
+# convert to onnx
+paddle2onnx --model_dir output_inference/ppyoloe_plus_crn_l_80e_coco --model_filename model.pdmodel --params_filename model.pdiparams --opset_version 11 --save_file ppyoloe_plus_crn_l_80e_coco.onnx
+
+```
+
+**Notes:** ONNX model only supports batch_size=1 now
+
+### Speed testing
+
+For fair comparison, the speed in [Model Zoo](#Model-Zoo) do not contains the time cost of data reading and post-processing(NMS), which is same as [YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet) in testing method. Thus, you should export model with extra `-o exclude_nms=True` setting.
+
+**Using Paddle Inference without TensorRT** to test speed, run following command
+
+```bash
+# export inference model
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams exclude_nms=True
+
+# speed testing with run_benchmark=True
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --run_mode=paddle --device=gpu --run_benchmark=True
+```
+
+**Using Paddle Inference with TensorRT** to test speed, run following command
+
+```bash
+# export inference model with trt=True
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams exclude_nms=True trt=True
+
+# speed testing with run_benchmark=True,run_mode=trt_fp32/trt_fp16
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --run_mode=trt_fp16 --device=gpu --run_benchmark=True
+
+```
+
+**Using TensorRT Inference with ONNX** to test speed, run following command
+
+```bash
+# export inference model with trt=True
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_s_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_80e_coco.pdparams exclude_nms=True trt=True
+
+# convert to onnx
+paddle2onnx --model_dir output_inference/ppyoloe_plus_crn_s_80e_coco --model_filename model.pdmodel --params_filename model.pdiparams --opset_version 12 --save_file ppyoloe_plus_crn_s_80e_coco.onnx
+
+# trt inference using fp16 and batch_size=1
+trtexec --onnx=./ppyoloe_plus_crn_s_80e_coco.onnx --saveEngine=./ppyoloe_s_bs1.engine --workspace=1024 --avgRuns=1000 --shapes=image:1x3x640x640,scale_factor:1x2 --fp16
+
+# trt inference using fp16 and batch_size=32
+trtexec --onnx=./ppyoloe_plus_crn_s_80e_coco.onnx --saveEngine=./ppyoloe_s_bs32.engine --workspace=1024 --avgRuns=1000 --shapes=image:32x3x640x640,scale_factor:32x2 --fp16
+
+# Using the above script, T4 and tensorrt 7.2 machine, the speed of PPYOLOE-s model is as follows,
+
+# batch_size=1, 2.80ms, 357fps
+# batch_size=32, 67.69ms, 472fps
+
+```
+
+
+### Deployment
+
+PP-YOLOE can be deployed by following approaches:
+  - Paddle Inference [Python](../../deploy/python) & [C++](../../deploy/cpp)
+  - [Paddle-TensorRT](../../deploy/TENSOR_RT.md)
+  - [PaddleServing](https://github.com/PaddlePaddle/Serving)
+  - [PaddleSlim](../slim)
+
+Next, we will introduce how to use Paddle Inference to deploy PP-YOLOE models in TensorRT FP16 mode.
+
+First, refer to [Paddle Inference Docs](https://www.paddlepaddle.org.cn/inference/master/user_guides/download_lib.html#python), download and install packages corresponding to CUDA, CUDNN and TensorRT version.
+
+Then, Exporting PP-YOLOE for Paddle Inference **with TensorRT**, use following command.
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+```
+
+Finally, inference in TensorRT FP16 mode.
+
+```bash
+# inference single image
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --device=gpu --run_mode=trt_fp16
+
+# inference all images in the directory
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_dir=demo/ --device=gpu  --run_mode=trt_fp16
+
+```
+
+**Notes:**
+- TensorRT will perform optimization for the current hardware platform according to the definition of the network, generate an inference engine and serialize it into a file. This inference engine is only applicable to the current hardware hardware platform. If your hardware and software platform has not changed, you can set `use_static=True` in [enable_tensorrt_engine](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.4/deploy/python/infer.py#L660). In this way, the serialized file generated will be saved in the `output_inference` folder, and the saved serialized file will be loaded the next time when TensorRT is executed.
+- PaddleDetection release/2.4 and later versions will support NMS calling TensorRT, which requires PaddlePaddle release/2.3 and later versions.
+
+### Other Datasets
+
+Model | AP | AP<sub>50</sub>
+---|---|---
+[YOLOX](https://github.com/Megvii-BaseDetection/YOLOX) | 22.6 | 37.5
+[YOLOv5](https://github.com/ultralytics/yolov5) | 26.0 | 42.7
+**PP-YOLOE** | **30.5** | **46.4**
+
+**Notes**
+- Here, we use [VisDrone](https://github.com/VisDrone/VisDrone-Dataset) dataset, and to detect 9 objects including `person, bicycles, car, van, truck, tricycle, awning-tricycle, bus, motor`.
+- Above models trained using official default config, and load pretrained parameters on COCO dataset.
+- *Due to the limited time, more verification results will be supplemented in the future. You are also welcome to contribute to PP-YOLOE*
+
+
+## Appendix
+
+Ablation experiments of PP-YOLOE.
+
+| NO.  |        Model                 | Box AP<sup>val</sup> | Params(M) | FLOPs(G) | V100 FP32 FPS |
+| :--: | :---------------------------: | :------------------: | :-------: | :------: | :-----------: |
+|  A   | PP-YOLOv2          |         49.1         |   54.58   |  115.77   |     68.9     |
+|  B   | A + Anchor-free    |         48.8         |   54.27   |  114.78   |     69.8     |
+|  C   | B + CSPRepResNet   |         49.5         |   47.42   |  101.87   |     85.5     |
+|  D   | C + TAL            |         50.4         |   48.32   |  104.75   |     84.0     |
+|  E   | D + ET-Head        |         50.9         |   52.20   |  110.07   |     78.1     |
--- a/paddle_detection/configs/ppyoloe/README_cn.md
+++ b/paddle_detection/configs/ppyoloe/README_cn.md
@@ -0,0 +1,332 @@
+简体中文 | [English](README.md)
+
+# PP-YOLOE
+
+## 最新动态
+- 发布PP-YOLOE+模型: **(2022.08)**
+  - 使用大规模数据集obj365预训练模型
+  - 在backbone中block分支中增加alpha参数
+  - 优化端到端推理速度，提升训练收敛速度
+
+## 历史版本模型
+- 详情请参考：[PP-YOLOE 2022.03版本](./README_legacy.md)
+
+## 内容
+- [简介](#简介)
+- [模型库](#模型库)
+- [使用说明](#使用说明)
+- [附录](#附录)
+
+## 简介
+PP-YOLOE是基于PP-YOLOv2的卓越的单阶段Anchor-free模型，超越了多种流行的YOLO模型。PP-YOLOE有一系列的模型，即s/m/l/x，可以通过width multiplier和depth multiplier配置。PP-YOLOE避免了使用诸如Deformable Convolution或者Matrix NMS之类的特殊算子，以使其能轻松地部署在多种多样的硬件上。更多细节可以参考我们的[report](https://arxiv.org/abs/2203.16250)。
+
+<div align="center">
+  <img src="../../docs/images/ppyoloe_plus_map_fps.png" width=500 />
+</div>
+
+PP-YOLOE+_l在COCO test-dev2017达到了53.3的mAP, 同时其速度在Tesla V100上达到了78.1 FPS。PP-YOLOE+_s/m/x同样具有卓越的精度速度性价比, 其精度速度可以在[模型库](#模型库)中找到。
+
+PP-YOLOE由以下方法组成
+- 可扩展的backbone和neck
+- [Task Alignment Learning](https://arxiv.org/abs/2108.07755)
+- Efficient Task-aligned head with [DFL](https://arxiv.org/abs/2006.04388)和[VFL](https://arxiv.org/abs/2008.13367)
+- [SiLU(Swish)激活函数](https://arxiv.org/abs/1710.05941)
+
+## 模型库
+
+### COCO数据集模型库
+
+|       模型        | Epoch |   GPU个数   | 每GPU图片个数 |  骨干网络  |    输入尺寸    | Box AP<sup>val<br>0.5:0.95 | Box AP<sup>test<br>0.5:0.95 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |                                         模型下载                                         |                    配置文件                     |
+|:---------------:|:-----:|:---------:|:--------:|:----------:|:----------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------:| :---------------------: |:------------------------------------------------------------------------------------:|:-------------------------------------------:|
+|   PP-YOLOE+_s   |  80   |     8     |    8     | cspresnet-s |    640     |            43.7            |            43.9             |   7.93    |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_s_80e_coco.yml) |
+|   PP-YOLOE+_m   |  80   |     8     |    8     | cspresnet-m |    640     |            49.8            |            50.0             |   23.43   |  49.91   |   123.4   |  208.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_m_80e_coco.yml) |
+|   PP-YOLOE+_m(distill)   |  80   |     8     |    8     | cspresnet-m |    640     |      **51.0**     |            51.2                |   23.43   |  49.91   |   123.4   |  208.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco_distill.pdparams) | [config](distill/ppyoloe_plus_crn_m_80e_coco_distill.yml) |
+|   PP-YOLOE+_l   |  80   |     8     |    8     | cspresnet-l |    640     |            52.9            |            53.3             |   52.20   |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_l_80e_coco.yml) |
+|   PP-YOLOE+_l(distill)   |  80   |     8     |    8     | cspresnet-l |    640     |      **54.0**     |            54.4                |   52.20   |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco_distill.pdparams) | [config](distill/ppyoloe_plus_crn_l_80e_coco_distill.yml) |
+|   PP-YOLOE+_x   |  80   |     8     |    8     | cspresnet-x |    640     |            54.7            |            54.9             |   98.42   |  206.59  |   45.0    |   95.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_x_80e_coco.pdparams) | [config](./ppyoloe_plus_crn_x_80e_coco.yml) |
+
+**注意:**:
+- M和L模型使用了蒸馏，具体请参考[distill](distill)。
+
+#### Tiny模型
+
+|    模型    | Epoch |  GPU个数  | 每GPU图片个数 |  骨干网络  |  输入尺寸  | Box AP<sup>val<br>0.5:0.95 | Box AP<sup>test<br>0.5:0.95 | Params(M) | FLOPs(G) | T4 TensorRT FP16(FPS) |  模型下载 |  配置文件 |
+|:----------:|:-----:|:--------:|:-----------:|:---------:|:--------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------------:| :------: |:--------:|
+|   PP-YOLOE+_t-aux(640) |  300   |     8     |    8     | cspresnet-t  |    640  |            39.9       |            56.6             |   4.85   |  19.15 |  344.8   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(640)-relu |  300   |     8     |    8     | cspresnet-t  |    640  |            36.4       |            53.0        |   3.60   |  12.17 |   476.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_relu_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_relu_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(320) |  300   |     8     |    8     | cspresnet-t  |    320  |            33.3       |            48.5             |   4.85   |  4.80 |   729.9   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_320_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_320_300e_coco.yml) |
+|   PP-YOLOE+_t-aux(320)-relu |  300   |     8     |    8     | cspresnet-t  |    320  |            30.1       |            44.7        |   3.60   |  3.04 |   984.8   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.pdparams) | [config](./ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.yml) |
+
+
+### 综合指标
+|            模型            | Epoch | AP<sup>0.5:0.95 | AP<sup>0.5 | AP<sup>0.75 | AP<sup>small | AP<sup>medium | AP<sup>large | AR<sup>small | AR<sup>medium | AR<sup>large |
+|:------------------------:|:-----:|:---------------:|:----------:|:-----------:|:------------:|:-------------:|:------------:|:------------:|:-------------:|:------------:|
+|       PP-YOLOE+_s        |  80   |      43.7       |    60.6    |    47.9     |     26.5     |     47.5      |     59.0     |     46.7     |     71.4      |     81.7     |
+|       PP-YOLOE+_m        |  80   |      49.8       |    67.1    |    54.5     |     31.8     |     53.9      |     66.2     |     53.3     |     75.0      |     84.6     |
+|       PP-YOLOE+_m(distill)|  80   |      51.0       |    68.1    |    55.8     |     32.5     |     55.7      |     67.4     |     51.9     |     76.1      |     86.4     |
+|       PP-YOLOE+_l        |  80   |      52.9       |    70.1    |    57.9     |     35.2     |     57.5      |     69.1     |     56.0     |     77.9      |     86.9     |
+|       PP-YOLOE+_l(distill)|  80   |      54.0       |    71.2    |    59.2     |     36.1     |     58.8      |     70.4     |     55.0     |     78.7      |     87.7     |
+|       PP-YOLOE+_x        |  80   |      54.7       |    72.0    |    59.9     |     37.9     |     59.3      |     70.4     |     57.0     |     78.7      |     87.2     |
+
+**注意:**:
+- M和L模型使用了蒸馏，具体请参考[distill](distill)。
+
+
+### 端到端速度
+|            模型            | AP<sup>0.5:0.95 | TRT-FP32(fps) | TRT-FP16(fps) |
+|:------------------------:|:---------------:|:-------------:|:-------------:|
+|       PP-YOLOE+_s        |      43.7       |     44.44     |     47.85     |
+|       PP-YOLOE+_m        |      49.8       |     39.06     |     43.86     |
+|       PP-YOLOE+_l        |      52.9       |     34.01     |     42.02     |
+|       PP-YOLOE+_x        |      54.7       |     26.88     |     36.76     |
+
+**注意:**
+
+- PP-YOLOE模型使用COCO数据集中train2017作为训练集，使用val2017和test-dev2017作为测试集。
+- 综合指标的表格与模型库的表格里的模型权重是**同一个权重**，综合指标是使用**val2017**作为验证精度的。
+- PP-YOLOE模型训练过程中使用8 GPUs进行混合精度训练，如果**GPU卡数**或者**batch size**发生了改变，你需要按照公式 **lr<sub>new</sub> = lr<sub>default</sub> * (batch_size<sub>new</sub> * GPU_number<sub>new</sub>) / (batch_size<sub>default</sub> * GPU_number<sub>default</sub>)** 调整学习率。
+- PP-YOLOE模型推理速度测试采用单卡V100，batch size=1进行测试，使用**CUDA 10.2**, **CUDNN 7.6.5**，TensorRT推理速度测试使用**TensorRT 6.0.1.8**。
+- 参考[速度测试](#速度测试)以复现PP-YOLOE推理速度测试结果。
+- 如果你设置了`--run_benchmark=True`, 你首先需要安装以下依赖`pip install pynvml psutil GPUtil`。
+- 端到端速度测试包含模型前处理 + 模型推理 + 模型后处理及NMS的时间，测试使用**Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz**, **单卡V100**, **CUDA 11.2**, **CUDNN 8.2.0**, **TensorRT 8.0.1.6**。
+
+### Objects365数据集模型库
+|       模型       | Epoch |  机器个数 | GPU个数   | 每GPU图片个数 |  骨干网络  |   输入尺寸   | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  模型下载  | 配置文件 |
+|:---------------:|:-----:|:-----------:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :--------:|:--------:|
+|   PP-YOLOE+_s   |  60   |  3 |  8     |    8     | cspresnet-s |    640     |   18.1  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_s_60e_objects365.yml) |
+|   PP-YOLOE+_m   |  60   |   4 |  8     |    8     | cspresnet-m |    640     |   25.0  |  23.43  |  49.91   |   123.4       |  208.3  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_m_60e_objects365.yml) |
+|   PP-YOLOE+_l   |  60   |   3 |  8     |    8     | cspresnet-l |    640     |   30.8  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_l_60e_objects365.yml) |
+|   PP-YOLOE+_x   |  60   |  4 |   8     |    8     | cspresnet-x |    640     |   32.7  |  98.42 |  206.59      |   45.0        |  95.2  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_x_obj365_pretrained.pdparams) | [config](./objects365/ppyoloe_plus_crn_x_60e_objects365.yml) |
+
+
+**注意:**
+- 多机训练细节见[文档](../../docs/tutorials/DistributedTraining_cn.md)
+- Objects365数据集下载请参考[objects365官网](http://www.objects365.org/overview.html)。具体种类列表可下载由PaddleDetection团队整理的[objects365_detection_label_list.txt](https://bj.bcebos.com/v1/paddledet/data/objects365/objects365_detection_label_list.txt)并存放在`dataset/objects365/`，每一行即表示第几个种类。inference或导出模型时需要读取到种类数，如果没有标注json文件时，可以进行如下更改`configs/datasets/objects365_detection.yml`：
+```
+TestDataset:
+  !ImageFolder
+    # anno_path: annotations/zhiyuan_objv2_val.json
+    anno_path: objects365_detection_label_list.txt
+    dataset_dir: dataset/objects365/
+```
+
+
+### VOC数据集模型库
+|       模型       | Epoch |   GPU个数   | 每GPU图片个数 |  骨干网络  |   输入尺寸   | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  模型下载  |  配置文件 |
+|:---------------:|:-----:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :-------: |:--------:|
+|   PP-YOLOE+_s   |  30   |     8     |    8     | cspresnet-s |    640     |   86.7  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_30e_voc.pdparams) | [config](./voc/ppyoloe_plus_crn_s_30e_voc.yml) |
+|   PP-YOLOE+_l   |  30   |     8     |    8     | cspresnet-l |    640     |   89.0  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_30e_voc.pdparams) | [config](./voc/ppyoloe_plus_crn_l_30e_voc.yml) |
+
+
+### 垂类应用模型
+
+PaddleDetection团队提供了基于PP-YOLOE的各种垂类检测模型的配置文件和权重，用户可以下载进行使用：
+
+|     场景    |    相关数据集    |    链接   |
+| :--------: | :---------: | :------: |
+|  行人检测   | CrowdHuman  |   [pphuman](../pphuman)  |
+|  车辆检测   | BDD100K、UA-DETRAC  |  [ppvehicle](../ppvehicle)   |
+|  小目标检测 | VisDrone、DOTA、xView |  [smalldet](../smalldet)   |
+| 密集目标检测 | SKU110k | [application](./application) |
+| 旋转框检测 | DOTA | [PP-YOLOE-R](../rotate/ppyoloe_r/) |
+
+
+## 使用说明
+
+### 数据集和评价指标
+
+下载PaddleDetection团队提供的**COCO和VOC数据**，并解压放置于`PaddleDetection/dataset/`下：
+
+```
+wget https://bj.bcebos.com/v1/paddledet/data/coco.tar
+# tar -xvf coco.tar
+
+wget https://bj.bcebos.com/v1/paddledet/data/voc.zip
+# unzip voc.zip
+```
+
+**注意:**
+ - COCO风格格式，请参考 [format-data](https://cocodataset.org/#format-data) 和 [format-results](https://cocodataset.org/#format-results)。
+ - COCO风格评测指标，请参考 [detection-eval](https://cocodataset.org/#detection-eval) ，并首先安装 [cocoapi](https://github.com/cocodataset/cocoapi)。
+ - VOC风格格式和评测指标，请参考 [VOC2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/index.html)。
+
+### 自定义数据集
+
+1.自定义数据集的标注制作，请参考 [DetAnnoTools](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/docs/tutorials/data/DetAnnoTools.md);
+2.自定义数据集的训练准备，请参考 [PrepareDataSet](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/docs/tutorials/data/PrepareDetDataSet.md).
+
+
+### 训练
+
+请执行以下指令训练PP-YOLOE+
+
+```bash
+python -m paddle.distributed.launch --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml --eval --amp
+```
+**注意:**
+- 如果需要边训练边评估，请添加`--eval`.
+- PP-YOLOE+支持混合精度训练，请添加`--amp`.
+- PaddleDetection支持多机训练，可以参考[多机训练教程](../../docs/tutorials/DistributedTraining_cn.md).
+
+### 评估
+
+执行以下命令在单个GPU上评估COCO val2017数据集
+
+```bash
+CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams
+```
+
+在coco test-dev2017上评估，请先从[COCO数据集下载](https://cocodataset.org/#download)下载COCO test-dev2017数据集，然后解压到COCO数据集文件夹并像`configs/ppyolo/ppyolo_test.yml`一样配置`EvalDataset`。
+
+### 推理
+
+使用以下命令在单张GPU上预测图片，使用`--infer_img`推理单张图片以及使用`--infer_dir`推理文件中的所有图片。
+
+
+```bash
+# 推理单张图片
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams --infer_img=demo/000000014439_640x640.jpg
+
+# 推理文件中的所有图片
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams --infer_dir=demo
+```
+
+### 模型导出
+
+PP-YOLOE+在GPU上部署或者速度测试需要通过`tools/export_model.py`导出模型。
+
+当你**使用Paddle Inference但不使用TensorRT**时，运行以下的命令导出模型
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams
+```
+
+当你**使用Paddle Inference且使用TensorRT**时，需要指定`-o trt=True`来导出模型。
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+```
+
+如果你想将PP-YOLOE模型导出为**ONNX格式**，参考
+[PaddleDetection模型导出为ONNX格式教程](../../deploy/EXPORT_ONNX_MODEL.md)，运行以下命令：
+
+```bash
+
+# 导出推理模型
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml --output_dir=output_inference -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+
+# 安装paddle2onnx
+pip install paddle2onnx
+
+# 转换成onnx格式
+paddle2onnx --model_dir output_inference/ppyoloe_plus_crn_l_80e_coco --model_filename model.pdmodel --params_filename model.pdiparams --opset_version 11 --save_file ppyoloe_plus_crn_l_80e_coco.onnx
+```
+
+**注意：** ONNX模型目前只支持batch_size=1
+
+### 速度测试
+
+为了公平起见，在[模型库](#模型库)中的速度测试结果均为不包含数据预处理和模型输出后处理(NMS)的数据(与[YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet)测试方法一致)，需要在导出模型时指定`-o exclude_nms=True`.
+
+**使用Paddle Inference但不使用TensorRT**进行测速，执行以下命令：
+
+```bash
+# 导出模型
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams exclude_nms=True
+
+# 速度测试，使用run_benchmark=True
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --run_mode=paddle --device=gpu --run_benchmark=True
+```
+
+**使用Paddle Inference且使用TensorRT**进行测速，执行以下命令：
+
+```bash
+# 导出模型，使用trt=True
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams exclude_nms=True trt=True
+
+# 速度测试，使用run_benchmark=True, run_mode=trt_fp32/trt_fp16
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --run_mode=trt_fp16 --device=gpu --run_benchmark=True
+
+```
+
+
+**使用 ONNX 和 TensorRT** 进行测速，执行以下命令：
+
+```bash
+# 导出模型
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_s_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_80e_coco.pdparams exclude_nms=True trt=True
+
+# 转化成ONNX格式
+paddle2onnx --model_dir output_inference/ppyoloe_plus_crn_s_80e_coco --model_filename model.pdmodel --params_filename model.pdiparams --opset_version 12 --save_file ppyoloe_plus_crn_s_80e_coco.onnx
+
+# 测试速度，半精度，batch_size=1
+trtexec --onnx=./ppyoloe_plus_crn_s_80e_coco.onnx --saveEngine=./ppyoloe_s_bs1.engine --workspace=1024 --avgRuns=1000 --shapes=image:1x3x640x640,scale_factor:1x2 --fp16
+
+# 测试速度，半精度，batch_size=32
+trtexec --onnx=./ppyoloe_plus_crn_s_80e_coco.onnx --saveEngine=./ppyoloe_s_bs32.engine --workspace=1024 --avgRuns=1000 --shapes=image:32x3x640x640,scale_factor:32x2 --fp16
+
+# 使用上边的脚本, 在T4 和 TensorRT 7.2的环境下，PPYOLOE-plus-s模型速度如下
+# batch_size=1, 2.80ms, 357fps
+# batch_size=32, 67.69ms, 472fps
+```
+
+
+
+### 部署
+
+PP-YOLOE可以使用以下方式进行部署：
+  - Paddle Inference [Python](../../deploy/python) & [C++](../../deploy/cpp)
+  - [Paddle-TensorRT](../../deploy/TENSOR_RT.md)
+  - [PaddleServing](https://github.com/PaddlePaddle/Serving)
+  - [PaddleSlim模型量化](../slim)
+
+接下来，我们将介绍PP-YOLOE如何使用Paddle Inference在TensorRT FP16模式下部署
+
+首先，参考[Paddle Inference文档](https://www.paddlepaddle.org.cn/inference/master/user_guides/download_lib.html#python)，下载并安装与你的CUDA, CUDNN和TensorRT相应的wheel包。
+
+然后，运行以下命令导出模型
+
+```bash
+python tools/export_model.py -c configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml -o weights=https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_coco.pdparams trt=True
+```
+
+最后，使用TensorRT FP16进行推理
+
+```bash
+# 推理单张图片
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_file=demo/000000014439_640x640.jpg --device=gpu --run_mode=trt_fp16
+
+# 推理文件夹下的所有图片
+CUDA_VISIBLE_DEVICES=0 python deploy/python/infer.py --model_dir=output_inference/ppyoloe_plus_crn_l_80e_coco --image_dir=demo/ --device=gpu  --run_mode=trt_fp16
+
+```
+
+**注意：**
+- TensorRT会根据网络的定义，执行针对当前硬件平台的优化，生成推理引擎并序列化为文件。该推理引擎只适用于当前软硬件平台。如果你的软硬件平台没有发生变化，你可以设置[enable_tensorrt_engine](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.4/deploy/python/infer.py#L660)的参数`use_static=True`，这样生成的序列化文件将会保存在`output_inference`文件夹下，下次执行TensorRT时将加载保存的序列化文件。
+- PaddleDetection release/2.4及其之后的版本将支持NMS调用TensorRT，需要依赖PaddlePaddle release/2.3及其之后的版本
+
+### 泛化性验证
+
+模型 | AP | AP<sub>50</sub>
+---|---|---
+[YOLOX](https://github.com/Megvii-BaseDetection/YOLOX) | 22.6 | 37.5
+[YOLOv5](https://github.com/ultralytics/yolov5) | 26.0 | 42.7
+**PP-YOLOE** | **30.5** | **46.4**
+
+**注意**
+- 试验使用[VisDrone](https://github.com/VisDrone/VisDrone-Dataset)数据集, 并且检测其中的9类，包括 `person, bicycles, car, van, truck, tricyle, awning-tricyle, bus, motor`.
+- 以上模型训练均采用官方提供的默认参数，并且加载COCO预训练参数
+- *由于人力/时间有限，后续将会持续补充更多验证结果，也欢迎各位开源用户贡献，共同优化PP-YOLOE*
+
+
+## 附录
+
+PP-YOLOE消融实验
+
+| 序号 |        模型                  | Box AP<sup>val</sup> | 参数量(M) | FLOPs(G) | V100 FP32 FPS |
+| :--: | :---------------------------: | :-------------------: | :-------: | :------: | :-----------: |
+|  A   | PP-YOLOv2          |         49.1         |   54.58   |  115.77   |     68.9     |
+|  B   | A + Anchor-free    |         48.8         |   54.27   |  114.78   |     69.8     |
+|  C   | B + CSPRepResNet   |         49.5         |   47.42   |  101.87   |     85.5     |
+|  D   | C + TAL            |         50.4         |   48.32   |  104.75   |     84.0     |
+|  E   | D + ET-Head        |         50.9         |   52.20   |  110.07   |     78.1     |
--- a/paddle_detection/configs/ppyoloe/README_legacy.md
+++ b/paddle_detection/configs/ppyoloe/README_legacy.md
@@ -0,0 +1,39 @@
+# PP-YOLOE Legacy Model Zoo (2022.03)
+
+## Legacy Model Zoo
+|          Model           | Epoch | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>val<br>0.5:0.95 | Box AP<sup>test<br>0.5:0.95 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) | download | config  |
+|:------------------------:|:-------:|:-------:|:--------:|:----------:| :-------:| :------------------: | :-------------------: |:---------:|:--------:|:---------------:| :---------------------: | :------: | :------: |
+| PP-YOLOE-s                  | 400 |     8      |    32    | cspresnet-s |     640     |       43.4        |        43.6         |   7.93    |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_s_400e_coco.pdparams) | [config](./ppyoloe_crn_s_400e_coco.yml)                   |
+| PP-YOLOE-s                  | 300 |     8      |    32    | cspresnet-s |     640     |       43.0        |        43.2         |   7.93    |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_s_300e_coco.pdparams) | [config](./ppyoloe_crn_s_300e_coco.yml)                   |
+| PP-YOLOE-m                  | 300 |     8      |    28    | cspresnet-m |     640     |       49.0        |        49.1         |   23.43   |  49.91   |   123.4   |  208.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams) | [config](./ppyoloe_crn_m_300e_coco.yml)                   |
+| PP-YOLOE-l                  | 300 |     8      |    20    | cspresnet-l |     640     |       51.4        |        51.6         |   52.20   |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_l_300e_coco.pdparams) | [config](./ppyoloe_crn_l_300e_coco.yml)                   |
+| PP-YOLOE-x                  | 300 |     8      |    16    | cspresnet-x |     640     |       52.3        |        52.4         |   98.42   |  206.59  |   45.0    |   95.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_x_300e_coco.pdparams) | [config](./ppyoloe_crn_x_300e_coco.yml)                   |
+
+### Comprehensive Metrics
+|          Model           | Epoch | AP<sup>0.5:0.95 | AP<sup>0.5 |  AP<sup>0.75  | AP<sup>small  | AP<sup>medium | AP<sup>large | AR<sup>small | AR<sup>medium | AR<sup>large | download | config  |
+|:----------------------:|:-----:|:---------------:|:----------:|:-------------:| :------------:| :-----------: | :----------: |:------------:|:-------------:|:------------:| :-----: | :-----: |
+| PP-YOLOE-s             | 400 |      43.4      |     60.0    |     47.5      |     25.7      |      47.8     |     59.2     |     43.9     |      70.8     |   81.9         | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_s_400e_coco.pdparams) | [config](./ppyoloe_crn_s_400e_coco.yml)|
+| PP-YOLOE-s             | 300 |      43.0      |     59.6    |     47.2      |     26.0      |      47.4     |     58.7     |     45.1     |      70.6     |   81.4         | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_s_300e_coco.pdparams) | [config](./ppyoloe_crn_s_300e_coco.yml)|
+| PP-YOLOE-m             | 300 |      49.0      |     65.9    |     53.8      |     30.9      |      53.5     |     65.3     |     50.9     |      74.4     |   84.7         | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams) | [config](./ppyoloe_crn_m_300e_coco.yml)|
+| PP-YOLOE-l             | 300 |      51.4      |     68.6    |     56.2      |     34.8      |      56.1     |     68.0     |     53.1     |      76.8     |   85.6         | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_l_300e_coco.pdparams) | [config](./ppyoloe_crn_l_300e_coco.yml)|
+| PP-YOLOE-x             | 300 |      52.3      |     69.5    |     56.8      |     35.1      |      57.0     |     68.6     |     55.5     |      76.9     |   85.7         | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_x_300e_coco.pdparams) | [config](./ppyoloe_crn_x_300e_coco.yml)|
+
+
+**Notes:**
+
+- PP-YOLOE is trained on COCO train2017 dataset and evaluated on val2017 & test-dev2017 dataset.
+- The model weights in the table of Comprehensive Metrics are **the same as** that in the original Model Zoo, and evaluated on **val2017**.
+- PP-YOLOE used 8 GPUs for training, if **GPU number** or **mini-batch size** is changed, **learning rate** should be adjusted according to the formula **lr<sub>new</sub> = lr<sub>default</sub> * (batch_size<sub>new</sub> * GPU_number<sub>new</sub>) / (batch_size<sub>default</sub> * GPU_number<sub>default</sub>)**.
+- PP-YOLOE inference speed is tesed on single Tesla V100 with batch size as 1, **CUDA 10.2**, **CUDNN 7.6.5**, **TensorRT 6.0.1.8** in TensorRT mode.
+
+## Appendix
+
+Ablation experiments of PP-YOLOE.
+
+| NO.  |        Model                 | Box AP<sup>val</sup> | Params(M) | FLOPs(G) | V100 FP32 FPS |
+| :--: | :---------------------------: | :------------------: | :-------: | :------: | :-----------: |
+|  A   | PP-YOLOv2          |         49.1         |   54.58   |  115.77   |     68.9     |
+|  B   | A + Anchor-free    |         48.8         |   54.27   |  114.78   |     69.8     |
+|  C   | B + CSPRepResNet   |         49.5         |   47.42   |  101.87   |     85.5     |
+|  D   | C + TAL            |         50.4         |   48.32   |  104.75   |     84.0     |
+|  E   | D + ET-Head        |         50.9         |   52.20   |  110.07   |     78.1     |
--- a/paddle_detection/configs/ppyoloe/_base_/optimizer_300e.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/optimizer_300e.yml
@@ -0,0 +1,18 @@
+epoch: 300
+
+LearningRate:
+  base_lr: 0.01
+  schedulers:
+    - name: CosineDecay
+      max_epochs: 360
+    - name: LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/_base_/optimizer_36e_xpu.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/optimizer_36e_xpu.yml
@@ -0,0 +1,18 @@
+epoch: 36
+
+LearningRate:
+  base_lr: 0.00125
+  schedulers:
+    - name: CosineDecay
+      max_epochs: 43
+    - name: LinearWarmup
+      start_factor: 0.001
+      steps: 2000
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/_base_/optimizer_400e.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/optimizer_400e.yml
@@ -0,0 +1,18 @@
+epoch: 400
+
+LearningRate:
+  base_lr: 0.01
+  schedulers:
+    - name: CosineDecay
+      max_epochs: 480
+    - name: LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/_base_/optimizer_60e.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/optimizer_60e.yml
@@ -0,0 +1,18 @@
+epoch: 60
+
+LearningRate:
+  base_lr: 0.001
+  schedulers:
+    - name: CosineDecay
+      max_epochs: 72
+    - name: LinearWarmup
+      start_factor: 0.
+      epochs: 1
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/_base_/optimizer_80e.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/optimizer_80e.yml
@@ -0,0 +1,18 @@
+epoch: 80
+
+LearningRate:
+  base_lr: 0.001
+  schedulers:
+    - name: CosineDecay
+      max_epochs: 96
+    - name: LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_crn.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_crn.yml
@@ -0,0 +1,47 @@
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+ema_black_list: ['proj_conv.weight']
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: 100
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_crn.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_crn.yml
@@ -0,0 +1,48 @@
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+ema_black_list: ['proj_conv.weight']
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: 30
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_crn_tiny_auxhead.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_crn_tiny_auxhead.yml
@@ -0,0 +1,60 @@
+architecture: PPYOLOEWithAuxHead
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+ema_black_list: ['proj_conv.weight']
+custom_black_list: ['reduce_mean']
+
+PPYOLOEWithAuxHead:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  aux_head: SimpleConvHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [384, 384, 384]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+
+SimpleConvHead:
+  feat_in: 288
+  feat_out: 288
+  num_convs: 1
+  fpn_strides: [32, 16, 8]
+  norm_type: 'gn'
+  act: 'LeakyReLU'
+  reg_max: 16
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: 100
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  attn_conv: 'repvgg' #
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+    is_close_gt: True #
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_reader.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_reader.yml
@@ -0,0 +1,40 @@
+worker_num: 4
+eval_height: &eval_height 640
+eval_width: &eval_width 640
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [320, 352, 384, 416, 448, 480, 512, 544, 576, 608, 640, 672, 704, 736, 768], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_reader_320.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_plus_reader_320.yml
@@ -0,0 +1,40 @@
+worker_num: 4
+eval_height: &eval_height 320
+eval_width: &eval_width 320
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [224, 256, 288, 320, 352, 384, 416, 448, 480, 512, 544], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
--- a/paddle_detection/configs/ppyoloe/_base_/ppyoloe_reader.yml
+++ b/paddle_detection/configs/ppyoloe/_base_/ppyoloe_reader.yml
@@ -0,0 +1,40 @@
+worker_num: 4
+eval_height: &eval_height 640
+eval_width: &eval_width 640
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [320, 352, 384, 416, 448, 480, 512, 544, 576, 608, 640, 672, 704, 736, 768], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: True}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: True}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: True}
+    - Permute: {}
+  batch_size: 1
--- a/paddle_detection/configs/ppyoloe/application/README.md
+++ b/paddle_detection/configs/ppyoloe/application/README.md
@@ -0,0 +1,69 @@
+# PP-YOLOE+ 下游任务
+
+我们验证了PP-YOLOE+模型强大的泛化能力，在农业、低光、工业等不同场景下游任务检测效果稳定提升!
+
+农业数据集采用[Embrapa WGISD](https://github.com/thsant/wgisd)，该数据集用于葡萄栽培中基于图像的监测和现场机器人技术，提供了来自5种不同葡萄品种的实地实例，
+处理后的COCO格式，包含图片训练集242张，测试集58张，5个类别，[Embrapa WGISD COCO格式下载](https://bj.bcebos.com/v1/paddledet/data/wgisd.zip)；
+
+低光数据集使用[ExDark](https://github.com/cs-chan/Exclusively-Dark-Image-Dataset/tree/master/Dataset)，该数据集是一个专门在低光照环境下拍摄出针对低光目标检测的数据集，包括从极低光环境到暮光环境等10种不同光照条件下的图片，
+处理后的COCO格式，包含图片训练集5891张，测试集1472张，12个类别，[ExDark COCO格式下载](https://bj.bcebos.com/v1/paddledet/data/Exdark.zip)；
+
+工业数据集使用[PKU-Market-PCB](https://robotics.pkusz.edu.cn/resources/dataset/)，该数据集用于印刷电路板（PCB）的瑕疵检测，提供了6种常见的PCB缺陷，
+处理后的COCO格式，包含图片训练集555张，测试集138张，6个类别，[PKU-Market-PCB COCO格式下载](https://bj.bcebos.com/v1/paddledet/data/PCB_coco.zip)。
+
+商超数据集[SKU110k](https://github.com/eg4000/SKU110K_CVPR19)是商品超市场景下的密集目标检测数据集，包含11,762张图片和超过170个实例。其中包括8,233张用于训练的图像、588张用于验证的图像和2,941张用于测试的图像。
+
+
+## 实验结果：
+
+|    模型  |       数据集     | mAP<sup>val<br>0.5:0.95 |  下载链接  | 配置文件 |
+|:---------|:---------------:|:-----------------------:|:---------:| :-----: |
+|PP-YOLOE_m|   Embrapa WGISD  |  52.7 | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_80e_wgisd.pdparams) | [配置文件](./ppyoloe_crn_m_80e_wgisd.yml) |
+|PP-YOLOE+_m<br>(obj365_pretrained)|   Embrapa WGISD  |  60.8(+8.1) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_obj365_pretrained_wgisd.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_obj365_pretrained_wgisd.yml) |
+|PP-YOLOE+_m<br>(coco_pretrained)|   Embrapa WGISD  |  59.7(+7.0) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco_pretrained_wgisd.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_coco_pretrained_wgisd.yml) |
+|PP-YOLOE_m|      ExDark      |  56.4 | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_80e_exdark.pdparams) | [配置文件](./ppyoloe_crn_m_80e_exdark.yml) |
+|PP-YOLOE+_m<br>(obj365_pretrained)|   ExDark  |  57.7(+1.3) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_obj365_pretrained_exdark.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_obj365_pretrained_exdark.yml) |
+|PP-YOLOE+_m<br>(coco_pretrained)|   ExDark  |  58.1(+1.7) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco_pretrained_exdark.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_coco_pretrained_exdark.yml) |
+|PP-YOLOE_m|      PKU-Market-PCB      |  50.8 | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_80e_pcb.pdparams) | [配置文件](./ppyoloe_crn_m_80e_pcb.yml) |
+|PP-YOLOE+_m<br>(obj365_pretrained)|   PKU-Market-PCB  |  52.7(+1.9) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_obj365_pretrained_pcb.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_obj365_pretrained_pcb.yml) |
+|PP-YOLOE+_m<br>(coco_pretrained)|   PKU-Market-PCB  |  52.4(+1.6) | [下载链接](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_coco_pretrained_pcb.pdparams) | [配置文件](./ppyoloe_plus_crn_m_80e_coco_pretrained_pcb.yml) |
+
+**注意:**
+- PP-YOLOE模型训练过程中使用8 GPUs进行训练，如果**GPU卡数**或者**batch size**发生了改变，你需要按照公式 **lr<sub>new</sub> = lr<sub>default</sub> * (batch_size<sub>new</sub> * GPU_number<sub>new</sub>) / (batch_size<sub>default</sub> * GPU_number<sub>default</sub>)** 调整学习率。
+- 具体使用教程请参考[ppyoloe](../ppyoloe#getting-start)。  
+
+
+## SKU110k Model ZOO
+|     Model      | Epoch | GPU number | images/GPU |  backbone  | input shape | Box AP<sup>val<br>0.5:0.95 (maxDets=300) | Box AP<sup>test<br>0.5:0.95 (maxDets=300) | download | config |
+|:--------------:|:-----:|:-------:|:----------:|:----------:| :-------:|:-------------------------:|:---------------------------:|:---------:|:------:|
+| PP-YOLOE+_s | 80 | 8 | 8 | cspresnet-s | 960 | 57.4 | 58.8 | [download](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_80e_sku110k.pdparams) | [config](./ppyoloe_plus_crn_s_80e_sku110k.yml) |
+| PP-YOLOE+_m | 80 | 8 | 8 | cspresnet-m | 960 | 58.2 | 59.7 | [download](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_m_80e_sku110k.pdparams) | [config](./ppyoloe_plus_crn_m_80e_sku110k.yml) |
+| PP-YOLOE+_l | 80 | 8 | 4 | cspresnet-l | 960 | 58.8 | 60.2 | [download](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_80e_sku110k.pdparams) | [config](./ppyoloe_plus_crn_l_80e_sku110k.yml) |
+| PP-YOLOE+_x | 80 | 8 | 4 | cspresnet-x | 960 | 59.0 | 60.3 | [download](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_x_80e_sku110k.pdparams) | [config](./ppyoloe_plus_crn_x_80e_sku110k.yml) |
+
+
+**注意:**
+- SKU110k系列模型训练过程中使用8 GPUs进行训练，如果**GPU卡数**或者**batch size**发生了改变，你需要按照公式 **lr<sub>new</sub> = lr<sub>default</sub> * (batch_size<sub>new</sub> * GPU_number<sub>new</sub>) / (batch_size<sub>default</sub> * GPU_number<sub>default</sub>)** 调整学习率。
+- SKU110k数据集使用**maxDets=300**的mAP值作为评估指标。
+- 具体使用教程请参考[ppyoloe](../ppyoloe#getting-start)。
+
+
+## 引用
+```
+@inproceedings{goldman2019dense,
+ author    = {Eran Goldman and Roei Herzig and Aviv Eisenschtat and Jacob Goldberger and Tal Hassner},
+ title     = {Precise Detection in Densely Packed Scenes},
+ booktitle = {Proc. Conf. Comput. Vision Pattern Recognition (CVPR)},
+ year      = {2019}
+}
+
+@article{Exdark,
+title={Getting to Know Low-light Images with The Exclusively Dark Dataset},
+author={Loh, Yuen Peng and Chan, Chee Seng},
+journal={Computer Vision and Image Understanding},
+volume={178},
+pages={30-42},
+year={2019},
+doi={https://doi.org/10.1016/j.cviu.2018.10.010}
+}
+```
--- a/paddle_detection/configs/ppyoloe/application/_base_/exdark_detection.yml
+++ b/paddle_detection/configs/ppyoloe/application/_base_/exdark_detection.yml
@@ -0,0 +1,20 @@
+metric: COCO
+num_classes: 12
+
+TrainDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: coco_annotations/train.json
+    dataset_dir: dataset/Exdark/
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: coco_annotations/val.json
+    dataset_dir: dataset/Exdark/
+
+TestDataset:
+  !ImageFolder
+    anno_path: coco_annotations/val.json # also support txt (like VOC's label_list.txt)
+    dataset_dir: dataset/Exdark/ # if set, anno_path will be 'dataset_dir/anno_path'
--- a/paddle_detection/configs/ppyoloe/application/_base_/pcb_detection.yml
+++ b/paddle_detection/configs/ppyoloe/application/_base_/pcb_detection.yml
@@ -0,0 +1,20 @@
+metric: COCO
+num_classes: 6
+
+TrainDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: pcb_cocoanno/train.json
+    dataset_dir: dataset/PCB_coco/
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: pcb_cocoanno/val.json
+    dataset_dir: dataset/PCB_coco/
+
+TestDataset:
+  !ImageFolder
+    anno_path: pcb_cocoanno/val.json # also support txt (like VOC's label_list.txt)
+    dataset_dir: dataset/PCB_coco/ # if set, anno_path will be 'dataset_dir/anno_path'
--- a/paddle_detection/configs/ppyoloe/application/_base_/sku110k.yml
+++ b/paddle_detection/configs/ppyoloe/application/_base_/sku110k.yml
@@ -0,0 +1,21 @@
+metric: COCO
+num_classes: 1
+
+TrainDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: annotations/annotations_train.json
+    dataset_dir: dataset/SKU110K_fixed
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'difficult']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: images
+    anno_path: annotations/annotations_val.json
+    dataset_dir: dataset/SKU110K_fixed
+    allow_empty: true
+
+TestDataset:
+  !ImageFolder
+    anno_path: annotations/annotations_test.json
+    dataset_dir: dataset/SKU110K_fixed
--- a/paddle_detection/configs/ppyoloe/application/_base_/wgisd_detection.yml
+++ b/paddle_detection/configs/ppyoloe/application/_base_/wgisd_detection.yml
@@ -0,0 +1,20 @@
+metric: COCO
+num_classes: 5
+
+TrainDataset:
+  !COCODataSet
+    image_dir: data
+    anno_path: coco_annotations/new_train_bbox_instances.json
+    dataset_dir: dataset/wgisd/
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: data
+    anno_path: coco_annotations/new_test_bbox_instances.json
+    dataset_dir: dataset/wgisd/
+
+TestDataset:
+  !ImageFolder
+    anno_path: coco_annotations/new_test_bbox_instances.json # also support txt (like VOC's label_list.txt)
+    dataset_dir: dataset/wgisd/ # if set, anno_path will be 'dataset_dir/anno_path'
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_exdark.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_exdark.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/exdark_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_crn.yml',
+  '../_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_crn_m_80e_exdark/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_pcb.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_pcb.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/pcb_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_crn.yml',
+  '../_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_crn_m_80e_pcb/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_wgisd.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_crn_m_80e_wgisd.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/wgisd_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_crn.yml',
+  '../_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_crn_m_80e_wgisd/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_l_80e_sku110k.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_l_80e_sku110k.yml
@@ -0,0 +1,127 @@
+_BASE_: [
+  './_base_/sku110k.yml',
+  '../../runtime.yml'
+]
+
+log_iter: 10
+snapshot_epoch: 20
+weights: output/ppyoloe_plus_crn_s_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams
+depth_mult: 1.0
+width_mult: 1.0
+
+
+# arch
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+  use_alpha: True
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: -1
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 3000
+    keep_top_k: 1000
+    score_threshold: 0.01
+    nms_threshold: 0.7
+
+
+# reader
+worker_num: 8
+eval_height: &eval_height 960
+eval_width: &eval_width 960
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [3000, 1800], keep_ratio: True, interp: 2}
+    - RandomDistort: {}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [480, 512, 544, 576, 608, 640, 672, 704, 736, 768, 800, 832, 864, 896, 928, 960, 992, 1024, 1056, 1088, 1120, 1152], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 4
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
+
+
+# optimizer
+epoch: 80
+
+LearningRate:
+  base_lr: 0.002
+  schedulers:
+    - !CosineDecay
+      max_epochs: 96
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_exdark.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_exdark.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/exdark_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_coco_pretrained_exdark/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_m_80e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_pcb.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_pcb.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/pcb_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_coco_pretrained_pcb/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_m_80e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_wgisd.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_coco_pretrained_wgisd.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/wgisd_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_coco_pretrained_wgisd/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_m_80e_coco.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_exdark.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_exdark.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/exdark_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_obj365_pretrained_exdark/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_pcb.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_pcb.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/pcb_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_obj365_pretrained_pcb/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_wgisd.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_obj365_pretrained_wgisd.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  './_base_/wgisd_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_obj365_pretrained_wgisd/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_sku110k.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_m_80e_sku110k.yml
@@ -0,0 +1,127 @@
+_BASE_: [
+  './_base_/sku110k.yml',
+  '../../runtime.yml'
+]
+
+log_iter: 10
+snapshot_epoch: 20
+weights: output/ppyoloe_plus_crn_s_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
+
+
+# arch
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+  use_alpha: True
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: -1
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 3000
+    keep_top_k: 1000
+    score_threshold: 0.01
+    nms_threshold: 0.7
+
+
+# reader
+worker_num: 8
+eval_height: &eval_height 960
+eval_width: &eval_width 960
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [3000, 1800], keep_ratio: True, interp: 2}
+    - RandomDistort: {}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [480, 512, 544, 576, 608, 640, 672, 704, 736, 768, 800, 832, 864, 896, 928, 960, 992, 1024, 1056, 1088, 1120, 1152], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
+
+
+# optimizer
+epoch: 80
+
+LearningRate:
+  base_lr: 0.004
+  schedulers:
+    - !CosineDecay
+      max_epochs: 96
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_s_80e_sku110k.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_s_80e_sku110k.yml
@@ -0,0 +1,127 @@
+_BASE_: [
+  './_base_/sku110k.yml',
+  '../../runtime.yml'
+]
+
+log_iter: 10
+snapshot_epoch: 20
+weights: output/ppyoloe_plus_crn_s_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.50
+
+
+# arch
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+  use_alpha: True
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: -1
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 3000
+    keep_top_k: 1000
+    score_threshold: 0.01
+    nms_threshold: 0.7
+
+
+# reader
+worker_num: 8
+eval_height: &eval_height 960
+eval_width: &eval_width 960
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [3000, 1800], keep_ratio: True, interp: 2}
+    - RandomDistort: {}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [480, 512, 544, 576, 608, 640, 672, 704, 736, 768, 800, 832, 864, 896, 928, 960, 992, 1024, 1056, 1088, 1120, 1152], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
+
+
+# optimizer
+epoch: 80
+
+LearningRate:
+  base_lr: 0.004
+  schedulers:
+    - !CosineDecay
+      max_epochs: 96
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_x_80e_sku110k.yml
+++ b/paddle_detection/configs/ppyoloe/application/ppyoloe_plus_crn_x_80e_sku110k.yml
@@ -0,0 +1,127 @@
+_BASE_: [
+  './_base_/sku110k.yml',
+  '../../runtime.yml'
+]
+
+log_iter: 10
+snapshot_epoch: 20
+weights: output/ppyoloe_plus_crn_s_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_x_obj365_pretrained.pdparams
+depth_mult: 1.33
+width_mult: 1.25
+
+
+# arch
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+  use_alpha: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+  use_alpha: True
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: -1
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 3000
+    keep_top_k: 1000
+    score_threshold: 0.01
+    nms_threshold: 0.7
+
+
+# reader
+worker_num: 8
+eval_height: &eval_height 960
+eval_width: &eval_width 960
+eval_size: &eval_size [*eval_height, *eval_width]
+
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [3000, 1800], keep_ratio: True, interp: 2}
+    - RandomDistort: {}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [480, 512, 544, 576, 608, 640, 672, 704, 736, 768, 800, 832, 864, 896, 928, 960, 992, 1024, 1056, 1088, 1120, 1152], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 4
+  shuffle: true
+  drop_last: true
+  use_shared_memory: true
+  collate_batch: true
+
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 2
+
+TestReader:
+  inputs_def:
+    image_shape: [3, *eval_height, *eval_width]
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: *eval_size, keep_ratio: False, interp: 2}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 1
+
+
+# optimizer
+epoch: 80
+
+LearningRate:
+  base_lr: 0.002
+  schedulers:
+    - !CosineDecay
+      max_epochs: 96
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 5
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2
--- a/paddle_detection/configs/ppyoloe/distill/README.md
+++ b/paddle_detection/configs/ppyoloe/distill/README.md
@@ -0,0 +1,46 @@
+# PPYOLOE+ Distillation(PPYOLOE+ 蒸馏)
+
+PaddleDetection提供了对PPYOLOE+ 进行模型蒸馏的方案，结合了logits蒸馏和feature蒸馏。更多蒸馏方案可以查看[slim/distill](../../slim/distill/)。
+
+## 模型库
+
+| 模型               |    方案     | 输入尺寸 | epochs |    Box mAP    |       配置文件    |     下载链接    |
+| ----------------- | ----------- | ------ | :----: | :-----------: | :--------------: | :------------: |
+|   PP-YOLOE+_x     |  teacher   |  640     | 80e   |      54.7     | [config](../ppyoloe_plus_crn_x_80e_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_x_80e_coco.pdparams) |
+|   PP-YOLOE+_l     |  student   |  640     | 80e   |      52.9     | [config](../ppyoloe_plus_crn_l_80e_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_l_80e_coco.pdparams) |
+|   PP-YOLOE+_l     |  distill   |  640     | 80e   |   **54.0(+1.1)**  | [config](./ppyoloe_plus_crn_l_80e_coco_distill.yml),[slim_config](../../slim/distill/ppyoloe_plus_distill_x_distill_l.yml)  | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_l_80e_coco_distill.pdparams) |
+|   PP-YOLOE+_l     |  teacher   |  640     | 80e   |      52.9     | [config](../ppyoloe_plus_crn_l_80e_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_l_80e_coco.pdparams) |
+|   PP-YOLOE+_m     |  student   |  640     | 80e   |      49.8     | [config](../ppyoloe_plus_crn_m_80e_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_m_80e_coco.pdparams) |
+|   PP-YOLOE+_m     |  distill   |  640     | 80e   |    **51.0(+1.2)**    | [config](./ppyoloe_plus_crn_m_80e_coco_distill.yml),[slim_config](../../slim/distill/ppyoloe_plus_distill_l_distill_m.yml)  | [model](https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_m_80e_coco_distill.pdparams) |
+
+## 快速开始
+
+### 训练
+```shell
+# 单卡
+python tools/train.py -c configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml --slim_config configs/slim/distill/ppyoloe_plus_distill_x_distill_l.yml
+# 多卡
+python -m paddle.distributed.launch --log_dir=ppyoloe_plus_distill_x_distill_l/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml --slim_config configs/slim/distill/ppyoloe_plus_distill_x_distill_l.yml
+```
+
+- `-c`: 指定模型配置文件，也是student配置文件。
+- `--slim_config`: 指定压缩策略配置文件，也是teacher配置文件。
+
+### 评估
+```shell
+python tools/eval.py -c configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml -o weights=output/ppyoloe_plus_crn_l_80e_coco_distill/model_final.pdparams
+```
+
+- `-c`: 指定模型配置文件，也是student配置文件。
+- `--slim_config`: 指定压缩策略配置文件，也是teacher配置文件。
+- `-o weights`: 指定压缩算法训好的模型路径。
+
+### 测试
+```shell
+python tools/infer.py -c configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml -o weights=output/ppyoloe_plus_crn_l_80e_coco_distill/model_final.pdparams --infer_img=demo/000000014439_640x640.jpg
+```
+
+- `-c`: 指定模型配置文件。
+- `--slim_config`: 指定压缩策略配置文件。
+- `-o weights`: 指定压缩算法训好的模型路径。
+- `--infer_img`: 指定测试图像路径。
--- a/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml
+++ b/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_l_80e_coco_distill.yml
@@ -0,0 +1,39 @@
+_BASE_: [
+  '../ppyoloe_plus_crn_l_80e_coco.yml',
+]
+for_distill: True
+architecture: PPYOLOE
+PPYOLOE:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+
+worker_num: 4
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [640], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: True
+  drop_last: True
+  use_shared_memory: True
+  collate_batch: True
+
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_l_80e_coco_distill/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams
+depth_mult: 1.0
+width_mult: 1.0
--- a/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_m_80e_coco_distill.yml
+++ b/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_m_80e_coco_distill.yml
@@ -0,0 +1,39 @@
+_BASE_: [
+  '../ppyoloe_plus_crn_m_80e_coco.yml',
+]
+for_distill: True
+architecture: PPYOLOE
+PPYOLOE:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+
+worker_num: 4
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [640], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: True
+  drop_last: True
+  use_shared_memory: True
+  collate_batch: True
+
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_coco_distill/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_s_80e_coco_distill.yml
+++ b/paddle_detection/configs/ppyoloe/distill/ppyoloe_plus_crn_s_80e_coco_distill.yml
@@ -0,0 +1,39 @@
+_BASE_: [
+  '../ppyoloe_plus_crn_s_80e_coco.yml',
+]
+for_distill: True
+architecture: PPYOLOE
+PPYOLOE:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+
+worker_num: 4
+TrainReader:
+  sample_transforms:
+    - Decode: {}
+    - RandomDistort: {}
+    - RandomExpand: {fill_value: [123.675, 116.28, 103.53]}
+    - RandomCrop: {}
+    - RandomFlip: {}
+  batch_transforms:
+    - BatchRandomResize: {target_size: [640], random_size: True, random_interp: True, keep_ratio: False}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+    - PadGT: {}
+  batch_size: 8
+  shuffle: True
+  drop_last: True
+  use_shared_memory: True
+  collate_batch: True
+
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_s_80e_coco_distill/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.50
--- a/paddle_detection/configs/ppyoloe/objects365/README_cn.md
+++ b/paddle_detection/configs/ppyoloe/objects365/README_cn.md
@@ -0,0 +1,23 @@
+# PP-YOLOE
+
+## 模型库
+
+### Objects365数据集模型库
+|       模型       | Epoch |  机器个数 | GPU个数   | 每GPU图片个数 |  骨干网络  |   输入尺寸   | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  模型下载  | 配置文件 |
+|:---------------:|:-----:|:-----------:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :--------:|:--------:|
+|   PP-YOLOE+_s   |  60   |  3 |  8     |    8     | cspresnet-s |    640     |   18.1  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams) | [config](./ppyoloe_plus_crn_s_60e_objects365.yml) |
+|   PP-YOLOE+_m   |  60   |   4 |  8     |    8     | cspresnet-m |    640     |   25.0  |  23.43  |  49.91   |   123.4       |  208.3  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams) | [config](./ppyoloe_plus_crn_m_60e_objects365.yml) |
+|   PP-YOLOE+_l   |  60   |   3 |  8     |    8     | cspresnet-l |    640     |   30.8  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams) | [config](./ppyoloe_plus_crn_l_60e_objects365.yml) |
+|   PP-YOLOE+_x   |  60   |  4 |   8     |    8     | cspresnet-x |    640     |   32.7  |  98.42 |  206.59      |   45.0        |  95.2  | [model](https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_x_obj365_pretrained.pdparams) | [config](./ppyoloe_plus_crn_x_60e_objects365.yml) |
+
+
+**注意:**
+- 多机训练细节见[文档](../../../docs/tutorials/DistributedTraining_cn.md)
+- Objects365数据集下载请参考[objects365官网](http://www.objects365.org/overview.html)。具体种类列表可下载由PaddleDetection团队整理的[objects365_detection_label_list.txt](https://bj.bcebos.com/v1/paddledet/data/objects365/objects365_detection_label_list.txt)并存放在`dataset/objects365/`，每一行即表示第几个种类。inference或导出模型时需要读取到种类数，如果没有标注json文件时，可以进行如下更改`configs/datasets/objects365_detection.yml`：
+```
+TestDataset:
+  !ImageFolder
+    # anno_path: annotations/zhiyuan_objv2_val.json
+    anno_path: objects365_detection_label_list.txt
+    dataset_dir: dataset/objects365/
+```
--- a/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_l_60e_objects365.yml
+++ b/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_l_60e_objects365.yml
@@ -0,0 +1,21 @@
+_BASE_: [
+  '../../datasets/objects365_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_60e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_l_60e_objects365/model_final
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_l_pretrained.pdparams
+
+CSPResNet:
+  use_alpha: False
+
+PPYOLOEHead:
+  static_assigner_epoch: 20
+
+depth_mult: 1.0
+width_mult: 1.0
--- a/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_m_60e_objects365.yml
+++ b/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_m_60e_objects365.yml
@@ -0,0 +1,21 @@
+_BASE_: [
+  '../../datasets/objects365_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_60e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_60e_objects365/model_final
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_m_pretrained.pdparams
+
+CSPResNet:
+  use_alpha: False
+
+PPYOLOEHead:
+  static_assigner_epoch: 20
+
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_s_60e_objects365.yml
+++ b/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_s_60e_objects365.yml
@@ -0,0 +1,21 @@
+_BASE_: [
+  '../../datasets/objects365_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_60e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_s_60e_objects365/model_final
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_s_pretrained.pdparams
+
+CSPResNet:
+  use_alpha: False
+
+PPYOLOEHead:
+  static_assigner_epoch: 20
+
+depth_mult: 0.33
+width_mult: 0.50
--- a/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_x_60e_objects365.yml
+++ b/paddle_detection/configs/ppyoloe/objects365/ppyoloe_plus_crn_x_60e_objects365.yml
@@ -0,0 +1,21 @@
+_BASE_: [
+  '../../datasets/objects365_detection.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_60e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_x_60e_objects365/model_final
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_x_pretrained.pdparams
+
+CSPResNet:
+  use_alpha: False
+
+PPYOLOEHead:
+  static_assigner_epoch: 20
+
+depth_mult: 1.33
+width_mult: 1.25
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_l_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_l_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_crn.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_crn_l_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_l_pretrained.pdparams
+depth_mult: 1.0
+width_mult: 1.0
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_l_36e_coco_xpu.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_l_36e_coco_xpu.yml
@@ -0,0 +1,71 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_36e_xpu.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+# note: these are default values (use_gpu = true and use_xpu = false) for CI.
+# set use_gpu = false and use_xpu = true for training.
+use_gpu: true
+use_xpu: false
+
+log_iter: 100
+snapshot_epoch: 1
+weights: output/ppyoloe_crn_l_36e_coco/model_final
+find_unused_parameters: True
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_l_pretrained.pdparams
+depth_mult: 1.0
+width_mult: 1.0
+
+TrainReader:
+  batch_size: 8
+
+architecture: YOLOv3
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+ema_black_list: ['proj_conv.weight']
+custom_black_list: ['reduce_mean']
+
+YOLOv3:
+  backbone: CSPResNet
+  neck: CustomCSPPAN
+  yolo_head: PPYOLOEHead
+  post_process: ~
+
+CSPResNet:
+  layers: [3, 6, 6, 3]
+  channels: [64, 128, 256, 512, 1024]
+  return_idx: [1, 2, 3]
+  use_large_stem: True
+
+CustomCSPPAN:
+  out_channels: [768, 384, 192]
+  stage_num: 1
+  block_num: 3
+  act: 'swish'
+  spp: true
+
+PPYOLOEHead:
+  fpn_strides: [32, 16, 8]
+  grid_cell_scale: 5.0
+  grid_cell_offset: 0.5
+  static_assigner_epoch: 4
+  use_varifocal_loss: True
+  loss_weight: {class: 1.0, iou: 2.5, dfl: 0.5}
+  static_assigner:
+    name: ATSSAssigner
+    topk: 9
+  assigner:
+    name: TaskAlignedAssigner
+    topk: 13
+    alpha: 1.0
+    beta: 6.0
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_m_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_m_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_crn.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_crn_m_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_m_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_s_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_s_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_crn.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_crn_s_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_s_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.50
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_s_400e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_s_400e_coco.yml
@@ -0,0 +1,18 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_400e.yml',
+  './_base_/ppyoloe_crn.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_crn_s_400e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_s_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.50
+
+PPYOLOEHead:
+  static_assigner_epoch: 133
--- a/paddle_detection/configs/ppyoloe/ppyoloe_crn_x_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_crn_x_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_crn.yml',
+  './_base_/ppyoloe_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_crn_x_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_x_pretrained.pdparams
+depth_mult: 1.33
+width_mult: 1.25
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_l_80e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_80e.yml',
+  './_base_/ppyoloe_plus_crn.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_l_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_l_obj365_pretrained.pdparams
+depth_mult: 1.0
+width_mult: 1.0
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_m_80e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_m_80e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_80e.yml',
+  './_base_/ppyoloe_plus_crn.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_m_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_m_obj365_pretrained.pdparams
+depth_mult: 0.67
+width_mult: 0.75
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_s_80e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_s_80e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_80e.yml',
+  './_base_/ppyoloe_plus_crn.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_s_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_s_obj365_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.50
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_plus_crn_tiny_auxhead.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_plus_crn_t_auxhead_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_t_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.375
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_320_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_320_300e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_plus_crn_tiny_auxhead.yml',
+  './_base_/ppyoloe_plus_reader_320.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_plus_crn_t_auxhead_320_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_t_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.375
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_relu_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_relu_300e_coco.yml
@@ -0,0 +1,26 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_plus_crn_tiny_auxhead.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_plus_crn_t_auxhead_relu_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_t_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.375
+
+
+CSPResNet:
+  act: 'relu'
+
+CustomCSPPAN:
+  act: 'relu'
+
+PPYOLOEHead:
+  act: 'relu'
+  attn_conv: None
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco.yml
@@ -0,0 +1,26 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_300e.yml',
+  './_base_/ppyoloe_plus_crn_tiny_auxhead.yml',
+  './_base_/ppyoloe_plus_reader_320.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 10
+weights: output/ppyoloe_plus_crn_t_auxhead_relu_320_300e_coco/model_final
+
+pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_t_pretrained.pdparams
+depth_mult: 0.33
+width_mult: 0.375
+
+
+CSPResNet:
+  act: 'relu'
+
+CustomCSPPAN:
+  act: 'relu'
+
+PPYOLOEHead:
+  act: 'relu'
+  attn_conv: None
--- a/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_x_80e_coco.yml
+++ b/paddle_detection/configs/ppyoloe/ppyoloe_plus_crn_x_80e_coco.yml
@@ -0,0 +1,15 @@
+_BASE_: [
+  '../datasets/coco_detection.yml',
+  '../runtime.yml',
+  './_base_/optimizer_80e.yml',
+  './_base_/ppyoloe_plus_crn.yml',
+  './_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_x_80e_coco/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/pretrained/ppyoloe_crn_x_obj365_pretrained.pdparams
+depth_mult: 1.33
+width_mult: 1.25
--- a/paddle_detection/configs/ppyoloe/voc/README_cn.md
+++ b/paddle_detection/configs/ppyoloe/voc/README_cn.md
@@ -0,0 +1,9 @@
+# PP-YOLOE
+
+## 模型库
+
+### VOC数据集模型库
+|       模型       | Epoch |   GPU个数   | 每GPU图片个数 |  骨干网络  |   输入尺寸   | Box AP<sup>0.5 | Params(M) | FLOPs(G) | V100 FP32(FPS) | V100 TensorRT FP16(FPS) |  模型下载  |  配置文件 |
+|:---------------:|:-----:|:-----------:|:-----------:|:---------:|:----------:|:--------------:|:---------:|:---------:|:-------------:|:-----------------------:| :-------: |:--------:|
+|   PP-YOLOE+_s   |  30   |     8     |    8     | cspresnet-s |    640     |   86.7  |  7.93  |  17.36   |   208.3   |  333.3   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_s_30e_voc.pdparams) | [config](./ppyoloe_plus_crn_s_30e_voc.yml) |
+|   PP-YOLOE+_l   |  30   |     8     |    8     | cspresnet-l |    640     |   89.0  |  52.20 |  110.07  |   78.1    |  149.2   | [model](https://paddledet.bj.bcebos.com/models/ppyoloe_plus_crn_l_30e_voc.pdparams) | [config](./ppyoloe_plus_crn_l_30e_voc.yml) |
--- a/paddle_detection/configs/ppyoloe/voc/ppyoloe_plus_crn_l_30e_voc.yml
+++ b/paddle_detection/configs/ppyoloe/voc/ppyoloe_plus_crn_l_30e_voc.yml
@@ -0,0 +1,43 @@
+_BASE_: [
+  '../../datasets/voc.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_l_30e_voc/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_l_80e_coco.pdparams
+depth_mult: 1.0
+width_mult: 1.0
+
+
+TrainReader:
+  batch_size: 8 # default 8 gpus, total bs = 64
+
+EvalReader:
+  batch_size: 4
+
+
+epoch: 30
+LearningRate:
+  base_lr: 0.001
+  schedulers:
+    - !CosineDecay
+      max_epochs: 36
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 1
+
+
+PPYOLOEHead:
+  static_assigner_epoch: -1
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7
--- a/paddle_detection/configs/ppyoloe/voc/ppyoloe_plus_crn_s_30e_voc.yml
+++ b/paddle_detection/configs/ppyoloe/voc/ppyoloe_plus_crn_s_30e_voc.yml
@@ -0,0 +1,43 @@
+_BASE_: [
+  '../../datasets/voc.yml',
+  '../../runtime.yml',
+  '../_base_/optimizer_80e.yml',
+  '../_base_/ppyoloe_plus_crn.yml',
+  '../_base_/ppyoloe_plus_reader.yml',
+]
+
+log_iter: 100
+snapshot_epoch: 5
+weights: output/ppyoloe_plus_crn_s_30e_voc/model_final
+
+pretrain_weights: https://bj.bcebos.com/v1/paddledet/models/ppyoloe_plus_crn_s_80e_coco.pdparams
+depth_mult: 0.33
+width_mult: 0.50
+
+
+TrainReader:
+  batch_size: 8 # default 8 gpus, total bs = 64
+
+EvalReader:
+  batch_size: 4
+
+
+epoch: 30
+LearningRate:
+  base_lr: 0.001
+  schedulers:
+    - !CosineDecay
+      max_epochs: 36
+    - !LinearWarmup
+      start_factor: 0.
+      epochs: 1
+
+
+PPYOLOEHead:
+  static_assigner_epoch: -1
+  nms:
+    name: MultiClassNMS
+    nms_top_k: 1000
+    keep_top_k: 300
+    score_threshold: 0.01
+    nms_threshold: 0.7