fix ppyoloe distill and docs (#7875)

* fix ppyoloe_distill and docs * fix docs, test=document_fix

fix ppyoloe distill and docs (#7875)
* fix ppyoloe_distill and docs * fix docs, test=document_fix
237b19d9 · Feng Ni · GitHub · d48a4bb9 · 237b19d9 · 237b19d9
6 changed file
--- a/configs/ppyoloe/README.md
+++ b/configs/ppyoloe/README.md
@@ -40,9 +40,13 @@ PP-YOLOE is composed of following methods:
 |:--------------:|:-----:|:-------:|:----------:|:----------:| :-------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------:| :---------------------: |:------------------------------------------------------------------------------------:|:-------------------------------------------:|
 |  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

@@ -57,10 +61,15 @@ PP-YOLOE is composed of following methods:
 ### 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+_l        |  80   |      52.9       |    70.1    |     57.9     |     35.2     |     57.5      |     69.1     |     56.0     |     77.9      |     86.9     |
-|       PP-YOLOE+_x        |  80   |      54.7       |    72.0    |     59.9     |     37.9     |     59.3      |     70.4     |     57.0     |     78.7      |     87.2     |
+|       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
@@ -92,6 +101,14 @@ PP-YOLOE is composed of following methods:

 **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

--- a/configs/ppyoloe/README_cn.md
+++ b/configs/ppyoloe/README_cn.md
@@ -40,9 +40,14 @@ PP-YOLOE由以下方法组成
 |:---------------:|:-----:|:---------:|:--------:|:----------:|:----------:|:--------------------------:|:---------------------------:|:---------:|:--------:|:---------------:| :---------------------: |:------------------------------------------------------------------------------------:|:-------------------------------------------:|
 |   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 |
@@ -58,9 +63,14 @@ PP-YOLOE由以下方法组成
 |:------------------------:|:-----:|:---------------:|:----------:|:-----------:|:------------:|:-------------:|:------------:|:------------:|:-------------:|:------------:|
 |       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) |
@@ -91,6 +101,14 @@ PP-YOLOE由以下方法组成

 **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

--- a/configs/ppyoloe/objects365/README_cn.md
+++ b/configs/ppyoloe/objects365/README_cn.md
@@ -13,3 +13,11 @@

 **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/configs/slim/README.md
+++ b/configs/slim/README.md
@@ -48,6 +48,7 @@ python tools/train.py -c configs/{MODEL.yml} --slim_config configs/slim/{SLIM_CO

 - `-c`: 指定模型配置文件。
 - `--slim_config`: 指定压缩策略配置文件。
+- 如果选择使用蒸馏，具体蒸馏方法和更多检测模型的蒸馏，请参考[蒸馏策略文档](distill/README.md)


 ### 评估
@@ -169,7 +170,7 @@ python3.7 tools/post_quant.py -c configs/ppyolo/ppyolo_mbv3_large_coco.yml --sli
 | YOLOv3-MobileNetV1 | baseline     | 608      |    29.4     | [下载链接](https://paddledet.bj.bcebos.com/models/yolov3_mobilenet_v1_270e_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/yolov3/yolov3_mobilenet_v1_270e_coco.yml) |                              -                               |
 | YOLOv3-MobileNetV1 | 蒸馏 | 608      | 31.0(+1.6) | [下载链接](https://paddledet.bj.bcebos.com/models/slim/yolov3_mobilenet_v1_coco_distill.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/yolov3/yolov3_mobilenet_v1_270e_coco.yml) | [slim配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/slim/distill/yolov3_mobilenet_v1_coco_distill.yml) |

- 具体蒸馏方法请参考[蒸馏策略文档](distill/README.md)
+- 具体蒸馏方法和更多检测模型的蒸馏，请参考[蒸馏策略文档](distill/README.md)

 ### 蒸馏剪裁联合策略


--- a/configs/slim/README_en.md
+++ b/configs/slim/README_en.md
@@ -46,6 +46,7 @@ python tools/train.py -c configs/{MODEL.yml} --slim_config configs/slim/{SLIM_CO

 - `-c`: Specify the model configuration file.
 - `--slim_config`: Specify the compression policy profile.
+- If you want to use distillation, please refer to [Distillation Doc](distill/README.md) for specific distillation methods and more distillation of detection models.


 ### Evaluation
@@ -156,7 +157,7 @@ Description:
 | YOLOv3-MobileNetV1 | baseline             | 608        |    29.4    |     [link](https://paddledet.bj.bcebos.com/models/yolov3_mobilenet_v1_270e_coco.pdparams)     | [Configuration File ](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/yolov3/yolov3_mobilenet_v1_270e_coco.yml) |                                                                         -                                                                         |
 | YOLOv3-MobileNetV1 | Distillation         | 608        | 31.0(+1.6) | [link](https://paddledet.bj.bcebos.com/models/slim/yolov3_mobilenet_v1_coco_distill.pdparams) | [Configuration File ](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/yolov3/yolov3_mobilenet_v1_270e_coco.yml) | [slimConfiguration File ](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/slim/distill/yolov3_mobilenet_v1_coco_distill.yml) |

- Please refer to the specific distillation method[Distillation Policy Document](distill/README.md)
+- For the specific distillation method and more distillation detection models, please refer to [distill](distill/README.md).

 ### Distillation Pruning Combined Strategy


--- a/ppdet/slim/distill_model.py
+++ b/ppdet/slim/distill_model.py
@@ -332,12 +332,11 @@ class PPYOLOEDistillModel(DistillModel):
            with paddle.no_grad():
                teacher_loss = self.teacher_model(inputs)
            if hasattr(self.teacher_model.yolo_head, "assigned_labels"):
-                self.student_model.yolo_head.assigned_labels, self.student_model.yolo_head.assigned_bboxes, self.student_model.yolo_head.assigned_scores, self.student_model.yolo_head.mask_positive = \
-                    self.teacher_model.yolo_head.assigned_labels, self.teacher_model.yolo_head.assigned_bboxes, self.teacher_model.yolo_head.assigned_scores, self.teacher_model.yolo_head.mask_positive
+                self.student_model.yolo_head.assigned_labels, self.student_model.yolo_head.assigned_bboxes, self.student_model.yolo_head.assigned_scores = \
+                    self.teacher_model.yolo_head.assigned_labels, self.teacher_model.yolo_head.assigned_bboxes, self.teacher_model.yolo_head.assigned_scores
                delattr(self.teacher_model.yolo_head, "assigned_labels")
                delattr(self.teacher_model.yolo_head, "assigned_bboxes")
                delattr(self.teacher_model.yolo_head, "assigned_scores")
-                delattr(self.teacher_model.yolo_head, "mask_positive")
            student_loss = self.student_model(inputs)

            logits_loss, feat_loss = self.distill_loss(self.teacher_model,