Skip to content

Commit

Permalink
add version info for examples
Browse files Browse the repository at this point in the history
  • Loading branch information
yuedongli1 committed Oct 9, 2024
1 parent c2f2e80 commit 72c7387
Show file tree
Hide file tree
Showing 2 changed files with 15 additions and 1 deletion.
9 changes: 8 additions & 1 deletion examples/finetune_SHWD/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,13 @@

本文以安全帽佩戴检测数据集(SHWD)为例,介绍自定义数据集在MindYOLO上进行finetune的主要流程。

#### 版本信息
* os: Linux EulerOS 2.0
* ascend driver: 23.0.6.b010
* ascend firmware: 7.1.0.11.220
* cann: 7.5.T7.0.B053
* mindspore: 2.3.1

#### 数据集格式转换

[SHWD数据集](https://github.com/njvisionpower/Safety-Helmet-Wearing-Dataset/tree/master)采用voc格式的数据标注,其文件目录如下所示:
Expand Down Expand Up @@ -122,7 +129,7 @@ optimizer:
```shell
python train.py --config ./examples/finetune_SHWD/yolov7-tiny_shwd.yaml
```
*注意:直接用yolov7-tiny默认参数在SHWD数据集上训练,可取得AP50 87.0的精度。将lr_init参数由0.01改为0.001,即可实现ap50为89.2的精度结果*
*注意:直接用yolov7-tiny默认参数在SHWD数据集上训练,可取得AP50 87.0的精度。将lr_init参数由0.01改为0.001,即可实现ap50为90.5的精度结果*

#### 可视化推理
使用/demo/predict.py即可用训练好的模型进行可视化推理,运行方式如下:
Expand Down
7 changes: 7 additions & 0 deletions examples/finetune_single_class_dataset/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,13 @@

本文以自制巧克力花生豆数据集为例,介绍单类别数据集使用MindYOLO进行训练的主要流程。

#### 版本信息
* os: Linux EulerOS 2.0
* ascend driver: 23.0.6.b010
* ascend firmware: 7.1.0.11.220
* cann: 7.5.T7.0.B053
* mindspore: 2.3.1

#### 数据集格式转换

巧克力花生豆数据集采用voc格式的数据标注,其文件目录如下所示:
Expand Down

0 comments on commit 72c7387

Please sign in to comment.