https://github.com/deepvac/mobilenetv3

DeepVAC-compliant MobileNetV3 implementation
https://github.com/deepvac/mobilenetv3

Last synced: 4 days ago
JSON representation

DeepVAC-compliant MobileNetV3 implementation

• Host: GitHub
• URL: https://github.com/deepvac/mobilenetv3
• Owner: DeepVAC
• License: gpl-3.0
• Created: 2021-01-28T08:47:40.000Z (almost 4 years ago)
• Default Branch: main
• Last Pushed: 2021-06-23T10:24:20.000Z (over 3 years ago)
• Last Synced: 2023-03-05T16:49:50.332Z (over 1 year ago)
• Language: Python
• Size: 46.9 KB
• Stars: 2
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# MobileNetV3

DeepVAC-compliant MobileNetV3 implementation

# 简介

本项目实现了符合DeepVAC规范的MobileNet V3

**项目依赖**

- deepvac >= 0.5.6

- pytorch >= 1.8.0

- opencv-python

- numpy

# 如何运行本项目

## 1. 阅读[DeepVAC规范](https://github.com/DeepVAC/deepvac)

可以粗略阅读，建立起第一印象。

## 2. 准备运行环境

使用Deepvac规范指定[Docker镜像](https://github.com/DeepVAC/deepvac#2-%E7%8E%AF%E5%A2%83%E5%87%86%E5%A4%87)

## 3. 准备数据集

自行准备。

## 4. 修改配置文件

修改config.py文件。主要修改内容：

- 指定训练集、验证集、测试集的图片目录前缀、对应的标注txt文件和分隔符

```python

# line 43-45

fileline_path = 'data/train_cls.txt'

delimiter = ' ' 

sample_path_prefix = 

# line 69

fileline_path = 'data/val_cls.txt'

# line 82

fileline_path = 'data/test_cls.txt'

```

- 修改分类数

```

config.core.MobileNetv3Train.cls_num = 4

```

## 5. 训练

执行命令：

```bash

python3 train.py

```

## 6. 测试

指定要测试模型和测试数据的路径，在config.py指定待测模型和数据路径：

```python

fileline_path = 'data/test_cls.txt'

config.core.MobileNetv3Test.model_path = "your test model dir / pretrained weights"

```

然后运行测试脚本：

```bash

python3 test.py

```

## 7， 更多功能

如果要在本项目中开启如下功能：

- 预训练模型加载

- checkpoint加载

- 使用tensorboard

- 启用TorchScript

- 转换ONNX

- 转换NCNN

- 转换CoreML

- 开启量化

- 开启自动混合精度训练

请参考[DeepVAC](https://github.com/DeepVAC/deepvac)。