图形分类比测模型,包括使用Keras框架的AlexNet模型、使用tensorflow框架的VGG16模型
|
3 months ago | |
---|---|---|
checkpoints | 3 months ago | |
dataset | 3 months ago | |
models | 3 months ago | |
.gitignore | 3 months ago | |
README.md | 3 months ago | |
export_onnx.py | 3 months ago | |
requirements.txt | 3 months ago | |
train_alexnet.py | 3 months ago | |
train_vgg16.py | 3 months ago |
此项目包含AlexNet模型的Keras框架实现和VGG16模型的tensorflow框架实现和与其对应的模型训练文件
classification-models-tensorflow
├── README.md
├── checkpoints # 保存所有的权重信息
├── export_onnx.py # 模型权重转换为onnx脚本
├── models # 模型定义
│ └── AlexNet.py
├── train_alexnet.py # AlexNet模型训练脚本
└── train_vgg16.py # VGG16模型训练脚本
python=3.9
pillow
scipy
numpy==1.24.4
tensorflow==2.10.0
shell
conda create -n tensorflow python=3.9
conda activate tensorflow
shell
pip install -r requirements.txt -i https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
shell
python train_alexnet.py --data-path dataset/imagenette2-320 --output-dir checkpoints/alexnet --batch-size 64 --epochs 90
shell
python train_vgg16.py --data-path dataset/imagenette2-320 --output-dir checkpoints/vgg16 --batch-size 64 --epochs 90 --lr 0.01 --opt sgd
python export_onnx.py --model_dir checkpoints/alexnet
python export_onnx.py --model_dir checkpoints/vgg16