A best practice for tensorflow project template architecture.
image.png
Fllder Structure
├── base
│ ├── base_model.py - this file contains the abstract class of the model.
│ └── base_train.py - this file contains the abstract class of the trainer.
│
│
├── model - this folder contains any model of your project.
│ └── example_model.py
│
│
├── trainer - this folder contains trainers of your project.
│ └── example_trainer.py
│
├── mains - here's the main(s) of your project (you may need more than one main).
│ └── example_main.py - here's an example of main that is responsible for the whole pipeline.
│
├── data _loader
│ └── data_generator.py - here's the data_generator that is responsible for all data handling.
│
└── utils
├── logger.py
└── any_other_utils_you_need
网友评论