nn-trainer is a library designed to provide an easy-to-use interface for rapid prototyping and training of neural networks. This tool aims to simplify the process of managing data, augmenting data, and implementing training pipelines using TensorFlow. The primary goal is to accelerate the workflow for machine learning practitioners by abstracting away common boilerplate code.

• Data handling and augmentation
• Serialization and deserialization of datasets
• Easy-to-use interface for creating TensorFlow datasets
• Predefined models and utilities for common neural network tasks

To install the nn-trainer library, you need to have Python version >=3.9 and <3.12. You can install the library using Poetry, a dependency manager for Python.

1. Clone the repository:

   git clone https://github.com/schneiderlo/nn-trainer.git
   cd nn-trainer

2. Install dependencies using Poetry:

   poetry install

Below is a basic example of how to use the nn-trainer library to prepare data, augment it, and use it for training a neural network.

-- To be completed --

To set up the project for development, ensure you have the necessary dependencies installed. The development dependencies include tools for linting, testing, and packaging.

Install development dependencies:

poetry install --with dev

Tests are written using pytest. To run the tests, execute the following command:

poetry run pytest

To run a specific test file, use the following command:

poetry run pytest -k test_name

Ensure the code adheres to the defined style guidelines by running linters:

poetry run pylint nn_trainer
poetry run flake8 nn_trainer