PyNeuralNet 1.2.0

PyNeuralNet is a python library for prototyping and building neural networks. PyNeuralNet uses PyTorch as a computational backend for deep learning models.

PyNeuralNet

Welcome to the PyNeuralNet repository! PyNeuralNet is a python library for prototyping and building neural networks. PyNeuralNet uses PyTorch as a computational backend for deep learning models.

Installation

1. First, make sure you have Python installed on your system.
2. Use this pip command to install the latest version of package.

   pip install pyneuralnet
   

Usage

from pyneuralnet import train

dataset_loader = 'local'
root_dir = 'path/to/root_diractory'
metadata_file = 'path/to/meta_info_file.txt'
network = 'usrcnn'
batchs = 4

train(datasetloader, metadata_file, root_dir, epochs=25, batch_size=batchs, network=network)

• Parameters
  • dataset_loader: Type of dataset loader, there is two type of dataloaders (locally - local and from internet - internet). In this example, it is set to 'local'.
  • metadata_file: Path to the metadata file. If you load your dataset from internet you should type an url like this.
  • root_dir: Path to the root directory where the dataset is located. If you load your dataset from internet you should type an url like this, example image.
  • network: Neural network architecture to be used (e.g., 'usrcnn'). There are 6 type of networks(for now) which are based on Convolutional Neural Network e.g., usrcnn, esrcnn, bsrcnn, isrcnn, rsrcnn and srcnn.
  • epochs: Number of training epochs.
  • batch_size: Size of each training batch.

Contributing

Contributions are welcome! If you'd like to contribute to this project, follow these steps:

1. Fork this repository.
2. Create a new branch for your feature or bug fix.
3. Make your changes and submit a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

Feel free to reach out to me at [e-mail] if you have any questions or feedback! Or just open an issue on PyNeuralNet's github page.

Happy coding! 🚀