Portal de Periódicos da CAPES

TorchGAN: A Flexible Framework for GAN Training and Evaluation

2021; Open Journals; Volume: 6; Issue: 66 Linguagem: Inglês

10.21105/joss.02606

2475-9066

Avik Pal, Aniket Das,

Model Reduction and Neural Networks

TorchGAN is a PyTorch based framework for writing succinct and comprehensible code for training and evaluation of Generative Adversarial Networks. The framework's modular design allows effortless customization of the model architecture, loss functions, training paradigms, and evaluation metrics. The key features of TorchGAN are its extensibility, built-in support for a large number of popular models, losses and evaluation metrics, and zero overhead compared to vanilla PyTorch. By using the framework to implement several popular GAN models, we demonstrate its extensibility and ease of use. We also benchmark the training time of our framework for said models against the corresponding baseline PyTorch implementations and observe that TorchGAN's features bear almost zero overhead.