This repository contains an implementation of CoatNet using PyTorch Lightning, allowing for efficient and scalable training.
CoatNet integrates the strengths of convolutional neural networks (CNNs) and Transformers for image classification tasks. For more details, refer to the original paper: CoatNet: Marrying Convolution and Attention for All Data Sizes.
- CoatNet model implementation
- Training routine using PyTorch Lightning
- Support for Distributed Data Parallel (DDP) training
To get started with training CoatNet on your dataset:
- Clone this repository.
- Install the required dependencies.
- Run the training script with your dataset.
- Python 3.x
- PyTorch
- PyTorch Lightning
For a detailed list of requirements, see requirements.txt.