resnet tiny mnist

Introduction

The ResNet-Tiny-MNIST is a compact ResNet model designed for the MNIST dataset. It achieves a high accuracy of 0.985 on the validation set, making it suitable for image classification tasks involving handwritten digits.

Architecture

This model is a scaled-down version of the ResNet architecture, tailored to efficiently handle the MNIST dataset's requirements. ResNet, or Residual Network, is known for its ability to train very deep networks by using skip connections to mitigate the vanishing gradient problem.

Training

The model was trained using the PyTorch framework and leverages the benefits of the ResNet architecture to achieve high accuracy on MNIST. The training process includes standard techniques used in deep learning to enhance performance, though specific details like epochs, batch size, and optimizer settings are not provided.

Guide: Running Locally

To run the ResNet-Tiny-MNIST model locally, follow these steps:

1. Clone the Repository: Download the code from the repository.

   git clone https://huggingface.co/fxmarty/resnet-tiny-mnist
   

2. Install Dependencies: Ensure you have PyTorch installed. Additional dependencies can be installed via:

   pip install -r requirements.txt
   

3. Run Inference: Use the model to perform inference on MNIST images.

For enhanced performance, consider using cloud GPUs offered by providers like AWS, Google Cloud, or Azure.

License

This project is licensed under the GPL-3.0 license, which allows for freedom to use, modify, and distribute the software, provided that all copies and derivatives are also licensed similarly.