PyTorch Mobile Edge Inference

Project Overview

This project provides an efficient framework for deploying PyTorch models on mobile devices and edge computing systems. The aim is to bring the capabilities of deep learning to the edge.

Features

• Lightweight model deployment
• Support for various mobile platforms
• Efficient memory and compute resources handling

Quick Start

1. Clone the repository:

   git clone https://github.com/sttadic/pytorch-mobile-edge-inference
   cd pytorch-mobile-edge-inference

2. Install the required dependencies:

   pip install -r requirements.txt

Basic Usage

• Import the necessary libraries and load your model in your application.
• Use the provided APIs to run inference on input data.

Model Export Steps

1. Train your model using PyTorch.
2. Use torch.jit.script to export your model to TorchScript for performance optimization.
3. Convert the scripted model for mobile compatibility following the guidelines provided in the documentation.

License

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