Welcome to the world of Burn, a comprehensive deep learning framework designed with performance, flexibility, and efficiency in mind. In this guide, we’ll navigate through the fundamental components that make Burn stand out, while ensuring you get started smoothly!

Is Burn Right for You?

Burn aims to provide a robust and efficient platform for deep learning, focusing specifically on:

• Dynamic deep learning frameworks.
• High-performance capabilities using Rust.
• Seamless integration and deployments across various hardware.

Let’s dive in and see how you can use Burn effectively!

Setting Up Burn

To get started, you should have Rust installed on your machine. If you haven’t installed Rust yet, you can do so by visiting the Rust installation page.

Understanding the Basics of Burn

The primary components of Burn encompass various elements such as tensors, modules, optimizers, and an overarching philosophy to facilitate efficient deep learning. For a clearer understanding, envision Burn as an artist’s toolkit:

• Tensors are your brushes—the essential tools that allow you to manipulate data.
• Modules represent your palette, where you mix and blend colors (or layers) to create masterpieces (models).
• Optimizers serve as your artistic vision, guiding your technique to perfect your work over time.

Code Snippet: Creating a Basic Module

Here’s a simple code snippet demonstrating how to create a simple neural network module using Burn’s syntax:

use burn::nn;
use burn::module::Module;
use burn::tensor::backend::Backend;

#[derive(Module, Debug)]
pub struct PositionWiseFeedForward {
    linear_inner: nn::Linear,
    linear_outer: nn::Linear,
    dropout: nn::Dropout,
    gelu: nn::Gelu,
}

impl PositionWiseFeedForward {
    pub fn forward(self, input: Tensor) -> Tensor {
        let x = self.linear_inner.forward(input);
        let x = self.gelu.forward(x);
        let x = self.dropout.forward(x);
        self.linear_outer.forward(x)
    }
}