This library strives to serve as a comprehensive deep learning framework, offering exceptional flexibility and written in Rust. Our objective is to cater to both researchers and practitioners by simplifying the process of experimenting, training, and deploying models.
- Customizable and user-friendly neural network module 🔥
- Comprehensive training capabilities, including
metrics
,logging
, andcheckpointing
📈 - Versatile Tensor crate with pluggable backends 🔧
- Dataset crate with a variety of utilities and sources 📚
- Import crate for seamless integration of pretrained models 📦
The best way to get started with burn
is to clone the repo and play with the
examples. This may also be a good idea to take a look the main
components of burn
to get a quick overview of the fundamental building blocks. If
you're interested in how the framework works, you can read our
architecture document.
- MNIST train a model on CPU/GPU using different backends.
- MNIST Inference Web run trained model in the browser for inference.
- Text Classification train a transformer encoder from scratch on GPU.
- Text Generation train an autoregressive transformer from scratch on GPU.
Understanding the key components and philosophy of burn
can greatly help when beginning to work
with the framework.
Nearly everything in burn
is based on the Backend
trait, which enables you to run tensor
operations using different implementations without having to modify your code. While a backend may
not necessarily have autodiff capabilities, the ADBackend
trait specifies when autodiff is needed.
This trait not only abstracts operations but also tensor, device and element types, providing each
backend the flexibility they need. It's worth noting that the trait assumes eager mode since burn
fully supports dynamic graphs. However, we may create another API to assist with integrating
graph-based backends, without requiring any changes to the user's code.
At the core of burn lies the Tensor
struct, which encompasses multiple types of tensors, including
Float
, Int
, and Bool
. The element types of these tensors are specified by the backend and are
usually designated as a generic argument (e.g., NdArrayBackend<f32>
). Although the same struct is
used for all tensors, the available methods differ depending on the tensor kind. You can specify the
desired tensor kind by setting the third generic argument, which defaults to Float
. The first
generic argument specifies the backend, while the second specifies the number of dimensions.
use burn::tensor::backend::Backend;
use burn::tensor::{Tensor, Int};
fn function<B: Backend>(tensor_float: Tensor<B, 2>) {
let _tensor_bool = tensor_float.clone().equal_elem(2.0); // Tensor<B, 2, Bool>
let _tensor_int = tensor_float.argmax(1) // Tensor<B, 2, Int>
}
As demonstrated in the previous example, nearly all operations require owned tensors as parameters,
which means that calling Clone
explicitly is necessary when reusing the same tensor multiple
times. However, there's no need to worry since the tensor's data won't be copied, it will be flagged
as readonly when multiple tensors use the same allocated memory. This enables backends to reuse
tensor data when possible, similar to a copy-on-write pattern, while remaining completely
transparent to the user.
The 'Backend' trait is highly flexible, enabling backpropagation to be implemented using a simple backend decorator, which makes any backend differentiable.
use burn::tensor::backend::{ADBackend, Backend};
use burn::tensor::{Distribution, Tensor};
use burn_autodiff::ADBackendDecorator;
use burn_ndarray::NdArrayBackend;
fn linear<B: Backend>(x: Tensor<B, 2>, weight: Tensor<B, 2>, bias: Tensor<B, 2>) -> Tensor<B, 2> {
x.matmul(weight) + bias
}
fn main() {
type Backend = NdArrayBackend<f32>;
let weight = Tensor::random([3, 3], Distribution::Standard);
let bias = Tensor::zeros([1, 3]);
let x = Tensor::random([3, 3], Distribution::Standard);
let y = linear::<Backend>(x.clone(), weight.clone(), bias.clone());
// y.backward() // Method backward doesn't exist
let y = linear::<ADBackendDecorator<Backend>>(
Tensor::from_inner(x),
Tensor::from_inner(weight).require_grad(),
Tensor::from_inner(bias).require_grad(),
);
let grads = y.backward(); // Method exists
}
The Module
derive allows you to create your own neural network modules, similar to PyTorch. The
derive function only generates the necessary methods to essentially act as a parameter container for
your type, it makes no assumptions about how the forward pass is declared.
use burn::nn;
use burn::module::Module;
use burn::tensor::backend::Backend;
#[derive(Module, Debug)]
pub struct PositionWiseFeedForward<B: Backend> {
linear_inner: Linear<B>,
linear_outer: Linear<B>,
dropout: Dropout,
gelu: GELU,
}
impl<B: Backend> PositionWiseFeedForward<B> {
pub fn forward<const D: usize>(&self, input: Tensor<B, D>) -> Tensor<B, D> {
let x = self.linear_inner.forward(input);
let x = self.gelu.forward(x);
let x = self.dropout.forward(x);
self.linear_outer.forward(x)
}
}
Note that all fields declared in the struct must also implement the Module
trait. The Tensor
struct doesn't implement Module
, but Param<Tensor<B, D>>
does.
The Config
derive lets you define serializable and deserializable configurations or
hyper-parameters for your modules or any components.
use burn::config::Config;
#[derive(Config)]
pub struct PositionWiseFeedForwardConfig {
pub d_model: usize,
pub d_ff: usize,
#[config(default = 0.1)]
pub dropout: f64,
}
The derive also adds useful methods to your config, similar to a builder pattern.
fn main() {
let config = PositionWiseFeedForwardConfig::new(512, 2048);
println!("{}", config.d_model); // 512
println!("{}", config.d_ff); // 2048
println!("{}", config.dropout); // 0.1
let config = config.with_dropout(0.2);
println!("{}", config.dropout); // 0.2
}
The Learner
is the main struct
that let you train a neural network with support for logging
,
metric
, checkpointing
and more. In order to create a learner, you must use the LearnerBuilder
.
use burn::train::LearnerBuilder;
use burn::train::metric::{AccuracyMetric, LossMetric};
use burn::record::DefaultRecordSettings;
fn main() {
let dataloader_train = ...;
let dataloader_valid = ...;
let model = ...;
let optim = ...;
let learner = LearnerBuilder::new("/tmp/artifact_dir")
.metric_train_plot(AccuracyMetric::new())
.metric_valid_plot(AccuracyMetric::new())
.metric_train(LossMetric::new())
.metric_valid(LossMetric::new())
.with_file_checkpointer::<DefaultRecordSettings>(2)
.num_epochs(10)
.build(model, optim);
let _model_trained = learner.fit(dataloader_train, dataloader_valid);
}
See this example for a real usage.
Burn supports no_std
with alloc
for the inference mode with the NDArray backend. Simply disable
the default features of the burn
and burn-ndarray
crates (minimum required to run the inference
mode). See the
burn-no-std-tests example as
a reference implementation.
Additionally burn-core
and burn-tensor
crates support no_std
with alloc
if needed to
direclty include them as dependencies (the burn
crates reexports burn-core
and burn-tensor
).
Note, under the no_std
mode, a random seed is generated during the build time if the seed is not
initialized by Backend::seed
method. Additionally,
spin::mutex::Mutex is used in place of
std::sync::Mutex under the no_std
mode.
Before contributing, please take a moment to review our code of conduct. It's also highly recommended to read our architecture document, which explains our architectural decisions. Please see more details in our contributing guide.
Burn is currently in active development, and there will be breaking changes. While any resulting issues are likely to be easy to fix, there are no guarantees at this stage.
You can sponsor the founder of Burn from his GitHub Sponsors profile. The Burn-rs organization doesn't yet have a fiscal entity, but other sponsor methods might become available as the project grows.
Thanks to all current sponsors 🙏.
Burn is distributed under the terms of both the MIT license and the Apache License (Version 2.0). See LICENSE-APACHE and LICENSE-MIT for details. Opening a pull request is assumed to signal agreement with these licensing terms.