Struct LpLoss 

pub struct LpLoss {
    pub p: f64,
}

Computes the Lp Loss between predictions and targets.

This loss function computes the element-wise p-th power of absolute errors, then reduces them via mean or sum.

Mathematical Definition

For predictions ŷ and targets y, the element-wise loss is:

Lᵢ = |ŷᵢ - yᵢ|ᵖ

With mean reduction (default), the final loss is:

L = (1/n) × Σᵢ |ŷᵢ - yᵢ|ᵖ

Notes

• This implementation computes |error|^p, not the Lp norm (Σ|error|^p)^(1/p).
• The p = 1 case uses an optimized abs() operation.
• The p = 2 case uses an optimized computation error * error instead of powf.

Example

use burn_nn::loss::{LpLossConfig, Reduction};
use burn::tensor::Tensor;

// Create L2 loss
let l2_loss = LpLossConfig::l2();

let predictions: Tensor<Backend, 2> = /* model output */;
let targets: Tensor<Backend, 2> = /* ground truth */;

// Compute loss with mean reduction (MSE)
let mse = l2_loss.forward(predictions.clone(), targets.clone(), Reduction::Mean);

// Compute loss with sum reduction (SSE)
let sse = l2_loss.forward(predictions.clone(), targets.clone(), Reduction::Sum);

// Compute loss with no reduction
let unreduced_l2_loss = l2_loss.forward_no_reduction(predictions, targets);

Fields

p: f64

The order of the norm (e.g., 1 for L1, 2 for L2). Equivalently, the exponent p for computing |error|^p.

Implementations

impl LpLoss

pub fn forward<const D: usize, B>( &self, predictions: Tensor<B, D>, targets: Tensor<B, D>, reduction: Reduction, ) -> Tensor<B, 1>

where B: Backend,

Computes the element-wise loss |error|^p with reduction.

Arguments

• predictions - The model’s predicted values.
• targets - The ground truth target values.
• reduction - Specifies how to reduce the element-wise losses:
  • Reduction::Mean or Reduction::Auto: Returns the mean of all element-wise losses.
  • Reduction::Sum: Returns the sum of all element-wise losses.

Returns

A scalar tensor containing the reduced loss value.

Shapes

• predictions: [...dims] - Any shape
• targets: [...dims] - Must match predictions shape
• output: [1] - Scalar loss value

pub fn forward_no_reduction<const D: usize, B>( &self, predictions: Tensor<B, D>, targets: Tensor<B, D>, ) -> Tensor<B, D>

where B: Backend,

Computes the element-wise loss |error|^p without reduction.

Arguments

• predictions - The model’s predicted values.
• targets - The ground truth target values.

Returns

A tensor of the same shape as the inputs, containing |prediction - target|^p for each element.

Shapes

• predictions: [...dims] - Any shape
• targets: [...dims] - Must match predictions shape
• output: [...dims] - Same shape as inputs

pub fn forward_reduce_dims<const D: usize, B>( &self, predictions: Tensor<B, D>, targets: Tensor<B, D>, dims: &[usize], ) -> Tensor<B, D>

where B: Backend,

Computes the element-wise loss |error|^p with reduction over specified dimensions.

Calculates element-wise |predictions - targets|^p, then takes the mean over the specified dimensions. Useful for per-sample or per-channel losses (e.g., when working with images).

Dimensions can be provided in any order. They are sorted internally and reduced from highest to lowest to ensure indices remain valid.

Arguments

• predictions - The model’s predicted values.
• targets - The ground truth target values.
• dims - Dimensions to reduce over.

Returns

A tensor with the specified dimensions reduced to size 1.

Example

// Image tensor: [batch, C, H, W]
let l2_loss = LpLossConfig::l2();

// Per-image MSE for PSNR: reduce over C, H, W → [batch, 1, 1, 1]
let mse_per_image = l2_loss.forward_reduce_dims(predictions, targets, &[1, 2, 3]);

Trait Implementations

impl<B> AutodiffModule<B> for LpLoss

where B: AutodiffBackend,

type InnerModule = LpLoss

Inner module without auto-differentiation.

fn valid(&self) -> <LpLoss as AutodiffModule<B>>::InnerModule

Returns the same module, but on the inner backend without auto-differentiation.

fn from_inner(module: <LpLoss as AutodiffModule<B>>::InnerModule) -> LpLoss

Wraps an inner module back into an auto-diff module.

impl Clone for LpLoss

fn clone(&self) -> LpLoss

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for LpLoss

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Display for LpLoss

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl<B> Module<B> for LpLoss

where B: Backend,

type Record = EmptyRecord

Type to save and load the module.

fn visit<V>(&self, _visitor: &mut V)

where V: ModuleVisitor<B>,

Visit each tensor parameter in the module with a visitor.

fn map<M>(self, _mapper: &mut M) -> LpLoss

where M: ModuleMapper<B>,

Map each tensor parameter in the module with a mapper.

fn load_record(self, _record: <LpLoss as Module<B>>::Record) -> LpLoss

Load the module state from a record.

fn into_record(self) -> <LpLoss as Module<B>>::Record

Convert the module into a record containing the state.

fn to_device(self, _: &<B as BackendTypes>::Device) -> LpLoss

Move the module and all of its sub-modules to the given device.

fn fork(self, _: &<B as BackendTypes>::Device) -> LpLoss

Fork the module and all of its sub-modules to the given device.

fn collect_devices( &self, devices: Vec<<B as BackendTypes>::Device>, ) -> Vec<<B as BackendTypes>::Device>

Return all the devices found in the underneath module tree added to the given vector without duplicates.

fn devices(&self) -> Vec<<B as BackendTypes>::Device>

Return all the devices found in the underneath module tree without duplicates.

fn no_grad(self) -> Self

Each tensor in the module tree will not require grad.

fn train<AB>(self) -> Self::TrainModule

where AB: AutodiffBackend<InnerBackend = B>, Self: HasAutodiffModule<AB>,

Move the module and all of its sub-modules to the autodiff backend.

fn num_params(&self) -> usize

Get the number of parameters the module has, including all of its sub-modules.

fn save_file<FR, PB>( self, file_path: PB, recorder: &FR, ) -> Result<(), RecorderError>

where FR: FileRecorder<B>, PB: Into<PathBuf>,

Save the module to a file using the provided file recorder.

fn load_file<FR, PB>( self, file_path: PB, recorder: &FR, device: &<B as BackendTypes>::Device, ) -> Result<Self, RecorderError>

where FR: FileRecorder<B>, PB: Into<PathBuf>,

Load the module from a file using the provided file recorder.

fn quantize_weights(self, quantizer: &mut Quantizer) -> Self

Quantize the weights of the module.

impl ModuleDisplay for LpLoss

fn format(&self, passed_settings: DisplaySettings) -> String

Formats the module with provided display settings.

fn custom_settings(&self) -> Option<DisplaySettings>

Custom display settings for the module.

fn custom_content(&self, _content: Content) -> Option<Content>

Custom attributes for the module.

impl ModuleDisplayDefault for LpLoss

fn content(&self, content: Content) -> Option<Content>

Attributes of the module used for display purposes.

fn num_params(&self) -> usize

Gets the number of the parameters of the module.