Struct PoissonNllLossConfig 

pub struct PoissonNllLossConfig {
    pub log_input: bool,
    pub full: bool,
    pub eps: f64,
}

Configuration for creating a PoissonNllLoss instance.

This configuration allows customization of the Poisson Negative Log Likelihood (NLL) loss behavior, such as whether the input is in log-space, whether to include the Stirling approximation term, and a small epsilon value to avoid numerical instability.

Fields

log_input: bool

If true, the predictions are expected to be in log-space.

When log_input is true, the loss is computed as:

L(predictions, target) = exp(predictions) - target * predictions

When log_input is false, the loss is computed as:

L(predictions, target) = predictions - target * log(predictions + eps)

full: bool

Whether to compute the full loss, including the Stirling approximation term.

When full is true, the Stirling approximation term is added to the loss:

target * log(target) - target + 0.5 * log(2 * PI * target)

eps: f64

A small value to avoid evaluation of log(0) when log_input is false.

This epsilon value is added to the predictions to ensure numerical stability when computing the logarithm.

Implementations

impl PoissonNllLossConfig

pub fn new() -> PoissonNllLossConfig

Create a new instance of the config.

Arguments

Default Arguments

log_input

If true, the predictions are expected to be in log-space.

When log_input is true, the loss is computed as:

L(predictions, target) = exp(predictions) - target * predictions

When log_input is false, the loss is computed as:

L(predictions, target) = predictions - target * log(predictions + eps)

• Defaults to true

full

Whether to compute the full loss, including the Stirling approximation term.

When full is true, the Stirling approximation term is added to the loss:

target * log(target) - target + 0.5 * log(2 * PI * target)

• Defaults to false

eps

A small value to avoid evaluation of log(0) when log_input is false.

This epsilon value is added to the predictions to ensure numerical stability when computing the logarithm.

• Defaults to 1e-8

impl PoissonNllLossConfig

pub fn with_log_input(self, log_input: bool) -> PoissonNllLossConfig

Sets the value for the field log_input.

If true, the predictions are expected to be in log-space.

When log_input is true, the loss is computed as:

L(predictions, target) = exp(predictions) - target * predictions

When log_input is false, the loss is computed as:

L(predictions, target) = predictions - target * log(predictions + eps)

• Defaults to true

pub fn with_full(self, full: bool) -> PoissonNllLossConfig

Sets the value for the field full.

Whether to compute the full loss, including the Stirling approximation term.

When full is true, the Stirling approximation term is added to the loss:

target * log(target) - target + 0.5 * log(2 * PI * target)

• Defaults to false

pub fn with_eps(self, eps: f64) -> PoissonNllLossConfig

Sets the value for the field eps.

A small value to avoid evaluation of log(0) when log_input is false.

This epsilon value is added to the predictions to ensure numerical stability when computing the logarithm.

• Defaults to 1e-8

impl PoissonNllLossConfig

pub fn init(&self) -> PoissonNllLoss

Initializes a PoissonNllLoss instance with the current configuration.

Panics

• Panics if eps is not a positive number.

Trait Implementations

impl Clone for PoissonNllLossConfig

fn clone(&self) -> PoissonNllLossConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Config for PoissonNllLossConfig

fn save<P>(&self, file: P) -> Result<(), Error>

where P: AsRef<Path>,

Saves the configuration to a file.

fn load<P>(file: P) -> Result<Self, ConfigError>

where P: AsRef<Path>,

Loads the configuration from a file.

fn load_binary(data: &[u8]) -> Result<Self, ConfigError>

Loads the configuration from a binary buffer.

impl Debug for PoissonNllLossConfig

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for PoissonNllLossConfig

fn deserialize<D>( deserializer: D, ) -> Result<PoissonNllLossConfig, <D as Deserializer<'de>>::Error>

where D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Display for PoissonNllLossConfig

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

Formats the value using the given formatter.

impl Serialize for PoissonNllLossConfig

fn serialize<S>( &self, serializer: S, ) -> Result<<S as Serializer>::Ok, <S as Serializer>::Error>

where S: Serializer,

Serialize this value into the given Serde serializer.