Enum burn_tensor::Distribution

pub enum Distribution<E> {
    Default,
    Bernoulli(f64),
    Uniform(E, E),
    Normal(f64, f64),
}

Distribution for random value of a tensor.

Variants

Default

Uniform distribution from 0 (inclusive) to 1 (exclusive).

Bernoulli(f64)

Bernoulli distribution with the given probability.

Uniform(E, E)

Uniform distribution. The range is inclusive.

Normal(f64, f64)

Normal distribution with the given mean and standard deviation.

Implementations

impl<E> Distribution<E>where Standard: Distribution<E>, E: SampleUniform,

pub fn sampler<R: RngCore>(self, rng: &mut R) -> DistributionSampler<'_, E, R>

Creates a new distribution sampler.

Arguments

• rng - The random number generator.

Returns

The distribution sampler.

impl<E> Distribution<E>where E: Element,

pub fn convert<EOther: Element>(self) -> Distribution<EOther>

Converts the distribution to a different element type.

Returns

The converted distribution.

Trait Implementations

impl<E: Clone> Clone for Distribution<E>

fn clone(&self) -> Distribution<E>

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<E: Copy> Copy for Distribution<E>