randn

randn_like(x: &Tensor<T>) -> Result<Tensor<T>, TensorError>

same as randn but the shape will be based on x.

Parameters:

x: input Tensor

Returns:

Tensor with type T

Examples:

use hpt::{error::TensorError, ops::Random, Tensor};

fn main() -> Result<(), TensorError> {
    let a = Tensor::<f32>::randn([10, 10])?;
    println!("{}", a.randn_like()?);
    Ok(())
}

Backend Support

Backend	Supported
CPU	✅
Cuda	✅