Trait dfdx::nn::LoadFromNpz

pub trait LoadFromNpz<E: Dtype + NumpyDtype, D: Device<E>>: TensorCollection<E, D> {
    // Provided methods
    fn load<P: AsRef<Path>>(&mut self, path: P) -> Result<(), NpzError> { ... }
    fn read<R>(&mut self, r: &mut ZipArchive<R>) -> Result<(), NpzError>
       where R: Read + Seek { ... }
}

Something that can be loaded from a .npz file (which is a zip file).

All super::Modules in nn implement LoadFromNpz, and the zips are formatted in a .npz fashion.

Provided Methods

fn load<P: AsRef<Path>>(&mut self, path: P) -> Result<(), NpzError>

Loads data from a .npz zip archive at the specified path.

Example:

let mut model: (Linear<5, 10>, Linear<10, 5>) = Default::default();
model.load("tst.npz")?;

fn read<R>(&mut self, r: &mut ZipArchive<R>) -> Result<(), NpzError>where R: Read + Seek,

Reads this object from a ZipArchive. r with a base filename of filename_prefix.

Example:

let mut model: Linear<5, 10> = Default::default();
let mut zip = ZipArchive::new(...);
model.read("0", &mut zip)?;

Will try to read data from the following files:

• 0.weight.npy
• 0.bias.npy

Implementors

impl<E: Dtype + NumpyDtype, D: Device<E>, T: TensorCollection<E, D>> LoadFromNpz<E, D> for T