Struct Placed 

pub struct Placed {
    pub name: String,
    pub ty: GgmlType,
    pub dims: Vec<u64>,
    pub at: u64,
    pub bytes: u64,
    pub taps: Option<usize>,
    pub streamed: bool,
}

A tensor of the weight image: its format, its dims, where its bytes start and how many there are.

Fields

name: String

ty: GgmlType

dims: Vec<u64>

at: u64

bytes: u64

taps: Option<usize>

For a convolution’s kernel, its taps: the image holds it tap by tap, each tap’s every channel together — the transpose of the checkpoint’s order, each channel’s taps together — so a tap is one contiguous read.

streamed: bool

A matrix of weight blocks a matvec job streams through the array (a float matrix runs on the vector unit).

Implementations

impl Placed

pub fn row_bytes(&self) -> u64

Bytes of one row of dims[0] elements.

Trait Implementations

impl Clone for Placed

fn clone(&self) -> Placed

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Placed

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

Formats the value using the given formatter.

impl PartialEq for Placed

fn eq(&self, other: &Placed) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for Placed