Struct ConvStats 

pub struct ConvStats {
    pub input_shape: Vec<usize>,
    pub output_shape: Vec<usize>,
    pub kernel_shape: Vec<usize>,
    pub num_parameters: usize,
    pub flops: u64,
    pub receptive_field: Vec<usize>,
}

Statistics about a convolution operation (parameter count, FLOPs, receptive field).

Fields

input_shape: Vec<usize>

Shape of the input tensor.

output_shape: Vec<usize>

Shape of the output tensor.

kernel_shape: Vec<usize>

Shape of the kernel/weight tensor.

num_parameters: usize

Total number of learnable parameters (weights + bias if present).

flops: u64

Estimated floating-point operations (multiply-accumulate counted as 2).

receptive_field: Vec<usize>

Receptive field size in each spatial dimension.

Implementations

impl ConvStats

pub fn compute( input_shape: &[usize], weight_shape: &[usize], config: &ConvConfig, ) -> Result<Self, ConvError>

Compute convolution statistics from input/weight shapes and config.

Input shape: [batch, in_channels, spatial...] Weight shape: [out_channels, in_channels/groups, kernel_spatial...]

pub fn summary(&self) -> String

Human-readable summary string.

Trait Implementations

impl Clone for ConvStats

fn clone(&self) -> ConvStats

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ConvStats

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

Formats the value using the given formatter.