Skip to main content
Using Deno in production at your company? Earn free Deno merch.
Give us feedback
Module

x/netsaur/backends/gpu/layers/dense.ts>DenseGPULayer

Powerful machine learning, accelerated by WebGPU
denosaurs/netsaur
Go to Latest
class DenseGPULayer
import { DenseGPULayer } from "https://deno.land/x/netsaur@0.1.5/backends/gpu/layers/dense.ts";

Base class for all layers.

Constructors

new
DenseGPULayer(config: DenseLayerConfig, backend: WebGPUBackend)

Properties

activationFn: GPUActivationFn
biases: GPUMatrix
cost: GPUMatrix
costFunction: GPUCostFunction
error: GPUMatrix
input: GPUMatrix
output: GPUMatrix
outputSize: number
weights: GPUMatrix

Methods

backPropagate(
error: GPUMatrix,
prev: GPUMatrix,
rate: number,
last: number,
costFn?: GPUCostFunction,
): Promise<GPUMatrix>
feedForward(input: GPUMatrix): Promise<GPUMatrix>
initialize(
inputSize: Size,
batches: number,
type: DataType,
)
reset(type: DataType, batches: number)
toJSON(): Promise<LayerJSON>