19. layers/residual

Part: III · Layers
Package: github.com/openfluke/welvet/layers/residual
Status: ok — ✅

When

Building or training a net that needs the layers/residual Op (also usable as a Parallel cam).

Where

import "github.com/openfluke/welvet/layers/residual"

cd 19-residual && source ../env.sh && go run .

Why

Skip connections stabilize deep stacks: y = F(x) + x with correct skip grads.

What

F is Dense Dim→Dim, or mixed Ops via NewFromOps (Dense, SwiGLU, RMSNorm, LayerNorm). Parallel as F is parallel.ResidualGraft (y = F(x)+x) — Residual cannot import parallel.

Sample output (captured)

16 <nil>

Live capture from go run ./cmd/runall on local ../../welvet (exit 0).

Source

Copied from the Welvet feature book examples (openfluke.github.io/welvet/examples/19-residual).

package main

import (
	"fmt"

	"github.com/openfluke/welvet/core"
	"github.com/openfluke/welvet/layers/residual"
)

func main() {
	l, err := residual.New(residual.Config{Dim: 16, Depth: 2})
	if err != nil {
		panic(err)
	}
	x := core.NewTensor[float32](1, 16)
	_, y, err := residual.Forward(l, x)
	fmt.Println(len(y.Data), err)
}