Jump to content

Connect SuperML | Leeroopedia MCP: Equip your AI agents with best practices, code verification, and debugging knowledge. Powered by Leeroo — building Organizational Superintelligence. Contact us at founders@leeroo.com.

Principle:Tensorflow Tfjs Layer API Surface

From Leeroopedia


Knowledge Sources
Domains Deep_Learning, API_Design, Layers_API
Last Updated 2026-02-10 06:00 GMT

Overview

Public API surface that exposes layer construction, metric computation, weight initialization, and utility functions through factory functions registered on the tf.layers, tf.metrics, tf.initializers, tf.constraints, and tf.regularizers namespaces.

Description

TensorFlow.js Layers uses a factory function pattern to expose its API. Rather than requiring users to directly instantiate classes, the library provides top-level functions (e.g., tf.layers.dense(), tf.metrics.binaryAccuracy()) that create and return configured layer/metric instances. These factory functions are defined in exports_*.ts files:

  • exports_layers.ts: All layer constructors (dense, conv2d, lstm, etc.)
  • exports_metrics.ts: Metric functions (accuracy, precision, recall, etc.)
  • exports_initializers.ts: Weight initializer factories (glorotUniform, heNormal, etc.)
  • exports_constraints.ts: Weight constraint factories (maxNorm, nonNeg, etc.)
  • exports_regularizers.ts: Weight regularizer factories (l1, l2, l1l2)

This pattern provides a consistent, discoverable API surface while keeping internal class hierarchies encapsulated.

Usage

Use the factory functions under the tf.layers, tf.metrics, tf.initializers, tf.constraints, and tf.regularizers namespaces when building models. These are the intended public entry points for constructing all TensorFlow.js Layers objects.

Theoretical Basis

Pseudo-code Logic:

// Factory pattern: user-facing function creates internal class instance
// tf.layers.dense({units: 10}) internally calls: new Dense({units: 10})
// tf.regularizers.l2({l2: 0.01}) internally calls: new L2({l2: 0.01})

// Registration pattern using @doc annotations:
function dense(args: DenseLayerArgs): Dense {
  return new Dense(args);
}

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment