KerasCustom

🧩 Custom layers and operations for Keras 3, designed to extend its capabilities with modular, composable components.

---

🌟 Overview

KerasCustom provides advanced and non-native Keras layers and operations that simplify the definition, transformation, and manipulation of neural network architectures. It is built to integrate seamlessly with the Keras Core backend system (TensorFlow, JAX, or PyTorch).

This library is part of a broader ecosystem including:

• JacobiNet – for Jacobian computation as a Keras layer
• Decomon – for LiRPA / formal verification methods
• Onnx2Keras3 – for ONNX-to-Keras model conversion
• Keras2Marabou – for expressing verification properties
• Airobas – for end-to-end formal verification pipelines

---

---
maxdepth: 2
caption: Contents
---
install
getting_started
tutorials
api/modules
contribute
Github  <https://github.com/ducoffeM/keras_custom>

🚀 Installation

```bash pip install git+https://github.com/ducoffeM/keras_custom.git

📊 Summary of keras_custom Functions

Category	Functions
Array Creation & Manipulation	append, arange, diag, diagonal, expand_dims, flip, full_like, get_item, hstack, identity, moveaxis, ones_like, repeat, roll, sort, split, squeeze, stack, swapaxes, transpose, tril, triu, zeros_like
Mathematical Operations	abs, absolute, add, ceil, clip, divide, expm1, floor, log, log10, log1p, log2, logaddexp, maximum, minimum, negative, power, reciprocal, round, sign, sqrt, square, true_divide, trunc
Reduction Operations	all, amax, amin, any, average, cumprod, cumsum, max, mean, min, prod, std, sum, var
Linear Algebra	cross, norm, trace
Trigonometric Functions	arccos, arccosh, arcsin, arcsinh, arctan, arctan2, arctanh, cos, cosh, sin, sinh, tan, tanh