Amber

Amber is a custom made Neural Network library written in Python. I am build this libray to learn more about Neural Networks and how they work. Maybe in the future I will use this library to build some cool projects.

Requirements 📦

• Python 3.6 or higher
• Numpy == 2.1.1

Components Implemented 🧩

• Models
• Layers
  • Input
  • Dense
  • Softmax
• Activation Functions
  • ReLU
  • Sigmoid

Usage 🚀

from amber.models import Models
from amber.layers import Dense, Input, Softmax

model = Models([
    Input(2),
    Dense(3, activation='relu'),
    Dense(2, activation='sigmoid'),
    Softmax()
])

model.compile()

x = np.array([1.3, 2.7])

y = model.forward(x)

Future Plans 🚀

• Implement more layers
• Implement more activation functions
• Implement Backpropagation
• Implement Optimizers
• Implement Loss Functions

Directory Structure 📁

├── amber
│   ├── layers.py
│   ├── models.py
│   ├── activations.py
│   ├── nurons.py
│   ├── utils.py

Support Me 🙏

If you like this project, consider supporting me. You can do this by:

• Starring this repository
• Sharing this repository with your friends
• Following me on Twitter