BareBonesNN is a lightweight neural network library implemented from scratch in Python, without relying on high-level machine learning libraries. This project aims to provide a clear and fundamental understanding of neural network concepts and operations, making it an excellent educational resource for learning and experimentation.

Features

• Basic Neural Network Components: Core classes including Value, Neuron, Layer, and MLP (Multi-Layer Perceptron).
• Automatic Differentiation: Built-in backpropagation for gradient computation.
• Customisable and Extensible: Easily modify and extend the code to experiment with various neural network architectures.
• Lightweight: Minimal dependencies, focusing on core principles.

Installation

To install BareBonesNN, use pip:

pip install barebonesnn==0.1.0

Usage

Here is a basic example demonstrating how to create and use a simple neural network using BareBonesNN:

Example: Creating and Using an MLP

from barebonesnn import Value, MLP

# Create a simple MLP with 3 input neurons and layers with [4, 4, 1] neurons
mlp = MLP(3, [4, 4, 1])

# Example input
x = [Value(1.0), Value(2.0), Value(3.0)]

# Forward pass
output = mlp(x)
print("Output:", output)

# Backward pass
output.backward()

# Inspect gradients
for param in mlp.parameters():
    print(param.label, param.grad)

Structure

• barebonesnn/value.py: Contains the Value class, which represents a value in a computational graph.
• barebonesnn/neuron.py: Contains the Neuron class, representing a single neuron.
• barebonesnn/layer.py: Contains the Layer class, representing a layer of neurons.
• barebonesnn/mlp.py: Contains the MLP class, representing a multi-layer perceptron.

Documentation

Value Class

Represents a value in a computational graph, supporting basic operations and automatic differentiation.

Neuron Class

Represents a single neuron in a neural network, with a set of weights and a bias.

Layer Class

Represents a layer in a neural network, consisting of multiple neurons.

MLP Class

Represents a multi-layer perceptron, consisting of multiple layers.

Examples

The examples/ directory contains scripts demonstrating various usages of the library. To run an example:

python examples/example.py

Contributing

Contributions are welcome! If you have suggestions, bug reports, or feature requests, please open an issue or submit a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.