Hammer
Hammer is an example application that demonstrates how to use the AI Neural Network Framework implemented in Go. This repository is used for testing and explaining the capabilities of the AI framework, showcasing how to define and run neural networks with various neuron types.
Overview
The main application sets up a neural network using the blueprint
package. It defines a network configuration in JSON format, initializes the network, and runs it with specified inputs over multiple timesteps.
The network includes various neuron types:
- Input Neurons
- Dense Neurons
- Recurrent Neurons (RNN)
- Long Short-Term Memory Neurons (LSTM)
- Convolutional Neurons (CNN)
- Attention Mechanism Neurons
- Output Neurons
Purpose
This repository serves as a testing ground for the AI Neural Network Framework. It allows for:
- Experimentation with different neuron configurations, network architectures, and inputs.
- Facilitating understanding and development of the framework.
- Demonstrating the framework’s ability to process inputs through a variety of neuron types.
Getting Started
-
Setup:
- Ensure that you have Go installed.
- Set up the AI Neural Network Framework (
blueprint
package).
-
Run the Application:
- Execute the main program to initialize and run the neural network.
-
Observe the Outputs:
- Review the outputs to see how the network processes the inputs through various neuron types over multiple timesteps.
License
This project is licensed under the Apache License 2.0.