myfirstNN

A simple TUI tool to create and train NN on random datas, to teach the model how to calculate squares (x^2)

Software Structure

The file has 4 packages:

• display: contains the module that handles the TUI with Textual. Also contains the sub-package components with the different TUI components.

• learning: contains the module with the functions to train the model once is created.

• neuraln: contains the module with the class to create the model with his NN.

• core: starts and runs all the software. His called in main.py to start the whole package

How it works

When you start the software, the TUI runs in your terminal. When you set the name and the number of inputs, you can press the button "Start Training" and let the magic starts! You'll see al the log of the training, with a live loss, and ** you're gonna prove how an AI model training works!**

How to install

To install this software, I need to have installed on your machine Python 3.11 or newest.

Then run

pip install myfirstnn

myfirstnn

to have a run-ready software. If you want to have an easier access to the software files on your device, run

git clone https://github.com/alessioparolini2011/myfirstNN.git

This command needs Git on your device.

Credits

This project was built using the following amazing open-source libraries:

• PyTorch - An open-source machine learning framework used for creating, structuring, and training the neural network models.
• Textual - A Python framework, used to build the beautiful and interactive Terminal User Interface (TUI).