Text-to-Image Neural Network Training Visualization

Example:

Xylene	ML-Xylene

Inspired and partly copied from https://github.com/MaxRobinsonTheGreat/mandelbrotnn

This repository contains code to visualize the training of a neural network model in PyTorch. The neural network is trained to learn 2D representations of a given text or picture. After training, these images are compiled into a video, visualizing the learning process.

Usage

1. Generate 2D Data Points from Text:

   • The text_to_points function in src/text_to_image.py generates 2D data points that visually represent a given text.

2. Train the Neural Network:

   • The train function in src/train.py contains the training loop, where the neural network is trained to learn the representations of the text.

3. Create a Video from Saved Images After Training:

   • After training, call the images_to_video function from src/video_creator.py to compile the saved images into a video.

Requirements

Install the necessary packages with:

pip install -r requirements.txt

Instructions

Clone the Repository:

git clone https://github.com/yourusername/Text-to-Image-NN-Training-Visualization.git

Navigate to the Project Directory:

cd Text-to-Image-NN-Training-Visualization

Install the Requirements:

pip install -r requirements.txt

Run Your Training Script:

Change the settings in the run_example.py file and start creating cool videos!