DataPredict Pretrained Neural Networks

This repository contains all neural network models for various tasks.

How Models are Organized in the Repository

Folder Structure

The repository is organized into different levels, each showing specific details. Here's what each level means, starting from the top:

1. Platform / Subject
2. Task
3. Model Type (like Reinforcement Learning, Online Learning, Genetic Algorithm, Combined)
4. Model (Contains Codes and Data for Input, Output, Model, and Model Parameters)
5. Model Variations (Contains Codes and Data for Model and Model Parameters)

For example, in the path "Roblox/Sword Fighting/Reinforcement Learning/Version Alpha/V1":

• "Roblox" is the top-level "Platform / Subject."
• "Sword Fighting" is the second level indicating the "Task."

Models Relationship

Imagine you have two models, Model A and Model B.

Model B is considered a variation of Model A only if:

• Both models take the same input types and have the same number of neurons.
• Both models produce the same output types and have the same number of neurons.
• Both models have the same type.

If these conditions aren't met, Model B is treated as different from Model A and gets its own folder.

Want To Contribute?

Please read and sign this Contributor License Agreement. Once you have signed it, please email it to me at [email protected] and notify me on LinkedIn or on DevForum.