You are not logged in.
Pages: 1
This is not my code, but I remade it from the original blog post for an easy download. Until some vast changes are made I wont be officially submitting this.
Blog Post:
https://highleapstudios.wordpress.com/2 … -studio-1/
My Git:
https://github.com/dsk80004/Gamemaker-S … al-Network
If anyone has any questions on integrating it into a project feel free to ask.
Some of the changes made are things like only drawing 1/10 of the simulations, and some scaling to view more arrows at once. Thinking about incorporating a save system for the project, so your neurons can be loaded from the last test. Allowing for you to send the program to a stronger computer to train on, then return the neuron information back to your testing computer. Also a few quality of life improvements for the structuring of code.
If anyone is more experience with RNNs I could use some help adding other learning methods such as Q learning systems as a scripted function.
Offline
I've added a bunch of functions and difference memory gates already. When the final version is complete I'll be posting a github release.
Offline
Pages: 1