Godot Deep RL using TorchSharp [Jul2024 update]

[GeorgeS2019]

Jul 2024 update

We now have two possibilities with godot and TorchSharp.

• Godot using TorchSharp for RL:

RL_Matrix

• https://github.com/asieradzk/RL_Matrix

Torchsharp: Gym-based Reinforcement Learning framework
dotnet/TorchSharp#1333

• Godot using Unity like Multigent like framework using Torchsharp for RL

AI4USharp

https://github.com/gilzamir18/AI4USharp
dotnet/TorchSharp#1337

Details

Currently there is an unprecedented **Deep RL using TorchSharp with QLearning** started [here](https://github.com/xin-pu/DeepSharp/issues/4).

I urge the Godot community here who are interested of exporting Onnx and subsequently using c# to feedback and pershap start contributing other RL algorithm and RL Environment to this project.

If there is sufficient interest, there is a good change to develop an early stage of SB3 for .NET using a modification of the project.

[GeorgeS2019]

[GeorgeS2019]

@edbeeching
Latest 2024 Godot4 and TorchSharp Deep Reinforcement Learning update

Using Godot, How would do you make a multi-agent scene

[GeorgeS2019]

asieradzk/RL_Matrix#13 (comment)

@edbeeching
Only c# Godot RL using Torchsharp through RLib rather than SD3

[GeorgeS2019]

@edbeeching

We now have two possibilities with godot and TorchSharp.

• Godot using TorchSharp for RL
• Godot using Unity like Multigent like framework using Torchsharp for RL