Ivan-267
Ivan-267
This PR proposes some changes to the description of the observation space. However, please verify the validity first since I'm not very familiar with this env. My current understanding after...
## Description Fixes the variable horizon concept link from the tutorial for training AIRL. ## Testing I've tried the original link in master branch and the link from this PR...
## Proposal: In addition to reward, having an overview of the current success rate can be useful in many envs. This can be a very important metric in envs that...
Note: The exact features of the sensor and whether to implement it is yet to be determined, it's just an idea at this point. The most basic functionality would include...
This is my attempt to get discrete only actions to work properly on onnx exported models with SB3 and also discrete actions > 2 to work with SB3. As I'm...
As we discussed previously on discord, one idea is to implement the functionality from the sb3 example to the gdrl command, I'm adding it here as a potential todo for...
- Fixes some issues with initial high rewards due static platforms detecting collisions with goal areas, - adjusts rewards, - other sight modifications. I was able to train this with...
TODO: - [x] Update plugin to read onnx actions properly.
This is a quickly made lightweight fast-training env without external assets or other dependencies for testing. It was made to have a very simple env to test how easy it...
> [!IMPORTANT] > This uses a mixture of plugin updates not currently in main branch including: https://github.com/edbeeching/godot_rl_agents_plugin/pull/37, https://github.com/edbeeching/godot_rl_agents_plugin/pull/40, and following modifications: - Removed the negative step reward (it seems to...