Counterfactual-Multi-Agent-Policy-Gradients icon indicating copy to clipboard operation
Counterfactual-Multi-Agent-Policy-Gradients copied to clipboard

Published 20 hours ago • verified publisher icon matteokarldonati
→

Metadata

PyTorch implementation of Foerster, Jakob N., et al. "Counterfactual multi-agent policy gradients."

Results 2 Counterfactual-Multi-Agent-Policy-Gradients issues
Sort by recently updated
recently updated
newest added

Why is there no GRU in the actor network? I saw that there was in the original COMA text.

![image](https://github.com/matteokarldonati/Counterfactual-Multi-Agent-Policy-Gradients/assets/125025612/40467c1e-aa00-4509-88a2-83eec0358a5f) ![image](https://github.com/matteokarldonati/Counterfactual-Multi-Agent-Policy-Gradients/assets/125025612/ea3cfa0e-e71c-4eda-a03d-30b934f2da8d)

zhangmazi123321 avatar zhangmazi123321

Bad results sometimes occurred

2 comment

![output](https://user-images.githubusercontent.com/9446592/181408032-84af230c-b2c4-486f-9b72-4c6420ecd12c.png) I repeat running the code and sometimes the learning curve drops during training.

hccz95 avatar hccz95

← Metadata

49
Stars
11
Forks
Watchers

Owner

verified publisher icon matteokarldonati

Metadata

PyTorch implementation of Foerster, Jakob N., et al. "Counterfactual multi-agent policy gradients."

Back