search

Theohhhu / UPDeT

Official Implementation of 'UPDeT: Universal Multi-agent Reinforcement Learning via Policy Decoupling with Transformers' ICLR 2021(spotlight)
MIT License
129 stars 17 forks source link

Unsatisfactory experimental results #14

Closed zichuan-liu closed 2 years ago

zichuan-liu commented 2 years ago

Hi, i wanna know why I run 5m_vs_6m win rate is only about 20%-40% in qmix, the same happens on 'rnn' and 'updet', I also use sc2 version 4.10, run 2M steps, torch=1.4.1, the code did not make any changes. Is there anyone else like this? How do I reproduce these results, if you can answer too thanks!

zichuan-liu commented 2 years ago

I solved it

CrazySssst commented 1 year ago

Hi, I have the same problem (QMIX 5m_vs_6m sc2 version 4.10, run 2M, perform very low). Can you tell me how to solve it?