It's a PyTorch implementation,why use tensorflow?

mlii / mfrl

Mean Field Multi-Agent Reinforcement Learning

MIT License

369 stars 101 forks source link

Closed qwq1082 closed 4 years ago

qwq1082 commented 4 years ago

It's not a PyTorch implementation of MF-Q and MF-AC, but it uses tensorflow.

mlii commented 4 years ago

Thanks for pointing out the typo!