AWACTrainer is just an implementation of AWR algorithm?

rail-berkeley / rlkit

Collection of reinforcement learning algorithms

MIT License

2.45k stars 550 forks source link

AWACTrainer is just an implementation of AWR algorithm? #122

Closed vinowan closed 3 years ago

vinowan commented 3 years ago

can't find implementation of awac algorithm in AWACTrainer，please let me know if i am wrong

ps: in section a.3 of the awac paper，awac algorithm is building on top of TD3，but the source code shows that it seems like its building on top of SAC?