RobvanGastel / rnn-sac

Implementation of the Discrete Soft Actor-Critic algorithm with RNN policy in PyTorch
17 stars 2 forks source link

continous sac #1

Open luo013 opened 2 years ago

luo013 commented 2 years ago

hello, my brother. Whether continuous SAC can be combined with LSTM?

RobvanGastel commented 2 years ago

Hi Luo013, thanks for showing interest! This modification is possible for sure. It would require changing the CategoricalPolicy of the SAC algorithm to the regular Gaussian Policy as is done in OpenAI SpinningUp and possibly a small part of the adjustment which is not contained within these classes. I will try to adjust this somewhere this week if I find time.

luo013 commented 2 years ago

Thanky you for your answer

sinaebrahimi commented 2 months ago

Hi,

Thank you for this awesome repo. Did you add continuous option as well?