[About SMAC results] - Githubissues

semitable / seac

The official code base of Shared Experience Actor-Critic (NeurIPS2020)

19 stars 23 forks source link

Open Jiwonjeon9603 opened 2 years ago

Jiwonjeon9603 commented 2 years ago

Hello,

In this paper, SMAC with sparse rewards needs many time steps (0.2*10^8) to converge.

I want to know how much time step does it need when I want to train SMAC with dense rewards.

( I am considering '3m')