StepNeverStop/RLs

Implement Model Saving Mechanism

Closed this issue · 0 comments

StepNeverStop commented

include save model based on:

  • training time cost
  • score performance of under-training policy
  • training timestep
  • ...