search

ShangtongZhang / DeepRL

Modularized Implementation of Deep RL Algorithms in PyTorch
MIT License
3.21k stars 686 forks source link

Fix for sync actor #38

Closed ashigirl96 closed 6 years ago

ashigirl96 commented 6 years ago

In sync case, the sync actor cannot finish program.

ShangtongZhang commented 6 years ago

Is this because you forget to set config.max_steps? This self.start() shouldn't be a problem as I have overwrite start here https://github.com/ShangtongZhang/DeepRL/blob/master/deep_rl/agent/BaseAgent.py#L70