Skip to content

Conversation

@JacobHA
Copy link
Owner

@JacobHA JacobHA commented Jul 2, 2024

TODOs:

  • add flag for using a new thread
  • understand stop event logic
  • ensure that the eval is happening at the correct timestep (and not using later, more up-to-date q nets)

@JacobHA JacobHA marked this pull request as draft July 2, 2024 15:49
@JacobHA JacobHA marked this pull request as ready for review July 2, 2024 16:27
@JacobHA
Copy link
Owner Author

JacobHA commented Jul 2, 2024

@BobMak How does this logic look for you? I'd still like to understand what the stop is doing, and whether the log interval is correct

@JacobHA JacobHA requested a review from BobMak July 2, 2024 16:31
train_fps = self.log_interval / train_time
self.log_history('time/train_fps', train_fps, self.learn_env_steps)
self.avg_eval_rwd = self.evaluate()
if self.use_threaded_eval:
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks like we are waiting for the finish of the old worker but not starting the new worker thread when self.use_threaded_eval is true

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants