You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using a SerialSampler with the default collector, and on some training iterations I'm getting 0 new completed trajectories along with 0 StepsInTrajWindow. Additionally, there simply isn't a DiscountedReturn line on the log during that iteration, along with all the other TrajInfo values.
Does anyone know what might cause this? I'm not getting any warnings or errors but I'm concerned I might be producing actions with nans in them.
I'm using the Hopper-v3 environment.
The text was updated successfully, but these errors were encountered:
Hmm, how long is your config["runner"]["log_interval_steps"] and how many environments are you using config["sampler"]["batch_B"]? If the first of those is small, and the second is large, it could simply be that none of the environment instances hit the end of their trajectory since the last logging event?
I'm using a SerialSampler with the default collector, and on some training iterations I'm getting 0 new completed trajectories along with 0 StepsInTrajWindow. Additionally, there simply isn't a DiscountedReturn line on the log during that iteration, along with all the other TrajInfo values.
Does anyone know what might cause this? I'm not getting any warnings or errors but I'm concerned I might be producing actions with nans in them.
I'm using the Hopper-v3 environment.
The text was updated successfully, but these errors were encountered: