Activity
Merge branch 'main' into vecnorm_fix
Merge branch 'main' into vecnorm_fix
Added _key_str method to to_observation_norm
Added _key_str method to to_observation_norm
Changed network device to CUDA in config
Changed network device to CUDA in config
Code functional; added a dummy make_value_estimator function
Code functional; added a dummy make_value_estimator function
NOTE: this code is broken; it shows the issue with using the value es…
NOTE: this code is broken; it shows the issue with using the value es…
Replay buffer now uses a temporary handler
Replay buffer now uses a temporary handler
Restructed TQCLoss; different losses are computed in separate methods
Restructed TQCLoss; different losses are computed in separate methods
Forgot one tensordict.get ...
Forgot one tensordict.get ...
Made quantile huber loss compatible with arbitrary batch sizes
Made quantile huber loss compatible with arbitrary batch sizes
Removed unnecessary copying on tensordict; since it is being modified…
Removed unnecessary copying on tensordict; since it is being modified…
Changed actor_loss comp to make compatible with arbitrary batch sizes
Changed actor_loss comp to make compatible with arbitrary batch sizes
Use .get method to access tensordict elements
Use .get method to access tensordict elements
Added TQC implementation (initial commit)
Added TQC implementation (initial commit)