cpid::OnlineZORBReplayBufferFrame Struct Reference

State, Action taken, reward Taking in an additional action taken allows you to not just take the max action, but use your own inference strategy, for example, if some actions are invalid. More...

#include <zeroordertrainer.h>

Inherits cpid::ReplayBufferFrame.

Public Member Functions
	OnlineZORBReplayBufferFrame (std::vector< torch::Tensor > state, std::vector< long > actions, double reward)
Public Member Functions inherited from cpid::ReplayBufferFrame
virtual	~ReplayBufferFrame ()=default
template<class Archive >
void	serialize (Archive &ar)

Public Attributes
std::vector< torch::Tensor >	state
std::vector< long >	actions
double	reward

Detailed Description

State, Action taken, reward Taking in an additional action taken allows you to not just take the max action, but use your own inference strategy, for example, if some actions are invalid.

Constructor & Destructor Documentation

cpid::OnlineZORBReplayBufferFrame::OnlineZORBReplayBufferFrame ( std::vector< torch::Tensor >  state, std::vector< long >  actions, double  reward  )

inline

Member Data Documentation

std::vector<long> cpid::OnlineZORBReplayBufferFrame::actions

double cpid::OnlineZORBReplayBufferFrame::reward

std::vector<torch::Tensor> cpid::OnlineZORBReplayBufferFrame::state

---

The documentation for this struct was generated from the following file:

• cpid/zeroordertrainer.h