categorical_policy_agent.hpp File Reference

Stochastic categorical (discrete-action) policy over a logit-producing Module. More...

#include <cstdint>
#include "pulsatrix/agent.hpp"
#include "pulsatrix/device_backend.hpp"
#include "pulsatrix/module.hpp"
#include "pulsatrix/tensor.hpp"

Include dependency graph for categorical_policy_agent.hpp:

Go to the source code of this file.

Classes
class	pulsatrix::CategoricalPolicyAgent
	Samples an action from softmax(policy_network(observation)) – the discrete-action stochastic policy every policy-gradient method in this phase (REINFORCE, A2C, PPO) is built on. More...

Namespaces
namespace	pulsatrix

Detailed Description

Stochastic categorical (discrete-action) policy over a logit-producing Module.