def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{ "q_values": self.q_values, })
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **self._extra_action_fetches)
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{"vf_preds": self.vf})
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{"behaviour_logits": self.model.outputs})
def extra_compute_action_fetches(self): out = {"behaviour_logits": self.model.outputs} if not self.config["vtrace"]: out["vf_preds"] = self.value_function return dict(TFPolicyGraph.extra_compute_action_fetches(self), **out)
def extra_compute_action_fetches(self): return dict( TFPolicyGraph.extra_compute_action_fetches(self), **{ SampleBatch.VF_PREDS: self.value_function, BEHAVIOUR_LOGITS: self.logits })
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{SampleBatch.VF_PREDS: self.vf})
def extra_compute_action_fetches(self): return dict( TFPolicyGraph.extra_compute_action_fetches(self), **{ "vf_preds": self.value_function, "logits": self.logits })
def extra_compute_action_fetches(self): out = {"behaviour_logits": self.model.outputs} if not self.config["vtrace"]: out["vf_preds"] = self.value_function return dict(TFPolicyGraph.extra_compute_action_fetches(self), **out)
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{"behavior_logp": self.sampled_logp})
def extra_compute_action_fetches(self): return dict( TFPolicyGraph.extra_compute_action_fetches(self), **{ "q_values": self.q_values, })
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{ "vf_preds": self.value_function, "logits": self.logits })
def extra_compute_action_fetches(self): return dict( TFPolicyGraph.extra_compute_action_fetches(self), **{"vf_preds": self.vf})
def extra_compute_action_fetches(self): return dict(TFPolicyGraph.extra_compute_action_fetches(self), **{BEHAVIOUR_LOGITS: self.model.outputs})
def extra_compute_action_fetches(self): return dict( TFPolicyGraph.extra_compute_action_fetches(self), **{"behaviour_logits": self.model.outputs})