def calc_pdparam(self, x, evaluate=True):
'''
Calculate pdparams for multi-action by chunking the network logits output
'''
x = torch.cat(torch.split(x, self.state_dims, dim=1)).unsqueeze_(dim=1)
pdparam = SARSA.calc_pdparam(self, x, evaluate=evaluate)
return pdparam
def calc_pdparam(self, x, evaluate=True):
'''
Calculate pdparams for multi-action by chunking the network logits output
'''
x = torch.cat(torch.split(x, self.state_dims, dim=1)).unsqueeze_(dim=1)
pdparam = SARSA.calc_pdparam(self, x, evaluate=evaluate)
return pdparam
def calc_pdparam(self, xs, evaluate=True, net=None):
'''
Calculate pdparams for multi-action by chunking the network logits output
'''
pdparam = SARSA.calc_pdparam(self, xs, evaluate=evaluate, net=net)
return pdparam