f2a7ac7952dfd93abdfdb3a2e1dbed96066ab191,catalyst/rl/core/policy_handler.py,PolicyHandler,_get_q_values,#PolicyHandler#Any#Any#Any#,38

Before Change

    @torch.no_grad()
    def _get_q_values(self, critic: CriticSpec, state: np.ndarray, device):
        states = torch.Tensor(state).to(device).unsqueeze(0)
        output = critic(states)
        // We use the last head to perform actions
        // This is the head corresponding to the largest gamma
        if self.value_distribution == "categorical":

After Change

    @torch.no_grad()
    def _get_q_values(self, critic: CriticSpec, state: np.ndarray, device):
        states = _state2device(state, device)
        output = critic(states)
        // We use the last head to perform actions
        // This is the head corresponding to the largest gamma
        if self.value_distribution == "categorical":

Instances