b5e848af46b4a6fc21e718803dbf3d7c72afd72a,ch08/lib/common.py,,calc_loss,#,92

Before Change

        next_states_v = next_states_v.cuda()
        actions_v = actions_v.cuda()
        rewards_v = rewards_v.cuda()
        done_mask = done_mask.cuda()

    state_action_values = net(states_v).gather(1, actions_v.unsqueeze(-1)).squeeze(-1)
    next_state_actions = net(next_states_v).max(1)[1]
    next_state_values = tgt_net(next_states_v).gather(1, next_state_actions.unsqueeze(-1)).squeeze(-1)

After Change

    next_state_values = tgt_net(next_states_v).gather(1, next_state_actions.unsqueeze(-1)).squeeze(-1)
    next_state_values[done_mask] = 0.0

    expected_state_action_values = next_state_values.detach() * gamma + rewards_v
    return nn.MSELoss()(state_action_values, expected_state_action_values)

Instances