RuntimeError:后端CPU的预期对象,但为参数#2'mat2'获得了后端CUDA

时间:2019-10-09 18:56:27

标签: python gpu cluster-computing pytorch

我有一个如下所示的代码段。它完美地在CPU中工作。但是,我想在GPU中运行代码,但出现以下错误。有什么想法吗?

def take_action(状态):

global steps_done    
sample = random.random()
eps_threshold = EPS_END + (EPS_START - EPS_END) * \
    math.exp(-1. * steps_done / EPS_DECAY)

steps_done += 1

if sample > eps_threshold:
    with torch.no_grad():

        action=torch.as_tensor(torch.argmax(policy_net(state)))

else:

    action=torch.tensor(random.randrange(0,7,1))

RuntimeError:后端CPU的预期对象,但参数#2'mat2'获得了后端CUDA。

0 个答案:

没有答案