PyTorch autograd——只能为标量输出隐式创建 grad

PyTorch autograd -- grad can be implicitly created only for scalar outputs

我在 PyTorch 中使用 autograd 工具,发现自己处于需要通过整数索引访问一维张量中的值的情况。像这样:

def basic_fun(x_cloned):
    res = []
    for i in range(len(x)):
        res.append(x_cloned[i] * x_cloned[i])
    print(res)
    return Variable(torch.FloatTensor(res))


def get_grad(inp, grad_var):
    A = basic_fun(inp)
    A.backward()
    return grad_var.grad


x = Variable(torch.FloatTensor([1, 2, 3, 4, 5]), requires_grad=True)
x_cloned = x.clone()
print(get_grad(x_cloned, x))

我收到以下错误消息:

[tensor(1., grad_fn=<ThMulBackward>), tensor(4., grad_fn=<ThMulBackward>), tensor(9., grad_fn=<ThMulBackward>), tensor(16., grad_fn=<ThMulBackward>), tensor(25., grad_fn=<ThMulBackward>)]
Traceback (most recent call last):
  File "/home/mhy/projects/pytorch-optim/predict.py", line 74, in <module>
    print(get_grad(x_cloned, x))
  File "/home/mhy/projects/pytorch-optim/predict.py", line 68, in get_grad
    A.backward()
  File "/home/mhy/.local/lib/python3.5/site-packages/torch/tensor.py", line 93, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph)
  File "/home/mhy/.local/lib/python3.5/site-packages/torch/autograd/__init__.py", line 90, in backward
    allow_unreachable=True)  # allow_unreachable flag
RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn

总的来说,我有点怀疑使用变量的克隆版本应该如何在梯度计算中保留该变量。变量本身实际上并未用于 A 的计算,因此当您调用 A.backward() 时,它不应该是该操作的一部分。

感谢您对这种方法的帮助,或者是否有更好的方法来避免丢失梯度历史并仍然通过 requires_grad=True!

通过一维张量进行索引

**编辑(9 月 15 日):**

res 是包含 1 到 5 平方值的零维张量列表。为了连接成包含 [1.0, 4.0, ..., 25.0] 的单个张量,我更改了 return Variable(torch.FloatTensor(res))torch.stack(res, dim=0),产生 tensor([ 1., 4., 9., 16., 25.], grad_fn=<StackBackward>).

但是,我收到了这个由 A.backward() 行引起的新错误。

Traceback (most recent call last):
  File "<project_path>/playground.py", line 22, in <module>
    print(get_grad(x_cloned, x))
  File "<project_path>/playground.py", line 16, in get_grad
    A.backward()
  File "/home/mhy/.local/lib/python3.5/site-packages/torch/tensor.py", line 93, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph)
  File "/home/mhy/.local/lib/python3.5/site-packages/torch/autograd/__init__.py", line 84, in backward
    grad_tensors = _make_grads(tensors, grad_tensors)
  File "/home/mhy/.local/lib/python3.5/site-packages/torch/autograd/__init__.py", line 28, in _make_grads
    raise RuntimeError("grad can be implicitly created only for scalar outputs")
RuntimeError: grad can be implicitly created only for scalar outputs

在basic_fun函数中,res变量已经是一个torch-autograd-Variable,不需要再转换。恕我直言

def basic_fun(x_cloned):
    res = []
    for i in range(len(x)):
        res.append(x_cloned[i] * x_cloned[i])
    print(res)
    #return Variable(torch.FloatTensor(res))
    return res[0]

def get_grad(inp, grad_var):
    A = basic_fun(inp)
    A.backward()
    return grad_var.grad


x = Variable(torch.FloatTensor([1, 2, 3, 4, 5]), requires_grad=True)
x_cloned = x.clone()
print(get_grad(x_cloned, x))

我将 basic_fun 更改为以下内容,这解决了我的问题:

def basic_fun(x_cloned):
    res = torch.FloatTensor([0])
    for i in range(len(x)):
        res += x_cloned[i] * x_cloned[i]
    return res

这个版本returns一个标量值。