def train_one_iter(grad):
grad_tensor = flow.Tensor(grad, requires_grad=False)
loss = x * grad_tensor
loss = flow.sum(x * grad_tensor)
loss.backward()
sgd.step()
sgd.zero_grad()
def train_one_iter(grad):
grad_tensor = flow.Tensor(grad,
requires_grad=False,
device=flow.device(device))
loss = flow.sum(x * grad_tensor)
loss.backward()
adam.step()
adam.zero_grad()
def _test_sum_impl(test_case, device):
input = flow.Tensor(np.random.randn(2, 3),
dtype=flow.float32,
device=flow.device(device))
of_out = flow.sum(input, dim=0)
np_out = np.sum(input.numpy(), axis=0)
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-5, 1e-5))
input = flow.Tensor(np.random.randn(2, 3),
dtype=flow.float32,
device=flow.device(device))
of_out = flow.sum(input, dim=0)
np_out = np.sum(input.numpy(), axis=0)
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-5, 1e-5))
input = flow.Tensor(np.random.randn(2, 3),
dtype=flow.float32,
device=flow.device(device))
of_out = flow.sum(input, dim=1)
of_out2 = input.sum(dim=1)
np_out = np.sum(input.numpy(), axis=1)
test_case.assertTrue(
np.allclose(of_out2.numpy(), of_out.numpy(), 1e-5, 1e-5))
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-5, 1e-5))
input = flow.Tensor(
np.random.randn(4, 5, 6),
dtype=flow.float32,
device=flow.device(device),
requires_grad=True,
)
of_out = flow.sum(input, dim=(2, 1))
np_out = np.sum(input.numpy(), axis=(2, 1))
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-5, 1e-5))
of_out = of_out.sum()
of_out.backward()
np_grad = np.ones((4, 5, 6))
test_case.assertTrue(np.allclose(input.grad.numpy(), np_grad, 1e-5, 1e-5))
def test_sum(test_case):
input = flow.Tensor(np.random.randn(2, 3), dtype=flow.float32)
of_out = flow.sum(input, dim=0)
np_out = np.sum(input.numpy(), axis=0)
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-4, 1e-4))
input = flow.Tensor(np.random.randn(2, 3), dtype=flow.float32)
of_out = flow.sum(input, dim=0)
np_out = np.sum(input.numpy(), axis=0)
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-4, 1e-4))
input = flow.Tensor(np.random.randn(2, 3), dtype=flow.float32)
of_out = flow.sum(input, dim=1)
of_out2 = input.sum(dim=1)
np_out = np.sum(input.numpy(), axis=1)
test_case.assertTrue(
np.allclose(of_out2.numpy(), of_out.numpy(), 1e-4, 1e-4))
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-4, 1e-4))
input = flow.Tensor(np.random.randn(4, 5, 6), dtype=flow.float32)
of_out = flow.sum(input, dim=(2, 1))
np_out = np.sum(input.numpy(), axis=(2, 1))
test_case.assertTrue(np.allclose(of_out.numpy(), np_out, 1e-4, 1e-4))
def training_step(self, batch, **kwargs):
return (flow.sum(self(batch)), list(self.m.parameters())[0])