Enable gradient tests for tf.linalg.qr in eager mode.
PiperOrigin-RevId: 312572186 Change-Id: I4d1e62478fa41b277bd5191210f2da5e5c090653
This commit is contained in:
parent
99350ea1f0
commit
e28b37be96
|
@ -28,7 +28,7 @@ from tensorflow.python.framework import ops
|
||||||
from tensorflow.python.framework import test_util
|
from tensorflow.python.framework import test_util
|
||||||
from tensorflow.python.ops import array_ops
|
from tensorflow.python.ops import array_ops
|
||||||
from tensorflow.python.ops import control_flow_ops
|
from tensorflow.python.ops import control_flow_ops
|
||||||
from tensorflow.python.ops import gradient_checker
|
from tensorflow.python.ops import gradient_checker_v2
|
||||||
from tensorflow.python.ops import linalg_ops
|
from tensorflow.python.ops import linalg_ops
|
||||||
from tensorflow.python.ops import math_ops
|
from tensorflow.python.ops import math_ops
|
||||||
from tensorflow.python.ops import stateless_random_ops
|
from tensorflow.python.ops import stateless_random_ops
|
||||||
|
@ -175,13 +175,16 @@ class QrGradOpTest(test.TestCase):
|
||||||
|
|
||||||
def _GetQrGradOpTest(dtype_, shape_, full_matrices_):
|
def _GetQrGradOpTest(dtype_, shape_, full_matrices_):
|
||||||
|
|
||||||
@test_util.run_v1_only("b/120545219")
|
def RandomInput():
|
||||||
def Test(self):
|
|
||||||
np.random.seed(42)
|
|
||||||
a = np.random.uniform(low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
a = np.random.uniform(low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
||||||
if dtype_ in [np.complex64, np.complex128]:
|
if dtype_ in [np.complex64, np.complex128]:
|
||||||
a += 1j * np.random.uniform(
|
a += 1j * np.random.uniform(
|
||||||
low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
||||||
|
return a
|
||||||
|
|
||||||
|
@test_util.run_in_graph_and_eager_modes(use_gpu=True)
|
||||||
|
def Test(self):
|
||||||
|
np.random.seed(42)
|
||||||
# Optimal stepsize for central difference is O(epsilon^{1/3}).
|
# Optimal stepsize for central difference is O(epsilon^{1/3}).
|
||||||
epsilon = np.finfo(dtype_).eps
|
epsilon = np.finfo(dtype_).eps
|
||||||
delta = 0.1 * epsilon**(1.0 / 3.0)
|
delta = 0.1 * epsilon**(1.0 / 3.0)
|
||||||
|
@ -189,23 +192,16 @@ def _GetQrGradOpTest(dtype_, shape_, full_matrices_):
|
||||||
tol = 3e-2
|
tol = 3e-2
|
||||||
else:
|
else:
|
||||||
tol = 1e-6
|
tol = 1e-6
|
||||||
with self.session(use_gpu=True):
|
# TODO(b/157171666): Sadly we have to double the computation because
|
||||||
tf_a = constant_op.constant(a)
|
# gradient_checker_v2.compute_gradient expects a list of functions.
|
||||||
tf_b = linalg_ops.qr(tf_a, full_matrices=full_matrices_)
|
funcs = [
|
||||||
for b in tf_b:
|
lambda a: linalg_ops.qr(a, full_matrices=full_matrices_)[0],
|
||||||
x_init = np.random.uniform(
|
lambda a: linalg_ops.qr(a, full_matrices=full_matrices_)[1]
|
||||||
low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
]
|
||||||
if dtype_ in [np.complex64, np.complex128]:
|
for f in funcs:
|
||||||
x_init += 1j * np.random.uniform(
|
theoretical, numerical = gradient_checker_v2.compute_gradient(
|
||||||
low=-1.0, high=1.0, size=shape_).astype(dtype_)
|
f, [RandomInput()], delta=delta)
|
||||||
theoretical, numerical = gradient_checker.compute_gradient(
|
self.assertAllClose(theoretical, numerical, atol=tol, rtol=tol)
|
||||||
tf_a,
|
|
||||||
tf_a.get_shape().as_list(),
|
|
||||||
b,
|
|
||||||
b.get_shape().as_list(),
|
|
||||||
x_init_value=x_init,
|
|
||||||
delta=delta)
|
|
||||||
self.assertAllClose(theoretical, numerical, atol=tol, rtol=tol)
|
|
||||||
|
|
||||||
return Test
|
return Test
|
||||||
|
|
||||||
|
|
Loading…
Reference in New Issue