experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Enable 1st order gradient tests for tf.linalg.svd in eager mode.

Browse Source 
PiperOrigin-RevId: 312756858
Change-Id: I20d73e8972014b96bc90952949820390ae77e08d
This commit is contained in:
A. Unique TensorFlower 2020-05-21 15:51:02 -07:00 committed by TensorFlower Gardener
parent 60fb5dcc7d
commit e312350702
1 changed files with 26 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

39
tensorflow/python/kernel_tests/svd_op_test.py
View File

@ -29,6 +29,7 @@ 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
from tensorflow.python.ops import gradient_checker_v2
from tensorflow.python.ops import gradients_impl from tensorflow.python.ops import gradients_impl
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
@ -225,45 +226,41 @@ def _NormalizingSvd(tf_a, full_matrices_):
def _GetSvdGradOpTest(dtype_, shape_, compute_uv_, full_matrices_): def _GetSvdGradOpTest(dtype_, shape_, compute_uv_, full_matrices_):
@test_util.run_v1_only("b/120545219") @test_util.run_in_graph_and_eager_modes(use_gpu=True)
def Test(self): def Test(self):
def RandomInput():
np.random.seed(42) 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
# Optimal stepsize for central difference is O(epsilon^{1/3}). # Optimal stepsize for central difference is O(epsilon^{1/3}).
# See Equation (21) in: # See Equation (21) in:
# http://www.karenkopecky.net/Teaching/eco613614/Notes_NumericalDifferentiation.pdf # http://www.karenkopecky.net/Teaching/eco613614/Notes_NumericalDifferentiation.pdf
# TODO(rmlarsen): Move step size control to gradient checker. # TODO(rmlarsen): Move step size control to gradient checker.
epsilon = np.finfo(dtype_).eps epsilon = np.finfo(dtype_).eps
delta = 0.1 * epsilon**(1.0 / 3.0) delta = 0.25 * epsilon**(1.0 / 3.0)
if dtype_ in [np.float32, np.complex64]: if dtype_ in [np.float32, np.complex64]:
tol = 3e-2 tol = 3e-2
else: else:
tol = 1e-6 tol = 1e-6
with self.session(use_gpu=True):
tf_a = constant_op.constant(a)
if compute_uv_: if compute_uv_:
tf_s, tf_u, tf_v = _NormalizingSvd(tf_a, full_matrices_) funcs = [
outputs = [tf_s, tf_u, tf_v] lambda a: _NormalizingSvd(a, full_matrices_)[0],
lambda a: _NormalizingSvd(a, full_matrices_)[1],
lambda a: _NormalizingSvd(a, full_matrices_)[2]
]
else: else:
tf_s = linalg_ops.svd(tf_a, compute_uv=False) funcs = [lambda a: linalg_ops.svd(a, compute_uv=False)]
outputs = [tf_s]
for b in outputs: for f in funcs:
x_init = np.random.uniform( theoretical, numerical = gradient_checker_v2.compute_gradient(
low=-1.0, high=1.0, size=shape_).astype(dtype_) f, [RandomInput()], delta=delta)
if dtype_ in [np.complex64, np.complex128]:
x_init += 1j * np.random.uniform(
low=-1.0, high=1.0, size=shape_).astype(dtype_)
theoretical, numerical = gradient_checker.compute_gradient(
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) self.assertAllClose(theoretical, numerical, atol=tol, rtol=tol)
return Test return Test