Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add tanhshrink_py #1146

Merged
merged 3 commits into from
Feb 25, 2020
Merged

add tanhshrink_py #1146

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2625187
add tanhshrink_py
fsx950223 Feb 24, 2020
f98db46
format code
fsx950223 Feb 24, 2020
04c853f
remove useless test cases
fsx950223 Feb 25, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions tensorflow_addons/activations/tanhshrink.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,3 +38,7 @@ def tanhshrink(x: types.TensorLike) -> tf.Tensor:
@tf.RegisterGradient("Addons>Tanhshrink")
def _tanhshrink_grad(op, grad):
return _activation_so.ops.addons_tanhshrink_grad(grad, op.inputs[0])


def _tanhshrink_py(x):
return x - tf.math.tanh(x)
53 changes: 37 additions & 16 deletions tensorflow_addons/activations/tanhshrink_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,30 +18,51 @@
import numpy as np
import tensorflow as tf
from tensorflow_addons.activations import tanhshrink
from tensorflow_addons.activations.tanhshrink import _tanhshrink_py
from tensorflow_addons.utils import test_utils


@test_utils.run_all_in_graph_and_eager_modes
class TanhshrinkTest(tf.test.TestCase, parameterized.TestCase):
# @parameterized.named_parameters(
# ("float16", np.float16), ("float32", np.float32), ("float64", np.float64)
# )
# def test_tanhshrink(self, dtype):
# x = tf.constant([-2.0, -1.0, 0.0, 1.0, 2.0], dtype=dtype)
# expected_result = tf.constant(
# [-1.0359724, -0.23840582, 0.0, 0.23840582, 1.0359724], dtype=dtype
# )

# self.assertAllCloseAccordingToType(tanhshrink(x), expected_result)

# @parameterized.named_parameters(("float32", np.float32), ("float64", np.float64))
# def test_theoretical_gradients(self, dtype):
# # Only test theoretical gradients for float32 and float64
# # because of the instability of float16 while computing jacobian
# x = tf.constant([-2.0, -1.0, 0.0, 1.0, 2.0], dtype=dtype)

# theoretical, numerical = tf.test.compute_gradient(tanhshrink, [x])
# self.assertAllCloseAccordingToType(theoretical, numerical, atol=1e-4)

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe we could actually remove all that? The check with random tensors is much more robust. What do you think?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good idea. I remove them later.

@parameterized.named_parameters(
("float16", np.float16), ("float32", np.float32), ("float64", np.float64)
)
def test_tanhshrink(self, dtype):
x = tf.constant([-2.0, -1.0, 0.0, 1.0, 2.0], dtype=dtype)
expected_result = tf.constant(
[-1.0359724, -0.23840582, 0.0, 0.23840582, 1.0359724], dtype=dtype
)

self.assertAllCloseAccordingToType(tanhshrink(x), expected_result)

@parameterized.named_parameters(("float32", np.float32), ("float64", np.float64))
def test_theoretical_gradients(self, dtype):
# Only test theoretical gradients for float32 and float64
# because of the instability of float16 while computing jacobian
x = tf.constant([-2.0, -1.0, 0.0, 1.0, 2.0], dtype=dtype)

theoretical, numerical = tf.test.compute_gradient(tanhshrink, [x])
self.assertAllCloseAccordingToType(theoretical, numerical, atol=1e-4)
def test_same_as_py_func(self, dtype):
np.random.seed(1234)
for _ in range(20):
self.verify_funcs_are_equivalent(dtype)

def verify_funcs_are_equivalent(self, dtype):
x_np = np.random.uniform(-10, 10, size=(4, 4)).astype(dtype)
x = tf.convert_to_tensor(x_np)
with tf.GradientTape(persistent=True) as t:
t.watch(x)
y_native = tanhshrink(x)
y_py = _tanhshrink_py(x)
self.assertAllCloseAccordingToType(y_native, y_py, atol=1e-4)
grad_native = t.gradient(y_native, x)
grad_py = t.gradient(y_py, x)
self.assertAllCloseAccordingToType(grad_native, grad_py, atol=1e-4)


if __name__ == "__main__":
Expand Down