lunny/keras
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
f94df3479a
keras/keras_core/layers/rnn/gru_test.py

287 lines
9.2 KiB
Python
Raw Normal View History

Adds CategoryEncoding layer, bincount op, and tests (#161) * Adds unit normalization and tests * Adds layer normalization and initial tests * Fixes formatting in docstrings * Fix type issues for JAX * Fix nits * Initial stash for group_normalization and spectral_normalization * Adds spectral normalization and tests * Adds group normalization and tests * Formatting fixes * Fix small nit in docstring * Fix docstring and tests * Adds RandomContrast and associated tests * Remove arithmetic comment * Adds RandomBrightness and tests * Fix docstring and format * Fix nits and add backend generator * Inlines random_contrast helper * Add bincount op * Add CategoryEncoding layer and tests * Fix formatting * Fix JAX issues * Fix JAX bincount * Formatting and small fix * Fix nits and docstrings * Add args to bincount op test
2023-05-14 00:07:43 +00:00
import numpy as np
from absl.testing import parameterized
from keras_core import initializers
from keras_core import layers
from keras_core import testing
class GRUTest(testing.TestCase, parameterized.TestCase):
def test_basics(self):
self.run_layer_test(
layers.GRU,
init_kwargs={"units": 3, "dropout": 0.5, "recurrent_dropout": 0.5},
input_shape=(3, 2, 4),
call_kwargs={"training": True},
expected_output_shape=(3, 3),
expected_num_trainable_weights=3,
expected_num_non_trainable_weights=0,
supports_masking=True,
)
self.run_layer_test(
layers.GRU,
init_kwargs={
"units": 3,
"return_sequences": True,
"bias_regularizer": "l1",
"kernel_regularizer": "l2",
"recurrent_regularizer": "l2",
},
input_shape=(3, 2, 4),
expected_output_shape=(3, 2, 3),
expected_num_losses=3,
expected_num_trainable_weights=3,
expected_num_non_trainable_weights=0,
supports_masking=True,
)
@parameterized.parameters([1, 2])
def test_correctness(self, implementation):
sequence = np.arange(72).reshape((3, 6, 4)).astype("float32")
layer = layers.GRU(
3,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.5217289, 0.5217289, 0.5217289],
[0.6371659, 0.6371659, 0.6371659],
[0.39384964, 0.39384964, 0.3938496],
]
),
output,
)
layer = layers.GRU(
3,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
go_backwards=True,
)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.24406259, 0.24406259, 0.24406259],
[0.611516, 0.611516, 0.611516],
[0.3928808, 0.3928808, 0.3928808],
]
),
output,
)
layer = layers.GRU(
3,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
unroll=True,
)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.5217289, 0.5217289, 0.5217289],
[0.6371659, 0.6371659, 0.6371659],
[0.39384964, 0.39384964, 0.3938496],
]
),
output,
)
layer = layers.GRU(
3,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
reset_after=False,
)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.51447755, 0.51447755, 0.51447755],
[0.6426879, 0.6426879, 0.6426879],
[0.40208298, 0.40208298, 0.40208298],
]
),
output,
)
layer = layers.GRU(
3,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
use_bias=False,
)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.49988455, 0.49988455, 0.49988455],
[0.64701194, 0.64701194, 0.64701194],
[0.4103359, 0.4103359, 0.4103359],
]
),
output,
)
def test_statefulness(self):
sequence = np.arange(24).reshape((2, 3, 4)).astype("float32")
layer = layers.GRU(
4,
stateful=True,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
)
layer(sequence)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.29542392, 0.29542392, 0.29542392, 0.29542392],
[0.5885018, 0.5885018, 0.5885018, 0.5885018],
]
),
output,
)
layer.reset_state()
layer(sequence)
output = layer(sequence)
self.assertAllClose(
np.array(
[
[0.29542392, 0.29542392, 0.29542392, 0.29542392],
[0.5885018, 0.5885018, 0.5885018, 0.5885018],
]
),
output,
)
Increase GRU + LSTM test coverage.
2023-05-16 20:25:48 +00:00
def test_pass_initial_state(self):
sequence = np.arange(24).reshape((2, 4, 3)).astype("float32")
initial_state = np.arange(4).reshape((2, 2)).astype("float32")
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
)
output = layer(sequence, initial_state=initial_state)
self.assertAllClose(
np.array([[0.23774096, 0.33508456], [0.83659905, 1.0227708]]),
output,
)
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
go_backwards=True,
)
output = layer(sequence, initial_state=initial_state)
self.assertAllClose(
np.array([[0.13486053, 0.23261218], [0.78257304, 0.9691353]]),
output,
)
def test_masking(self):
sequence = np.arange(24).reshape((2, 4, 3)).astype("float32")
mask = np.array([[True, True, False, True], [True, False, False, True]])
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
unroll=True,
)
output = layer(sequence, mask=mask)
self.assertAllClose(
np.array([[0.19393763, 0.19393763], [0.30818558, 0.30818558]]),
output,
)
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
return_sequences=True,
)
output = layer(sequence, mask=mask)
self.assertAllClose(
np.array(
[
[0.03606692, 0.03606692],
[0.09497581, 0.09497581],
[0.09497581, 0.09497581],
[0.19393763, 0.19393763],
],
),
output[0],
)
self.assertAllClose(
np.array(
[
[0.16051409, 0.16051409],
[0.16051409, 0.16051409],
[0.16051409, 0.16051409],
[0.30818558, 0.30818558],
],
),
output[1],
)
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
return_sequences=True,
zero_output_for_mask=True,
)
output = layer(sequence, mask=mask)
self.assertAllClose(
np.array(
[
[0.03606692, 0.03606692],
[0.09497581, 0.09497581],
[0.0, 0.0],
[0.19393763, 0.19393763],
],
),
output[0],
)
self.assertAllClose(
np.array(
[
[0.16051409, 0.16051409],
[0.0, 0.0],
[0.0, 0.0],
[0.30818558, 0.30818558],
],
),
output[1],
)
layer = layers.GRU(
2,
kernel_initializer=initializers.Constant(0.01),
recurrent_initializer=initializers.Constant(0.02),
bias_initializer=initializers.Constant(0.03),
go_backwards=True,
)
output = layer(sequence, mask=mask)
self.assertAllClose(
np.array([[0.11669192, 0.11669192], [0.28380975, 0.28380975]]),
output,
)
Reference in New Issue Copy Permalink