def brute_force_decode(sequence_lengths, inputs, transition_params):
num_words = inputs.shape[0]
num_tags = inputs.shape[1]
all_sequence_scores = []
all_sequences = []
tag_indices_iterator = itertools.product(range(num_tags), repeat=sequence_lengths)
inputs = tf.expand_dims(inputs, 0)
sequence_lengths = tf.expand_dims(sequence_lengths, 0)
transition_params = tf.constant(transition_params)
# Compare the dynamic program with brute force computation.
for tag_indices in tag_indices_iterator:
tag_indices = list(tag_indices)
tag_indices.extend([0] * (num_words - sequence_lengths))
all_sequences.append(tag_indices)
sequence_score = text.crf_sequence_score(
inputs=inputs,
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=sequence_lengths,
transition_params=transition_params,
)
sequence_score = tf.squeeze(sequence_score, [0])
all_sequence_scores.append(sequence_score)
expected_max_sequence_index = np.argmax(all_sequence_scores)
expected_max_sequence = all_sequences[expected_max_sequence_index]
expected_max_score = all_sequence_scores[expected_max_sequence_index]
return expected_max_sequence, expected_max_score
def testCrfSequenceScore(self):
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]],
dtype=np.float32)
# Test both the length-1 and regular cases.
sequence_lengths_list = [
np.array(3, dtype=np.int32),
np.array(1, dtype=np.int32)
]
inputs_list = [
np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]],
dtype=np.float32),
np.array([[4, 5, -3]], dtype=np.float32),
]
tag_indices_list = [
np.array([1, 2, 1, 0], dtype=np.int32),
np.array([1], dtype=np.int32)
]
for sequence_lengths, inputs, tag_indices in zip(
sequence_lengths_list, inputs_list, tag_indices_list):
sequence_score = text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params))
sequence_score = tf.squeeze(sequence_score, [0])
tf_sequence_score = self.evaluate(sequence_score)
expected_sequence_score = self.calculateSequenceScore(
inputs, transition_params, tag_indices, sequence_lengths)
self.assertAllClose(tf_sequence_score, expected_sequence_score)
def test_viterbi_decode(dtype):
inputs = np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]], dtype=dtype)
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]], dtype=dtype)
sequence_lengths = np.array(3, dtype=np.int32)
num_words = inputs.shape[0]
num_tags = inputs.shape[1]
all_sequence_scores = []
all_sequences = []
# Compare the dynamic program with brute force computation.
for tag_indices in itertools.product(range(num_tags), repeat=sequence_lengths):
tag_indices = list(tag_indices)
tag_indices.extend([0] * (num_words - sequence_lengths))
all_sequences.append(tag_indices)
sequence_score = text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params),
)
sequence_score = tf.squeeze(sequence_score, [0])
all_sequence_scores.append(sequence_score)
expected_max_sequence_index = np.argmax(all_sequence_scores)
expected_max_sequence = all_sequences[expected_max_sequence_index]
expected_max_score = all_sequence_scores[expected_max_sequence_index]
actual_max_sequence, actual_max_score = text.viterbi_decode(
inputs[:sequence_lengths], transition_params
)
test_utils.assert_allclose_according_to_type(actual_max_score, expected_max_score)
assert actual_max_sequence == expected_max_sequence[:sequence_lengths]
def test_crf_sequence_score(dtype):
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]], dtype=dtype)
# Test both the length-1 and regular cases.
sequence_lengths_list = [
np.array(3, dtype=np.int32),
np.array(1, dtype=np.int32),
]
inputs_list = [
np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]], dtype=dtype),
np.array([[4, 5, -3]], dtype=dtype),
]
tag_indices_list = [
np.array([1, 2, 1, 0], dtype=np.int32),
np.array([1], dtype=np.int32),
]
for sequence_lengths, inputs, tag_indices in zip(
sequence_lengths_list, inputs_list, tag_indices_list
):
sequence_score = text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params),
)
sequence_score = tf.squeeze(sequence_score, [0])
expected_sequence_score = calculate_sequence_score(
inputs, transition_params, tag_indices, sequence_lengths
)
test_utils.assert_allclose_according_to_type(
sequence_score, expected_sequence_score
)
def testCrfDecode(self):
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]],
dtype=np.float32)
# Test both the length-1 and regular cases.
sequence_lengths_list = [
np.array(3, dtype=np.int32),
np.array(1, dtype=np.int64)
]
inputs_list = [
np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]],
dtype=np.float32),
np.array([[-1, 2, 1]], dtype=np.float32),
]
tag_indices_list = [
np.array([1, 2, 1, 0], dtype=np.int32),
np.array([2], dtype=np.int32)
]
for sequence_lengths, inputs, tag_indices in zip(
sequence_lengths_list, inputs_list, tag_indices_list):
num_words = inputs.shape[0]
num_tags = inputs.shape[1]
all_sequence_scores = []
all_sequences = []
# Compare the dynamic program with brute force computation.
for tag_indices in itertools.product(
range(num_tags), repeat=sequence_lengths):
tag_indices = list(tag_indices)
tag_indices.extend([0] * (num_words - sequence_lengths))
all_sequences.append(tag_indices)
sequence_score = text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params))
sequence_score = tf.squeeze(sequence_score, [0])
all_sequence_scores.append(sequence_score)
tf_all_sequence_scores = self.evaluate(all_sequence_scores)
expected_max_sequence_index = np.argmax(tf_all_sequence_scores)
expected_max_sequence = all_sequences[expected_max_sequence_index]
expected_max_score = tf_all_sequence_scores[
expected_max_sequence_index]
actual_max_sequence, actual_max_score = text.crf_decode(
tf.expand_dims(inputs, 0), tf.constant(transition_params),
tf.expand_dims(sequence_lengths, 0))
actual_max_sequence = tf.squeeze(actual_max_sequence, [0])
actual_max_score = tf.squeeze(actual_max_score, [0])
tf_actual_max_sequence, tf_actual_max_score = self.evaluate(
[actual_max_sequence, actual_max_score])
self.assertAllClose(tf_actual_max_score, expected_max_score)
self.assertEqual(
list(tf_actual_max_sequence[:sequence_lengths]),
expected_max_sequence[:sequence_lengths])
def testCrfLogNorm(self):
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]],
dtype=np.float32)
# Test both the length-1 and regular cases.
sequence_lengths_list = [
np.array(3, dtype=np.int32),
np.array(1, dtype=np.int64),
]
inputs_list = [
np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]],
dtype=np.float32),
np.array([[3, -1, 3]], dtype=np.float32),
]
tag_indices_list = [
np.array([1, 2, 1, 0], dtype=np.int32),
np.array([2], dtype=np.int32),
]
for sequence_lengths, inputs, tag_indices in zip(
sequence_lengths_list, inputs_list, tag_indices_list):
num_words = inputs.shape[0]
num_tags = inputs.shape[1]
all_sequence_scores = []
# Compare the dynamic program with brute force computation.
for tag_indices in itertools.product(range(num_tags),
repeat=sequence_lengths):
tag_indices = list(tag_indices)
tag_indices.extend([0] * (num_words - sequence_lengths))
all_sequence_scores.append(
text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params),
))
brute_force_log_norm = tf.reduce_logsumexp(all_sequence_scores)
log_norm = text.crf_log_norm(
inputs=tf.expand_dims(inputs, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params),
)
log_norm = tf.squeeze(log_norm, [0])
tf_brute_force_log_norm, tf_log_norm = self.evaluate(
[brute_force_log_norm, log_norm])
self.assertAllClose(tf_log_norm, tf_brute_force_log_norm)
def testViterbiDecode(self):
inputs = np.array([[4, 5, -3], [3, -1, 3], [-1, 2, 1], [0, 0, 0]],
dtype=np.float32)
transition_params = np.array([[-3, 5, -2], [3, 4, 1], [1, 2, 1]],
dtype=np.float32)
sequence_lengths = np.array(3, dtype=np.int32)
# TODO: https://github.com/PyCQA/pylint/issues/3139
# pylint: disable=E1136
num_words = inputs.shape[0]
num_tags = inputs.shape[1]
# pylint: enable=E1136
all_sequence_scores = []
all_sequences = []
# Compare the dynamic program with brute force computation.
for tag_indices in itertools.product(range(num_tags),
repeat=sequence_lengths):
tag_indices = list(tag_indices)
tag_indices.extend([0] * (num_words - sequence_lengths))
all_sequences.append(tag_indices)
sequence_score = text.crf_sequence_score(
inputs=tf.expand_dims(inputs, 0),
tag_indices=tf.expand_dims(tag_indices, 0),
sequence_lengths=tf.expand_dims(sequence_lengths, 0),
transition_params=tf.constant(transition_params),
)
sequence_score = tf.squeeze(sequence_score, [0])
all_sequence_scores.append(sequence_score)
tf_all_sequence_scores = self.evaluate(all_sequence_scores)
expected_max_sequence_index = np.argmax(tf_all_sequence_scores)
expected_max_sequence = all_sequences[expected_max_sequence_index]
expected_max_score = tf_all_sequence_scores[
expected_max_sequence_index]
actual_max_sequence, actual_max_score = text.viterbi_decode(
inputs[:sequence_lengths], transition_params)
self.assertAllClose(actual_max_score, expected_max_score)
self.assertEqual(actual_max_sequence,
expected_max_sequence[:sequence_lengths])
