def test_read_embedding_file_inside_archive(self):
token2vec = {
"think": torch.Tensor([0.143, 0.189, 0.555, 0.361, 0.472]),
"make": torch.Tensor([0.878, 0.651, 0.044, 0.264, 0.872]),
"difference": torch.Tensor([0.053, 0.162, 0.671, 0.110, 0.259]),
"àèìòù": torch.Tensor([1.0, 2.0, 3.0, 4.0, 5.0])
}
vocab = Vocabulary()
for token in token2vec:
vocab.add_token_to_namespace(token)
params = Params({
'pretrained_file':
str(self.FIXTURES_ROOT / 'embeddings/multi-file-archive.zip'),
'embedding_dim':
5
})
with pytest.raises(
ValueError,
message=
"No ValueError when pretrained_file is a multi-file archive"):
Embedding.from_params(vocab, params)
for ext in ['.zip', '.tar.gz']:
archive_path = str(
self.FIXTURES_ROOT / 'embeddings/multi-file-archive') + ext
file_uri = format_embeddings_file_uri(
archive_path, 'folder/fake_embeddings.5d.txt')
params = Params({'pretrained_file': file_uri, 'embedding_dim': 5})
embeddings = Embedding.from_params(vocab, params).weight.data
for tok, vec in token2vec.items():
i = vocab.get_token_index(tok)
assert torch.equal(embeddings[i],
vec), 'Problem with format ' + archive_path
def test_make_vocab_without_extension(self):
existing_serialization_dir = self.TEST_DIR / 'existing'
extended_serialization_dir = self.TEST_DIR / 'extended'
existing_vocab_path = existing_serialization_dir / 'vocabulary'
extended_vocab_path = extended_serialization_dir / 'vocabulary'
vocab = Vocabulary()
vocab.add_token_to_namespace('some_weird_token_1', namespace='tokens')
vocab.add_token_to_namespace('some_weird_token_2', namespace='tokens')
# if extend is False, its users responsibility to make sure that dataset instances
# will be indexible by provided vocabulary. At least @@UNKNOWN@@ should be present in
# namespace for which there could be OOV entries seen in dataset during indexing.
# For `tokens` ns, new words will be seen but `tokens` has @@UNKNOWN@@ token.
# but for 'labels' ns, there is no @@UNKNOWN@@ so required to add 'N', 'V' upfront.
vocab.add_token_to_namespace('N', namespace='labels')
vocab.add_token_to_namespace('V', namespace='labels')
os.makedirs(existing_serialization_dir, exist_ok=True)
vocab.save_to_files(existing_vocab_path)
self.params['vocabulary'] = {}
self.params['vocabulary']['directory_path'] = existing_vocab_path
self.params['vocabulary']['extend'] = False
make_vocab_from_params(self.params, extended_serialization_dir)
with open(extended_vocab_path / 'tokens.txt') as f:
tokens = [line.strip() for line in f]
assert tokens[0] == '@@UNKNOWN@@'
assert tokens[1] == 'some_weird_token_1'
assert tokens[2] == 'some_weird_token_2'
assert len(tokens) == 3
def test_get_embedding_layer_uses_correct_embedding_dim(self):
vocab = Vocabulary()
vocab.add_token_to_namespace('word1')
vocab.add_token_to_namespace('word2')
embeddings_filename = str(self.TEST_DIR / "embeddings.gz")
with gzip.open(embeddings_filename, 'wb') as embeddings_file:
embeddings_file.write("word1 1.0 2.3 -1.0\n".encode('utf-8'))
embeddings_file.write("word2 0.1 0.4 -4.0\n".encode('utf-8'))
embedding_weights = _read_pretrained_embeddings_file(
embeddings_filename, 3, vocab)
assert tuple(embedding_weights.size()) == (
4, 3) # 4 because of padding and OOV
with pytest.raises(ConfigurationError):
_read_pretrained_embeddings_file(embeddings_filename, 4, vocab)
class TestDataset(AllenNlpTestCase):
def setUp(self):
self.vocab = Vocabulary()
self.vocab.add_token_to_namespace("this")
self.vocab.add_token_to_namespace("is")
self.vocab.add_token_to_namespace("a")
self.vocab.add_token_to_namespace("sentence")
self.vocab.add_token_to_namespace(".")
self.token_indexer = {"tokens": SingleIdTokenIndexer()}
self.instances = self.get_instances()
super(TestDataset, self).setUp()
def test_instances_must_have_homogeneous_fields(self):
instance1 = Instance({"tag": (LabelField(1, skip_indexing=True))})
instance2 = Instance({"words": TextField([Token("hello")], {})})
with pytest.raises(ConfigurationError):
_ = Batch([instance1, instance2])
def test_padding_lengths_uses_max_instance_lengths(self):
dataset = Batch(self.instances)
dataset.index_instances(self.vocab)
padding_lengths = dataset.get_padding_lengths()
assert padding_lengths == {"text1": {"num_tokens": 5}, "text2": {"num_tokens": 6}}
def test_as_tensor_dict(self):
dataset = Batch(self.instances)
dataset.index_instances(self.vocab)
padding_lengths = dataset.get_padding_lengths()
tensors = dataset.as_tensor_dict(padding_lengths)
text1 = tensors["text1"]["tokens"].detach().cpu().numpy()
text2 = tensors["text2"]["tokens"].detach().cpu().numpy()
numpy.testing.assert_array_almost_equal(text1, numpy.array([[2, 3, 4, 5, 6],
[1, 3, 4, 5, 6]]))
numpy.testing.assert_array_almost_equal(text2, numpy.array([[2, 3, 4, 1, 5, 6],
[2, 3, 1, 0, 0, 0]]))
def get_instances(self):
field1 = TextField([Token(t) for t in ["this", "is", "a", "sentence", "."]],
self.token_indexer)
field2 = TextField([Token(t) for t in ["this", "is", "a", "different", "sentence", "."]],
self.token_indexer)
field3 = TextField([Token(t) for t in ["here", "is", "a", "sentence", "."]],
self.token_indexer)
field4 = TextField([Token(t) for t in ["this", "is", "short"]],
self.token_indexer)
instances = [Instance({"text1": field1, "text2": field2}),
Instance({"text1": field3, "text2": field4})]
return instances
def test_tokens_to_indices_uses_pos_tags(self):
tokens = self.tokenizer.split_words("This is a sentence.")
tokens = [t for t in tokens] + [Token("</S>")]
vocab = Vocabulary()
root_index = vocab.add_token_to_namespace('ROOT',
namespace='dep_labels')
none_index = vocab.add_token_to_namespace('NONE',
namespace='dep_labels')
indexer = DepLabelIndexer()
assert indexer.tokens_to_indices([tokens[1]], vocab, "tokens1") == {
"tokens1": [root_index]
}
assert indexer.tokens_to_indices([tokens[-1]], vocab, "tokens-1") == {
"tokens-1": [none_index]
}
def test_tokens_to_indices_uses_ner_tags(self):
tokens = self.tokenizer.split_words("Larry Page is CEO of Google.")
tokens = [t for t in tokens] + [Token("</S>")]
vocab = Vocabulary()
person_index = vocab.add_token_to_namespace('PERSON',
namespace='ner_tags')
none_index = vocab.add_token_to_namespace('NONE', namespace='ner_tags')
vocab.add_token_to_namespace('ORG', namespace='ner_tags')
indexer = NerTagIndexer()
assert indexer.tokens_to_indices([tokens[1]], vocab, "tokens1") == {
"tokens1": [person_index]
}
assert indexer.tokens_to_indices([tokens[-1]], vocab, "tokens-1") == {
"tokens-1": [none_index]
}
def test_get_embedding_layer_initializes_unseen_words_randomly_not_zero(
self):
vocab = Vocabulary()
vocab.add_token_to_namespace("word")
vocab.add_token_to_namespace("word2")
embeddings_filename = str(self.TEST_DIR / "embeddings.gz")
with gzip.open(embeddings_filename, 'wb') as embeddings_file:
embeddings_file.write("word 1.0 2.3 -1.0\n".encode('utf-8'))
params = Params({
'pretrained_file': embeddings_filename,
'embedding_dim': 3,
})
embedding_layer = Embedding.from_params(vocab, params)
word_vector = embedding_layer.weight.data[vocab.get_token_index(
"word2")]
assert not numpy.allclose(word_vector.numpy(),
numpy.array([0.0, 0.0, 0.0]))
def test_read_hdf5_raises_on_invalid_shape(self):
vocab = Vocabulary()
vocab.add_token_to_namespace("word")
embeddings_filename = str(self.TEST_DIR / "embeddings.hdf5")
embeddings = numpy.random.rand(vocab.get_vocab_size(), 10)
with h5py.File(embeddings_filename, 'w') as fout:
_ = fout.create_dataset('embedding',
embeddings.shape,
dtype='float32',
data=embeddings)
params = Params({
'pretrained_file': embeddings_filename,
'embedding_dim': 5,
})
with pytest.raises(ConfigurationError):
_ = Embedding.from_params(vocab, params)
class TestTokenCharactersEncoder(AllenNlpTestCase):
def setUp(self):
super(TestTokenCharactersEncoder, self).setUp()
self.vocab = Vocabulary()
self.vocab.add_token_to_namespace("1", "token_characters")
self.vocab.add_token_to_namespace("2", "token_characters")
self.vocab.add_token_to_namespace("3", "token_characters")
self.vocab.add_token_to_namespace("4", "token_characters")
params = Params({
"embedding": {
"embedding_dim": 2,
"vocab_namespace": "token_characters"
},
"encoder": {
"type": "cnn",
"embedding_dim": 2,
"num_filters": 4,
"ngram_filter_sizes": [1, 2],
"output_dim": 3
}
})
self.encoder = TokenCharactersEncoder.from_params(
vocab=self.vocab, params=deepcopy(params))
self.embedding = Embedding.from_params(vocab=self.vocab,
params=params["embedding"])
self.inner_encoder = Seq2VecEncoder.from_params(params["encoder"])
constant_init = lambda tensor: torch.nn.init.constant_(tensor, 1.)
initializer = InitializerApplicator([(".*", constant_init)])
initializer(self.encoder)
initializer(self.embedding)
initializer(self.inner_encoder)
def test_get_output_dim_uses_encoder_output_dim(self):
assert self.encoder.get_output_dim() == 3
def test_forward_applies_embedding_then_encoder(self):
numpy_tensor = numpy.random.randint(6, size=(3, 4, 7))
inputs = torch.from_numpy(numpy_tensor)
encoder_output = self.encoder(inputs)
reshaped_input = inputs.view(12, 7)
embedded = self.embedding(reshaped_input)
mask = (inputs != 0).long().view(12, 7)
reshaped_manual_output = self.inner_encoder(embedded, mask)
manual_output = reshaped_manual_output.view(3, 4, 3)
assert_almost_equal(encoder_output.data.numpy(),
manual_output.data.numpy())
def test_read_hdf5_format_file(self):
vocab = Vocabulary()
vocab.add_token_to_namespace("word")
vocab.add_token_to_namespace("word2")
embeddings_filename = str(self.TEST_DIR / "embeddings.hdf5")
embeddings = numpy.random.rand(vocab.get_vocab_size(), 5)
with h5py.File(embeddings_filename, 'w') as fout:
_ = fout.create_dataset('embedding',
embeddings.shape,
dtype='float32',
data=embeddings)
params = Params({
'pretrained_file': embeddings_filename,
'embedding_dim': 5,
})
embedding_layer = Embedding.from_params(vocab, params)
assert numpy.allclose(embedding_layer.weight.data.numpy(), embeddings)
def test_forward_works_with_projection_layer(self):
vocab = Vocabulary()
vocab.add_token_to_namespace('the')
vocab.add_token_to_namespace('a')
params = Params({
'pretrained_file':
str(self.FIXTURES_ROOT / 'embeddings/glove.6B.300d.sample.txt.gz'),
'embedding_dim':
300,
'projection_dim':
20
})
embedding_layer = Embedding.from_params(vocab, params)
input_tensor = torch.LongTensor([[3, 2, 1, 0]])
embedded = embedding_layer(input_tensor).data.numpy()
assert embedded.shape == (1, 4, 20)
input_tensor = torch.LongTensor([[[3, 2, 1, 0]]])
embedded = embedding_layer(input_tensor).data.numpy()
assert embedded.shape == (1, 1, 4, 20)
def test_make_vocab_with_extension(self):
existing_serialization_dir = self.TEST_DIR / 'existing'
extended_serialization_dir = self.TEST_DIR / 'extended'
existing_vocab_path = existing_serialization_dir / 'vocabulary'
extended_vocab_path = extended_serialization_dir / 'vocabulary'
vocab = Vocabulary()
vocab.add_token_to_namespace('some_weird_token_1', namespace='tokens')
vocab.add_token_to_namespace('some_weird_token_2', namespace='tokens')
os.makedirs(existing_serialization_dir, exist_ok=True)
vocab.save_to_files(existing_vocab_path)
self.params['vocabulary'] = {}
self.params['vocabulary']['directory_path'] = existing_vocab_path
self.params['vocabulary']['extend'] = True
self.params['vocabulary']['min_count'] = {"tokens": 3}
make_vocab_from_params(self.params, extended_serialization_dir)
vocab_files = os.listdir(extended_vocab_path)
assert set(vocab_files) == {
'labels.txt', 'non_padded_namespaces.txt', 'tokens.txt'
}
with open(extended_vocab_path / 'tokens.txt') as f:
tokens = [line.strip() for line in f]
assert tokens[0] == '@@UNKNOWN@@'
assert tokens[1] == 'some_weird_token_1'
assert tokens[2] == 'some_weird_token_2'
tokens.sort()
assert tokens == [
'.', '@@UNKNOWN@@', 'animals', 'are', 'some_weird_token_1',
'some_weird_token_2'
]
with open(extended_vocab_path / 'labels.txt') as f:
labels = [line.strip() for line in f]
labels.sort()
assert labels == ['N', 'V']
def test_embedding_layer_actually_initializes_word_vectors_correctly(self):
vocab = Vocabulary()
vocab.add_token_to_namespace("word")
vocab.add_token_to_namespace("word2")
unicode_space = "\xa0\xa0\xa0\xa0\xa0\xa0\xa0\xa0"
vocab.add_token_to_namespace(unicode_space)
embeddings_filename = str(self.TEST_DIR / "embeddings.gz")
with gzip.open(embeddings_filename, 'wb') as embeddings_file:
embeddings_file.write("word 1.0 2.3 -1.0\n".encode('utf-8'))
embeddings_file.write(
f"{unicode_space} 3.4 3.3 5.0\n".encode('utf-8'))
params = Params({
'pretrained_file': embeddings_filename,
'embedding_dim': 3,
})
embedding_layer = Embedding.from_params(vocab, params)
word_vector = embedding_layer.weight.data[vocab.get_token_index(
"word")]
assert numpy.allclose(word_vector.numpy(),
numpy.array([1.0, 2.3, -1.0]))
word_vector = embedding_layer.weight.data[vocab.get_token_index(
unicode_space)]
assert numpy.allclose(word_vector.numpy(), numpy.array([3.4, 3.3,
5.0]))
word_vector = embedding_layer.weight.data[vocab.get_token_index(
"word2")]
assert not numpy.allclose(word_vector.numpy(),
numpy.array([1.0, 2.3, -1.0]))
def test_index_converts_field_correctly(self):
vocab = Vocabulary()
sentence_index = vocab.add_token_to_namespace("sentence", namespace='words')
capital_a_index = vocab.add_token_to_namespace("A", namespace='words')
capital_a_char_index = vocab.add_token_to_namespace("A", namespace='characters')
s_index = vocab.add_token_to_namespace("s", namespace='characters')
e_index = vocab.add_token_to_namespace("e", namespace='characters')
n_index = vocab.add_token_to_namespace("n", namespace='characters')
t_index = vocab.add_token_to_namespace("t", namespace='characters')
c_index = vocab.add_token_to_namespace("c", namespace='characters')
field = TextField([Token(t) for t in ["A", "sentence"]],
{"words": SingleIdTokenIndexer(namespace="words")})
field.index(vocab)
# pylint: disable=protected-access
assert field._indexed_tokens["words"] == [capital_a_index, sentence_index]
field1 = TextField([Token(t) for t in ["A", "sentence"]],
{"characters": TokenCharactersIndexer(namespace="characters")})
field1.index(vocab)
assert field1._indexed_tokens["characters"] == [[capital_a_char_index],
[s_index, e_index, n_index, t_index,
e_index, n_index, c_index, e_index]]
field2 = TextField([Token(t) for t in ["A", "sentence"]],
token_indexers={"words": SingleIdTokenIndexer(namespace="words"),
"characters": TokenCharactersIndexer(namespace="characters")})
field2.index(vocab)
assert field2._indexed_tokens["words"] == [capital_a_index, sentence_index]
assert field2._indexed_tokens["characters"] == [[capital_a_char_index],
[s_index, e_index, n_index, t_index,
e_index, n_index, c_index, e_index]]
def test_label_field_can_index_with_vocab(self):
vocab = Vocabulary()
vocab.add_token_to_namespace("entailment", namespace="labels")
vocab.add_token_to_namespace("contradiction", namespace="labels")
vocab.add_token_to_namespace("neutral", namespace="labels")
label = LabelField("entailment")
label.index(vocab)
tensor = label.as_tensor(label.get_padding_lengths())
assert tensor.item() == 0
class IteratorTest(AllenNlpTestCase):
def setUp(self):
super(IteratorTest, self).setUp()
self.token_indexers = {"tokens": SingleIdTokenIndexer()}
self.vocab = Vocabulary()
self.this_index = self.vocab.add_token_to_namespace('this')
self.is_index = self.vocab.add_token_to_namespace('is')
self.a_index = self.vocab.add_token_to_namespace('a')
self.sentence_index = self.vocab.add_token_to_namespace('sentence')
self.another_index = self.vocab.add_token_to_namespace('another')
self.yet_index = self.vocab.add_token_to_namespace('yet')
self.very_index = self.vocab.add_token_to_namespace('very')
self.long_index = self.vocab.add_token_to_namespace('long')
instances = [
self.create_instance(["this", "is", "a", "sentence"]),
self.create_instance(["this", "is", "another", "sentence"]),
self.create_instance(["yet", "another", "sentence"]),
self.create_instance([
"this", "is", "a", "very", "very", "very", "very", "long",
"sentence"
]),
self.create_instance(["sentence"]),
]
class LazyIterable:
def __iter__(self):
return (instance for instance in instances)
self.instances = instances
self.lazy_instances = LazyIterable()
def create_instance(self, str_tokens: List[str]):
tokens = [Token(t) for t in str_tokens]
instance = Instance({'text': TextField(tokens, self.token_indexers)})
instance.index_fields(self.vocab)
return instance
def assert_instances_are_correct(self, candidate_instances):
# First we need to remove padding tokens from the candidates.
# pylint: disable=protected-access
candidate_instances = [
tuple(w for w in instance if w != 0)
for instance in candidate_instances
]
expected_instances = [
tuple(instance.fields["text"]._indexed_tokens["tokens"])
for instance in self.instances
]
assert set(candidate_instances) == set(expected_instances)
def test_tokens_to_indices_produces_correct_characters(self):
vocab = Vocabulary()
vocab.add_token_to_namespace("A", namespace='characters')
vocab.add_token_to_namespace("s", namespace='characters')
vocab.add_token_to_namespace("e", namespace='characters')
vocab.add_token_to_namespace("n", namespace='characters')
vocab.add_token_to_namespace("t", namespace='characters')
vocab.add_token_to_namespace("c", namespace='characters')
indexer = TokenCharactersIndexer("characters")
indices = indexer.tokens_to_indices([Token("sentential")], vocab, "char")
assert indices == {"char": [[3, 4, 5, 6, 4, 5, 6, 1, 1, 1]]}
def test_tokens_to_indices_uses_pos_tags(self):
tokens = self.tokenizer.split_words("This is a sentence.")
tokens = [t for t in tokens] + [Token("</S>")]
vocab = Vocabulary()
verb_index = vocab.add_token_to_namespace('VERB', namespace='pos_tags')
cop_index = vocab.add_token_to_namespace('VBZ', namespace='pos_tags')
none_index = vocab.add_token_to_namespace('NONE', namespace='pos_tags')
# Have to add other tokens too, since we're calling `tokens_to_indices` on all of them
vocab.add_token_to_namespace('DET', namespace='pos_tags')
vocab.add_token_to_namespace('NOUN', namespace='pos_tags')
vocab.add_token_to_namespace('PUNCT', namespace='pos_tags')
indexer = PosTagIndexer(coarse_tags=True)
indices = indexer.tokens_to_indices(tokens, vocab, "tokens")
assert len(indices) == 1
assert "tokens" in indices
assert indices["tokens"][1] == verb_index
assert indices["tokens"][-1] == none_index
indexer._coarse_tags = False # pylint: disable=protected-access
assert indexer.tokens_to_indices([tokens[1]], vocab, "coarse") == {
"coarse": [cop_index]
}
class TestTextField(AllenNlpTestCase):
def setUp(self):
self.vocab = Vocabulary()
self.vocab.add_token_to_namespace("sentence", namespace='words')
self.vocab.add_token_to_namespace("A", namespace='words')
self.vocab.add_token_to_namespace("A", namespace='characters')
self.vocab.add_token_to_namespace("s", namespace='characters')
self.vocab.add_token_to_namespace("e", namespace='characters')
self.vocab.add_token_to_namespace("n", namespace='characters')
self.vocab.add_token_to_namespace("t", namespace='characters')
self.vocab.add_token_to_namespace("c", namespace='characters')
super(TestTextField, self).setUp()
def test_field_counts_vocab_items_correctly(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words")})
namespace_token_counts = defaultdict(lambda: defaultdict(int))
field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["words"]["This"] == 1
assert namespace_token_counts["words"]["is"] == 1
assert namespace_token_counts["words"]["a"] == 1
assert namespace_token_counts["words"]["sentence"] == 1
assert namespace_token_counts["words"]["."] == 1
assert list(namespace_token_counts.keys()) == ["words"]
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"characters": TokenCharactersIndexer("characters")})
namespace_token_counts = defaultdict(lambda: defaultdict(int))
field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["characters"]["T"] == 1
assert namespace_token_counts["characters"]["h"] == 1
assert namespace_token_counts["characters"]["i"] == 2
assert namespace_token_counts["characters"]["s"] == 3
assert namespace_token_counts["characters"]["a"] == 1
assert namespace_token_counts["characters"]["e"] == 3
assert namespace_token_counts["characters"]["n"] == 2
assert namespace_token_counts["characters"]["t"] == 1
assert namespace_token_counts["characters"]["c"] == 1
assert namespace_token_counts["characters"]["."] == 1
assert list(namespace_token_counts.keys()) == ["characters"]
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words"),
"characters": TokenCharactersIndexer("characters")})
namespace_token_counts = defaultdict(lambda: defaultdict(int))
field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["characters"]["T"] == 1
assert namespace_token_counts["characters"]["h"] == 1
assert namespace_token_counts["characters"]["i"] == 2
assert namespace_token_counts["characters"]["s"] == 3
assert namespace_token_counts["characters"]["a"] == 1
assert namespace_token_counts["characters"]["e"] == 3
assert namespace_token_counts["characters"]["n"] == 2
assert namespace_token_counts["characters"]["t"] == 1
assert namespace_token_counts["characters"]["c"] == 1
assert namespace_token_counts["characters"]["."] == 1
assert namespace_token_counts["words"]["This"] == 1
assert namespace_token_counts["words"]["is"] == 1
assert namespace_token_counts["words"]["a"] == 1
assert namespace_token_counts["words"]["sentence"] == 1
assert namespace_token_counts["words"]["."] == 1
assert set(namespace_token_counts.keys()) == {"words", "characters"}
def test_index_converts_field_correctly(self):
vocab = Vocabulary()
sentence_index = vocab.add_token_to_namespace("sentence", namespace='words')
capital_a_index = vocab.add_token_to_namespace("A", namespace='words')
capital_a_char_index = vocab.add_token_to_namespace("A", namespace='characters')
s_index = vocab.add_token_to_namespace("s", namespace='characters')
e_index = vocab.add_token_to_namespace("e", namespace='characters')
n_index = vocab.add_token_to_namespace("n", namespace='characters')
t_index = vocab.add_token_to_namespace("t", namespace='characters')
c_index = vocab.add_token_to_namespace("c", namespace='characters')
field = TextField([Token(t) for t in ["A", "sentence"]],
{"words": SingleIdTokenIndexer(namespace="words")})
field.index(vocab)
# pylint: disable=protected-access
assert field._indexed_tokens["words"] == [capital_a_index, sentence_index]
field1 = TextField([Token(t) for t in ["A", "sentence"]],
{"characters": TokenCharactersIndexer(namespace="characters")})
field1.index(vocab)
assert field1._indexed_tokens["characters"] == [[capital_a_char_index],
[s_index, e_index, n_index, t_index,
e_index, n_index, c_index, e_index]]
field2 = TextField([Token(t) for t in ["A", "sentence"]],
token_indexers={"words": SingleIdTokenIndexer(namespace="words"),
"characters": TokenCharactersIndexer(namespace="characters")})
field2.index(vocab)
assert field2._indexed_tokens["words"] == [capital_a_index, sentence_index]
assert field2._indexed_tokens["characters"] == [[capital_a_char_index],
[s_index, e_index, n_index, t_index,
e_index, n_index, c_index, e_index]]
# pylint: enable=protected-access
def test_get_padding_lengths_raises_if_no_indexed_tokens(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words")})
with pytest.raises(ConfigurationError):
field.get_padding_lengths()
def test_padding_lengths_are_computed_correctly(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
assert padding_lengths == {"num_tokens": 5}
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"characters": TokenCharactersIndexer("characters")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
assert padding_lengths == {"num_tokens": 5, "num_token_characters": 8}
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"characters": TokenCharactersIndexer("characters"),
"words": SingleIdTokenIndexer("words")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
assert padding_lengths == {"num_tokens": 5, "num_token_characters": 8}
def test_as_tensor_handles_words(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
tensor_dict = field.as_tensor(padding_lengths)
numpy.testing.assert_array_almost_equal(tensor_dict["words"].detach().cpu().numpy(),
numpy.array([1, 1, 1, 2, 1]))
def test_as_tensor_handles_longer_lengths(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
padding_lengths["num_tokens"] = 10
tensor_dict = field.as_tensor(padding_lengths)
numpy.testing.assert_array_almost_equal(tensor_dict["words"].detach().cpu().numpy(),
numpy.array([1, 1, 1, 2, 1, 0, 0, 0, 0, 0]))
def test_as_tensor_handles_characters(self):
field = TextField([Token(t) for t in ["This", "is", "a", "sentence", "."]],
token_indexers={"characters": TokenCharactersIndexer("characters")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
tensor_dict = field.as_tensor(padding_lengths)
expected_character_array = numpy.array([[1, 1, 1, 3, 0, 0, 0, 0],
[1, 3, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0],
[3, 4, 5, 6, 4, 5, 7, 4],
[1, 0, 0, 0, 0, 0, 0, 0]])
numpy.testing.assert_array_almost_equal(tensor_dict["characters"].detach().cpu().numpy(),
expected_character_array)
def test_as_tensor_handles_words_and_characters_with_longer_lengths(self):
field = TextField([Token(t) for t in ["a", "sentence", "."]],
token_indexers={"words": SingleIdTokenIndexer("words"),
"characters": TokenCharactersIndexer("characters")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
padding_lengths["num_tokens"] = 5
padding_lengths["num_token_characters"] = 10
tensor_dict = field.as_tensor(padding_lengths)
numpy.testing.assert_array_almost_equal(tensor_dict["words"].detach().cpu().numpy(),
numpy.array([1, 2, 1, 0, 0]))
numpy.testing.assert_array_almost_equal(tensor_dict["characters"].detach().cpu().numpy(),
numpy.array([[1, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[3, 4, 5, 6, 4, 5, 7, 4, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]))
def test_printing_doesnt_crash(self):
field = TextField([Token(t) for t in ["A", "sentence"]],
{"words": SingleIdTokenIndexer(namespace="words")})
print(field)
def test_token_embedder_returns_dict(self):
field = TextField([Token(t) for t in ["A", "sentence"]],
token_indexers={"field_with_dict": DictReturningTokenIndexer(),
"words": SingleIdTokenIndexer("words"),
"characters": TokenCharactersIndexer("characters")})
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
assert padding_lengths == {
'token_ids': 5,
'additional_key': 2,
'words': 2,
'characters': 2,
'num_token_characters': 8
}
padding_lengths['token_ids'] = 7
padding_lengths['additional_key'] = 3
padding_lengths['words'] = 4
padding_lengths['characters'] = 4
tensors = field.as_tensor(padding_lengths)
assert list(tensors['token_ids'].shape) == [7]
assert list(tensors['additional_key'].shape) == [3]
assert list(tensors['words'].shape) == [4]
assert list(tensors['characters'].shape) == [4, 8]
class TestListField(AllenNlpTestCase):
def setUp(self):
self.vocab = Vocabulary()
self.vocab.add_token_to_namespace("this", "words")
self.vocab.add_token_to_namespace("is", "words")
self.vocab.add_token_to_namespace("a", "words")
self.vocab.add_token_to_namespace("sentence", 'words')
self.vocab.add_token_to_namespace("s", 'characters')
self.vocab.add_token_to_namespace("e", 'characters')
self.vocab.add_token_to_namespace("n", 'characters')
self.vocab.add_token_to_namespace("t", 'characters')
self.vocab.add_token_to_namespace("c", 'characters')
for label in ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k']:
self.vocab.add_token_to_namespace(label, 'labels')
self.word_indexer = {"words": SingleIdTokenIndexer("words")}
self.words_and_characters_indexers = {
"words": SingleIdTokenIndexer("words"),
"characters": TokenCharactersIndexer("characters")
}
self.field1 = TextField(
[Token(t) for t in ["this", "is", "a", "sentence"]],
self.word_indexer)
self.field2 = TextField(
[Token(t) for t in ["this", "is", "a", "different", "sentence"]],
self.word_indexer)
self.field3 = TextField(
[Token(t) for t in ["this", "is", "another", "sentence"]],
self.word_indexer)
self.empty_text_field = self.field1.empty_field()
self.index_field = IndexField(1, self.field1)
self.empty_index_field = self.index_field.empty_field()
self.sequence_label_field = SequenceLabelField([1, 1, 0, 1],
self.field1)
self.empty_sequence_label_field = self.sequence_label_field.empty_field(
)
super(TestListField, self).setUp()
def test_get_padding_lengths(self):
list_field = ListField([self.field1, self.field2, self.field3])
list_field.index(self.vocab)
lengths = list_field.get_padding_lengths()
assert lengths == {"num_fields": 3, "list_num_tokens": 5}
def test_list_field_can_handle_empty_text_fields(self):
list_field = ListField(
[self.field1, self.field2, self.empty_text_field])
list_field.index(self.vocab)
tensor_dict = list_field.as_tensor(list_field.get_padding_lengths())
numpy.testing.assert_array_equal(
tensor_dict["words"].detach().cpu().numpy(),
numpy.array([[2, 3, 4, 5, 0], [2, 3, 4, 1, 5], [0, 0, 0, 0, 0]]))
def test_list_field_can_handle_empty_index_fields(self):
list_field = ListField(
[self.index_field, self.index_field, self.empty_index_field])
list_field.index(self.vocab)
tensor = list_field.as_tensor(list_field.get_padding_lengths())
numpy.testing.assert_array_equal(tensor.detach().cpu().numpy(),
numpy.array([[1], [1], [-1]]))
def test_list_field_can_handle_empty_sequence_label_fields(self):
list_field = ListField([
self.sequence_label_field, self.sequence_label_field,
self.empty_sequence_label_field
])
list_field.index(self.vocab)
tensor = list_field.as_tensor(list_field.get_padding_lengths())
numpy.testing.assert_array_equal(
tensor.detach().cpu().numpy(),
numpy.array([[1, 1, 0, 1], [1, 1, 0, 1], [0, 0, 0, 0]]))
def test_all_fields_padded_to_max_length(self):
list_field = ListField([self.field1, self.field2, self.field3])
list_field.index(self.vocab)
tensor_dict = list_field.as_tensor(list_field.get_padding_lengths())
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][0].detach().cpu().numpy(),
numpy.array([2, 3, 4, 5, 0]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][1].detach().cpu().numpy(),
numpy.array([2, 3, 4, 1, 5]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][2].detach().cpu().numpy(),
numpy.array([2, 3, 1, 5, 0]))
def test_nested_list_fields_are_padded_correctly(self):
nested_field1 = ListField(
[LabelField(c) for c in ['a', 'b', 'c', 'd', 'e']])
nested_field2 = ListField(
[LabelField(c) for c in ['f', 'g', 'h', 'i', 'j', 'k']])
list_field = ListField(
[nested_field1.empty_field(), nested_field1, nested_field2])
list_field.index(self.vocab)
padding_lengths = list_field.get_padding_lengths()
assert padding_lengths == {'num_fields': 3, 'list_num_fields': 6}
tensor = list_field.as_tensor(padding_lengths).detach().cpu().numpy()
numpy.testing.assert_almost_equal(
tensor, [[-1, -1, -1, -1, -1, -1], [0, 1, 2, 3, 4, -1],
[5, 6, 7, 8, 9, 10]])
def test_fields_can_pad_to_greater_than_max_length(self):
list_field = ListField([self.field1, self.field2, self.field3])
list_field.index(self.vocab)
padding_lengths = list_field.get_padding_lengths()
padding_lengths["list_num_tokens"] = 7
padding_lengths["num_fields"] = 5
tensor_dict = list_field.as_tensor(padding_lengths)
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][0].detach().cpu().numpy(),
numpy.array([2, 3, 4, 5, 0, 0, 0]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][1].detach().cpu().numpy(),
numpy.array([2, 3, 4, 1, 5, 0, 0]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][2].detach().cpu().numpy(),
numpy.array([2, 3, 1, 5, 0, 0, 0]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][3].detach().cpu().numpy(),
numpy.array([0, 0, 0, 0, 0, 0, 0]))
numpy.testing.assert_array_almost_equal(
tensor_dict["words"][4].detach().cpu().numpy(),
numpy.array([0, 0, 0, 0, 0, 0, 0]))
def test_as_tensor_can_handle_multiple_token_indexers(self):
# pylint: disable=protected-access
self.field1._token_indexers = self.words_and_characters_indexers
self.field2._token_indexers = self.words_and_characters_indexers
self.field3._token_indexers = self.words_and_characters_indexers
list_field = ListField([self.field1, self.field2, self.field3])
list_field.index(self.vocab)
padding_lengths = list_field.get_padding_lengths()
tensor_dict = list_field.as_tensor(padding_lengths)
words = tensor_dict["words"].detach().cpu().numpy()
characters = tensor_dict["characters"].detach().cpu().numpy()
numpy.testing.assert_array_almost_equal(
words,
numpy.array([[2, 3, 4, 5, 0], [2, 3, 4, 1, 5], [2, 3, 1, 5, 0]]))
numpy.testing.assert_array_almost_equal(
characters[0],
numpy.array([[5, 1, 1, 2, 0, 0, 0, 0, 0],
[1, 2, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[2, 3, 4, 5, 3, 4, 6, 3, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0]]))
numpy.testing.assert_array_almost_equal(
characters[1],
numpy.array([[5, 1, 1, 2, 0, 0, 0, 0, 0],
[1, 2, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 1, 1, 1, 3, 1, 3, 4, 5],
[2, 3, 4, 5, 3, 4, 6, 3, 0]]))
numpy.testing.assert_array_almost_equal(
characters[2],
numpy.array([[5, 1, 1, 2, 0, 0, 0, 0, 0],
[1, 2, 0, 0, 0, 0, 0, 0, 0],
[1, 4, 1, 5, 1, 3, 1, 0, 0],
[2, 3, 4, 5, 3, 4, 6, 3, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0]]))
def test_as_tensor_can_handle_multiple_token_indexers_and_empty_fields(
self):
# pylint: disable=protected-access
self.field1._token_indexers = self.words_and_characters_indexers
self.field2._token_indexers = self.words_and_characters_indexers
self.field3._token_indexers = self.words_and_characters_indexers
list_field = ListField(
[self.field1.empty_field(), self.field1, self.field2])
list_field.index(self.vocab)
padding_lengths = list_field.get_padding_lengths()
tensor_dict = list_field.as_tensor(padding_lengths)
words = tensor_dict["words"].detach().cpu().numpy()
characters = tensor_dict["characters"].detach().cpu().numpy()
numpy.testing.assert_array_almost_equal(
words,
numpy.array([[0, 0, 0, 0, 0], [2, 3, 4, 5, 0], [2, 3, 4, 1, 5]]))
numpy.testing.assert_array_almost_equal(characters[0],
numpy.zeros([5, 9]))
numpy.testing.assert_array_almost_equal(
characters[1],
numpy.array([[5, 1, 1, 2, 0, 0, 0, 0, 0],
[1, 2, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[2, 3, 4, 5, 3, 4, 6, 3, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0]]))
numpy.testing.assert_array_almost_equal(
characters[2],
numpy.array([[5, 1, 1, 2, 0, 0, 0, 0, 0],
[1, 2, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 1, 1, 1, 3, 1, 3, 4, 5],
[2, 3, 4, 5, 3, 4, 6, 3, 0]]))
def test_printing_doesnt_crash(self):
list_field = ListField([self.field1, self.field2])
print(list_field)
def setUp(self):
super(SpanBasedF1Test, self).setUp()
vocab = Vocabulary()
vocab.add_token_to_namespace("O", "tags")
vocab.add_token_to_namespace("B-ARG1", "tags")
vocab.add_token_to_namespace("I-ARG1", "tags")
vocab.add_token_to_namespace("B-ARG2", "tags")
vocab.add_token_to_namespace("I-ARG2", "tags")
vocab.add_token_to_namespace("B-V", "tags")
vocab.add_token_to_namespace("I-V", "tags")
vocab.add_token_to_namespace("U-ARG1", "tags")
vocab.add_token_to_namespace("U-ARG2", "tags")
vocab.add_token_to_namespace("B-C-ARG1", "tags")
vocab.add_token_to_namespace("I-C-ARG1", "tags")
vocab.add_token_to_namespace("B-ARGM-ADJ", "tags")
vocab.add_token_to_namespace("I-ARGM-ADJ", "tags")
self.vocab = vocab
class KnowledgeGraphFieldTest(AllenNlpTestCase):
def setUp(self):
self.tokenizer = WordTokenizer(SpacyWordSplitter(pos_tags=True))
self.utterance = self.tokenizer.tokenize("where is mersin?")
self.token_indexers = {"tokens": SingleIdTokenIndexer("tokens")}
json = {
'question': self.utterance,
'columns': ['Name in English', 'Location in English'],
'cells': [['Paradeniz', 'Mersin'], ['Lake Gala', 'Edirne']]
}
self.graph = TableQuestionKnowledgeGraph.read_from_json(json)
self.vocab = Vocabulary()
self.name_index = self.vocab.add_token_to_namespace("name",
namespace='tokens')
self.in_index = self.vocab.add_token_to_namespace("in",
namespace='tokens')
self.english_index = self.vocab.add_token_to_namespace(
"english", namespace='tokens')
self.location_index = self.vocab.add_token_to_namespace(
"location", namespace='tokens')
self.paradeniz_index = self.vocab.add_token_to_namespace(
"paradeniz", namespace='tokens')
self.mersin_index = self.vocab.add_token_to_namespace(
"mersin", namespace='tokens')
self.lake_index = self.vocab.add_token_to_namespace("lake",
namespace='tokens')
self.gala_index = self.vocab.add_token_to_namespace("gala",
namespace='tokens')
self.negative_one_index = self.vocab.add_token_to_namespace(
"-1", namespace='tokens')
self.zero_index = self.vocab.add_token_to_namespace("0",
namespace='tokens')
self.one_index = self.vocab.add_token_to_namespace("1",
namespace='tokens')
self.oov_index = self.vocab.get_token_index('random OOV string',
namespace='tokens')
self.edirne_index = self.oov_index
self.field = KnowledgeGraphField(self.graph, self.utterance,
self.token_indexers, self.tokenizer)
super(KnowledgeGraphFieldTest, self).setUp()
def test_count_vocab_items(self):
namespace_token_counts = defaultdict(lambda: defaultdict(int))
self.field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["tokens"] == {
'-1': 1,
'0': 1,
'1': 1,
'name': 1,
'in': 2,
'english': 2,
'location': 1,
'paradeniz': 1,
'mersin': 1,
'lake': 1,
'gala': 1,
'edirne': 1,
}
def test_index_converts_field_correctly(self):
# pylint: disable=protected-access
self.field.index(self.vocab)
assert self.field._indexed_entity_texts.keys() == {'tokens'}
# Note that these are sorted by their _identifiers_, not their cell text, so the
# `fb:row.rows` show up after the `fb:cells`.
expected_array = [[self.negative_one_index], [self.zero_index],
[self.one_index], [self.edirne_index],
[self.lake_index, self.gala_index],
[self.mersin_index], [self.paradeniz_index],
[
self.location_index, self.in_index,
self.english_index
],
[self.name_index, self.in_index, self.english_index]]
assert self.field._indexed_entity_texts['tokens'] == expected_array
def test_get_padding_lengths_raises_if_not_indexed(self):
with pytest.raises(AssertionError):
self.field.get_padding_lengths()
def test_padding_lengths_are_computed_correctly(self):
# pylint: disable=protected-access
self.field.index(self.vocab)
assert self.field.get_padding_lengths() == {
'num_entities': 9,
'num_entity_tokens': 3,
'num_utterance_tokens': 4
}
self.field._token_indexers[
'token_characters'] = TokenCharactersIndexer()
self.field.index(self.vocab)
assert self.field.get_padding_lengths() == {
'num_entities': 9,
'num_entity_tokens': 3,
'num_utterance_tokens': 4,
'num_token_characters': 9
}
def test_as_tensor_produces_correct_output(self):
self.field.index(self.vocab)
padding_lengths = self.field.get_padding_lengths()
padding_lengths['num_utterance_tokens'] += 1
padding_lengths['num_entities'] += 1
tensor_dict = self.field.as_tensor(padding_lengths)
assert tensor_dict.keys() == {'text', 'linking'}
expected_text_tensor = [
[self.negative_one_index, 0, 0], [self.zero_index, 0, 0],
[self.one_index, 0, 0], [self.edirne_index, 0, 0],
[self.lake_index, self.gala_index, 0], [self.mersin_index, 0, 0],
[self.paradeniz_index, 0, 0],
[self.location_index, self.in_index, self.english_index],
[self.name_index, self.in_index, self.english_index], [0, 0, 0]
]
assert_almost_equal(
tensor_dict['text']['tokens'].detach().cpu().numpy(),
expected_text_tensor)
linking_tensor = tensor_dict['linking'].detach().cpu().numpy()
expected_linking_tensor = [
[
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # -1, "where"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # -1, "is"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # -1, "mersin"
[0, 0, 0, 0, 0, -1, 0, 0, 0, 0]
], # -1, "?"
[
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0, "where"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0, "is"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0, "mersin"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # 0, "?"
[
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 1, "where"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 1, "is"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 1, "mersin"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # 1, "?"
[
[0, 0, 0, 0, 0, .2, 0, 0, 0, 0], # fb:cell.edirne, "where"
[0, 0, 0, 0, 0, -1.5, 0, 0, 0, 0], # fb:cell.edirne, "is"
[0, 0, 0, 0, 0, .1666, 0, 0, 0, 0], # fb:cell.edirne, "mersin"
[0, 0, 0, 0, 0, -5, 0, 0, 0, 0], # fb:cell.edirne, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:cell.edirne, padding
[
[0, 0, 0, 0, 0, -.6, 0, 0, 0, 0], # fb:cell.lake_gala, "where"
[0, 0, 0, 0, 0, -3.5, 0, 0, 0, 0], # fb:cell.lake_gala, "is"
[0, 0, 0, 0, 0, -.3333, 0, 0, 0,
0], # fb:cell.lake_gala, "mersin"
[0, 0, 0, 0, 0, -8, 0, 0, 0, 0], # fb:cell.lake_gala, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:cell.lake_gala, padding
[
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # fb:cell.mersin, "where"
[0, 0, 0, 0, 0, -1.5, 0, 0, 0, 0], # fb:cell.mersin, "is"
[0, 1, 1, 1, 1, 1, 0, 0, 1, 1], # fb:cell.mersin, "mersin"
[0, 0, 0, 0, 0, -5, 0, 0, 0, 0], # fb:cell.mersin, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:cell.mersin, padding
[
[0, 0, 0, 0, 0, -.6, 0, 0, 0, 0], # fb:cell.paradeniz, "where"
[0, 0, 0, 0, 0, -3, 0, 0, 0, 0], # fb:cell.paradeniz, "is"
[0, 0, 0, 0, 0, -.1666, 0, 0, 0,
0], # fb:cell.paradeniz, "mersin"
[0, 0, 0, 0, 0, -8, 0, 0, 0, 0], # fb:cell.paradeniz, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:cell.paradeniz, padding
[
[0, 0, 0, 0, 0, -2.6, 0, 0, 0,
0], # fb:row.row.name_in_english, "where"
[0, 0, 0, 0, 0, -7.5, 0, 0, 0,
0], # fb:row.row.name_in_english, "is"
[0, 0, 0, 0, 0, -1.8333, 1, 1, 0,
0], # fb:row.row.name_in_english, "mersin"
[0, 0, 0, 0, 0, -18, 0, 0, 0,
0], # fb:row.row.name_in_english, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:row.row.name_in_english, padding
[
[0, 0, 0, 0, 0, -1.6, 0, 0, 0,
0], # fb:row.row.location_in_english, "where"
[0, 0, 0, 0, 0, -5.5, 0, 0, 0,
0], # fb:row.row.location_in_english, "is"
[0, 0, 0, 0, 0, -1, 0, 0, 0,
0], # fb:row.row.location_in_english, "mersin"
[0, 0, 0, 0, 0, -14, 0, 0, 0,
0], # fb:row.row.location_in_english, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
], # fb:row.row.location_in_english, padding
[
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # padding, "where"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # padding, "is"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # padding, "mersin"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], # padding, "?"
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
]
] # padding, padding
for entity_index, entity_features in enumerate(
expected_linking_tensor):
for question_index, feature_vector in enumerate(entity_features):
assert_almost_equal(linking_tensor[entity_index,
question_index],
feature_vector,
decimal=4,
err_msg=f"{entity_index} {question_index}")
def test_lemma_feature_extractor(self):
# pylint: disable=protected-access
utterance = self.tokenizer.tokenize("Names in English")
field = KnowledgeGraphField(self.graph, self.utterance,
self.token_indexers, self.tokenizer)
entity = 'fb:row.row.name_in_english'
lemma_feature = field._contains_lemma_match(
entity, field._entity_text_map[entity], utterance[0], 0, utterance)
assert lemma_feature == 1
def test_span_overlap_fraction(self):
# pylint: disable=protected-access
utterance = self.tokenizer.tokenize(
"what is the name in english of mersin?")
field = KnowledgeGraphField(self.graph, self.utterance,
self.token_indexers, self.tokenizer)
entity = 'fb:row.row.name_in_english'
entity_text = field._entity_text_map[entity]
feature_values = [
field._span_overlap_fraction(entity, entity_text, token, i,
utterance)
for i, token in enumerate(utterance)
]
assert feature_values == [0, 0, 0, 1, 2 / 3, 1 / 3, 0, 0, 0]
def test_batch_tensors(self):
self.field.index(self.vocab)
padding_lengths = self.field.get_padding_lengths()
tensor_dict1 = self.field.as_tensor(padding_lengths)
tensor_dict2 = self.field.as_tensor(padding_lengths)
batched_tensor_dict = self.field.batch_tensors(
[tensor_dict1, tensor_dict2])
assert batched_tensor_dict.keys() == {'text', 'linking'}
expected_single_tensor = [
[self.negative_one_index, 0, 0], [self.zero_index, 0, 0],
[self.one_index, 0, 0], [self.edirne_index, 0, 0],
[self.lake_index, self.gala_index, 0], [self.mersin_index, 0, 0],
[self.paradeniz_index, 0, 0],
[self.location_index, self.in_index, self.english_index],
[self.name_index, self.in_index, self.english_index]
]
expected_batched_tensor = [
expected_single_tensor, expected_single_tensor
]
assert_almost_equal(
batched_tensor_dict['text']['tokens'].detach().cpu().numpy(),
expected_batched_tensor)
expected_linking_tensor = torch.stack(
[tensor_dict1['linking'], tensor_dict2['linking']])
assert_almost_equal(
batched_tensor_dict['linking'].detach().cpu().numpy(),
expected_linking_tensor.detach().cpu().numpy())
class TestBasicTextFieldEmbedder(AllenNlpTestCase):
def setUp(self):
super(TestBasicTextFieldEmbedder, self).setUp()
self.vocab = Vocabulary()
self.vocab.add_token_to_namespace("1")
self.vocab.add_token_to_namespace("2")
self.vocab.add_token_to_namespace("3")
self.vocab.add_token_to_namespace("4")
params = Params({
"words1": {
"type": "embedding",
"embedding_dim": 2
},
"words2": {
"type": "embedding",
"embedding_dim": 5
},
"words3": {
"type": "embedding",
"embedding_dim": 3
}
})
self.token_embedder = BasicTextFieldEmbedder.from_params(
vocab=self.vocab, params=params)
self.inputs = {
"words1": torch.LongTensor([[0, 2, 3, 5]]),
"words2": torch.LongTensor([[1, 4, 3, 2]]),
"words3": torch.LongTensor([[1, 5, 1, 2]])
}
def test_get_output_dim_aggregates_dimension_from_each_embedding(self):
assert self.token_embedder.get_output_dim() == 10
def test_forward_asserts_input_field_match(self):
self.inputs['words4'] = self.inputs['words3']
del self.inputs['words3']
with pytest.raises(ConfigurationError):
self.token_embedder(self.inputs)
self.inputs['words3'] = self.inputs['words4']
del self.inputs['words4']
def test_forward_concats_resultant_embeddings(self):
assert self.token_embedder(self.inputs).size() == (1, 4, 10)
def test_forward_works_on_higher_order_input(self):
params = Params({
"words": {
"type": "embedding",
"num_embeddings": 20,
"embedding_dim": 2,
},
"characters": {
"type": "character_encoding",
"embedding": {
"embedding_dim": 4,
"num_embeddings": 15,
},
"encoder": {
"type": "cnn",
"embedding_dim": 4,
"num_filters": 10,
"ngram_filter_sizes": [3],
},
}
})
token_embedder = BasicTextFieldEmbedder.from_params(vocab=self.vocab,
params=params)
inputs = {
'words': (torch.rand(3, 4, 5, 6) * 20).long(),
'characters': (torch.rand(3, 4, 5, 6, 7) * 15).long(),
}
assert token_embedder(inputs,
num_wrapping_dims=2).size() == (3, 4, 5, 6, 12)
def test_forward_runs_with_non_bijective_mapping(self):
elmo_fixtures_path = self.FIXTURES_ROOT / 'elmo'
options_file = str(elmo_fixtures_path / 'options.json')
weight_file = str(elmo_fixtures_path / 'lm_weights.hdf5')
params = Params({
"words": {
"type": "embedding",
"num_embeddings": 20,
"embedding_dim": 2,
},
"elmo": {
"type": "elmo_token_embedder",
"options_file": options_file,
"weight_file": weight_file
},
"embedder_to_indexer_map": {
"words": ["words"],
"elmo": ["elmo", "words"]
}
})
token_embedder = BasicTextFieldEmbedder.from_params(self.vocab, params)
inputs = {
'words': (torch.rand(3, 6) * 20).long(),
'elmo': (torch.rand(3, 6, 50) * 15).long(),
}
token_embedder(inputs)
def test_old_from_params_new_from_params(self):
old_params = Params({
"words1": {
"type": "embedding",
"embedding_dim": 2
},
"words2": {
"type": "embedding",
"embedding_dim": 5
},
"words3": {
"type": "embedding",
"embedding_dim": 3
}
})
with pytest.warns(DeprecationWarning):
BasicTextFieldEmbedder.from_params(params=old_params,
vocab=self.vocab)
new_params = Params({
"token_embedders": {
"words1": {
"type": "embedding",
"embedding_dim": 2
},
"words2": {
"type": "embedding",
"embedding_dim": 5
},
"words3": {
"type": "embedding",
"embedding_dim": 3
}
}
})
token_embedder = BasicTextFieldEmbedder.from_params(params=new_params,
vocab=self.vocab)
assert token_embedder(self.inputs).size() == (1, 4, 10)
class TestProductionRuleField(AllenNlpTestCase):
def setUp(self):
super(TestProductionRuleField, self).setUp()
self.vocab = Vocabulary()
self.s_rule_index = self.vocab.add_token_to_namespace("S -> [NP, VP]", namespace='rule_labels')
self.np_index = self.vocab.add_token_to_namespace("NP -> test", namespace='rule_labels')
def test_field_counts_vocab_items_correctly(self):
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
namespace_token_counts = defaultdict(lambda: defaultdict(int))
field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["rule_labels"]["S -> [NP, VP]"] == 1
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=False)
namespace_token_counts = defaultdict(lambda: defaultdict(int))
field.count_vocab_items(namespace_token_counts)
assert namespace_token_counts["rule_labels"]["S -> [NP, VP]"] == 0
def test_index_converts_field_correctly(self):
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
field.index(self.vocab)
assert field._rule_id == self.s_rule_index
def test_padding_lengths_are_computed_correctly(self):
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
field.index(self.vocab)
assert field.get_padding_lengths() == {}
def test_as_tensor_produces_correct_output(self):
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
field.index(self.vocab)
tensor_tuple = field.as_tensor(field.get_padding_lengths())
assert isinstance(tensor_tuple, tuple)
assert len(tensor_tuple) == 3
assert tensor_tuple[0] == 'S -> [NP, VP]'
assert tensor_tuple[1] is True
assert_almost_equal(tensor_tuple[2].detach().cpu().numpy(), [self.s_rule_index])
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=False)
field.index(self.vocab)
tensor_tuple = field.as_tensor(field.get_padding_lengths())
assert isinstance(tensor_tuple, tuple)
assert len(tensor_tuple) == 3
assert tensor_tuple[0] == 'S -> [NP, VP]'
assert tensor_tuple[1] is False
assert tensor_tuple[2] is None
def test_batch_tensors_does_not_modify_list(self):
field = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
tensor_dict1 = field.as_tensor(padding_lengths)
field = ProductionRuleField('NP -> test', is_global_rule=True)
field.index(self.vocab)
padding_lengths = field.get_padding_lengths()
tensor_dict2 = field.as_tensor(padding_lengths)
tensor_list = [tensor_dict1, tensor_dict2]
assert field.batch_tensors(tensor_list) == tensor_list
def test_doubly_nested_field_works(self):
field1 = ProductionRuleField('S -> [NP, VP]', is_global_rule=True)
field2 = ProductionRuleField('NP -> test', is_global_rule=True)
field3 = ProductionRuleField('VP -> eat', is_global_rule=False)
list_field = ListField([ListField([field1, field2, field3]),
ListField([field1, field2])])
list_field.index(self.vocab)
padding_lengths = list_field.get_padding_lengths()
tensors = list_field.as_tensor(padding_lengths)
assert isinstance(tensors, list)
assert len(tensors) == 2
assert isinstance(tensors[0], list)
assert len(tensors[0]) == 3
assert isinstance(tensors[1], list)
assert len(tensors[1]) == 3
tensor_tuple = tensors[0][0]
assert tensor_tuple[0] == 'S -> [NP, VP]'
assert tensor_tuple[1] is True
assert_almost_equal(tensor_tuple[2].detach().cpu().numpy(), [self.s_rule_index])
tensor_tuple = tensors[0][1]
assert tensor_tuple[0] == 'NP -> test'
assert tensor_tuple[1] is True
assert_almost_equal(tensor_tuple[2].detach().cpu().numpy(), [self.np_index])
tensor_tuple = tensors[0][2]
assert tensor_tuple[0] == 'VP -> eat'
assert tensor_tuple[1] is False
assert tensor_tuple[2] is None
tensor_tuple = tensors[1][0]
assert tensor_tuple[0] == 'S -> [NP, VP]'
assert tensor_tuple[1] is True
assert_almost_equal(tensor_tuple[2].detach().cpu().numpy(), [self.s_rule_index])
tensor_tuple = tensors[1][1]
assert tensor_tuple[0] == 'NP -> test'
assert tensor_tuple[1] is True
assert_almost_equal(tensor_tuple[2].detach().cpu().numpy(), [self.np_index])
# This item was just padding.
tensor_tuple = tensors[1][2]
assert tensor_tuple[0] == ''
assert tensor_tuple[1] is False
assert tensor_tuple[2] is None