def test_load_embedding_has_all_words(self, instances, embedding_type):
single_instance = instances["single_instance"]
MAX_NUM_WORDS = 100
vocab = Vocab(
instances=single_instance,
max_num_tokens=MAX_NUM_WORDS,
embedding_type=embedding_type,
)
vocab.build_vocab()
embedding = vocab.load_embedding()
assert embedding.size(0) == vocab.get_vocab_len()
def test_random_embeddinng_has_2dimensions(self, instances):
single_instance = instances["single_instance"]
MAX_NUM_WORDS = 100
vocab = Vocab(
instances=single_instance,
max_num_tokens=MAX_NUM_WORDS,
embedding_type=None,
embedding_dimension=300,
)
vocab.build_vocab()
embeddings = vocab.load_embedding()
assert embeddings.ndimension() == 2
