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