class FastTextEmbedding(WordEmbedding):
"""FastText word embedding"""
def __init__(self,
sentences=None,
corpus_file_path: str = None,
dim: int = 100,
saved_model_path: str = None,
save_path=None) -> None:
"""Constructor for FastTextEmbedding classes (RAII)"""
super().__init__()
if sentences is not None:
self._impl = FastText(size=dim)
self._impl.build_vocab(sentences=sentences)
self._impl.train(sentences=sentences,
total_words=self._impl.corpus_total_words,
epochs=10)
elif corpus_file_path is not None:
self._impl = FastText(size=dim)
self._impl.build_vocab(corpus_file=corpus_file_path)
self._impl.train(corpus_file=corpus_file_path,
total_words=self._impl.corpus_total_words,
epochs=10)
elif saved_model_path is not None:
# load from saved FastText embedding file
self._impl = load_facebook_model(saved_model_path)
else:
raise AssertionError(
'sentences or corpus_file_path should be given as not None')
self.vocab = self._impl.wv.index2word
self.vocab_size = len(self.vocab)
self.embedding_matrix = []
for word in self.vocab:
self.embedding_matrix.append(self._impl.wv[word])
def word2vec(self, words):
return [self._impl.wv.get_vector(w) for w in words]
def vec2word(self, vectors):
return [self._impl.similar_by_vector(v, topn=1)[0] for v in vectors]
def get_embedding_matrix(self) -> np.ndarray:
return np.array(self.embedding_matrix)
