(self, words_list=None, update=True)
| 38 | self.tokenizer = Tokenizer(dict_path, do_lower_case=True) |
| 39 | |
| 40 | def init(self, words_list=None, update=True): |
| 41 | if words_list!=None: |
| 42 | token_ids_list, segment_ids_list = [], [] |
| 43 | for words in words_list: |
| 44 | token_ids, segment_ids = self.tokenizer.encode(words) |
| 45 | token_ids_list.append(token_ids) |
| 46 | segment_ids_list.append(segment_ids) |
| 47 | token_ids_list = sequence_padding(token_ids_list) |
| 48 | segment_ids_list = sequence_padding(segment_ids_list) |
| 49 | self.words_list_pre = self.encoder.predict([token_ids_list, segment_ids_list]) |
| 50 | self.words_list_pre = self._normalize(self.words_list_pre) |
| 51 | return self |
| 52 | |
| 53 | def _predict(self, words): |
| 54 | with self.session.as_default(): |
nothing calls this directly
no test coverage detected