def _get_sentence_positions_to_span(doc: Document, start_token: int,
end_token: int, wrt_span):
_, _, sent_idx = wrt_span
return [
sent_idx - doc.get_token_sent_idx(idx)
for idx in range(start_token, end_token)
]
def _get_attention_features(self, doc: Document, sent: Sentence,
t1_idx: int, t2_idx: int) -> dict:
attention_features = {}
for namespace, idx in zip(['head', 'dep'], [t1_idx, t2_idx]):
position_features = self.token_position_fe.extract_features_from_doc(
doc, sent.start_token, sent.end_token,
(idx, idx + 1, doc.get_token_sent_idx(idx)))
attention_features.update(namespaced(position_features, namespace))
attention_features.update(
self._get_arc_features(doc, sent, t1_idx, t2_idx,
self.attention_features_converters,
"attention"))
return attention_features