def test_token_selection(self):
prediction = {
'input_ids': np.array([0, 1, 2, 3, 4]),
'probabilities': np.array([1.0, 0.5, 0.2, 0.5, 0.3]),
'column_ids': np.array([0, 1, 2, 3, 4]),
'row_ids': np.array([0, 1, 1, 2, 2]),
'segment_ids': np.array([0, 1, 1, 1, 1]),
}
answers = prediction_utils._get_token_answers(
prediction,
cell_classification_threshold=0.49999,
)
logging.info(answers)
self.assertEqual(answers, [
prediction_utils.TokenAnswer(column_index=0,
row_index=0,
begin_token_index=0,
end_token_index=1,
token_ids=[1],
score=0.5),
prediction_utils.TokenAnswer(column_index=2,
row_index=1,
begin_token_index=0,
end_token_index=1,
token_ids=[3],
score=0.5),
])
def test_span_selection_with_row_boundary(self):
prediction = {
'input_ids': np.array([1, 2, 3, 4]),
'span_indexes': np.array([[1, 1], [1, 2], [2, 1]]),
'span_logits': np.array([-100.0, 10.0, 5.0]),
'column_ids': np.array([2, 2, 2, 2]),
'row_ids': np.array([1, 2, 2, 2]),
}
answers = prediction_utils._get_token_answers(
prediction, cell_classification_threshold=0.5)
self.assertEqual(answers, [
prediction_utils.TokenAnswer(
column_index=1,
row_index=1,
begin_token_index=0,
end_token_index=2,
token_ids=[2, 3],
score=10.0,
)
])
def test_end_to_end(self, span_prediction):
estimator = self._create_estimator(span_prediction=span_prediction)
def _input_fn(params):
return table_dataset_test_utils.create_random_dataset(
num_examples=params['batch_size'] * 2,
batch_size=params['batch_size'],
repeat=False,
generator_kwargs=self._generator_kwargs())
result = estimator.predict(_input_fn)
num_examples = 0
for prediction in result:
if span_prediction != _SpanPredictionMode.NONE:
self.assertIn('span_logits', prediction)
self.assertIn('span_indexes', prediction)
logging.info('prediction: %s', prediction)
_ = prediction_utils._get_token_answers(
prediction,
cell_classification_threshold=0.5,
)
num_examples += 1
self.assertEqual(num_examples, _BATCH_SIZE * 2)
