squad_lib.py

# Copyright 2024 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

"""Library to process data for SQuAD 1.1 and SQuAD 2.0."""
# pylint: disable=g-bad-import-order
import collections
import copy
import json
import math
import os

import six

from absl import logging
import tensorflow as tf, tf_keras

from official.nlp.tools import tokenization


class SquadExample(object):
  """A single training/test example for simple sequence classification.

  For examples without an answer, the start and end position are -1.

  Attributes:
    qas_id: ID of the question-answer pair.
    question_text: Original text for the question.
    doc_tokens: The list of tokens in the context obtained by splitting on
      whitespace only.
    orig_answer_text: Original text for the answer.
    start_position: Starting index of the answer in `doc_tokens`.
    end_position: Ending index of the answer in `doc_tokens`.
    is_impossible: Whether the question is impossible to answer given the
      context. Only used in SQuAD 2.0.
  """

  def __init__(self,
               qas_id,
               question_text,
               doc_tokens,
               orig_answer_text=None,
               start_position=None,
               end_position=None,
               is_impossible=False):
    self.qas_id = qas_id
    self.question_text = question_text
    self.doc_tokens = doc_tokens
    self.orig_answer_text = orig_answer_text
    self.start_position = start_position
    self.end_position = end_position
    self.is_impossible = is_impossible

  def __str__(self):
    return self.__repr__()

  def __repr__(self):
    s = ""
    s += "qas_id: %s" % (tokenization.printable_text(self.qas_id))
    s += ", question_text: %s" % (
        tokenization.printable_text(self.question_text))
    s += ", doc_tokens: [%s]" % (" ".join(self.doc_tokens))
    if self.start_position:
      s += ", start_position: %d" % (self.start_position)
    if self.start_position:
      s += ", end_position: %d" % (self.end_position)
    if self.start_position:
      s += ", is_impossible: %r" % (self.is_impossible)
    return s


class InputFeatures(object):
  """A single set of features of data."""

  def __init__(self,
               unique_id,
               example_index,
               doc_span_index,
               tokens,
               token_to_orig_map,
               token_is_max_context,
               input_ids,
               input_mask,
               segment_ids,
               paragraph_mask=None,
               class_index=None,
               start_position=None,
               end_position=None,
               is_impossible=None):
    self.unique_id = unique_id
    self.example_index = example_index
    self.doc_span_index = doc_span_index
    self.tokens = tokens
    self.token_to_orig_map = token_to_orig_map
    self.token_is_max_context = token_is_max_context
    self.input_ids = input_ids
    self.input_mask = input_mask
    self.segment_ids = segment_ids
    self.start_position = start_position
    self.end_position = end_position
    self.is_impossible = is_impossible
    self.paragraph_mask = paragraph_mask
    self.class_index = class_index


class FeatureWriter(object):
  """Writes InputFeature to TF example file."""

  def __init__(self, filename, is_training):
    self.filename = filename
    self.is_training = is_training
    self.num_features = 0
    tf.io.gfile.makedirs(os.path.dirname(filename))
    self._writer = tf.io.TFRecordWriter(filename)

  def process_feature(self, feature):
    """Write a InputFeature to the TFRecordWriter as a tf.train.Example."""
    self.num_features += 1

    def create_int_feature(values):
      feature = tf.train.Feature(
          int64_list=tf.train.Int64List(value=list(values)))
      return feature

    features = collections.OrderedDict()
    features["unique_ids"] = create_int_feature([feature.unique_id])
    features["input_ids"] = create_int_feature(feature.input_ids)
    features["input_mask"] = create_int_feature(feature.input_mask)
    features["segment_ids"] = create_int_feature(feature.segment_ids)

    if feature.paragraph_mask is not None:
      features["paragraph_mask"] = create_int_feature(feature.paragraph_mask)
    if feature.class_index is not None:
      features["class_index"] = create_int_feature([feature.class_index])

    if self.is_training:
      features["start_positions"] = create_int_feature([feature.start_position])
      features["end_positions"] = create_int_feature([feature.end_position])
      impossible = 0
      if feature.is_impossible:
        impossible = 1
      features["is_impossible"] = create_int_feature([impossible])

    tf_example = tf.train.Example(features=tf.train.Features(feature=features))
    self._writer.write(tf_example.SerializeToString())

  def close(self):
    self._writer.close()


def read_squad_examples(input_file, is_training,
                        version_2_with_negative,
                        translated_input_folder=None):
  """Read a SQuAD json file into a list of SquadExample."""
  with tf.io.gfile.GFile(input_file, "r") as reader:
    input_data = json.load(reader)["data"]

  if translated_input_folder is not None:
    translated_files = tf.io.gfile.glob(
        os.path.join(translated_input_folder, "*.json"))
    for file in translated_files:
      with tf.io.gfile.GFile(file, "r") as reader:
        input_data.extend(json.load(reader)["data"])

  def is_whitespace(c):
    if c == " " or c == "\t" or c == "\r" or c == "\n" or ord(c) == 0x202F:
      return True
    return False

  examples = []
  for entry in input_data:
    for paragraph in entry["paragraphs"]:
      paragraph_text = paragraph["context"]
      doc_tokens = []
      char_to_word_offset = []
      prev_is_whitespace = True
      for c in paragraph_text:
        if is_whitespace(c):
          prev_is_whitespace = True
        else:
          if prev_is_whitespace:
            doc_tokens.append(c)
          else:
            doc_tokens[-1] += c
          prev_is_whitespace = False
        char_to_word_offset.append(len(doc_tokens) - 1)

      for qa in paragraph["qas"]:
        qas_id = qa["id"]
        question_text = qa["question"]
        start_position = None
        end_position = None
        orig_answer_text = None
        is_impossible = False
        if is_training:

          if version_2_with_negative:
            is_impossible = qa["is_impossible"]
          if (len(qa["answers"]) != 1) and (not is_impossible):
            raise ValueError(
                "For training, each question should have exactly 1 answer.")
          if not is_impossible:
            answer = qa["answers"][0]
            orig_answer_text = answer["text"]
            answer_offset = answer["answer_start"]
            answer_length = len(orig_answer_text)
            start_position = char_to_word_offset[answer_offset]
            end_position = char_to_word_offset[answer_offset + answer_length -
                                               1]
            # Only add answers where the text can be exactly recovered from the
            # document. If this CAN'T happen it's likely due to weird Unicode
            # stuff so we will just skip the example.
            #
            # Note that this means for training mode, every example is NOT
            # guaranteed to be preserved.
            actual_text = " ".join(doc_tokens[start_position:(end_position +
                                                              1)])
            cleaned_answer_text = " ".join(
                tokenization.whitespace_tokenize(orig_answer_text))
            if actual_text.find(cleaned_answer_text) == -1:
              logging.warning("Could not find answer: '%s' vs. '%s'",
                              actual_text, cleaned_answer_text)
              continue
          else:
            start_position = -1
            end_position = -1
            orig_answer_text = ""

        example = SquadExample(
            qas_id=qas_id,
            question_text=question_text,
            doc_tokens=doc_tokens,
            orig_answer_text=orig_answer_text,
            start_position=start_position,
            end_position=end_position,
            is_impossible=is_impossible)
        examples.append(example)

  return examples


def convert_examples_to_features(examples,
                                 tokenizer,
                                 max_seq_length,
                                 doc_stride,
                                 max_query_length,
                                 is_training,
                                 output_fn,
                                 xlnet_format=False,
                                 batch_size=None):
  """Loads a data file into a list of `InputBatch`s."""

  base_id = 1000000000
  unique_id = base_id
  feature = None
  for (example_index, example) in enumerate(examples):
    query_tokens = tokenizer.tokenize(example.question_text)

    if len(query_tokens) > max_query_length:
      query_tokens = query_tokens[0:max_query_length]

    tok_to_orig_index = []
    orig_to_tok_index = []
    all_doc_tokens = []
    for (i, token) in enumerate(example.doc_tokens):
      orig_to_tok_index.append(len(all_doc_tokens))
      sub_tokens = tokenizer.tokenize(token)
      for sub_token in sub_tokens:
        tok_to_orig_index.append(i)
        all_doc_tokens.append(sub_token)

    tok_start_position = None
    tok_end_position = None
    if is_training and example.is_impossible:
      tok_start_position = -1
      tok_end_position = -1
    if is_training and not example.is_impossible:
      tok_start_position = orig_to_tok_index[example.start_position]
      if example.end_position < len(example.doc_tokens) - 1:
        tok_end_position = orig_to_tok_index[example.end_position + 1] - 1
      else:
        tok_end_position = len(all_doc_tokens) - 1
      (tok_start_position, tok_end_position) = _improve_answer_span(
          all_doc_tokens, tok_start_position, tok_end_position, tokenizer,
          example.orig_answer_text)

    # The -3 accounts for [CLS], [SEP] and [SEP]
    max_tokens_for_doc = max_seq_length - len(query_tokens) - 3

    # We can have documents that are longer than the maximum sequence length.
    # To deal with this we do a sliding window approach, where we take chunks
    # of the up to our max length with a stride of `doc_stride`.
    _DocSpan = collections.namedtuple(  # pylint: disable=invalid-name
        "DocSpan", ["start", "length"])
    doc_spans = []
    start_offset = 0
    while start_offset < len(all_doc_tokens):
      length = len(all_doc_tokens) - start_offset
      if length > max_tokens_for_doc:
        length = max_tokens_for_doc
      doc_spans.append(_DocSpan(start=start_offset, length=length))
      if start_offset + length == len(all_doc_tokens):
        break
      start_offset += min(length, doc_stride)

    for (doc_span_index, doc_span) in enumerate(doc_spans):
      tokens = []
      token_to_orig_map = {}
      token_is_max_context = {}
      segment_ids = []

      # Paragraph mask used in XLNet.
      # 1 represents paragraph and class tokens.
      # 0 represents query and other special tokens.
      paragraph_mask = []

      # pylint: disable=cell-var-from-loop
      def process_query(seg_q):
        for token in query_tokens:
          tokens.append(token)
          segment_ids.append(seg_q)
          paragraph_mask.append(0)
        tokens.append("[SEP]")
        segment_ids.append(seg_q)
        paragraph_mask.append(0)

      def process_paragraph(seg_p):
        for i in range(doc_span.length):
          split_token_index = doc_span.start + i
          token_to_orig_map[len(tokens)] = tok_to_orig_index[split_token_index]

          is_max_context = _check_is_max_context(doc_spans, doc_span_index,
                                                 split_token_index)
          token_is_max_context[len(tokens)] = is_max_context
          tokens.append(all_doc_tokens[split_token_index])
          segment_ids.append(seg_p)
          paragraph_mask.append(1)
        tokens.append("[SEP]")
        segment_ids.append(seg_p)
        paragraph_mask.append(0)

      def process_class(seg_class):
        class_index = len(segment_ids)
        tokens.append("[CLS]")
        segment_ids.append(seg_class)
        paragraph_mask.append(1)
        return class_index

      if xlnet_format:
        seg_p, seg_q, seg_class, seg_pad = 0, 1, 2, 3
        process_paragraph(seg_p)
        process_query(seg_q)
        class_index = process_class(seg_class)
      else:
        seg_p, seg_q, seg_class, seg_pad = 1, 0, 0, 0
        class_index = process_class(seg_class)
        process_query(seg_q)
        process_paragraph(seg_p)

      input_ids = tokenizer.convert_tokens_to_ids(tokens)

      # The mask has 1 for real tokens and 0 for padding tokens. Only real
      # tokens are attended to.
      input_mask = [1] * len(input_ids)

      # Zero-pad up to the sequence length.
      while len(input_ids) < max_seq_length:
        input_ids.append(0)
        input_mask.append(0)
        segment_ids.append(seg_pad)
        paragraph_mask.append(0)

      assert len(input_ids) == max_seq_length
      assert len(input_mask) == max_seq_length
      assert len(segment_ids) == max_seq_length
      assert len(paragraph_mask) == max_seq_length

      start_position = 0
      end_position = 0
      span_contains_answer = False

      if is_training and not example.is_impossible:
        # For training, if our document chunk does not contain an annotation
        # we throw it out, since there is nothing to predict.
        doc_start = doc_span.start
        doc_end = doc_span.start + doc_span.length - 1
        span_contains_answer = (tok_start_position >= doc_start and
                                tok_end_position <= doc_end)
        if span_contains_answer:
          doc_offset = 0 if xlnet_format else len(query_tokens) + 2
          start_position = tok_start_position - doc_start + doc_offset
          end_position = tok_end_position - doc_start + doc_offset

      if example_index < 20:
        logging.info("*** Example ***")
        logging.info("unique_id: %s", (unique_id))
        logging.info("example_index: %s", (example_index))
        logging.info("doc_span_index: %s", (doc_span_index))
        logging.info("tokens: %s",
                     " ".join([tokenization.printable_text(x) for x in tokens]))
        logging.info(
            "token_to_orig_map: %s", " ".join([
                "%d:%d" % (x, y) for (x, y) in six.iteritems(token_to_orig_map)
            ]))
        logging.info(
            "token_is_max_context: %s", " ".join([
                "%d:%s" % (x, y)
                for (x, y) in six.iteritems(token_is_max_context)
            ]))
        logging.info("input_ids: %s", " ".join([str(x) for x in input_ids]))
        logging.info("input_mask: %s", " ".join([str(x) for x in input_mask]))
        logging.info("segment_ids: %s", " ".join([str(x) for x in segment_ids]))
        logging.info("paragraph_mask: %s", " ".join(
            [str(x) for x in paragraph_mask]))
        logging.info("class_index: %d", class_index)
        if is_training:
          if span_contains_answer:
            answer_text = " ".join(tokens[start_position:(end_position + 1)])
            logging.info("start_position: %d", (start_position))
            logging.info("end_position: %d", (end_position))
            logging.info("answer: %s", tokenization.printable_text(answer_text))
          else:
            logging.info("document span doesn't contain answer")

      feature = InputFeatures(
          unique_id=unique_id,
          example_index=example_index,
          doc_span_index=doc_span_index,
          tokens=tokens,
          paragraph_mask=paragraph_mask,
          class_index=class_index,
          token_to_orig_map=token_to_orig_map,
          token_is_max_context=token_is_max_context,
          input_ids=input_ids,
          input_mask=input_mask,
          segment_ids=segment_ids,
          start_position=start_position,
          end_position=end_position,
          is_impossible=not span_contains_answer)

      # Run callback
      if is_training:
        output_fn(feature)
      else:
        output_fn(feature, is_padding=False)

      unique_id += 1

  if not is_training and feature:
    assert batch_size
    num_padding = 0
    num_examples = unique_id - base_id
    if unique_id % batch_size != 0:
      num_padding = batch_size - (num_examples % batch_size)
    logging.info("Adding padding examples to make sure no partial batch.")
    logging.info("Adds %d padding examples for inference.", num_padding)
    dummy_feature = copy.deepcopy(feature)
    for _ in range(num_padding):
      dummy_feature.unique_id = unique_id

      # Run callback
      output_fn(feature, is_padding=True)
      unique_id += 1
  return unique_id - base_id


def _improve_answer_span(doc_tokens, input_start, input_end, tokenizer,
                         orig_answer_text):
  """Returns tokenized answer spans that better match the annotated answer."""

  # The SQuAD annotations are character based. We first project them to
  # whitespace-tokenized words. But then after WordPiece tokenization, we can
  # often find a "better match". For example:
  #
  #   Question: What year was John Smith born?
  #   Context: The leader was John Smith (1895-1943).
  #   Answer: 1895
  #
  # The original whitespace-tokenized answer will be "(1895-1943).". However
  # after tokenization, our tokens will be "( 1895 - 1943 ) .". So we can match
  # the exact answer, 1895.
  #
  # However, this is not always possible. Consider the following:
  #
  #   Question: What country is the top exporter of electronics?
  #   Context: The Japanese electronics industry is the lagest in the world.
  #   Answer: Japan
  #
  # In this case, the annotator chose "Japan" as a character sub-span of
  # the word "Japanese". Since our WordPiece tokenizer does not split
  # "Japanese", we just use "Japanese" as the annotation. This is fairly rare
  # in SQuAD, but does happen.
  tok_answer_text = " ".join(tokenizer.tokenize(orig_answer_text))

  for new_start in range(input_start, input_end + 1):
    for new_end in range(input_end, new_start - 1, -1):
      text_span = " ".join(doc_tokens[new_start:(new_end + 1)])
      if text_span == tok_answer_text:
        return (new_start, new_end)

  return (input_start, input_end)


def _check_is_max_context(doc_spans, cur_span_index, position):
  """Check if this is the 'max context' doc span for the token."""

  # Because of the sliding window approach taken to scoring documents, a single
  # token can appear in multiple documents. E.g.
  #  Doc: the man went to the store and bought a gallon of milk
  #  Span A: the man went to the
  #  Span B: to the store and bought
  #  Span C: and bought a gallon of
  #  ...
  #
  # Now the word 'bought' will have two scores from spans B and C. We only
  # want to consider the score with "maximum context", which we define as
  # the *minimum* of its left and right context (the *sum* of left and
  # right context will always be the same, of course).
  #
  # In the example the maximum context for 'bought' would be span C since
  # it has 1 left context and 3 right context, while span B has 4 left context
  # and 0 right context.
  best_score = None
  best_span_index = None
  for (span_index, doc_span) in enumerate(doc_spans):
    end = doc_span.start + doc_span.length - 1
    if position < doc_span.start:
      continue
    if position > end:
      continue
    num_left_context = position - doc_span.start
    num_right_context = end - position
    score = min(num_left_context, num_right_context) + 0.01 * doc_span.length
    if best_score is None or score > best_score:
      best_score = score
      best_span_index = span_index

  return cur_span_index == best_span_index


def write_predictions(all_examples,
                      all_features,
                      all_results,
                      n_best_size,
                      max_answer_length,
                      do_lower_case,
                      output_prediction_file,
                      output_nbest_file,
                      output_null_log_odds_file,
                      version_2_with_negative=False,
                      null_score_diff_threshold=0.0,
                      verbose=False):
  """Write final predictions to the json file and log-odds of null if needed."""
  logging.info("Writing predictions to: %s", (output_prediction_file))
  logging.info("Writing nbest to: %s", (output_nbest_file))

  all_predictions, all_nbest_json, scores_diff_json = (
      postprocess_output(
          all_examples=all_examples,
          all_features=all_features,
          all_results=all_results,
          n_best_size=n_best_size,
          max_answer_length=max_answer_length,
          do_lower_case=do_lower_case,
          version_2_with_negative=version_2_with_negative,
          null_score_diff_threshold=null_score_diff_threshold,
          verbose=verbose))

  write_to_json_files(all_predictions, output_prediction_file)
  write_to_json_files(all_nbest_json, output_nbest_file)
  if version_2_with_negative:
    write_to_json_files(scores_diff_json, output_null_log_odds_file)


def postprocess_output(all_examples,
                       all_features,
                       all_results,
                       n_best_size,
                       max_answer_length,
                       do_lower_case,
                       version_2_with_negative=False,
                       null_score_diff_threshold=0.0,
                       xlnet_format=False,
                       verbose=False):
  """Postprocess model output, to form predicton results."""

  example_index_to_features = collections.defaultdict(list)
  for feature in all_features:
    example_index_to_features[feature.example_index].append(feature)
  unique_id_to_result = {}
  for result in all_results:
    unique_id_to_result[result.unique_id] = result

  _PrelimPrediction = collections.namedtuple(  # pylint: disable=invalid-name
      "PrelimPrediction",
      ["feature_index", "start_index", "end_index", "start_logit", "end_logit"])

  all_predictions = collections.OrderedDict()
  all_nbest_json = collections.OrderedDict()
  scores_diff_json = collections.OrderedDict()

  for (example_index, example) in enumerate(all_examples):
    features = example_index_to_features[example_index]

    prelim_predictions = []
    # keep track of the minimum score of null start+end of position 0
    score_null = 1000000  # large and positive
    min_null_feature_index = 0  # the paragraph slice with min mull score
    null_start_logit = 0  # the start logit at the slice with min null score
    null_end_logit = 0  # the end logit at the slice with min null score
    for (feature_index, feature) in enumerate(features):
      if feature.unique_id not in unique_id_to_result:
        logging.info("Skip eval example %s, not in pred.", feature.unique_id)
        continue
      result = unique_id_to_result[feature.unique_id]

      # if we could have irrelevant answers, get the min score of irrelevant
      if version_2_with_negative:
        if xlnet_format:
          feature_null_score = result.class_logits
        else:
          feature_null_score = result.start_logits[0] + result.end_logits[0]
        if feature_null_score < score_null:
          score_null = feature_null_score
          min_null_feature_index = feature_index
          null_start_logit = result.start_logits[0]
          null_end_logit = result.end_logits[0]
      for (start_index, start_logit,
           end_index, end_logit) in _get_best_indexes_and_logits(
               result=result,
               n_best_size=n_best_size,
               xlnet_format=xlnet_format):
        # We could hypothetically create invalid predictions, e.g., predict
        # that the start of the span is in the question. We throw out all
        # invalid predictions.
        if start_index >= len(feature.tokens):
          continue
        if end_index >= len(feature.tokens):
          continue
        if start_index not in feature.token_to_orig_map:
          continue
        if end_index not in feature.token_to_orig_map:
          continue
        if not feature.token_is_max_context.get(start_index, False):
          continue
        if end_index < start_index:
          continue
        length = end_index - start_index + 1
        if length > max_answer_length:
          continue
        prelim_predictions.append(
            _PrelimPrediction(
                feature_index=feature_index,
                start_index=start_index,
                end_index=end_index,
                start_logit=start_logit,
                end_logit=end_logit))

    if version_2_with_negative and not xlnet_format:
      prelim_predictions.append(
          _PrelimPrediction(
              feature_index=min_null_feature_index,
              start_index=0,
              end_index=0,
              start_logit=null_start_logit,
              end_logit=null_end_logit))
    prelim_predictions = sorted(
        prelim_predictions,
        key=lambda x: (x.start_logit + x.end_logit),
        reverse=True)

    _NbestPrediction = collections.namedtuple(  # pylint: disable=invalid-name
        "NbestPrediction", ["text", "start_logit", "end_logit"])

    seen_predictions = {}
    nbest = []
    for pred in prelim_predictions:
      if len(nbest) >= n_best_size:
        break
      feature = features[pred.feature_index]
      if pred.start_index > 0 or xlnet_format:  # this is a non-null prediction
        tok_tokens = feature.tokens[pred.start_index:(pred.end_index + 1)]
        orig_doc_start = feature.token_to_orig_map[pred.start_index]
        orig_doc_end = feature.token_to_orig_map[pred.end_index]
        orig_tokens = example.doc_tokens[orig_doc_start:(orig_doc_end + 1)]
        tok_text = " ".join(tok_tokens)

        # De-tokenize WordPieces that have been split off.
        tok_text = tok_text.replace(" ##", "")
        tok_text = tok_text.replace("##", "")

        # Clean whitespace
        tok_text = tok_text.strip()
        tok_text = " ".join(tok_text.split())
        orig_text = " ".join(orig_tokens)

        final_text = get_final_text(
            tok_text, orig_text, do_lower_case, verbose=verbose)
        if final_text in seen_predictions:
          continue

        seen_predictions[final_text] = True
      else:
        final_text = ""
        seen_predictions[final_text] = True

      nbest.append(
          _NbestPrediction(
              text=final_text,
              start_logit=pred.start_logit,
              end_logit=pred.end_logit))

    # if we didn't include the empty option in the n-best, include it
    if version_2_with_negative and not xlnet_format:
      if "" not in seen_predictions:
        nbest.append(
            _NbestPrediction(
                text="", start_logit=null_start_logit,
                end_logit=null_end_logit))
    # In very rare edge cases we could have no valid predictions. So we
    # just create a nonce prediction in this case to avoid failure.
    if not nbest:
      nbest.append(
          _NbestPrediction(text="empty", start_logit=0.0, end_logit=0.0))

    assert len(nbest) >= 1

    total_scores = []
    best_non_null_entry = None
    for entry in nbest:
      total_scores.append(entry.start_logit + entry.end_logit)
      if not best_non_null_entry:
        if entry.text:
          best_non_null_entry = entry

    probs = _compute_softmax(total_scores)

    nbest_json = []
    for (i, entry) in enumerate(nbest):
      output = collections.OrderedDict()
      output["text"] = entry.text
      output["probability"] = probs[i]
      output["start_logit"] = entry.start_logit
      output["end_logit"] = entry.end_logit
      nbest_json.append(output)

    assert len(nbest_json) >= 1

    if not version_2_with_negative:
      all_predictions[example.qas_id] = nbest_json[0]["text"]
    else:
      # pytype: disable=attribute-error
      # predict "" iff the null score - the score of best non-null > threshold
      if best_non_null_entry is not None:
        if xlnet_format:
          score_diff = score_null
          scores_diff_json[example.qas_id] = score_diff
          all_predictions[example.qas_id] = best_non_null_entry.text
        else:
          score_diff = score_null - best_non_null_entry.start_logit - (
              best_non_null_entry.end_logit)
          scores_diff_json[example.qas_id] = score_diff
          if score_diff > null_score_diff_threshold:
            all_predictions[example.qas_id] = ""
          else:
            all_predictions[example.qas_id] = best_non_null_entry.text
      else:
        logging.warning("best_non_null_entry is None")
        scores_diff_json[example.qas_id] = score_null
        all_predictions[example.qas_id] = ""
      # pytype: enable=attribute-error

    all_nbest_json[example.qas_id] = nbest_json

  return all_predictions, all_nbest_json, scores_diff_json


def write_to_json_files(json_records, json_file):
  with tf.io.gfile.GFile(json_file, "w") as writer:
    writer.write(json.dumps(json_records, indent=4) + "\n")


def get_final_text(pred_text, orig_text, do_lower_case, verbose=False):
  """Project the tokenized prediction back to the original text."""

  # When we created the data, we kept track of the alignment between original
  # (whitespace tokenized) tokens and our WordPiece tokenized tokens. So
  # now `orig_text` contains the span of our original text corresponding to the
  # span that we predicted.
  #
  # However, `orig_text` may contain extra characters that we don't want in
  # our prediction.
  #
  # For example, let's say:
  #   pred_text = steve smith
  #   orig_text = Steve Smith's
  #
  # We don't want to return `orig_text` because it contains the extra "'s".
  #
  # We don't want to return `pred_text` because it's already been normalized
  # (the SQuAD eval script also does punctuation stripping/lower casing but
  # our tokenizer does additional normalization like stripping accent
  # characters).
  #
  # What we really want to return is "Steve Smith".
  #
  # Therefore, we have to apply a semi-complicated alignment heruistic between
  # `pred_text` and `orig_text` to get a character-to-character alignment. This
  # can fail in certain cases in which case we just return `orig_text`.

  def _strip_spaces(text):
    ns_chars = []
    ns_to_s_map = collections.OrderedDict()
    for (i, c) in enumerate(text):
      if c == " ":
        continue
      ns_to_s_map[len(ns_chars)] = i
      ns_chars.append(c)
    ns_text = "".join(ns_chars)
    return (ns_text, ns_to_s_map)

  # We first tokenize `orig_text`, strip whitespace from the result
  # and `pred_text`, and check if they are the same length. If they are
  # NOT the same length, the heuristic has failed. If they are the same
  # length, we assume the characters are one-to-one aligned.
  tokenizer = tokenization.BasicTokenizer(do_lower_case=do_lower_case)

  tok_text = " ".join(tokenizer.tokenize(orig_text))

  start_position = tok_text.find(pred_text)
  if start_position == -1:
    if verbose:
      logging.info("Unable to find text: '%s' in '%s'", pred_text, orig_text)
    return orig_text
  end_position = start_position + len(pred_text) - 1

  (orig_ns_text, orig_ns_to_s_map) = _strip_spaces(orig_text)
  (tok_ns_text, tok_ns_to_s_map) = _strip_spaces(tok_text)

  if len(orig_ns_text) != len(tok_ns_text):
    if verbose:
      logging.info("Length not equal after stripping spaces: '%s' vs '%s'",
                   orig_ns_text, tok_ns_text)
    return orig_text

  # We then project the characters in `pred_text` back to `orig_text` using
  # the character-to-character alignment.
  tok_s_to_ns_map = {}
  for (i, tok_index) in six.iteritems(tok_ns_to_s_map):
    tok_s_to_ns_map[tok_index] = i

  orig_start_position = None
  if start_position in tok_s_to_ns_map:
    ns_start_position = tok_s_to_ns_map[start_position]
    if ns_start_position in orig_ns_to_s_map:
      orig_start_position = orig_ns_to_s_map[ns_start_position]

  if orig_start_position is None:
    if verbose:
      logging.info("Couldn't map start position")
    return orig_text

  orig_end_position = None
  if end_position in tok_s_to_ns_map:
    ns_end_position = tok_s_to_ns_map[end_position]
    if ns_end_position in orig_ns_to_s_map:
      orig_end_position = orig_ns_to_s_map[ns_end_position]

  if orig_end_position is None:
    if verbose:
      logging.info("Couldn't map end position")
    return orig_text

  output_text = orig_text[orig_start_position:(orig_end_position + 1)]
  return output_text


def _get_best_indexes_and_logits(result,
                                 n_best_size,
                                 xlnet_format=False):
  """Generates the n-best indexes and logits from a list."""
  if xlnet_format:
    for i in range(n_best_size):
      for j in range(n_best_size):
        j_index = i * n_best_size + j
        yield (result.start_indexes[i], result.start_logits[i],
               result.end_indexes[j_index], result.end_logits[j_index])
  else:
    start_index_and_score = sorted(enumerate(result.start_logits),
                                   key=lambda x: x[1], reverse=True)
    end_index_and_score = sorted(enumerate(result.end_logits),
                                 key=lambda x: x[1], reverse=True)
    for i in range(len(start_index_and_score)):
      if i >= n_best_size:
        break
      for j in range(len(end_index_and_score)):
        if j >= n_best_size:
          break
        yield (start_index_and_score[i][0], start_index_and_score[i][1],
               end_index_and_score[j][0], end_index_and_score[j][1])


def _compute_softmax(scores):
  """Compute softmax probability over raw logits."""
  if not scores:
    return []

  max_score = None
  for score in scores:
    if max_score is None or score > max_score:
      max_score = score

  exp_scores = []
  total_sum = 0.0
  for score in scores:
    x = math.exp(score - max_score)
    exp_scores.append(x)
    total_sum += x

  probs = []
  for score in exp_scores:
    probs.append(score / total_sum)
  return probs


def generate_tf_record_from_json_file(input_file_path,
                                      vocab_file_path,
                                      output_path,
                                      translated_input_folder=None,
                                      max_seq_length=384,
                                      do_lower_case=True,
                                      max_query_length=64,
                                      doc_stride=128,
                                      version_2_with_negative=False,
                                      xlnet_format=False):
  """Generates and saves training data into a tf record file."""
  train_examples = read_squad_examples(
      input_file=input_file_path,
      is_training=True,
      version_2_with_negative=version_2_with_negative,
      translated_input_folder=translated_input_folder)
  tokenizer = tokenization.FullTokenizer(
      vocab_file=vocab_file_path, do_lower_case=do_lower_case)
  train_writer = FeatureWriter(filename=output_path, is_training=True)
  number_of_examples = convert_examples_to_features(
      examples=train_examples,
      tokenizer=tokenizer,
      max_seq_length=max_seq_length,
      doc_stride=doc_stride,
      max_query_length=max_query_length,
      is_training=True,
      output_fn=train_writer.process_feature,
      xlnet_format=xlnet_format)
  train_writer.close()

  meta_data = {
      "task_type": "bert_squad",
      "train_data_size": number_of_examples,
      "max_seq_length": max_seq_length,
      "max_query_length": max_query_length,
      "doc_stride": doc_stride,
      "version_2_with_negative": version_2_with_negative,
  }

  return meta_data