run_classifier.py

# coding=utf-8
# Copyright 2018 The Google AI Language Team Authors.
#
# 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.
"""BERT finetuning runner."""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import collections
import csv
import os
import modeling
import optimization
import tokenization
import tensorflow as tf
import numpy as np

flags = tf.flags

FLAGS = flags.FLAGS

## Required parameters
flags.DEFINE_string(
    "data_dir", None,
    "The input data dir. Should contain the .tsv files (or other data files) "
    "for the task.")

flags.DEFINE_string(
    "bert_config_file", None,
    "The config json file corresponding to the pre-trained BERT model. "
    "This specifies the model architecture.")

flags.DEFINE_string("task_name", None, "The name of the task to train.")

flags.DEFINE_string("vocab_file", None,
                    "The vocabulary file that the BERT model was trained on.")

flags.DEFINE_string(
    "output_dir", None,
    "The output directory where the model checkpoints will be written.")

## Other parameters

flags.DEFINE_string(
    "init_checkpoint", None,
    "Initial checkpoint (usually from a pre-trained BERT model).")

flags.DEFINE_bool(
    "do_lower_case", True,
    "Whether to lower case the input text. Should be True for uncased "
    "models and False for cased models.")

flags.DEFINE_integer(
    "max_seq_length", 128,
    "The maximum total input sequence length after WordPiece tokenization. "
    "Sequences longer than this will be truncated, and sequences shorter "
    "than this will be padded.")

flags.DEFINE_bool("do_train", False, "Whether to run training.")

flags.DEFINE_bool("do_eval", False, "Whether to run eval on the dev set.")

flags.DEFINE_bool(
    "do_predict", False,
    "Whether to run the model in inference mode on the test set.")

flags.DEFINE_integer("train_batch_size", 32, "Total batch size for training.")

flags.DEFINE_integer("eval_batch_size", 8, "Total batch size for eval.")

flags.DEFINE_integer("predict_batch_size", 1, "Total batch size for predict.")

flags.DEFINE_float("learning_rate", 5e-5, "The initial learning rate for Adam.")

flags.DEFINE_float("num_train_epochs", 3.0,
                   "Total number of training epochs to perform.")

flags.DEFINE_float(
    "warmup_proportion", 0.1,
    "Proportion of training to perform linear learning rate warmup for. "
    "E.g., 0.1 = 10% of training.")

flags.DEFINE_integer("save_checkpoints_steps", 100000,
                     "How often to save the model checkpoint.")

flags.DEFINE_integer("iterations_per_loop", 1000,
                     "How many steps to make in each estimator call.")

flags.DEFINE_bool("use_tpu", False, "Whether to use TPU or GPU/CPU.")

tf.flags.DEFINE_string(
    "tpu_name", None,
    "The Cloud TPU to use for training. This should be either the name "
    "used when creating the Cloud TPU, or a grpc://ip.address.of.tpu:8470 "
    "url.")

tf.flags.DEFINE_string(
    "tpu_zone", None,
    "[Optional] GCE zone where the Cloud TPU is located in. If not "
    "specified, we will attempt to automatically detect the GCE project from "
    "metadata.")

tf.flags.DEFINE_string(
    "gcp_project", None,
    "[Optional] Project name for the Cloud TPU-enabled project. If not "
    "specified, we will attempt to automatically detect the GCE project from "
    "metadata.")

tf.flags.DEFINE_string("master", None, "[Optional] TensorFlow master URL.")

flags.DEFINE_integer(
    "num_tpu_cores", 8,
    "Only used if `use_tpu` is True. Total number of TPU cores to use.")

flags.DEFINE_integer(
    "max_num_relations", 10,
    "Maximum number of relation within the text")

flags.DEFINE_integer(
    "max_distance", 2,
    "The max_distance for entity relative position")

class Extras(object):
  """
  Extra objact to pass entity related features to the model function
  """

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

  def __init__(self, guid, text_a, text_b, locations, labels, num_relations):
    """Constructs a InputExample.

    Args:
      guid: Unique id for the example.
      text_a: string. The untokenized text of the first sequence. For single
        sequence tasks, only this sequence must be specified.
      text_b: unused in the entity task
      locations: entity localtion
      labels: relation label
      num_relations: number of relations in the text
      label: (Optional) string. The label of the example. This should be
        specified for train and dev examples, but not for test examples.
    """
    self.guid = guid
    self.text_a = text_a
    self.text_b = text_b
    self.locations = locations
    self.labels = labels
    self.num_relations = num_relations


class PaddingInputExample(object):
  """Fake example so the num input examples is a multiple of the batch size.

  When running eval/predict on the TPU, we need to pad the number of examples
  to be a multiple of the batch size, because the TPU requires a fixed batch
  size. The alternative is to drop the last batch, which is bad because it means
  the entire output data won't be generated.

  We use this class instead of `None` because treating `None` as padding
  battches could cause silent errors.
  """


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

  def __init__(self, input_ids, input_mask, segment_ids,
               loc, mas, e1_mas, e2_mas, cls_mask,
               label_id):
    self.input_ids = input_ids
    self.input_mask = input_mask
    self.segment_ids = segment_ids
    self.loc = loc
    self.mas = mas
    self.e1_mas = e1_mas
    self.e2_mas = e2_mas
    self.cls_mask = cls_mask
    self.label_id = label_id


class DataProcessor(object):
  """Base class for data converters for sequence classification data sets."""

  def get_train_examples(self, data_dir):
    """Gets a collection of `InputExample`s for the train set."""
    raise NotImplementedError()

  def get_dev_examples(self, data_dir):
    """Gets a collection of `InputExample`s for the dev set."""
    raise NotImplementedError()

  def get_test_examples(self, data_dir):
    """Gets a collection of `InputExample`s for prediction."""
    raise NotImplementedError()

  def get_labels(self):
    """Gets the list of labels for this data set."""
    raise NotImplementedError()

  @classmethod
  def _read_tsv(cls, input_file, quotechar=None):
    """Reads a tab separated value file."""
    with tf.gfile.Open(input_file, "r") as f:
      reader = csv.reader(f, delimiter="\t", quotechar=quotechar)
      lines = []
      for line in reader:
        lines.append(line)
      return lines

class SemEvalProcessor(DataProcessor):
    """Processor for the SemEval data set (GLUE version)."""

    def get_train_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "SemEval.train.tsv")), "train")

    def get_dev_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "SemEval.test.tsv")), "dev")

    def get_test_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "SemEval.test.tsv")), "test")

    def get_labels(self, data_dir):
        """See base class."""
        label_list = []
        filein = open(os.path.join(data_dir, "SemEval.label.tsv"))
        for line in filein:
            label = line.strip()
            label_list.append(tokenization.convert_to_unicode(label))
        return label_list

    def _create_examples(self, lines, set_type):
        """Creates examples for the training and dev sets."""
        examples = []

        for (i, line) in enumerate(lines):
            guid = "%s-%s" % (set_type, i)
            text_a = tokenization.convert_to_unicode(line[0])
            num_relations = int((len(line)-1)/7)
            locations = list()
            labels = list()
            for j in range(num_relations):
              label = tokenization.convert_to_unicode(line[j*7+1])
              labels.append(label)
              # (lo, hi)
              entity_pos1 = (int(line[j*7+2])+1, int(line[j*7+3])+1)
              entity_pos2 = (int(line[j*7+5])+1, int(line[j*7+6])+1)
              # [((lo1,hi1), (lo2, hi2))]
              locations.append((entity_pos1, entity_pos2))
            examples.append(
                InputExample(guid=guid, text_a=text_a, text_b=None,
                  locations=locations, labels=labels, num_relations=num_relations))

        return examples

class ACEProcessor(DataProcessor):
    """Processor for the SemEval data set (GLUE version)."""

    def get_train_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "ACE.train.tsv")), "train")

    def get_dev_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "ACE.dev.tsv")), "dev")

    def get_test_examples(self, data_dir):
        """See base class."""
        return self._create_examples(
            self._read_tsv(os.path.join(data_dir, "ACE.test.tsv")), "test")
    def get_labels(self, data_dir):
        """See base class."""
        label_list = []
        filein = open(os.path.join(data_dir, "ACE.label.tsv"))
        for line in filein:
            label = line.strip()
            label_list.append(tokenization.convert_to_unicode(label))
        return label_list

    def _create_examples(self, lines, set_type):
        """Creates examples for the training and dev sets."""
        examples = []

        for (i, line) in enumerate(lines):
            # if i == 0:
            #   continue
            guid = "%s-%s" % (set_type, i)
            text_a = tokenization.convert_to_unicode(line[0])
            num_relations = int((len(line)-1)/7)
            locations = list()
            labels = list()
            for j in range(num_relations):
              label = tokenization.convert_to_unicode(line[j*7+1])
              labels.append(label)
              # (lo, hi)
              entity_pos1 = (int(line[j*7+2])+1, int(line[j*7+3])+1)
              entity_pos2 = (int(line[j*7+5])+1, int(line[j*7+6])+1)
              # [((lo1,hi1), (lo2, hi2))]
              locations.append((entity_pos1, entity_pos2))
            examples.append(
                InputExample(guid=guid, text_a=text_a, text_b=None,
                  locations=locations, labels=labels, num_relations=num_relations))

        return examples

def find_lo_hi(mapping, value):
  """
  find the boundary of a value in a list
  will return (0,0) if no such value in the list
  """
  try:
    lo = mapping.index(value)
    hi = min(len(mapping) - 1 - mapping[::-1].index(value), FLAGS.max_seq_length)
    return (lo, hi)
  except:
    return (0,0)

def convert_entity_row(mapping, loc, max_distance):
  """
  convert an entity span(lo,hi) to a relative distance vector of shape [max_seq_length]
  """
  lo, hi = loc
  res = [max_distance] * FLAGS.max_seq_length
  mas = [0] * FLAGS.max_seq_length
  for i in range(FLAGS.max_seq_length):
    if i < len(mapping):
      val = mapping[i]
      if val < lo - max_distance:
        res[i] = max_distance
      elif val < lo:
        res[i] = lo - val
      elif val <= hi:
        res[i] = 0
        mas[i] = 1
      elif val <= hi + max_distance:
        res[i] = val - hi + max_distance
      else:
        res[i] = 2 * max_distance
    else:
      res[i] = 2 * max_distance
  return res, mas

def prepare_extra_data(mapping, locs, max_distance):
  res = np.zeros([FLAGS.max_seq_length, FLAGS.max_seq_length], dtype=np.int8)
  mas = np.zeros([FLAGS.max_seq_length, FLAGS.max_seq_length], dtype=np.int8)
  
  e1_mas = np.zeros([FLAGS.max_num_relations, FLAGS.max_seq_length], dtype=np.int8)
  e2_mas = np.zeros([FLAGS.max_num_relations, FLAGS.max_seq_length], dtype=np.int8)

  entities = set()
  for loc in locs:
    entities.add(loc[0])
    entities.add(loc[1])

  for e in entities:
    (lo, hi) = e
    relative_position, _ = convert_entity_row(mapping, e, max_distance)
    sub_lo1, sub_hi1 = find_lo_hi(mapping, lo)
    sub_lo2, sub_hi2 = find_lo_hi(mapping, hi)
    if sub_lo1 == 0 and sub_hi1 == 0:
      continue
    if sub_lo2 == 0 and sub_hi2 == 0:
      continue
    # col
    res[:, sub_lo1:sub_hi2+1] = np.expand_dims(relative_position, -1)
    mas[1:, sub_lo1:sub_hi2+1] = 1

  for e in entities:
    (lo, hi) = e
    relative_position, _ = convert_entity_row(mapping, e, max_distance)
    sub_lo1, sub_hi1 = find_lo_hi(mapping, lo)
    sub_lo2, sub_hi2 = find_lo_hi(mapping, hi)
    if sub_lo1 == 0 and sub_hi1 == 0:
      continue
    if sub_lo2 == 0 and sub_hi2 == 0:
      continue
    # row
    res[sub_lo1:sub_hi2+1, :] = relative_position
    mas[sub_lo1:sub_hi2+1, 1:] = 1

  for idx, (e1,e2) in enumerate(locs):
    # e1
    (lo, hi) = e1
    _, mask = convert_entity_row(mapping, e1, max_distance)
    e1_mas[idx] = mask
    # e2
    (lo, hi) = e2
    _, mask = convert_entity_row(mapping, e2, max_distance)
    e2_mas[idx] = mask

  return res, mas, e1_mas, e2_mas

def convert_single_example(ex_index, example, label_list, max_seq_length,
                           tokenizer):
  """Converts a single `InputExample` into a single `InputFeatures`."""

  if isinstance(example, PaddingInputExample):
    return InputFeatures(
        input_ids=[0] * max_seq_length,
        input_mask=[0] * max_seq_length,
        segment_ids=[0] * max_seq_length,
        label_id=0,
        is_real_example=False)

  label_map = {}
  for (i, label) in enumerate(label_list):
    label_map[label] = i

  tokens_a, mapping_a = tokenizer.tokenize(example.text_a)
  tokens_b = None
  if example.text_b:
    tokens_b = tokenizer.tokenize(example.text_b)

  if tokens_b:
    # Modifies `tokens_a` and `tokens_b` in place so that the total
    # length is less than the specified length.
    # Account for [CLS], [SEP], [SEP] with "- 3"
    _truncate_seq_pair(tokens_a, tokens_b, max_seq_length - 3)
  else:
    # Account for [CLS] and [SEP] with "- 2"
    if len(tokens_a) > max_seq_length - 2:
      tokens_a = tokens_a[0:(max_seq_length - 2)]

  # The convention in BERT is:
  # (a) For sequence pairs:
  #  tokens:   [CLS] is this jack ##son ##ville ? [SEP] no it is not . [SEP]
  #  type_ids: 0     0  0    0    0     0       0 0     1  1  1  1   1 1
  # (b) For single sequences:
  #  tokens:   [CLS] the dog is hairy . [SEP]
  #  type_ids: 0     0   0   0  0     0 0
  #
  # Where "type_ids" are used to indicate whether this is the first
  # sequence or the second sequence. The embedding vectors for `type=0` and
  # `type=1` were learned during pre-training and are added to the wordpiece
  # embedding vector (and position vector). This is not *strictly* necessary
  # since the [SEP] token unambiguously separates the sequences, but it makes
  # it easier for the model to learn the concept of sequences.
  #
  # For classification tasks, the first vector (corresponding to [CLS]) is
  # used as the "sentence vector". Note that this only makes sense because
  # the entire model is fine-tuned.
  tokens = []
  segment_ids = []
  tokens.append("[CLS]")
  segment_ids.append(0)
  for token in tokens_a:
    tokens.append(token)
    segment_ids.append(0)
  tokens.append("[SEP]")
  segment_ids.append(0)

  if tokens_b:
    for token in tokens_b:
      tokens.append(token)
      segment_ids.append(1)
    tokens.append("[SEP]")
    segment_ids.append(1)

  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(0)

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

  loc, mas, e1_mas, e2_mas = prepare_extra_data(mapping_a, example.locations, FLAGS.max_distance)
  label_id = [label_map[label] for label in example.labels]
  label_id = label_id + [0] * (FLAGS.max_num_relations - len(label_id))
  cls_mask = [1] * example.num_relations + [0] * (FLAGS.max_num_relations - example.num_relations)

  np.set_printoptions(edgeitems=15)
  if ex_index < 5:
    tf.logging.info("*** Example ***")
    tf.logging.info("guid: %s" % (example.guid))
    tf.logging.info("tokens: %s" % " ".join(
        [tokenization.printable_text(x) for x in tokens]))
    tf.logging.info("input_ids: %s" % " ".join([str(x) for x in input_ids]))
    tf.logging.info("input_mask: %s" % " ".join([str(x) for x in input_mask]))
    tf.logging.info("segment_ids: %s" % " ".join([str(x) for x in segment_ids]))
    tf.logging.info("loc:")
    tf.logging.info("\n" + str(loc))
    tf.logging.info("mas:")
    tf.logging.info("\n" + str(mas))
    tf.logging.info("e1_mas:")
    tf.logging.info("\n" + str(e1_mas))
    tf.logging.info("e2_mas:")
    tf.logging.info("\n" + str(e2_mas))
    tf.logging.info("cls_mask:")
    tf.logging.info("\n" + str(cls_mask))
    tf.logging.info("labels: %s" % " ".join([str(x) for x in label_id]))

  feature = InputFeatures(
      input_ids=input_ids,
      input_mask=input_mask,
      segment_ids=segment_ids,
      loc=loc.flatten(),
      mas=mas.flatten(),
      e1_mas=e1_mas.flatten(),
      e2_mas=e2_mas.flatten(),
      cls_mask=cls_mask,
      label_id=label_id)
  return feature


def file_based_convert_examples_to_features(
    examples, label_list, max_seq_length, tokenizer, output_file):
  """Convert a set of `InputExample`s to a TFRecord file."""

  writer = tf.python_io.TFRecordWriter(output_file)

  for (ex_index, example) in enumerate(examples):
    if ex_index % 10000 == 0:
      tf.logging.info("Writing example %d of %d" % (ex_index, len(examples)))

    feature = convert_single_example(ex_index, example, label_list,
                                     max_seq_length, tokenizer)

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

    features = collections.OrderedDict()
    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)
    features["loc"] = create_int_feature(feature.loc)
    features["mas"] = create_int_feature(feature.mas)
    features["e1_mas"] = create_int_feature(feature.e1_mas)
    features["e2_mas"] = create_int_feature(feature.e2_mas)
    features["cls_mask"] = create_int_feature(feature.cls_mask)
    features["label_ids"] = create_int_feature(feature.label_id)

    tf_example = tf.train.Example(features=tf.train.Features(feature=features))
    writer.write(tf_example.SerializeToString())
  writer.close()


def file_based_input_fn_builder(input_file, seq_length, is_training,
                                drop_remainder):
  """Creates an `input_fn` closure to be passed to TPUEstimator."""

  name_to_features = {
      "input_ids": tf.FixedLenFeature([seq_length], tf.int64),
      "input_mask": tf.FixedLenFeature([seq_length], tf.int64),
      "segment_ids": tf.FixedLenFeature([seq_length], tf.int64),
      "loc": tf.FixedLenFeature([seq_length * seq_length], tf.int64),
      "mas": tf.FixedLenFeature([seq_length * seq_length], tf.int64),
      "e1_mas": tf.FixedLenFeature([FLAGS.max_num_relations * seq_length], tf.int64),
      "e2_mas": tf.FixedLenFeature([FLAGS.max_num_relations * seq_length], tf.int64),
      "cls_mask": tf.FixedLenFeature([FLAGS.max_num_relations], tf.int64),
      "label_ids": tf.FixedLenFeature([FLAGS.max_num_relations], tf.int64)
  }

  def _decode_record(record, name_to_features):
    """Decodes a record to a TensorFlow example."""
    example = tf.parse_single_example(record, name_to_features)

    # tf.Example only supports tf.int64, but the TPU only supports tf.int32.
    # So cast all int64 to int32.
    for name in list(example.keys()):
      t = example[name]
      if t.dtype == tf.int64:
        t = tf.to_int32(t)
      example[name] = t

    return example

  def input_fn(params):
    """The actual input function."""
    batch_size = params["batch_size"]

    # For training, we want a lot of parallel reading and shuffling.
    # For eval, we want no shuffling and parallel reading doesn't matter.
    d = tf.data.TFRecordDataset(input_file)
    if is_training:
      d = d.repeat()
      d = d.shuffle(buffer_size=100)

    d = d.apply(
        tf.contrib.data.map_and_batch(
            lambda record: _decode_record(record, name_to_features),
            batch_size=batch_size,
            drop_remainder=drop_remainder))

    return d

  return input_fn


def _truncate_seq_pair(tokens_a, tokens_b, max_length):
  """Truncates a sequence pair in place to the maximum length."""

  # This is a simple heuristic which will always truncate the longer sequence
  # one token at a time. This makes more sense than truncating an equal percent
  # of tokens from each, since if one sequence is very short then each token
  # that's truncated likely contains more information than a longer sequence.
  while True:
    total_length = len(tokens_a) + len(tokens_b)
    if total_length <= max_length:
      break
    if len(tokens_a) > len(tokens_b):
      tokens_a.pop()
    else:
      tokens_b.pop()


def create_model(bert_config, is_training, input_ids, input_mask, segment_ids,
                 labels, num_labels, use_one_hot_embeddings, extras):
  """Creates a classification model."""
  model = modeling.BertModel(
      config=bert_config,
      is_training=is_training,
      input_ids=input_ids,
      input_mask=input_mask,
      token_type_ids=segment_ids,
      use_one_hot_embeddings=use_one_hot_embeddings,
      extras=extras)

  output_layer = model.get_sequence_output()

  from_seq_length = output_layer.shape[1].value
  hidden_size = output_layer.shape[2].value

  # B 10 F 768
  output_layer = tf.stack([output_layer] * FLAGS.max_num_relations, axis=1)
  # B 10 F 1
  e1_mas = tf.reshape(extras.e1_mas, [-1, FLAGS.max_num_relations, from_seq_length, 1])
  # B 10 F 768
  e1 = tf.multiply(output_layer, tf.to_float(e1_mas))
  # B 10 768
  e1 = tf.reduce_sum(e1, axis=-2) / tf.maximum(1.0, tf.reduce_sum(tf.to_float(e1_mas), axis=-2))
  # B*10 768
  e1 = tf.reshape(e1, [-1, hidden_size])
    # B 10 F 1
  e2_mas = tf.reshape(extras.e2_mas, [-1, FLAGS.max_num_relations, from_seq_length, 1])
  # B 10 F 768
  e2 = tf.multiply(output_layer, tf.to_float(e2_mas))
  # B 10 768
  e2 = tf.reduce_sum(e2, axis=-2) / tf.maximum(1.0, tf.reduce_sum(tf.to_float(e2_mas), axis=-2))
  # B*10 768
  e2 = tf.reshape(e2, [-1, hidden_size])
  # B*10 768*2
  output_layer = tf.concat([e1, e2], axis=-1)

  output_weights = tf.get_variable(
      "cls/entity/output_weights", [num_labels, hidden_size*2],
      initializer=tf.truncated_normal_initializer(stddev=0.02))

  output_bias = tf.get_variable(
      "cls/entity/output_bias", [num_labels], initializer=tf.zeros_initializer())

  with tf.variable_scope("loss"):
    if is_training:
      # I.e., 0.1 dropout
      output_layer = tf.nn.dropout(output_layer, keep_prob=0.9)
    # B*10 num_label
    logits = tf.matmul(output_layer, output_weights, transpose_b=True)
    # B*10 num_label
    logits = tf.nn.bias_add(logits, output_bias)
    # B*10 num_label
    probabilities = tf.nn.softmax(logits, axis=-1)
    # B*10 num_label
    log_probs = tf.nn.log_softmax(logits, axis=-1)
    # B*10
    labels = tf.reshape(labels, [-1])
    # B*10 num_label
    one_hot_labels = tf.one_hot(labels, depth=num_labels, dtype=tf.float32)
    # B*10
    per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1)
    # B*10
    cls_mask = tf.reshape(tf.to_float(extras.cls_mask), [-1])
    # B*10
    per_example_loss = per_example_loss * cls_mask

    loss = tf.reduce_sum(per_example_loss) / tf.reduce_sum(cls_mask)

    return (loss, per_example_loss, logits, probabilities)


def model_fn_builder(bert_config, num_labels, init_checkpoint, learning_rate,
                     num_train_steps, num_warmup_steps, use_tpu,
                     use_one_hot_embeddings):
  """Returns `model_fn` closure for TPUEstimator."""

  def model_fn(features, labels, mode, params):  # pylint: disable=unused-argument
    """The `model_fn` for TPUEstimator."""

    tf.logging.info("*** Features ***")
    for name in sorted(features.keys()):
      tf.logging.info("  name = %s, shape = %s" % (name, features[name].shape))

    input_ids = features["input_ids"]
    input_mask = features["input_mask"]
    segment_ids = features["segment_ids"]
    label_ids = features["label_ids"]
    extras = Extras()
    extras.loc = features["loc"]
    extras.mas = features["mas"]
    extras.e1_mas = features["e1_mas"]
    extras.e2_mas = features["e2_mas"]
    extras.cls_mask = features["cls_mask"]
    extras.max_distance = FLAGS.max_distance

    is_training = (mode == tf.estimator.ModeKeys.TRAIN)

    (total_loss, per_example_loss, logits, probabilities) = create_model(
        bert_config, is_training, input_ids, input_mask, segment_ids, label_ids,
        num_labels, use_one_hot_embeddings, extras)

    tvars = tf.trainable_variables()
    initialized_variable_names = {}
    scaffold_fn = None
    if init_checkpoint:
      (assignment_map, initialized_variable_names
      ) = modeling.get_assignment_map_from_checkpoint(tvars, init_checkpoint)
      if use_tpu:

        def tpu_scaffold():
          tf.train.init_from_checkpoint(init_checkpoint, assignment_map)
          return tf.train.Scaffold()

        scaffold_fn = tpu_scaffold
      else:
        tf.train.init_from_checkpoint(init_checkpoint, assignment_map)

    tf.logging.info("**** Trainable Variables ****")
    for var in tvars:
      init_string = ""
      if var.name in initialized_variable_names:
        init_string = ", *INIT_FROM_CKPT*"
      tf.logging.info("  name = %s, shape = %s%s", var.name, var.shape,
                      init_string)

    output_spec = None
    if mode == tf.estimator.ModeKeys.TRAIN:

      train_op = optimization.create_optimizer(
          total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu)

      output_spec = tf.contrib.tpu.TPUEstimatorSpec(
          mode=mode,
          loss=total_loss,
          train_op=train_op,
          scaffold_fn=scaffold_fn)
    elif mode == tf.estimator.ModeKeys.EVAL:

      def metric_fn(per_example_loss, label_ids, logits, is_real_example):
        predictions = tf.argmax(logits, axis=-1, output_type=tf.int32)
        accuracy = tf.metrics.accuracy(
            labels=label_ids, predictions=predictions, weights=is_real_example)
        loss = tf.metrics.mean(values=per_example_loss, weights=is_real_example)
        return {
            "eval_accuracy": accuracy,
            "eval_loss": loss,
        }

      eval_metrics = (metric_fn,
                      [per_example_loss, label_ids, logits, is_real_example])
      output_spec = tf.contrib.tpu.TPUEstimatorSpec(
          mode=mode,
          loss=total_loss,
          eval_metrics=eval_metrics,
          scaffold_fn=scaffold_fn)
    else:
      # B 10 num_labels
      logits = tf.reshape(logits, [-1, FLAGS.max_num_relations, num_labels])
      # B 10
      predictions = tf.argmax(logits, axis=-1, output_type=tf.int32)
      output_spec = tf.contrib.tpu.TPUEstimatorSpec(
          mode=mode,
          predictions=predictions,
          scaffold_fn=scaffold_fn)
    return output_spec

  return model_fn


# This function is not used by this file but is still used by the Colab and
# people who depend on it.
def input_fn_builder(features, seq_length, is_training, drop_remainder):
  """Creates an `input_fn` closure to be passed to TPUEstimator."""

  all_input_ids = []
  all_input_mask = []
  all_segment_ids = []
  all_label_ids = []

  for feature in features:
    all_input_ids.append(feature.input_ids)
    all_input_mask.append(feature.input_mask)
    all_segment_ids.append(feature.segment_ids)
    all_label_ids.append(feature.label_id)

  def input_fn(params):
    """The actual input function."""
    batch_size = params["batch_size"]

    num_examples = len(features)

    # This is for demo purposes and does NOT scale to large data sets. We do
    # not use Dataset.from_generator() because that uses tf.py_func which is
    # not TPU compatible. The right way to load data is with TFRecordReader.
    d = tf.data.Dataset.from_tensor_slices({
        "input_ids":
            tf.constant(
                all_input_ids, shape=[num_examples, seq_length],
                dtype=tf.int32),
        "input_mask":
            tf.constant(
                all_input_mask,
                shape=[num_examples, seq_length],
                dtype=tf.int32),
        "segment_ids":
            tf.constant(
                all_segment_ids,
                shape=[num_examples, seq_length],
                dtype=tf.int32),
        "label_ids":
            tf.constant(all_label_ids, shape=[num_examples], dtype=tf.int32),
    })

    if is_training:
      d = d.repeat()
      d = d.shuffle(buffer_size=100)

    d = d.batch(batch_size=batch_size, drop_remainder=drop_remainder)
    return d

  return input_fn


# This function is not used by this file but is still used by the Colab and
# people who depend on it.
def convert_examples_to_features(examples, label_list, max_seq_length,
                                 tokenizer):
  """Convert a set of `InputExample`s to a list of `InputFeatures`."""

  features = []
  for (ex_index, example) in enumerate(examples):
    if ex_index % 10000 == 0:
      tf.logging.info("Writing example %d of %d" % (ex_index, len(examples)))

    feature = convert_single_example(ex_index, example, label_list,
                                     max_seq_length, tokenizer)

    features.append(feature)
  return features


def main(_):
  tf.logging.set_verbosity(tf.logging.INFO)

  processors = {
      "semeval": SemEvalProcessor,
      "ace": ACEProcessor
  }

  tokenization.validate_case_matches_checkpoint(FLAGS.do_lower_case,
                                                FLAGS.init_checkpoint)

  if not FLAGS.do_train and not FLAGS.do_eval and not FLAGS.do_predict:
    raise ValueError(
        "At least one of `do_train`, `do_eval` or `do_predict' must be True.")

  bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file)

  if FLAGS.max_seq_length > bert_config.max_position_embeddings:
    raise ValueError(
        "Cannot use sequence length %d because the BERT model "
        "was only trained up to sequence length %d" %
        (FLAGS.max_seq_length, bert_config.max_position_embeddings))

  tf.gfile.MakeDirs(FLAGS.output_dir)

  task_name = FLAGS.task_name.lower()

  if task_name not in processors:
    raise ValueError("Task not found: %s" % (task_name))

  processor = processors[task_name]()

  label_list = processor.get_labels(FLAGS.data_dir)

  tokenizer = tokenization.FullTokenizer(
      vocab_file=FLAGS.vocab_file, do_lower_case=FLAGS.do_lower_case)

  tpu_cluster_resolver = None
  if FLAGS.use_tpu and FLAGS.tpu_name:
    tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver(
        FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project)

  is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2
  run_config = tf.contrib.tpu.RunConfig(
      cluster=tpu_cluster_resolver,
      master=FLAGS.master,
      model_dir=FLAGS.output_dir,
      save_checkpoints_steps=FLAGS.save_checkpoints_steps,
      tpu_config=tf.contrib.tpu.TPUConfig(
          iterations_per_loop=FLAGS.iterations_per_loop,
          num_shards=FLAGS.num_tpu_cores,
          per_host_input_for_training=is_per_host))

  train_examples = None
  num_train_steps = None
  num_warmup_steps = None
  if FLAGS.do_train:
    train_examples = processor.get_train_examples(FLAGS.data_dir)
    num_train_steps = int(
        len(train_examples) / FLAGS.train_batch_size * FLAGS.num_train_epochs)
    num_warmup_steps = int(num_train_steps * FLAGS.warmup_proportion)

  model_fn = model_fn_builder(
      bert_config=bert_config,
      num_labels=len(label_list),
      init_checkpoint=FLAGS.init_checkpoint,
      learning_rate=FLAGS.learning_rate,
      num_train_steps=num_train_steps,
      num_warmup_steps=num_warmup_steps,
      use_tpu=FLAGS.use_tpu,
      use_one_hot_embeddings=FLAGS.use_tpu)

  # If TPU is not available, this will fall back to normal Estimator on CPU
  # or GPU.
  estimator = tf.contrib.tpu.TPUEstimator(
      use_tpu=FLAGS.use_tpu,
      model_fn=model_fn,
      config=run_config,
      train_batch_size=FLAGS.train_batch_size,
      eval_batch_size=FLAGS.eval_batch_size,
      predict_batch_size=FLAGS.predict_batch_size)

  if FLAGS.do_train:
    train_file = os.path.join(FLAGS.output_dir, "train.tf_record")
    file_based_convert_examples_to_features(
        train_examples, label_list, FLAGS.max_seq_length, tokenizer, train_file)
    tf.logging.info("***** Running training *****")
    tf.logging.info("  Num examples = %d", len(train_examples))
    tf.logging.info("  Batch size = %d", FLAGS.train_batch_size)
    tf.logging.info("  Num steps = %d", num_train_steps)
    train_input_fn = file_based_input_fn_builder(
        input_file=train_file,
        seq_length=FLAGS.max_seq_length,
        is_training=True,
        drop_remainder=True)
    estimator.train(input_fn=train_input_fn, max_steps=num_train_steps)

  if FLAGS.do_eval:
    eval_examples = processor.get_dev_examples(FLAGS.data_dir)
    num_actual_eval_examples = len(eval_examples)
    if FLAGS.use_tpu:
      # TPU requires a fixed batch size for all batches, therefore the number
      # of examples must be a multiple of the batch size, or else examples
      # will get dropped. So we pad with fake examples which are ignored
      # later on. These do NOT count towards the metric (all tf.metrics
      # support a per-instance weight, and these get a weight of 0.0).
      while len(eval_examples) % FLAGS.eval_batch_size != 0:
        eval_examples.append(PaddingInputExample())

    eval_file = os.path.join(FLAGS.output_dir, "eval.tf_record")
    file_based_convert_examples_to_features(
        eval_examples, label_list, FLAGS.max_seq_length, tokenizer, eval_file)

    tf.logging.info("***** Running evaluation *****")
    tf.logging.info("  Num examples = %d (%d actual, %d padding)",
                    len(eval_examples), num_actual_eval_examples,
                    len(eval_examples) - num_actual_eval_examples)
    tf.logging.info("  Batch size = %d", FLAGS.eval_batch_size)

    # This tells the estimator to run through the entire set.
    eval_steps = None
    # However, if running eval on the TPU, you will need to specify the
    # number of steps.
    if FLAGS.use_tpu:
      assert len(eval_examples) % FLAGS.eval_batch_size == 0
      eval_steps = int(len(eval_examples) // FLAGS.eval_batch_size)

    eval_drop_remainder = True if FLAGS.use_tpu else False
    eval_input_fn = file_based_input_fn_builder(
        input_file=eval_file,
        seq_length=FLAGS.max_seq_length,
        is_training=False,
        drop_remainder=eval_drop_remainder)

    result = estimator.evaluate(input_fn=eval_input_fn, steps=eval_steps)

    output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt")
    with tf.gfile.GFile(output_eval_file, "w") as writer:
      tf.logging.info("***** Eval results *****")
      for key in sorted(result.keys()):
        tf.logging.info("  %s = %s", key, str(result[key]))
        writer.write("%s = %s\n" % (key, str(result[key])))

  if FLAGS.do_predict:
    predict_examples = processor.get_test_examples(FLAGS.data_dir)
    num_actual_predict_examples = len(predict_examples)
    if FLAGS.use_tpu:
      # TPU requires a fixed batch size for all batches, therefore the number
      # of examples must be a multiple of the batch size, or else examples
      # will get dropped. So we pad with fake examples which are ignored
      # later on.
      while len(predict_examples) % FLAGS.predict_batch_size != 0:
        predict_examples.append(PaddingInputExample())

    predict_file = os.path.join(FLAGS.output_dir, "predict.tf_record")
    file_based_convert_examples_to_features(predict_examples, label_list,
                                            FLAGS.max_seq_length, tokenizer,
                                            predict_file)

    tf.logging.info("***** Running prediction*****")
    tf.logging.info("  Num examples = %d (%d actual, %d padding)",
                    len(predict_examples), num_actual_predict_examples,
                    len(predict_examples) - num_actual_predict_examples)
    tf.logging.info("  Batch size = %d", FLAGS.predict_batch_size)

    predict_drop_remainder = True if FLAGS.use_tpu else False
    predict_input_fn = file_based_input_fn_builder(
        input_file=predict_file,
        seq_length=FLAGS.max_seq_length,
        is_training=False,
        drop_remainder=True)

    result = estimator.predict(input_fn=predict_input_fn)
    output_predict_file = os.path.join(FLAGS.output_dir, "test_results.tsv")
    with tf.gfile.GFile(output_predict_file, "w") as writer:
      tf.logging.info("***** Predict results *****")
      for prediction, example in zip(result, predict_examples):
        for x in prediction[:example.num_relations]:
          writer.write(label_list[int(x)] + '\n')

if __name__ == "__main__":
  flags.mark_flag_as_required("data_dir")
  flags.mark_flag_as_required("task_name")
  flags.mark_flag_as_required("vocab_file")
  flags.mark_flag_as_required("bert_config_file")
  flags.mark_flag_as_required("output_dir")
  tf.app.run()