run_race.py

# coding=utf-8
# Copyright 2018 The Google AI 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.
"""ALBERT finetuning runner with sentence piece tokenization."""

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

import os
import time
import classifier_utils
import fine_tuning_utils
import modeling
import race_utils
import tokenization
import tensorflow.compat.v1 as tf
from tensorflow.contrib import cluster_resolver as contrib_cluster_resolver
from tensorflow.contrib import tpu as contrib_tpu

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(
    "albert_config_file", None,
    "The config json file corresponding to the pre-trained ALBERT model. "
    "This specifies the model architecture.")

flags.DEFINE_string("task_name", "race", "The name of the task to train.")

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

flags.DEFINE_string("train_file", None,
                    "path to preprocessed tfrecord file. "
                    "The file will be generated if not exst.")

flags.DEFINE_string("eval_file", None,
                    "path to preprocessed tfrecord file. "
                    "The file will be generated if not exst.")

flags.DEFINE_string("predict_file", None,
                    "path to preprocessed tfrecord file. "
                    "The file will be generated if not exst.")

flags.DEFINE_string("spm_model_file", None,
                    "The model file for sentence piece tokenization.")

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 ALBERT model).")

flags.DEFINE_string(
    "albert_hub_module_handle", None,
    "If set, the ALBERT hub module to use.")

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_float("dropout_prob", 0.1, "dropout probability.")

flags.DEFINE_integer(
    "max_seq_length", 512,
    "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_integer(
    "max_qa_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_integer(
    "num_keep_checkpoint", 5,
    "maximum number of keep checkpoints")


flags.DEFINE_bool(
    "high_only", False,
    "Whether to only run the model on the high school set.")

flags.DEFINE_bool(
    "middle_only", False,
    "Whether to only run the model on the middle school set.")

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

flags.DEFINE_bool("do_eval", True, "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", 8, "Total batch size for predict.")

flags.DEFINE_float("learning_rate", 1e-5, "The initial learning rate for Adam.")

flags.DEFINE_integer("train_step", 12000,
                     "Total number of training epochs to perform.")

flags.DEFINE_integer(
    "warmup_step", 1000,
    "number of steps to perform linear learning rate warmup for.")

flags.DEFINE_integer("save_checkpoints_steps", 100,
                     "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.")


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

  processors = {
      "race": race_utils.RaceProcessor
  }

  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.")

  albert_config = modeling.AlbertConfig.from_json_file(FLAGS.albert_config_file)

  if FLAGS.max_seq_length > albert_config.max_position_embeddings:
    raise ValueError(
        "Cannot use sequence length %d because the ALBERT model "
        "was only trained up to sequence length %d" %
        (FLAGS.max_seq_length, albert_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](
      use_spm=True if FLAGS.spm_model_file else False,
      do_lower_case=FLAGS.do_lower_case,
      high_only=FLAGS.high_only,
      middle_only=FLAGS.middle_only)

  label_list = processor.get_labels()

  tokenizer = fine_tuning_utils.create_vocab(
      vocab_file=FLAGS.vocab_file,
      do_lower_case=FLAGS.do_lower_case,
      spm_model_file=FLAGS.spm_model_file,
      hub_module=FLAGS.albert_hub_module_handle)

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

  is_per_host = contrib_tpu.InputPipelineConfig.PER_HOST_V2
  if FLAGS.do_train:
    iterations_per_loop = int(min(FLAGS.iterations_per_loop,
                                  FLAGS.save_checkpoints_steps))
  else:
    iterations_per_loop = FLAGS.iterations_per_loop
  run_config = contrib_tpu.RunConfig(
      cluster=tpu_cluster_resolver,
      master=FLAGS.master,
      model_dir=FLAGS.output_dir,
      save_checkpoints_steps=int(FLAGS.save_checkpoints_steps),
      keep_checkpoint_max=0,
      tpu_config=contrib_tpu.TPUConfig(
          iterations_per_loop=iterations_per_loop,
          num_shards=FLAGS.num_tpu_cores,
          per_host_input_for_training=is_per_host))

  train_examples = None
  if FLAGS.do_train:
    train_examples = processor.get_train_examples(FLAGS.data_dir)

  model_fn = race_utils.model_fn_builder(
      albert_config=albert_config,
      num_labels=len(label_list),
      init_checkpoint=FLAGS.init_checkpoint,
      learning_rate=FLAGS.learning_rate,
      num_train_steps=FLAGS.train_step,
      num_warmup_steps=FLAGS.warmup_step,
      use_tpu=FLAGS.use_tpu,
      use_one_hot_embeddings=FLAGS.use_tpu,
      max_seq_length=FLAGS.max_seq_length,
      dropout_prob=FLAGS.dropout_prob,
      hub_module=FLAGS.albert_hub_module_handle)

  # If TPU is not available, this will fall back to normal Estimator on CPU
  # or GPU.
  estimator = 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:
    if not tf.gfile.Exists(FLAGS.train_file):
      race_utils.file_based_convert_examples_to_features(
          train_examples, label_list, FLAGS.max_seq_length, tokenizer,
          FLAGS.train_file, FLAGS.max_qa_length)
    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", FLAGS.train_step)
    train_input_fn = classifier_utils.file_based_input_fn_builder(
        input_file=FLAGS.train_file,
        seq_length=FLAGS.max_seq_length,
        is_training=True,
        drop_remainder=True,
        task_name=task_name,
        use_tpu=FLAGS.use_tpu,
        bsz=FLAGS.train_batch_size,
        multiple=len(label_list))
    estimator.train(input_fn=train_input_fn, max_steps=FLAGS.train_step)

  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(classifier_utils.PaddingInputExample())

    if not tf.gfile.Exists(FLAGS.eval_file):
      race_utils.file_based_convert_examples_to_features(
          eval_examples, label_list, FLAGS.max_seq_length, tokenizer,
          FLAGS.eval_file, FLAGS.max_qa_length)

    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 = classifier_utils.file_based_input_fn_builder(
        input_file=FLAGS.eval_file,
        seq_length=FLAGS.max_seq_length,
        is_training=False,
        drop_remainder=eval_drop_remainder,
        task_name=task_name,
        use_tpu=FLAGS.use_tpu,
        bsz=FLAGS.eval_batch_size,
        multiple=len(label_list))

    def _find_valid_cands(curr_step):
      filenames = tf.gfile.ListDirectory(FLAGS.output_dir)
      candidates = []
      for filename in filenames:
        if filename.endswith(".index"):
          ckpt_name = filename[:-6]
          idx = ckpt_name.split("-")[-1]
          if idx != "best" and int(idx) > curr_step:
            candidates.append(filename)
      return candidates

    output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt")
    checkpoint_path = os.path.join(FLAGS.output_dir, "model.ckpt-best")
    key_name = "eval_accuracy"
    if tf.gfile.Exists(checkpoint_path + ".index"):
      result = estimator.evaluate(
          input_fn=eval_input_fn,
          steps=eval_steps,
          checkpoint_path=checkpoint_path)
      best_perf = result[key_name]
      global_step = result["global_step"]
    else:
      global_step = -1
      best_perf = -1
      checkpoint_path = None
    writer = tf.gfile.GFile(output_eval_file, "w")
    while global_step < FLAGS.train_step:
      steps_and_files = {}
      filenames = tf.gfile.ListDirectory(FLAGS.output_dir)
      for filename in filenames:
        if filename.endswith(".index"):
          ckpt_name = filename[:-6]
          cur_filename = os.path.join(FLAGS.output_dir, ckpt_name)
          if cur_filename.split("-")[-1] == "best":
            continue
          gstep = int(cur_filename.split("-")[-1])
          if gstep not in steps_and_files:
            tf.logging.info("Add {} to eval list.".format(cur_filename))
            steps_and_files[gstep] = cur_filename
      tf.logging.info("found {} files.".format(len(steps_and_files)))
      # steps_and_files = sorted(steps_and_files, key=lambda x: x[0])
      if not steps_and_files:
        tf.logging.info("found 0 file, global step: {}. Sleeping."
                        .format(global_step))
        time.sleep(1)
      else:
        for ele in sorted(steps_and_files.items()):
          step, checkpoint_path = ele
          if global_step >= step:
            if len(_find_valid_cands(step)) > 1:
              for ext in ["meta", "data-00000-of-00001", "index"]:
                src_ckpt = checkpoint_path + ".{}".format(ext)
                tf.logging.info("removing {}".format(src_ckpt))
                tf.gfile.Remove(src_ckpt)
            continue
          result = estimator.evaluate(
              input_fn=eval_input_fn,
              steps=eval_steps,
              checkpoint_path=checkpoint_path)
          global_step = result["global_step"]
          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])))
          writer.write("best = {}\n".format(best_perf))
          if result[key_name] > best_perf:
            best_perf = result[key_name]
            for ext in ["meta", "data-00000-of-00001", "index"]:
              src_ckpt = checkpoint_path + ".{}".format(ext)
              tgt_ckpt = checkpoint_path.rsplit("-", 1)[0] + "-best.{}".format(ext)
              tf.logging.info("saving {} to {}".format(src_ckpt, tgt_ckpt))
              tf.gfile.Copy(src_ckpt, tgt_ckpt, overwrite=True)
              writer.write("saved {} to {}\n".format(src_ckpt, tgt_ckpt))

          if len(_find_valid_cands(global_step)) > 1:
            for ext in ["meta", "data-00000-of-00001", "index"]:
              src_ckpt = checkpoint_path + ".{}".format(ext)
              tf.logging.info("removing {}".format(src_ckpt))
              tf.gfile.Remove(src_ckpt)
          writer.write("=" * 50 + "\n")
    writer.close()
  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(classifier_utils.PaddingInputExample())
      assert len(predict_examples) % FLAGS.predict_batch_size == 0
      predict_steps = int(len(predict_examples) // FLAGS.predict_batch_size)

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

    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 = classifier_utils.file_based_input_fn_builder(
        input_file=predict_file,
        seq_length=FLAGS.max_seq_length,
        is_training=False,
        drop_remainder=predict_drop_remainder,
        task_name=task_name,
        use_tpu=FLAGS.use_tpu,
        bsz=FLAGS.predict_batch_size,
        multiple=len(label_list))

    checkpoint_path = os.path.join(FLAGS.output_dir, "model.ckpt-best")
    result = estimator.evaluate(
        input_fn=predict_input_fn,
        steps=predict_steps,
        checkpoint_path=checkpoint_path)

    output_predict_file = os.path.join(FLAGS.output_dir, "predict_results.txt")
    with tf.gfile.GFile(output_predict_file, "w") as pred_writer:
      # num_written_lines = 0
      tf.logging.info("***** Predict results *****")
      pred_writer.write("***** Predict results *****\n")
      for key in sorted(result.keys()):
        tf.logging.info("  %s = %s", key, str(result[key]))
        pred_writer.write("%s = %s\n" % (key, str(result[key])))
      pred_writer.write("best = {}\n".format(best_perf))


if __name__ == "__main__":
  flags.mark_flag_as_required("data_dir")
  flags.mark_flag_as_required("spm_model_file")
  flags.mark_flag_as_required("albert_config_file")
  flags.mark_flag_as_required("output_dir")
  tf.app.run()