830a663e996b7b95c45acddd58ba6e6532599c0c,tensorflow_datasets/text/glue.py,Glue,_split_generators,#Glue#Any#,394

Before Change


            "mrpc_files": mrpc_files,
        })
    if self.builder_config.name == "mnli":
      return [
          train_split,
          tfds.core.SplitGenerator(
              name="validation_matched",
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir, "dev_matched.tsv"),
                  "split": "dev",
                  "mrpc_files": None,
              }),
          tfds.core.SplitGenerator(
              name="validation_mismatched",
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir, "dev_mismatched.tsv"),
                  "split": "dev",
                  "mrpc_files": None,
              }),
          tfds.core.SplitGenerator(
              name="test_matched",
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir, "test_matched.tsv"),
                  "split": "test",
                  "mrpc_files": None,
              }),
          tfds.core.SplitGenerator(
              name="test_mismatched",
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir, "test_mismatched.tsv"),
                  "split": "test",
                  "mrpc_files": None,
              }),
      ]
    else:
      return [
          train_split,
          tfds.core.SplitGenerator(

After Change


            "split": "train",
            "mrpc_files": mrpc_files,
        })
    if self.builder_config.name == "mnli":
      return [
          train_split,
          _mnli_split_generator(
              "validation_matched", data_dir, "dev", matched=True),
          _mnli_split_generator(
              "validation_mismatched", data_dir, "dev", matched=False),
          _mnli_split_generator("test_matched", data_dir, "test", matched=True),
          _mnli_split_generator(
              "test_mismatched", data_dir, "test", matched=False)
      ]
    elif self.builder_config.name == "mnli_matched":
      return [
          _mnli_split_generator("validation", data_dir, "dev", matched=True),
          _mnli_split_generator("test", data_dir, "test", matched=True)
      ]
    elif self.builder_config.name == "mnli_mismatched":
      return [
          _mnli_split_generator("validation", data_dir, "dev", matched=False),
          _mnli_split_generator("test", data_dir, "test", matched=False)
      ]
    else:
      return [
          train_split,
          tfds.core.SplitGenerator(
              name=tfds.Split.VALIDATION,
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir or "", "dev.tsv"),
                  "split": "dev",
                  "mrpc_files": mrpc_files,
              }),
          tfds.core.SplitGenerator(
              name=tfds.Split.TEST,
              num_shards=1,
              gen_kwargs={
                  "data_file": os.path.join(data_dir or "", "test.tsv"),
                  "split": "test",
                  "mrpc_files": mrpc_files,
              }),
      ]

  def _generate_examples(self, data_file, split, mrpc_files):
    if self.builder_config.name == "mrpc":
      // We have to prepare the MRPC dataset from the original sources ourselves.
      examples = self._generate_example_mrpc_files(
Italian Trulli
In pattern: SUPERPATTERN

Frequency: 3

Non-data size: 9

Instances


Project Name: tensorflow/datasets
Commit Name: 830a663e996b7b95c45acddd58ba6e6532599c0c
Time: 2019-07-16
Author: adarob@google.com
File Name: tensorflow_datasets/text/glue.py
Class Name: Glue
Method Name: _split_generators


Project Name: tensorflow/datasets
Commit Name: 830a663e996b7b95c45acddd58ba6e6532599c0c
Time: 2019-07-16
Author: adarob@google.com
File Name: tensorflow_datasets/text/glue.py
Class Name: Glue
Method Name: _split_generators


Project Name: tensorflow/datasets
Commit Name: c4b3882498d4c76ce2d824f458f06481a294e78a
Time: 2019-09-30
Author: ouwen.huang@duke.edu
File Name: tensorflow_datasets/image/duke_ultrasound.py
Class Name: DukeUltrasound
Method Name: _split_generators


Project Name: tensorflow/datasets
Commit Name: c79105b01b07fb2299fe6e69730b5ae5f6e1352b
Time: 2020-02-21
Author: no-reply@google.com
File Name: tensorflow_datasets/audio/librispeech.py
Class Name: Librispeech
Method Name: _split_generators