tf.train.BytesList

Used in tf.train.Example protos. Holds a list of byte-strings.

Used in the notebooks

Used in the tutorials

An Example proto is a representation of the following python type:

Dict[str,
     Union[List[bytes],
           List[int64],
           List[float]]]

This proto implements the List[bytes] portion.

from google.protobuf import text_format
example = text_format.Parse('''
  features {
    feature {key: "my_feature"
             value {bytes_list {value: ['abc', '12345' ]} } }
  }''',
  tf.train.Example())

example.features.feature['my_feature'].bytes_list.value
["abc", "12345"]

Use tf.io.parse_example to extract tensors from a serialized Example proto:

tf.io.parse_example(
    example.SerializeToString(),
    features = {'my_feature': tf.io.RaggedFeature(dtype=tf.string)})
{'my_feature': <tf.Tensor: shape=(2,), dtype=string,
                           numpy=array([b'abc', b'12345'], dtype=object)>}

See the tf.train.Example guide for usage details.

value repeated bytes value