tf.train.Example | TensorFlow v2.16.1 (original) (raw)

tf.train.Example

Stay organized with collections Save and categorize content based on your preferences.

An Example is a standard proto storing data for training and inference.

View aliases

Compat aliases for migration

tf.compat.v1.train.Example

Used in the notebooks

Used in the guide	Used in the tutorials
tf.data: Build TensorFlow input pipelines Estimators Ragged tensors	TFRecord and tf.train.Example Graph-based Neural Structured Learning in TFX TensorFlow Ranking Keras pipeline for distributed training FaceSSD Fairness Indicators Example Colab Graph regularization for sentiment classification using synthesized graphs

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

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

It contains a key-value store Example.features where each key (string) maps to a tf.train.Feature message which contains a fixed-type list. This flexible and compact format allows the storage of large amounts of typed data, but requires that the data shape and use be determined by the configuration files and parsers that are used to read and write this format (refer totf.io.parse_example for details).

from google.protobuf import text_format example = text_format.Parse(''' features { feature {key: "my_feature" value {int64_list {value: [1, 2, 3, 4]} } } }''', tf.train.Example())

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.int64)}) {'my_feature': <tf.Tensor: shape=(4,), dtype=float32, numpy=array([1, 2, 3, 4], dtype=int64)>}

While the list of keys, and the contents of each key could be different for every Example, TensorFlow expects a fixed list of keys, each with a fixedtf.dtype. A conformant Example dataset obeys the following conventions:

If a Feature K exists in one example with data type T, it must be of type T in all other examples when present. It may be omitted.
The number of instances of Feature K list data may vary across examples, depending on the requirements of the model.
If a Feature K doesn't exist in an example, a K-specific default will be used, if configured.
If a Feature K exists in an example but contains no items, the intent is considered to be an empty tensor and no default will be used.

Attributes
features	Features features