View source on GitHub
|
Input data parsing for tf-ranking library.
Supports the following data formats:
- tf.train.SequenceExample
- tf.train.Example in tf.train.Example.
Functions
build_ranking_dataset(...): Builds a ranking tf.dataset with a standard data format.
build_ranking_dataset_with_parsing_fn(...): Builds a ranking tf.dataset using the provided parsing_fn.
build_ranking_serving_input_receiver_fn(...): Returns a serving input receiver fn for a standard data format.
build_ranking_serving_input_receiver_fn_with_parsing_fn(...): Returns a receiver function with the provided parsing_fn.
build_sequence_example_serving_input_receiver_fn(...): Creates a serving_input_receiver_fn for SequenceExample inputs.
build_tf_example_serving_input_receiver_fn(...): Builds a serving input fn for tensorflow.training.Example.
make_parsing_fn(...): Returns a parsing fn for a standard data format.
parse_from_example_in_example(...): Parses an ExampleInExample batch to a feature map.
parse_from_example_list(...): Parses an ExampleListWithContext batch to a feature map.
parse_from_sequence_example(...): Parses SequenceExample to feature maps.
parse_from_tf_example(...): Parse function to convert tf.train.Example to feature maps.
read_batched_sequence_example_dataset(...): Returns a Dataset of features from SequenceExample.