Module: tf.contrib.lite

Defined in tensorflow/contrib/lite/python/

TensorFlow Lite tooling helper functionality.

EXPERIMENTAL: APIs here are unstable and likely to change without notice.


constants module: Constants for TFLite.


class ConverterMode: Enum class defining the converters available to generate TFLite models.

class DecodeError: Common base class for all non-exit exceptions.

class Interpreter: Interpreter inferace for TF-Lite Models.

class OpHint: A class that helps build tflite function invocations.

class TFLiteConverter: Convert a TensorFlow model into output_format using TOCO.

class TocoConverter: Convert a TensorFlow model into output_format using TOCO.


build_toco_convert_protos(...): Builds protocol buffers describing a conversion of a model using TOCO.

convert_op_hints_to_stubs(...): Converts a graphdef with LiteOp hints into stub operations.

toco_convert(...): Convert a model using TOCO. (deprecated)

toco_convert_protos(...): Convert input_data_str according to model and toco parameters.

Other Members