Link Search Menu Expand Document
enuSpace-tensorflow Manual

RecordInput

tensorflow C++ API

tensorflow::ops::RecordInput

Emits randomized records.

Summary

Arguments:

  • scope: A Scope object
  • file_pattern: Glob pattern for the data files.

Optional attributes (seeAttrs):

  • file_random_seed: Random seeds used to produce randomized records.
  • file_shuffle_shift_ratio: Shifts the list of files after the list is randomly shuffled.
  • file_buffer_size: The randomization shuffling buffer.
  • file_parallelism: How many sstables are opened and concurrently iterated over.
  • batch_size: The batch size.

Returns:

  • Output : A tensor of shape [batch_size].

Constructor

  • RecordInput(const ::tensorflow::Scope & scope, StringPiece file_pattern, const RecordInput::Attrs & attrs)).

Public attributes

  • tensorflow::Output records.

RecordInput block

Source link : https://github.com/EXPNUNI/enuSpace-Tensorflow/blob/master/enuSpaceTensorflow/tf_data_flow_ops.cpp

Argument:

  • Scope scope : A Scope object (A scope is generated automatically each page. A scope is not connected.)

  • component_types: Input DataType List in accordance with each input.

  • RecordInput::Attrs attrs : input attrs data. ex) file_random_seed_ = 301; file_shuffle_shift_ratio_ = 0.0f;file_buffer_size_ = 10000; file_parallelism_ = 16; batch_size_ = 32;

Return:

  • Output records: Output object of RecordInput class object.

Result:

  • std::vector(Tensor) product_result : Returned object of executed result by calling session.

Using Method

Need sample data file.